Quantum

Probability and
Randomness
Edited by
Andrei Khrennikov and Karl Svozil
Printed Edition of the Special Issue Published in Entropy

www.mdpi.com/journal/entropy
Quantum Probability and Randomness

www.pdfgrip.com
www.pdfgrip.com
Quantum Probability and Randomness

Special Issue Editors

Andrei Khrennikov
Karl Svozil

MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade

www.pdfgrip.com
Special Issue Editors
Andrei Khrennikov Karl Svozil
Linnaeus University Institute for Theoretical Physics of the
Sweden Vienna Technical University
Austria

Editorial Office
MDPI
St. Alban-Anlage 66
4052 Basel, Switzerland

This is a reprint of articles from the Special Issue published online in the open access journal Entropy
(ISSN 1099-4300) from 2018 to 2019 (available at: https://www.mdpi.com/journal/entropy/special
issues/Probability Randomness)

For citation purposes, cite each article independently as indicated on the article page online and as
indicated below:

LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. Journal Name Year, Article Number,
Page Range.

ISBN 978-3-03897-714-8 (Pbk)

ISBN 978-3-03897-715-5 (PDF)

Cover image courtesy of R.C.-Z. Quehenberger.


c 2019 by the authors. Articles in this book are Open Access and distributed under the Creative
Commons Attribution (CC BY) license, which allows users to download, copy and build upon
published articles, as long as the author and publisher are properly credited, which ensures maximum
dissemination and a wider impact of our publications.
The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons
license CC BY-NC-ND.

www.pdfgrip.com
Contents

About the Special Issue Editors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

Andrei Khrennikov and Karl Svozil

Quantum Probability and Randomness
Reprinted from: Entropy 2019, 21, 35, doi:10.3390/e21010035 . . . . . . . . . . . . . . . . . . . . . 1

Mladen Pavičić and Norman D. Megill

Vector Generation of Quantum Contextual Sets in Even Dimensional Hilbert Spaces
Reprinted from: Entropy 2018, 20, 928, doi:10.3390/e20120928 . . . . . . . . . . . . . . . . . . . . . 6

Aldo C. Martı́nez, Aldo Solı́s, Rafael Dı́az Hernández Rojas, Alfred B. U’Ren, Jorge G. Hirsch
and Isaac Pérez Castillo
Advanced Statistical Testing of Quantum Random Number Generators
Reprinted from: Entropy 2018, 20, 886, doi:10.3390/e20110886 . . . . . . . . . . . . . . . . . . . . . 18

Maria Luisa Dalla Chiara, Hector Freytes, Roberto Giuntini, Roberto Leporini and
Giuseppe Sergioli
Probabilities and Epistemic Operations in the Logics of Quantum Computation
Reprinted from: Entropy 2018, 20, 837, doi:10.3390/e20110837 . . . . . . . . . . . . . . . . . . . . . 31

Xiaomin Guo, Ripeng Liu, Pu Li, Chen Cheng, Mingchuan Wu and Yanqiang Guo
Enhancing Extractable Quantum Entropy in Vacuum-Based Quantum Random Number
Generato
Reprinted from: Entropy 2018, 20, 819, doi:10.3390/e20110819 . . . . . . . . . . . . . . . . . . . . . 53

Marco Enrı́quez, Francisco Delgado and Karol Życzkowski

Entanglement of Three-Qubit Random Pure States
Reprinted from: Entropy 2018, 20, 745, doi:10.3390/e20100745 . . . . . . . . . . . . . . . . . . . . . 66

Margarita A. Man’ko and Vladimir I. Man’ko

New Entropic Inequalities and Hidden Correlations in Quantum Suprematism Picture of Qudit
States
Reprinted from: Entropy 2018, 20, 692, doi:10.3390/e20090692 . . . . . . . . . . . . . . . . . . . . . 85

Arkady Plotnitsky
“The Heisenberg Method”: Geometry, Algebra, and Probability in Quantum Theory
Reprinted from: Entropy 2018, 20, 656, doi:10.3390/e20090656 . . . . . . . . . . . . . . . . . . . . . 102

Francisco Delgado
SU (2) Decomposition for the Quantum Information Dynamics in 2d-Partite Two-Level
Quantum Systems
Reprinted from: Entropy 2018, 20, 610, doi:10.3390/e20080610 . . . . . . . . . . . . . . . . . . . . . 148

Marius Nagy and Naya Nagy

An Information-Theoretic Perspective on the Quantum Bit Commitment Impossibility Theorem
Reprinted from: Entropy 2018, 20, 193, doi:10.3390/e20030193 . . . . . . . . . . . . . . . . . . . . . 189

Gregg Jaeger
Developments in Quantum Probability and the Copenhagen Approach
Reprinted from: Entropy 2018, 20, 420, doi:10.3390/e20060420 . . . . . . . . . . . . . . . . . . . . . 205

www.pdfgrip.com
Hans Havlicek and Karl Svozil
Dimensional Lifting through the Generalized Gram–Schmidt Process
Reprinted from: Entropy 2018, 20, 284, doi:10.3390/e20040284 . . . . . . . . . . . . . . . . . . . . . 224

Andrei Khrennikov, Alexander Alodjants, Anastasiia Trofimova and Dmitry Tsarev

On Interpretational Questions for Quantum-Like Modeling of Social Lasing
Reprinted from: Entropy 2018, 20, 921, doi:10.3390/e20120921 . . . . . . . . . . . . . . . . . . . . . 229

Paul Ballonoff
Paths of Cultural Systems
Reprinted from: Entropy 2018, 20, 8, doi:10.3390/e20010008 . . . . . . . . . . . . . . . . . . . . . . 253

vi

www.pdfgrip.com
About the Special Issue Editors
Andrei Khrennikov was born in 1958 in Volgorad and spent his childhood in the town of Bratsk,
in Siberia, north from the lake Baikal. In the period between 1975–1980, he studied at Moscow
State University, department of Mechanics and Mathematics, and in 1983, he received his PhD
in mathematical physics (quantum field theory) at the same department. In 1990, he became
full professor at Moscow University for Electronic Engineering. Since 1997, he has been a
professor of applied mathematics at Linnaueus University, South-East Sweden, and since 2002,
the director of the multidisciplinary research center at this university, as well as the International
Center for Mathematical Modeling in Physics, Engineering, Economics, and Cognitive Science.
His research interests are multidisciplinary, e.g., foundations of quantum physics, information, and
probability, cognitive modeling, ultrametric (non-Archimedean) mathematics, dynamical systems,
infinite-dimensional analysis, quantum-like models in psychology, and economics and finances.
He is the author of approcimately 500 papers and 20 monographs in mathematics, physics, biology,
psychology, cognitive science, economics, and finances.

Karl Svozil is a professor of theoretical physics at Vienna’s University of Technology. He earned a

Dr. Phil. while studying philosophy and sciences in the old, “Humboldtian” tradition in Heidelberg
and Vienna, emphasizing the unity of knowledge. After attending the Lawrence Berkeley Laboratory
and UC Berkeley, he worked as a physicist in Vienna, with many shorter stays abroad—among them,
the Lomonosov Moscow State University, Lebedev Physical Institute and ICPT Trieste. He recently
held an honorary position at the University of Auckland and served as president of the International
Quantum Structures Association. Svozil’s main interests include quantum logic, issues related to
(in)determininsism in physics, and “relativizing” relativity theory in the spirit of Alexandrov’s
theorem of incidence geometry.

vii

www.pdfgrip.com
www.pdfgrip.com
 entropy
Editorial
Quantum Probability and Randomness
Andrei Khrennikov 1,∗ and Karl Svozil 2
1 International Center for Mathematical Modeling in Physics, Engineering, Economics, and Cognitive Science,
Linnaeus University, 351 95 Växjö, Sweden
2 Institute for Theoretical Physics, Vienna University of Technology, Wiedner Hauptstrasse 8-10/136,
1040 Vienna, Austria; svozil@tuwien.ac.at
* Correspondence: Andrei.Khrennikov@lnu.se

Received: 2 January 2019; Accepted: 3 January 2019; Published: 7 January 2019

Keywords: quantum foundations; probability; irreducible randomness; random number generators;

quantum technology; entanglement; quantum-like models for social stochasticity; contextuality

The recent quantum information revolution has stimulated interest in the quantum foundations
by perceiving and re-evaluating the theory from a novel information-theoretical viewpoint [1–5].
Quantum probability and randomness play the crucial role in foundations of quantum mechanics.
It might not be totally unreasonable to claim that, already starting from some of the earliest
(in hindsight) indications of quanta in the 1902 Rutherford–Soddy exponential decay law and the
small aberrations predicted by Schweidler [6], the tide of indeterminism [7,8] was rolling against
chartered territories of fin de siécle mechanistic determinism. Riding the waves were researchers like
Exner, who already in his 1908 inaugural lecture as rector magnificus [9] postulated that irreducible
randomness is, and probability theory therefore needs to be, at the heart of all sciences; natural as well
as social. Exner [10] was forgotten but cited in Schrödinger’s alike “Zürcher Antrittsvorlesung” of
1922 [11]. Not much later Born expressed his inclinations to give up determinism in the world of the
atoms [12], thereby denying the existence of some inner properties of the quanta which condition a
definite outcome for, say, the scattering after collisions.
Von Neumann [13] was among the first who emphasized this new feature which was very different
from the “in principle knowable unknowns” grounded in epistemology alone. Quantum randomness
was treated as individual randomness; that is, as if single electrons or photons are sometimes capable of
behaving acausally and irreducibly randomly. Such randomness cannot be reduced to a variability
of properties of systems in some ensemble. Therefore, quantum randomness is often considered as
irreducible randomness.
Von Neumann understood well that it is difficult, if not outright impossible in general, to check
empirically the randomness for individual systems, say for electrons or photons. In particular, he
proceeded with the statistical interpretation of probability based on the mathematical model of von
Mises [14,15] based upon relative frequencies after admissible place selections.
At the same time, it is just and fair to note that the aforementioned tendencies to ground
physics, and by reductionism, all of science, in ontological indeterminism, have been strongly
contested and fiercely denied by eminent physicists; most prominently by Einstein. Planck [16]
(p. 539) (see also Earman [17] (p. 1372)) believed that causality could be neither generally proved nor
generally disproved. He suggested to postulate causality as a working hypothesis, a heuristic principle,
a sign-post (and for Planck the most valuable sign-post we possess) “to guide us in the motley confusion
of events”.
This is a good place to remark that random features of an individual system can be discussed in
the framework of subjective probability theory. The individual (irreducible) interpretation of quantum
randomness due to von Neumann matches well with the subjective probability interpretation of
quantum mechanics (QBism, see, e.g., [18,19]).

Entropy 2019, 21, 35; doi:10.3390/e21010035 1 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2019, 21, 35

The main reason for keeping the statistical interpretation was that the aforementioned individual
randomness of quantum systems was considered by von Neumann as one of the basic features of
nature (and not of the human mind!). Von Neumann was sure that such a natural phenomenon must
be treated statistically (by the same reason Bohr also treated quantum randomness statistically, see [20]
for details).
In particular, von Neumann remarked [13] (pp. 301–302), that, for measurement of some quantity
R for an ensemble of systems (of any origin),

It is not surprising that R does not have a sharp value . . ., and that a positive dispersion exists.
However, two different reasons for this behavior a priori conceivable:

1. The individual systems S1, . . . , SN of our ensemble can be in different states, so that the ensemble
[S1, . . . , SN ] is defined by their relative frequencies. The fact that we do not obtain sharp values
for the physical quantities in this case is caused by our lack of information: we do not know in
which state we are measuring, and therefore we cannot predict the results.
2. All individual systems S1, . . . , SN are in the same state, but the laws of nature are not causal.
Then, the cause of the dispersion is not our lack of information, but nature itself, which has
disregarded the principle of sufficient cause.

These are characterizations of epistemic and ontic indeterminism, respectively. Von Neumann
favored the second, ontic, case which he considered “important and new” (and which he believed
to be able to corroborate [21]). Therefore, for von Neumann, quantum randomness is essentially a
statistical exhibition of violation of causality, a violation of the principle of sufficient cause.
We compare this kind of randomness with classical interpretations of randomness, see, e.g.,
Chapter 2 [22]:

1. unpredictability (von Mises),

2. complexity-incompressibility (Kolmogorov, Solomonof, Chaitin),
3. typicality (Martin-Löf).

It seems that the interpretation of randomness as unpredictability (von Mises) is very close to the
interpretation of quantum randomness as an exhibition of acausality.
The article by Pavicic and Megill [23], Vector Generation of Quantum Contextual Sets in Even
Dimensional Hilbert Spaces, is a novel contribution to quantum contextuality theory. As is well
known, the most elaborated contextual sets, which offer blueprints for contextual experiments and
computational gates, are the Kochen–Specker sets. In this paper, a method of vector generation that
supersedes previous methods is presented. It is implemented by means of algorithms and programs
that generate hypergraphs embodying the Kochen-Specker property and that are designed to be
carried out on supercomputers.
Recent years were characterized by the tremendous development of quantum technology.
Quantum random generators are among the most important outputs of this development. As is
pointed out in the review by Martínez et al. [24], Advanced Statistical Testing of Quantum Random
Number Generators, the natural laws of the microscopic realm provide a fairly simple method to
generate non-deterministic sequences of random numbers, based on measurements of quantum states.
In practice, however, the experimental devices on which quantum random number generators are
based are often unable to pass some tests of randomness. In this review, two such tests are briefly
discussed, the challenges that have to be encountered in experimental implementations are pointed
out. Finally, the authors present a fairly simple method that successfully generates non-deterministic
maximally random sequences.
The connection between quantum logic and quantum probability is highlighted by
Dalla Chiara et al. [25] in the paper entitled Probabilities and Epistemic Operations in the Logics of Quantum
Computation. The authors stress that quantum computation theory has inspired new forms of quantum

www.pdfgrip.com
Entropy 2019, 21, 35

logic, called quantum computational logics. In this article, they investigate the epistemic operation
(which is informally used in a number of interesting quantum situations): the operation “being
probabilistically informed”.
In the paper entitled Enhancing Extractable Quantum Entropy in Vacuum-Based Quantum Random
Number Generator, Guo et al. [26] commit to enhancing quantum entropy content in the vacuum
noise based quantum RNG. They have taken into account main factors in this proposal to establish
the theoretical model of quantum entropy content, including the effects of classical noise, the
optimum dynamical analog-digital convertor (ADC) range, the local gain and the electronic gain
of the homodyne system.
The work by Enríquez et al. [27], Entanglement of Three-Qubit Random Pure States, is devoted
to studying entanglement properties of generic three-qubit pure states. There are obtained the
distributions of both the coefficients and the only phase in the five-term decomposition of Acín et al. for
an ensemble of random pure states generated by the Haar measure on U (8). Furthermore, the authors
analyze the probability distributions of two sets of polynomial invariants. One of these sets allows
us to classify three-qubit pure states into four classes. Entanglement in each class is characterized
using the minimal Renyi–Ingarden–Urbanik entropy. The numerical findings suggest some conjectures
relating some of those invariants with entanglement properties to be ground in future analytical work.
In the article New Entropic Inequalities and Hidden Correlations in Quantum Suprematism Picture
of Qubit States, Margarita A. Man’ko and Vladimir I. Man’ko [28] considered an analog of Bayes’
formula and the nonnegativity property of mutual information for systems with one random variable.
For single-qubit states, they presented new entropic inequalities in the form of the subadditivity and
condition corresponding to hidden correlations in quantum systems. Qubit states are represented in
the quantum suprematism picture, where these states are identified with three probability distributions,
describing the states of three classical coins, and illustrating the states by Triada of Malevich’s squares
with areas satisfying the quantum constraints.
In the article by Plotnitsky [29], “The Heisenberg Method”: Geometry, Algebra, and Probability in
Quantum Theory, quantum theory is reconsidered in terms of the following principle, which can be
symbolically represented as QUANTUMNESS→PROBABILITY→ALGEBRA. The principle states
that the quantumness of physical phenomena, that is, the specific character of physical phenomena
known as quantum, implies that our predictions concerning them are irreducibly probabilistic, even in
dealing with quantum phenomena resulting from the elementary individual quantum behavior (such
as that of elementary particles), which in turn implies that our theories concerning these phenomena
are fundamentally algebraic, in contrast to more geometrical classical or relativistic theories, although
these theories, too, have an algebraic component to them.
The work by Delgado [30], SU (2) Decomposition for the Quantum Information Dynamics in 2d-Partite
Two-Level Quantum Systems, presents a formalism to decompose the quantum information dynamics in
SU (22d ) for 2d-partite two-level systems into 2d−1 SU (2) quantum subsystems. It generates an easier
and more direct physical implementation of quantum processing developments for qubits.
The paper by Marius Nagy and Naya Nagy [31], An Information-Theoretic Perspective on the
Quantum Bit Commitment Impossibility Theorem, proposes a different approach to pinpoint the causes
for which an unconditionally secure quantum bit commitment protocol cannot be realized, beyond the
technical details on which the proof of Mayers’ no-go theorem is constructed.
In the Copenhagen approach to quantum mechanics as characterized by Heisenberg, probabilities
relate to the statistics of measurement outcomes on ensembles of systems and to individual
measurement events via the actualization of quantum potentiality. In the review by Jaeger [32],
Developments in Quantum Probability and the Copenhagen Approach, brief summaries are given of a series
of key results of different sorts that have been obtained since the final elements of the Copenhagen
interpretation were offered and it was explicitly named so by Heisenberg—in particular, results
from the investigation of the behavior of quantum probability since that time, the mid-1950s. This
review shows that these developments have increased the value to physics of notions characterizing

www.pdfgrip.com
Entropy 2019, 21, 35

the approach which were previously either less precise or mainly symbolic in character, including
complementarity, indeterminism, and unsharpness.
A new way of orthogonalizing ensembles of vectors by “lifting” them to higher dimensions is
introduced by Havlicek and Svozil [33] entitled Dimensional Lifting through the Generalized Gram-Schmidt
Process. This method can potentially be utilized for solving quantum decision and computing problems.
Recently the mathematical formalism and methodology of quantum theory started to be widely
applied outside of physics, especially in psychology, decision making, social and political science
(see, e.g., [34]). This special issue contains one paper belonging to this area of research, the article of
Khrennikov et al. [35], On Interpretational Questions for Quantum-Like Modeling of Social Lasing. The
formalisms of quantum field theory and theory of open quantum systems are applied to modeling
socio-political processes on the basis of the social laser model describing stimulated amplification of
social actions. The main aim of this paper is establishing the socio-psychological interpretations of the
quantum notions playing the basic role in lasing modeling.
The article by Paul Ballonoff [36], Paths of Cultural Systems, is also devoted to applications outside
physics, namely to anthropology. A theory of cultural structures predicts the objects observed by
anthropologists. A viable history (defined using pdqs) states how an individual in a population
following such history may perform culturally allowed associations, which allows a viable history to
continue to survive. The vector states on sets of viable histories identify demographic observables on
descent sequences.
We hope that the reader will enjoy the present issue, which will be useful to experts working in
all domains of quantum physics and quantum information theory, ranging from experimenters, to
theoreticians and philosophers.
The cover of this electronic book was created by Renate Quehenberg and the editors would like to
thank her for the graphical contribution to this special issue.

Acknowledgments: We express our thanks to the authors of the above contributions, and to the journal Entropy
and MDPI for their support during this work.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Khrennikov, A.; Weihs, G. Preface of the special issue Quantum foundations: Theory and experiment. Found.
Phys. 2012, 42, 721–724. [CrossRef]
2. Bengtsson, I.; Khrennikov, A. Preface. Found. Phys. 2011, 41, 281–281. [CrossRef]
3. D’Ariano, G.M.; Jaeger, G.; Khrennikov, A.; Plotnitsky, A. Preface of the special issue Quantum theory:
Advances and problems. Phys. Scr. 2014, T163, 010301.
4. Khrennikov, A.; de Raedt, H.; Plotnitsky, A.; Polyakov, S. Preface of the special issue Probing the
limits of quantum mechanics: Theory and experiment, Volume 1. Found. Phys. 2015, 45, 707–710;
doi:10.1007/s10701-015-9911-8. [CrossRef]
5. Khrennikov, A.; de Raedt, H.; Plotnitsky, A.; Polyakov, S. Preface of the special issue Probing the limits of
quantum mechanics: Theory and experiment, Volume 2. Found. Phys. 2015, doi:10.1007/s10701-015-9950-1.
[CrossRef]
6. Von Schweidler, E. Über Schwankungen der Radioaktiven Umwandlung; H. Dunod & E. Pinat: Paris, France,
1906; pp. 1–3. (In German)
7. Hiebert, E.N. Common frontiers of the exact sciences and the humanities. Phys. Perspect. 2000, 2, 6–29.
[CrossRef]
8. Stöltzner, M., Vienna indeterminism: Mach, Boltzmann, Exner. Synthese 1999, 119, 85–111. [CrossRef]
9. Exner, F.S. Über Gesetze in Naturwissenschaft und Humanistik: Inaugurationsrede Gehalten am 15. Oktober 1908;
A. Hölder: Wien, Austria, 2016.
10. Exner, F.S. Vorlesungen über die Physikalischen Grundlagen der Naturwissenschaften; F. Deuticke: Leipzig und
Wien, Germany, 1922.
11. Schrödinger, E. Was ist ein Naturgesetz? Naturwissenschaften 1929, 17, 1. [CrossRef]

www.pdfgrip.com
Entropy 2019, 21, 35

12. Born, M. Zur Quantenmechanik der Stoßvorgänge. Z. Phys. 1926, 37, 863–867. [CrossRef]
13. Von Neuman, J. Mathematical Foundations of Quantum Mechanics; Princeton University Press: Princeton, NJ,
USA, 1955.
14. Von Mises, R. Grundlagen der Wahrscheinlichkeitsrechnung. Math. Z. 1919, 5, 52–99. [CrossRef]
15. Von Mises, R. The Mathematical Theory of Probability and Statistics; Academic Press: London, UK, 1964.
16. Planck, M. The concept of causality. Proc. Phys. Soc. 1932, 44, 529–539. [CrossRef]
17. Earman, J. Aspects of determinism in modern physics. In Part B: Philosophy of Physics, Handbook of the
Philosophy of Science; Butterfield, J., Earman, J., Eds.; North-Holland: Amsterdam, The Netherlands, 2007;
pp. 1369–1434.
18. Fuchs, C.A. Quantum mechanics as quantum information (and only a little more). In Quantum Theory:
Reconsideration of Foundations; Växjö University Press: Växjö, Schweden, 2002; pp. 463–543.
19. Fuchs, C.A.; Schack, R. QBism and the Greeks: why a quantum state does not represent an element of
physical reality. Phys. Scr. 2014, 90, 015104. [CrossRef]
20. Plotnitsky, A.; Khrennikov, A. Reality without realism: On the ontological and epistemological architecture
of quantum mechanics. Found. Phys. 2015, 45, 1269–1300. [CrossRef]
21. Dieks, D. Von Neumann’s impossibility proof: Mathematics in the service of rhetorics. Stud. Hist. Philos.
Mod. Phys. 2017, 60, 136–148. [CrossRef]
22. Khrennikov, A. Probability and Randomness: Quantum Versus Classical; Imperial College Press: London,
UK, 2016.
23. Pavicic, M.; Megill, N.D. Vector Generation of Quantum Contextual Sets in Even Dimensional Hilbert Spaces.
Entropy 2018, 20, 928. [CrossRef]
24. Martínez, A.C.; Solis, A.; Rojas, R.D.H.; U’ Ren, A.B.; Hirsch, J.G.; Castillo, I.P. Advanced Statistical Testing
of Quantum Random Number Generators. Entropy 2018, 20, 886. [CrossRef]
25. Dalla Chiara, M.L.; Freytes, H.; Giuntini, R.; Leporini, R.; Sergioli, G. Probabilities and Epistemic Operations
in the Logics of Quantum Computation. Entropy 2018, 20, 837. [CrossRef]
26. Guo, X.; Liu, R.; Li, P.; Cheng, C.; Wu, M.; Guo, Y. Enhancing Extractable Quantum Entropy in Vacuum-Based
Quantum Random Number Generator. Entropy 2018, 20, 819. [CrossRef]
27. Enríquez, M.; Delgado, F.; Zyczkowski, K. Entanglement of Three-Qubit Random Pure States. Entropy 2018,
20, 745. [CrossRef]
28. Man’ko, M.A.; Man’ko, V.I. New Entropic Inequalities and Hidden Correlations in Quantum Suprematism
Picture of Qubit States. Entropy 2018, 20, 692. [CrossRef]
29. Plotnitsky, A. The Heisenberg Method: Geometry, Algebra, and Probability in Quantum Theory. Entropy
2018, 20, 656. [CrossRef]
30. Delgado, F. SU (2) Decomposition for the Quantum Information Dynamics in 2d-Partite Two-Level Quantum
Systems. Entropy 2018, 20, 610. [CrossRef]
31. Nagy, M.; Nagy, N. An Information-Theoretic Perspective on the Quantum Bit Commitment Impossibility
Theorem. Entropy 2018, 20, 193. [CrossRef]
32. Jaeger, G. Developments in Quantum Probability and the Copenhagen Approach. Entropy 2018, 20, 420.
[CrossRef]
33. Havlicek, H.; Karl Svozil, K. Dimensional Lifting through the Generalized Gram-Schmidt Process. Entropy
2018, 20, 284. [CrossRef]
34. Khrennikov, A. Ubiquitous Quantum Structure: From Psychology to Finances; Springer: Berlin/Heidelberg,
Germany; New York, NY, USA, 2010.
35. Khrennikov, A.; Alodjants, A.; Trofimova, A.; Tsarev, D. On Interpretational Questions for Quantum-Like
Modeling of Social Lasing. Entropy 2018, 20, 921. [CrossRef]
36. Ballonoff, P. Paths of Cultural Systems. Entropy 2018, 20, 8. [CrossRef]

c 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

www.pdfgrip.com
 entropy
Article
Vector Generation of Quantum Contextual Sets in
Even Dimensional Hilbert Spaces
Mladen Pavičić 1,2, *,† and Norman D. Megill 3,†
1 Nano Optics, Department of Physics, Humboldt University, 12489 Berlin, Germany
2 Center of Excellence for Advanced Materials and Sensors, Research Unit Photonics and Quantum Optics,
Institute Ruder Bošković, 10000 Zagreb, Croatia
3 Boston Information Group, Lexington, MA 02420, USA; nm@alum.mit.edu
* Correspondence: mpavicic@irb.hr
† These authors contributed equally to this work.

Received: 29 October 2018; Accepted: 24 November 2018; Published: 5 December 2018

Abstract: Recently, quantum contextuality has been proved to be the source of quantum
computation’s power. That, together with multiple recent contextual experiments, prompts improving
the methods of generation of contextual sets and finding their features. The most elaborated
contextual sets, which offer blueprints for contextual experiments and computational gates, are
the Kochen–Specker (KS) sets. In this paper, we show a method of vector generation that supersedes
previous methods. It is implemented by means of algorithms and programs that generate hypergraphs
embodying the Kochen–Specker property and that are designed to be carried out on supercomputers.
We show that vector component generation of KS hypergraphs exhausts all possible vectors that can
be constructed from chosen vector components, in contrast to previous studies that used incomplete
lists of vectors and therefore missed a majority of hypergraphs. Consequently, this unified method
is far more efficient for generations of KS sets and their implementation in quantum computation
and quantum communication. Several new KS classes and their features have been found and are
elaborated on in the paper. Greechie diagrams are discussed.

Keywords: quantum contextuality; Kochen–Specker sets; MMP hypergraphs; Greechie diagrams

1. Introduction
Recently, it has been discovered that quantum contextuality might have a significant place in
a development quantum communication [1,2], quantum computation [3,4], and lattice theory [5,6].
This has prompted experimental implementation of 4-, 6-, and 8-dimensional contextual experiments
with photons [7–13], neutrons [14–16], trapped ions [17], solid state molecular nuclear spins [18],
and paths [19,20].
Experimental contextual tests involve subtle issues, such as the possibility of noncontextual hidden
variable models that can reproduce quantum mechanical predictions up to arbitrary precision [21].
These models are important because they show how assignments of predetermined values to dense
sets of projection operators are precluded by any quantum model. Thus, Spekkens [22] introduces
generalised noncontextuality in an attempt to make precise the distinction between classical and
quantum theories, distinguishing the notions of preparation, transformation, and measurement of
noncontextuality and by doing so demonstrates that even the 2D Hilbert space is not inherently
noncontextual. Kunjwal and Spekkens [23] derive an inequality that does not assume that the value
assignments are deterministic, showing that noncontextuality cannot be salvaged by abandoning
determinism. Kunjwal [24] shows how to compute a noncontextuality inequality from an invariant
derived from a contextual set/configuration representing an experimental Kochen-Specker (KS) setup.
This opens up the possibility of finding contextual sets that provide the best noise robustness in

Entropy 2018, 20, 928; doi:10.3390/e20120928 6 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2018, 20, 928

demonstrating contextuality. The large number of such sets that we show in the present work can
provide a rich source for such an effort.
Quantum contextual configurations that have been elaborated on the most in the literature are the
KS sets, and, in this paper, we consider just them. In order to obtain KS sets, so far, various methods of
exploiting correlations, symmetries, geometry, qubit states, Pauli states, Lie algebras, etc., have been
found and used for generating master sets i.e., big sets which contain all smaller contextual sets [25–37].
All of these methods boil down either to finding a list of vectors and their n-tuples of
orthogonalities from which a master set can be read off or finding a structure, e.g., a polytope,
from which again a list of vectors and orthogonalities can be read off as well as a master set they build.
In the present paper, we take the simplest possible vector components within an n-dimensional Hilbert
space, e.g., {0, ±1}, and via our algorithms and programs exhaustively build all possible vectors and
their orthogonal n-tuples and then filter out KS sets from the sets in which the vectors are organized.
For a particular choice of components, the chances of getting KS sets are very high. We generate KS sets
for even-dimensional spaces, up to 32, that properly contain all previously obtained and known KS
sets, present their features and distributions, give examples of previously unknown sets, and present a
blueprint for implementation of a simple set with a complex coordinatization.

2. Results
The main results presented in this paper concern generation of contextual sets from several basic
vector components. Previous contextual sets from the literature made use of often complicated
sets of vectors that the authors arrived at, following particular symmetries, or geometries,
or polytope correlations, or Pauli operators, or qubit states, etc. In contrast, our approach considers
McKay–Megill–Pavičić (MMP) hypergraphs (defined in Section 2.1) from n-dimensional (nD) Hilbert
space (H n , n ≥ 3) originally consisting of n-tuples (in our approach represented by MMP hypergraph
edges) of orthogonal vectors (MMP hypergraph vertices) which exhaust themselves in forming
configurations/sets of vectors (MMP hypergraphs). Already in [38], we realised that hypergraphs
massively generated by their non-isomorphic upward construction might satisfy the Kochen–Specker
theorem even when there were no vectors by means of which they might be represented (see
Theorem 1), and finding coordinatizations for those hypergraphs which might have them, via standard
methods of solving systems of non-linear equations, is an exponentially complex task solvable only for
the smallest hypergraphs [38]. It was, therefore, rather surprising to us to discover that the hypergraphs
formed by very simple vector components often satisfied the Kochen–Specker theorem. In this paper,
we present a method of generation of KS MMP hypergraphs, also called KS hypergraphs, via such
simple sets of vector components.

Theorem 1 (MMP hypergraph reformulation of the Kochen–Specker theorem).

There are nD MMP hypergraphs, i.e., those whose each edge contains n vertices, called KS MMP hypergraphs,
to which it is impossible to assign 1s and 0s in such a way that

(α) No two vertices within any of its edges are both assigned the value 1;
(β) In any of its edges, not all of the vertices are assigned the value 0.

In Figure 1, we show the smallest possible 4D KS MMP hypergraph with six vertices and three
edges. We can easily verify that it is impossible to assign 1 and 0 to its vertices so as to satisfy the
conditions (α) and (β) from Theorem 1. For instance, if we assign 1 to the top green-blue vertex, then,
according to the condition (α), all of the other vertices contained in the blue and green edges must be
assigned value 0, but, herewith, all four vertices in the red edge are assigned 0s in violation of the
condition (β), or, if we assign 1 to the top red-blue vertex, then, according to the condition (α), all the
other vertices contained in the blue and red edges must be assigned value 0, but, herewith, all four
vertices in the green edge are assigned 0s in violation of the condition (β). Analogous verifications go
through for the remaining four vertices. We verified that there is neither a real nor complex vector

www.pdfgrip.com
Entropy 2018, 20, 928

solution of a corresponding system of nonlinear equations [38]. We have not tried quaternions as
of yet.

Figure 1. The smallest 4D KS MMP hypergraph without a coordinatization.

When a coordinatization of a KS MMP hypergraph exists, its vertices denote n-dimensional

vectors in H n , n ≥ 3, and edges designate orthogonal n-tuples of vectors containing the corresponding
vertices. In our present approach, a coordinatization is automatically assigned to each hypergraph by
the very procedure of its generation from the basic vector components. A KS MMP hypergraph with a
given coordinatization of whatever origin we often simply call a KS set.

2.1. Formalism
MMP hypergraphs are those whose edges (of size n) intersect each other in at most n − 2
vertices [26,37]. They are encoded by means of printable ASCII characters. Vertices are denoted by one
of the following characters: 1 2 ...9 A B ...Z a b ...z ! " # $ % & ’ ( ) * - / : ; < = > ? @ [ \ ] ˆ _ ‘
{ | } ~ [26]. When all of them are exhausted, one reuses them prefixed by ‘+’, then again by ‘++’, and so
forth. An n-dimensional KS set with k vectors and m n-tuples is represented by an MMP hypergraph
with k vertices and m edges which we denote as a k-m set. In its graphical representation, vertices are
depicted as dots and edges as straight or curved lines connecting m orthogonal vertices. We handle
MMP hypergraphs by means of algorithms in the programs SHORTD, MMPSTRIP, MMPSUBGRAPH,
VECFIND, STATES01, and others [5,30,38–41]. In its numerical representation (used for computer
processing), each MMP hypergraph is encoded in a single line in which all m edges are successively
given, separated by commas, and followed by assignments of coordinatization to k vertices (see 18-9
in Section 2.2).

2.2. KS Vector Lists vs. Vector Component MMP Hypergraphs

In Table 1, we give an overview of most of the k-m KS sets (KS hypergraphs with m vertices
and k edges) as defined via lists and tables of vectors used to build the KS master sets that one can
find in the literature. These master sets serve us to obtain billions of non-isomorphic smaller KS sets
(KS subsets, subhypergraphs) which define k-m classes. In doing so (via the aforementioned algorithms
and programs), we keep to minimal, critical, KS subhypergraphs in the sense that a removal of any of
their edges turns them into non-KS sets. Critical KS hypergraphs are all we need for an experimental
implementation: additional orthogonalities that bigger KS sets (containing critical ones) might possess
do not add any new property to the ones that the minimal critical core already has. The smallest
hypergraphs we give in the table are therefore the smallest criticals. Many more of them, as well as their
distributions, the reader can find in the cited references. Some coordinatizations are over-complicated
in the original literature. For example (as shown in [37]), for the 4D 148-265 master, components

www.pdfgrip.com
Entropy 2018, 20, 928

{0, ±i, ±1, ±ω, ±ω 2 }, where ω = e2πi/3 , suffice for building the coordinatization, and for the 6D 21-7
components {0, 1, ω } suffice. In addition, {0, ±1} suffice for building the 6D 236-1216.

Table 1. Vector lists from the literature; we call their masters list-masters. We shall make use of their
vector components from the last column to generate master hypergraphs in Section 2.3 which we call
component-masters. ω is a cubic root of unity: ω = e2πi/3 .

Smallest
dim Master Size Vector List List Origin Vector Components
Hypergraph
D C B A

E 9

symmetry, F 8

4D 24-24 [25,42,43] G
18−9 7 {0,±1}
geometry H 6
I 5

1 2 3 4

D C B A

E 9

Pauli F 8

4D 60-105 [28,37] G
18−9 7 {0,±1, ±i}
operators H 6
I 5

1 2 3 4

regular √
{0, ±(√ 5 − 1)/2, ±1,
4D 60-75 [27,30,37,41] polytope 26−13 ±( 5 + 1)/2, 2}
600-cell
O M N P
G 9
Z I
T 8 E U
X

{0, ±i,√±1, ±ω, ±√ω 2 ,

L Y

Witting A
d e
F

4D 148-265 [36,37] S Q J
R

polytope W
b 40−23 V
K
±iω 1/ 3 , ±iω 2/ 3 }
a
C H 7
c 5
B
D 6
3 1 2 4

21−7
6D 21-7 [19] symmetry {0, 1, ω, ω 2 }

Aravind & hypercube √

{0, ±1/2,
ñ1/ 3,
6D 236-1216 Waegell →hexaract
2016, [37] Schäfli {4, 34 }
34−16 ±1/ 2, 1}

36−9
8D 36-9 [37] symmetry {0, ±1}

Lie
as given
8D 120-2025 [35,37] algebra
E8
34−9 in [35]

r
n s Mx
y
p o z
q t $
N "#
O u (
’
!
Qubit l mR v
W
e

{0, ±1}
S w
16D 80-265 [37,44,45] f U V

72−11
&%
g H T
states h K
i
j − )
L
c
d

* a b
k C J G F Z
P AB Y
I E
Q X
9 1 2 3 4 5 6 7 8 D
Z
+P ut
+Q *
x ) z
" +R j +l ’ +j
+V +U r +S v +i +J
+W +T
+X q a e +Y +L +G
+Z p +e +q +F
+m +a +g +p+s
+d +c+b y + +r
Qubit Q O K 2 +3 +1 D {

{0, ±1}
144−11
3 EI P +C
+E

32D 160-661 [37,46]

+5
+2

46 w M +B
+4

78 ( +I +o +A

states 9B
F
D
+h GJ H
+N k
h
f
+K
+k +9
+8
+7
+6
L ~
N d ] +n +M }
R S +f > |
+H T U A C & ?
! ‘
V 5 +O
: _
W1 /XY b # < = l n @ s [ ^
;
g i %
c m −$ o \

www.pdfgrip.com
Entropy 2018, 20, 928

Some of the smallest KS hypergraphs in the table have ASCII characters assigned and some do
not. This is to stress that we can assign them in an arbitrary and random way to any hypergraph
and then the program VECFIND will provide them with a coordinatization in a fraction of a second.
For instance,
18-9: 1234,4567,789A,ABCD,DEFG,GHI1,I29B,35CE,68FH.
{1={0,0,0,1},2={0,0,1,0},3={1,1,0,0},4={1,-1,0,0},5={0,0,1,1},6={1,1,1,-1},
7={1,1,-1,1}, 8={1,-1,1,1},9={1,0,0,-1},A={0,1,1,0},B={1,0,0,1},C={1,-1,1,-1},
D={1,1,-1,-1},E={1,-1,-1,1},F={0,1,0,1},G={1,0,1,0},H={1,0,-1,0},I={0,1,0,0}}.
(To simplify parsing, this notation delineates vectors with braces instead of traditional parentheses in
order to reserve parentheses for component expressions.)

However, a real finding is that we can go the other way round and determine the KS sets from
nothing but vector components {0, ±1}.

2.3. Vector-Component-Generated Hypergraph Masters

We put simplest possible vector components, which might build vectors and therefore provide
a coordinatization to MMP hypergraphs, into our program VECFIND. Via its option -master,
the program builds an internal list of all possible non-zero vectors containing these components.
From this list, it finds all possible edges of the hypergraph, which it then generates. MMPSTRIP via
its option -U separates unconnected MMP subgraphs. We pipe the obtained hypergraphs through
the program STATES01 to keep those that possess the KS property. We can use other programs of
ours, MMPSTRIP, MMPSHUFFLE, SHORTD, STATES01, LOOP, etc., to obtain smaller KS subsets and
analyze their features.
The likelihood that chosen components will give us a KS master hypergraph and the speed
with which it does so depends on particular features they possess. Here, we will elaborate on
some of them and give a few examples. Features are based on statistics obtained in the process of
generating hypergraphs:

(i) the input set of components for generating two-qubit KS hypergraphs (4D) should contain number
pairs of opposite signs, e.g., ±1, and zero (0); we conjecture that the same holds for 3, 4, . . . qubits;
with 6D it does not hold literally; e.g., {0, 1, ω } generate a KS master; however, the following
combination of ω’s gives the opposite sign to 1: ω + ω 2 = −1;
(ii) mixing real and complex components gives a denser distribution of smaller KS hypergraphs;
(iii) reducing the number of components shortens the time needed to generate smaller hypergraphs
and apparently does not affect their distribution.

Feature (i) means that, no matter how many different numbers we use as our input components,
we will not get a KS master if at least to one of the numbers, the same number with the opposite
√ e.g., {0, 1, −i, 2, −3, 4, 5} or a similar string does not give any, while {0, ±1},
sign is not added. Thus,
or {0, ±i }, or {0, ±( 5 − 1)/2} do. Of course, the latter strings all give mutually isomorphic KS
masters, i.e., one and the same KS master, if used alone. More specifically, they yield a 40-32 master
with 40 vertices and 32 edges as shown in Table 2. When any of them are used together with other
components, although they would generate different component-masters, all the latter masters of a
particular dimension would have a common smallest hypergraph as also shown in Table 2.

10

www.pdfgrip.com
Entropy 2018, 20, 928

Table 2. Component-masters we obtained. List-masters are given in Table 1. In the last two rows of
all but the last column, we refer to the result [33] that there are 16D and 32D criticals with just nine
edges. According to the conjectured feature (i) above, the masters generated by {0, ±1} should contain
those criticals; they did not come out in [37], so, we do not know how many vertices they have. The
smallest ones we obtained are given in Table 1. The number of criticals given in the 4th column refer to
the number of them we successfully generated although there are many more of them except in the
40-32 class.

No of KS
Component-Master Smallest Contains
dim Vector Components Criticals
Size Hypergraph List-Masters
in Master
D C B A

{0,±√
1} or {0,±i} or
E 9
F 8

4D 40-32 6 18−9 24-24

{0, ±( 5 − 1)/2} or . . .
G 7

H 6
I 5

1 2 3 4

D C B A

E 9
F 8

4D {0,±1, ±i} 156-249 7.7 × 106 G

18−9 7
24-24, 60-105
H 6
I 5

1 2 3 4

√ D C B A

{0, ±(√ 5 − 1)/2, ±1,

E 9
F 8

4D 2316-3052 1.5 × 109 G

18−9 7
24-24, 60-75
±( 5 + 1)/2, 2} H
I 5
6

1 2 3 4

D C B A

E 9
F 8
24-24, 60-105
4D {0, ±1, ±i, ±ω, ±ω 2 } 400-1012 8 × 106 G
18−9 7

H
I 5
6 148-265
1 2 3 4

21−7
6D {0, ±1, ω, ω 2 } 11808-314446 3 × 107 21-7, 236-1216

8D {0, ±1} 3280-1361376 7 × 106 34−9 36-9, 120-2025

computationally ?−9
16D {0, ±1} 4 × 106 80-265
too demanding [33].
computationally ?−9
32D {0, ±1} 2.5 × 105 160-661
too demanding [33].

We obtained the following particular results which show the extent to which component-masters
give a more populated distribution of KS criticals than list-masters. We also closed several
open questions:

• As for the features (ii) and (iii) above, components {0, ±1, ω } generate the master 180-203 which
has the following smallest criticals 18-9, 20. . . 22-11, 22. . . 26-13, 24. . . 30-15, 30. . . 31-16, 28. . . 35-17,
33. . . 37-18, etc. This distribution is much denser than that of, e.g., the list-master 24-24 with
real vectors which in the same span of edges consists only of 18-9, 20-11, 22-13, and 24-15
criticals or of the list-master 60-75 which starts with the 26-13 critical. In Appendix A, we give a
detailed description of a 21-11 critical with a complex coordinatization and give a blueprint for its
experimental implementation;
• In [19], the reader is challenged to find a master set which would contain the "seven context star"
21-7 KS critical (shown in Tables 1 and 2). We find that {0, 1, ω } generate the 216-153 6D master
which contains just three criticals 21-7, 27-9, and 33-11, {0, 1, ω, ω 2 } generate 834-1609 master
from which we obtained 2.5 × 107 criticals, and {0, ±1, ω, ω 2 } generate 11808-314446 master from
which we obtained 3 × 107 criticals, all of them containing the seven context star. Some of the
obtained criticals are given in Appendix B;

11

www.pdfgrip.com
Entropy 2018, 20, 928

• The 60-75 list-master contains criticals with up to 41 edges and 60 vertices, while the 2316-3052
component-master generated from the same vector components contains criticals with up to close
to 200 edges and 300 vertices;
• The 60-105 list-master contains criticals with up to 40 edges and 60 vertices, while the 156-249
component-master generated from the same vector components contains criticals with up to at
least 58 edges and 88 vertices;
• Components {0, ±1} generate 332-1408 √ 6D master
√ which contains the 236-1216 list-master while
originally components {0, ±1/2, ±1/ 3, ±1/ 2, 1} were used;
• In [37], we generated 6D criticals with up to 177 vertices and 87 edges from the list-master 236-1216,
while, now, from the component-master 11808-314446, we obtain criticals with up to 201 vertices
and 107 edges;
• We did not generate 16D and 32D masters because that would take too many CPU days and we
already generated a huge number of criticals from submasters which are also defined by means of
the same vector components in [37]. See also Section 3.

3. Methods
Our methods for obtaining quantum contextual sets boil down to algorithms and programs
within the MMP language we developed to generate and handle KS MMP hypergraphs as the
most elaborated and implemented kind of these sets. The programs we make use of, VECFIND,
STATES01, MMPSTRIP, MMPSHUFFLE, SUBGRAPH, LOOP, SHORTD, etc., are freely available from
our repository http://goo.gl/xbx8U2. They are developed in [5,29,30,38–40,47,48] and extended for
the present elaboration. Each MMP hypergraph can be represented as a figure for a visualisation
but more importantly as a string of ASCII characters with one line per hypergraph, enabling us to
process millions of them simultaneously by inputting them into supercomputers and clusters. For the
latter elaboration, we developed other dynamical programs specifically for a supercomputer or cluster,
which enable piping of our files through our programs in order to parallelize jobs. The programs have
the flexibility of handling practically unlimited number of MMP hypergraph vertices and edges as we
can see from Table 2. The fact that we did not let our supercomputer run to generate 16D and 36D
masters and our remark that it would be "computationally too demanding" do not mean that such
runs are not feasible with the current computers, but that they would require too many CPU days on
the supercomputer and that we decided not to burden it with such a task at the present stage of our
research; see the explanation in Section 2.3.

4. Conclusions
The main result we obtain is that our vector component generation of KS hypergraphs (sets)
exhaustively use all possible vectors that can be constructed from chosen vector components. This is
in contrast to previous studies, which made use of serendipitously obtained lists of vectors curtailed
in number due to various methods applied to obtain them. Hence, we obtain a thorough and
maximally dense distribution of KS classes in all dimensions whose critical sets can therefore be
much more effectively used for possible implementation in quantum computation and communication.
A comparison of Tables 1 and 2 vividly illustrates the difference.
In Appendix A, we present a possible experimental implementation of a KS critical with complex
coordinatization generated from {0, ±1, ω }. What we immediately notice about the 21-11 critical from
Figure A1 is that the edges are interwoven in more intricate way than in the 18-9 (which has been
implemented already in several experiments), exhibiting the so-called δ-feature of the edges forming
the biggest loop within a KS hypergraph. The δ-feature refers to two neighbouring edges which share
two vertices, i.e., intersect each other at two vertices [37]. It stems directly from the representation
of KS configuration with MMP hypergraphs. Notice that the δ-feature precludes interpretation of
practically any KS hypergraph in an even dimensional Hilbert space by means of so-called Greechie
diagrams, which by definition require that two blocks (similar to hypergraph edges) do not share more

12

www.pdfgrip.com
Entropy 2018, 20, 928

than one atom (similar to a vertex) [6], on the one hand, and that the loops made by the blocks must
be of order five or higher (which is hardly ever realised in even dimensional KS hypergraphs—see
examples in [37]), on the other.
Our future engagement would be to tackle odd dimensional KS hypergraphs. Notice that, in a 3D
Hilbert space, it is possible to explore similarities between Greechie diagrams and MMP hypergraphs
because then neither of them can have edges/blocks which share more than one vertex/atom (via their
respective definitions) and loops in both of them are of the order five or higher [26,39].

Author Contributions: Conceptualization, M.P.; Data Curation, M.P.; Formal Analysis, M.P. and N.D.M.; Funding
Acquisition, M.P.; Investigation, M.P. and N.D.M.; Methodology, M.P. and N.D.M.; Project Administration, M.P.;
Resources, M.P.; Software, M.P. and N.D.M.; Supervision, M.P.; Validation, M.P. and N.D.M.; Visualization, M.P.;
Writing—Original Draft, M.P.; Writing—Review and Editing, M.P. and N.D.M.
Funding: Supported by the Croatian Science Foundation through project IP-2014-09-7515, the Ministry of Science
and Education (MSE) of Croatia through the Center of Excellence for Advanced Materials and Sensing Devices
(CEMS) funding, and by grants Nos. KK.01.1.1.01.0001 and 533-19-15-0022. This project was also supported by the
Alexander or Humboldt Foundation. Computational support was provided by the cluster Isabella of the Zagreb
University Computing Centre, by the Croatian National Grid Infrastructure (CRO-NGI), and by the Center for
Advanced Computing and Modelling (CNRM) for providing computing resources of the supercomputer Bura at
the University of Rijeka in Rijeka, Croatia. The supercomputer Bura and other information and communication
technology (ICT) research infrastructure were acquired through the project Development of research infrastructure for
laboratories of the University of Rijeka Campus, which is co-funded by the European regional development fund.
Acknowledgments: Technical supports of Emir Imamagić and Daniel Vrčić from Isabella and CRO-NGI and of
Miroslav Puškarić from CNRM are gratefully acknowledged.
Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations
The following abbreviations are used in this manuscript:

KS Kochen–Specker; defined in Section 1

MMP McKay-Megill-Pavičić; defined in Section 2.1

Appendix A. 21-11 KS Critical with Complex States from H2 ⊗ H2

Below, we present a possible implementation of a KS critical 21-11 with complex coordinatization
shown in Figure A1.
The vector components of the first qubit on a photon correspond to a linear (horizontal, H, vertical,
V, diagonal, D, antidiagonal A) and circular (right, R, left L) polarization, and those of the second qubit
to an angular momentum of the photon (+2, −2) and (h, v). One-to-one correspondence between
them is given below.

D B B=(1,1,0,0)
1=(1,1,1,−1) C A
E C=(1,−1,i,−i)
2=(1,−1,−1,−1)
L 9 D=(1,1,−1,−1)
3=(1,0,0,1) F
E=(1,1,1,1)
4=(0,1,−1,0)
5=(0,1,1,0) G 21−11−a 8 F=(1,−1,1,−1)
K G=(0,1,0,−1)
6=(0,0,0,1) H 7
H=(1,0,−1,0)
7=(1,0,0,0) I
6 I=(0,1,0,1)
8=(0,1,0,0)
J J=(1,−1,1,1)
9=(0,0,1,−1) 5
1 2 3 4 K=(0,0,1,0)
A=(0,0,1,1)
L=(1,−1,−i,i)

Figure A1. 21-11 KS set with complex coordinatization.

13

www.pdfgrip.com
Entropy 2018, 20, 928

An example of a tensor product of two vectors/states from H2 ⊗ H2 is:

⎛  ⎞
0 ⎛ ⎞
    ⎜1 ⎟ 0
⎜ 1 ⎟ ⎜ ⎟
1 0 ⎜ ⎟ ⎜1⎟
|01 = |0, 1 = |01 ⊗ |12 = ⊗ ⎜ ⎟
= ⎜  ⎟ = ⎜ ⎟ .
0 1 ⎜ ⎟ ⎝0⎠
1 2 ⎝ 0 ⎠ 0
0
1

This is our vector 8 from Figure A1. Since we are interested in the qubit states, we are going to
proceed in reverse—from 4-vectors to tensor products of polarization and angular momentum states.
Let us first define them:

         
1 0 1 1 1 −1 1 1
|H = ; |V  = ; | D = √ ; | A = √ ; | R = √ ;
0 1 2 1 2 1 2 i
1 1 1 1 1

         
1 1 1 0 1 1 1 1
| L = √ ; | + 2 = ; | − 2 = ; |h = √ ; |v = √ .
2 −i 0 1 2 1 2 −1
1 2 2 2 2

Now, one can read off our vertex states as follows:

⎛ ⎛ ⎞⎞ ⎛ ⎛ ⎞⎞
⎛ ⎞ ⎛ ⎞ ⎛ ⎞
⎛ ⎞ 1 1
1 1 0 1 ⎜1 ⎝ ⎠ ⎟ ⎜0 ⎝ ⎠ ⎟
⎜ ⎟ ⎜ ⎟ ⎜ ⎟ ⎜ ⎟ ⎜ 1 ⎟ ⎜ −1 ⎟
⎜ 1 ⎟ ⎜ ⎟ ⎜ ⎟
⎜ ⎟ 1⎜
⎜ 1 ⎟
⎟ 1 ⎜ ⎟ ⎜ ⎟
⎜1⎟ ⎜ 0 ⎟ 1 1 ⎜
⎜
⎟
⎟ 1 ⎜⎜
⎟
⎟
1 = ⎜ ⎟ → ⎜ ⎟ = (⎜ ⎟ + ⎜ ⎟) = √ ( √ ⎜ ⎛ ⎞ ⎟ + √2 ⎜ ⎛ ⎞ ⎟ )
⎜ 1 ⎟ 2⎜ 1 ⎟ 2 ⎜0⎟ ⎜ 1 ⎟ 2 2 ⎜ ⎟ ⎜ ⎟
⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎜ ⎟ ⎜ ⎟
1
⎝0 ⎝ ⎠ ⎠ 1
⎝1 ⎝ ⎠ ⎠
−1 −1 0 −1
1 −1
⎛ ⎞ ⎛ ⎞ ⎛ ⎞ ⎛ ⎞
1 1 1 1 0 1 1 1 1
= √ (⎝ ⎠ ⊗ √ ⎝ ⎠ + ⎝ ⎠ ⊗ √ ⎝ ⎠ = √ (| H |h + |V |v) = √ (| D | + 2 − | A| − 2).
2 0 2 1 1 2 −1 2 2
1 2 1 2

We will now skip real states and go directly to those with imaginary components, C and L, to
illustrate how they can be implemented via circular polarization:

⎞ ⎛ 
⎛1 ⎞
1 ⎜1 ⎟    
⎜ −1⎟ ⎜ −1 ⎟
⎜ ⎟ 1⎜ ⎟
⎟ = √1 1 1 1
C=⎜ ⎟→ ⎜ ⊗ √ = | R|v,
⎝ i ⎠ 2⎜
⎜  ⎟
⎟ 2 i 1 2 −1 2
−i ⎝ 1 ⎠
i
−1

⎞ ⎛ 
⎛ 1 ⎞
1 ⎜1 ⎟    
⎜ −1⎟ ⎜ −1 ⎟
⎜ ⎟ 1⎜ ⎟
⎟ = √1 1 1 1
L=⎜ ⎟→ ⎜ ⊗ √ = | L|v.
⎝ −i ⎠ 2⎜
⎜  ⎟
⎟ 2 −i 1 2 −1 2
⎝ 1 ⎠
i −i
−1

Thus, in order to handle a complex coordinatization, we need a fifth degree of freedom (circular
polarization), but, as we can see, it is manageable.

14

www.pdfgrip.com
Entropy 2018, 20, 928

Appendix B. 6D Criticals from the Masters Containing the Seven Context Star.
The 216-153 KS master generated from {0, 1, ω } contains 21-7 and 27-9, which can be viewed as
21-7 with a pair of δ-triplets interwoven with 21-7, as shown in Figure A2. The 834-1609 KS master
generated from {0, 1, ω, ω 2 }, which were used for a construction of 21-7 in [19], contains 39-13 as well.
Equally so, the 11808-314446 master generated from {0, ±1, ω, ω 2 }.

21−7 27−9 39−13

Figure A2. 21-11 KS set from [19] and 27-9 are contained in three different master sets, 39-13 in two
(together with 21-11 and 27-9); see the text.

References
1. Cabello, A.; D’Ambrosio, V.; Nagali, E.; Sciarrino, F. Hybrid Ququart-Encoded Quantum Cryptography
Protected by Kochen-Specker Contextuality. Phys. Rev. A 2011, 84, 030302(R). [CrossRef]
2. Nagata, K. Kochen-Specker Theorem as a Precondition for Secure Quantum Key Distribution. Phys. Rev. A
2005, 72, 012325. [CrossRef]
3. Howard, M.; Wallman, J.; Veitech, V.; Emerson, J. Contextuality Supplies the ‘Magic’ for Quantum
Computation. Nature 2014, 510, 351–355. [CrossRef] [PubMed]
4. Bartlett, S.D. Powered by Magic. Nature 2014, 510, 345–346. [CrossRef] [PubMed]
5. Pavičić, M.; McKay, B.D.; Megill, N.D.; Fresl, K. Graph Approach to Quantum Systems. J. Math. Phys. 2010,
51, 102103. [CrossRef]
6. Megill, N.D.; Pavičić, M. Kochen-Specker Sets and Generalized Orthoarguesian Equations.
Ann. Henri Poincare 2011, 12, 1417–1429. [CrossRef]
7. Simon, C.; Żukowski, M.; Weinfurter, H.; Zeilinger, A. Feasible Kochen-Specker Experiment with Single
Particles. Phys. Rev. Lett. 2000, 85, 1783–1786. [CrossRef] [PubMed]
8. Michler, M.; Weinfurter, H.; Żukowski, M. Experiments towards Falsification of Noncontextual Hidden
Variables. Phys. Rev. Lett. 2000, 84, 5457–5461. [CrossRef]
9. Amselem, E.; Rådmark, M.; Bourennane, M.; Cabello, A. State-Independent Quantum Contextuality with
Single Photons. Phys. Rev. Lett. 2009, 103, 160405. [CrossRef]
10. Liu, B.H.; Huang, Y.F.; Gong, Y.X.; Sun, F.W.; Zhang, Y.S.; Li, C.F.; Guo, G.C. Experimental Demonstration of
Quantum Contextuality with Nonentangled Photons. Phys. Rev. A 2009, 80, 044101. [CrossRef]
11. D’Ambrosio, V.; Herbauts, I.; Amselem, E.; Nagali, E.; Bourennane, M.; Sciarrino, F.; Cabello, A. Experimental
Implementation of a Kochen-Specker Set of Quantum Tests. Phys. Rev. X 2013, 3, 011012. [CrossRef]
12. Huang, Y.F.; Li, C.F.; Zhang, Y.S.; Pan, J.W.; Guo, G.C. Experimental Test of the Kochen-Specker Theorem
with Single Photons. Phys. Rev. Lett. 2003, 90, 250401. [CrossRef] [PubMed]
13. Cañas, G.; Etcheverry, S.; Gómez, E.S.; Saavedra, C.; Xavier, G.B.; Lima, G.; Cabello, A. Experimental
Implementation of an Eight-Dimensional Kochen-Specker Set and Observation of Its Connection with the
Greenberger-Horne-Zeilinger Theorem. Phys. Rev. A 2014, 90, 012119. [CrossRef]
14. Hasegawa, Y.; Loidl, R.; Badurek, G.; Baron, M.; Rauch, H. Quantum Contextuality in a Single-Neutron
Optical Experiment. Phys. Rev. Lett. 2006, 97, 230401. [CrossRef] [PubMed]
15. Cabello, A.; Filipp, S.; Rauch, H.; Hasegawa, Y. Proposed Experiment for Testing Quantum Contextuality
with Neutrons. Phys. Rev. Lett. 2008, 100, 130404. [CrossRef] [PubMed]

15

www.pdfgrip.com
Entropy 2018, 20, 928

16. Bartosik, H.; Klepp, J.; Schmitzer, C.; Sponar, S.; Cabello, A.; Rauch, H.; Hasegawa, Y. Experimental Test of
Quantum Contextuality in Neutron Interferometry. Phys. Rev. Lett. 2009, 103, 040403. [CrossRef] [PubMed]
17. Kirchmair, G.; Zähringer, F.; Gerritsma, R.; Kleinmann, M.; Gühne, O.; Cabello, A.; Blatt, R.; Roos, C.F.
State-Independent Experimental Test of Quantum Contextuality. Nature 2009, 460, 494–497. [CrossRef]
18. Moussa, O.; Ryan, C.A.; Cory, D.G.; Laflamme, R. Testing Contextuality on Quantum Ensembles with One
Clean Qubit. Phys. Rev. Lett. 2010, 104, 160501. [CrossRef]
19. Lisoněk, P.; Badziag, P.; Portillo, J.R.; Cabello, A. Kochen-Specker Set with Seven Contexts. Phys. Rev. A
2014, 89, 042101. [CrossRef]
20. Cañas, G.; Arias, M.; Etcheverry, S.; Gómez, E.S.; Cabello, A.; Saavedra, C.; Xavier, G.B.; Lima, G. Applying
the Simplest Kochen-Specker Set for Quantum Information Processing. Phys. Rev. Lett. 2014, 113, 090404.
[CrossRef]
21. Barrett, J.; Kent, A. Noncontextuality, Finite Precision Measurement and the Kochen-Specker. Stud. Hist.
Philos. Mod. Phys. 2004, 35, 151–176. [CrossRef]
22. Spekkens, R.W. Contextuality for Preparations, Transformations, and Unsharp Measurements. Phys. Rev. A
2005, 71, 052108. [CrossRef]
23. Kunjwal, R.; Spekkens, R.W. From the Kochen-Specker Theorem to Noncontextuality Inequalities without
Assuming Determinism. Phys. Rev. Lett. 2015, 115, 110403. [CrossRef] [PubMed]
24. Kunjwal, R. Hypergraph Framework for Irreducible Noncontextuality Inequalities from Logical Proofs of
the Kochen-Specker Theorem. arXiv 2018, arXiv:1805.02083.
25. Cabello, A.; Estebaranz, J.M.; García-Alcaine, G. Bell-Kochen-Specker Theorem: A Proof with 18 Vectors.
Phys. Lett. A 1996, 212, 183–187. [CrossRef]
26. Pavičić, M.; Merlet, J.P.; McKay, B.D.; Megill, N.D. Kochen-Specker Vectors. arXiv 2005, arXiv:quant-ph/0409014.
27. Waegell, M.; Aravind, P.K. Critical Noncolorings of the 600-Cell Proving the Bell-Kochen-Specker Theorem.
J. Phys. A 2010, 43, 105304. [CrossRef]
28. Waegell, M.; Aravind, P.K. Parity Proofs of the Kochen-Specker Theorem Based on 60 Complex Rays in Four
Dimensions. J. Phys. A 2011, 44, 505303. [CrossRef]
29. Megill, N.D.; Fresl, K.; Waegell, M.; Aravind, P.K.; Pavičić, M. Probabilistic Generation of Quantum
Contextual Sets. Phys. Lett. A 2011, 375, 3419–3424. [CrossRef]
30. Pavičić, M.; Megill, N.D.; Aravind, P.K.; Waegell, M. New Class of 4-Dim Kochen-Specker Sets. J. Math. Phys.
2011, 52, 022104. [CrossRef]
31. Waegell, M.; Aravind, P.K.; Megill, N.D.; Pavičić, M. Parity Proofs of the Bell-Kochen-Specker Theorem
Based on the 600-cell. Found. Phys. 2011, 41, 883–904. [CrossRef]
32. Waegell, M.; Aravind, P.K. Proofs of Kochen-Specker Theorem Based on a System of Three Qubits. J. Phys. A
2012, 45, 405301. [CrossRef]
33. Waegell, M.; Aravind, P.K. Proofs of the Kochen-Specker Theorem Based on the N-Qubit Pauli Group.
Phys. Rev. A 2013, 88, 012102. [CrossRef]
34. Waegell, M.; Aravind, P.K. Parity Proofs of the Kochen-Specker Theorem Based on 120-Cell. Found. Phys.
2014, 44, 1085–1095. [CrossRef]
35. Waegell, M.; Aravind, P.K. Parity Proofs of the Kochen-Specker Theorem Based on the Lie Algebra E8.
J. Phys. A 2015, 48, 225301. [CrossRef]
36. Waegell, M.; Aravind, P.K. The Penrose Dodecahedron and the Witting Polytope Are Identical in CP3 .
Phys. Lett. A 2017, 381, 1853–1857. [CrossRef]
37. Pavičić, M. Arbitrarily Exhaustive Hypergraph Generation of 4-, 6-, 8-, 16-, and 32-Dimensional Quantum
Contextual Sets. Phys. Rev. A 2017, 95, 062121. [CrossRef]
38. Pavičić, M.; Merlet, J.P.; McKay, B.D.; Megill, N.D. Kochen-Specker Vectors. J. Phys. A 2005, 38, 1577–1592;
3709 (corrigendum). [CrossRef]
39. McKay, B.D.; Megill, N.D.; Pavičić, M. Algorithms for Greechie Diagrams. Int. J. Theor. Phys. 2000, 39,
2381–2406. [CrossRef]
40. Pavičić, M.; Megill, N.D.; Merlet, J.P. New Kochen-Specker Sets in Four Dimensions. Phys. Lett. A 2010,
374, 2122–2128. [CrossRef]
41. Megill, N.D.; Fresl, K.; Waegell, M.; Aravind, P.K.; Pavičić, M. Probabilistic Generation of Quantum
Contextual Sets. arXiv 2011, arXiv:1105.1840.
42. Peres, A. Two Simple Proofs of the Bell-Kochen-Specker Theorem. J. Phys. A 1991, 24, L175–L178. [CrossRef]

16

www.pdfgrip.com
Entropy 2018, 20, 928

43. Kernaghan, M. Bell-Kochen-Specker Theorem for 20 Vectors. J. Phys. A 1994, 27, L829–L830. [CrossRef]
44. Harvey, C.; Chryssanthacopoulos, J. BKS Theorem and Bell’s Theorem in 16 Dimensions; Technical Report
PH-PKA-JC08; Worcester Polytechnic Institute: Worcester, MA, USA, 2012. Available online: https://web.
wpi.edu/Pubs/E-project/Available/E-project-042108-171725/unrestricted/MQPReport.pdf (accessed on
26 November 2018).
45. Planat, M. On Small Proofs of the Bell-Kochen-Specker Theorem for Two, Three and Four Qubits. Eur. Phys.
J. Plus 2012, 127, 86. [CrossRef]
46. Planat, M.; Saniga, M. Five-Qubit Contextuality, Noise-Like Distribution of Distances Between Maximal
Bases and Finite Geometry. Phys. Lett. A 2012, 376, 3485–3490. [CrossRef]
47. Pavičić, M.; Megill, N.D. Quantum Logic and Quantum Computation. In Handbook of Quantum Logic and
Quantum Structures; Engesser, K., Gabbay, D., Lehmann, D., Eds.; Elsevier: Amsterdam, The Netherlands,
2007; pp. 751–787.
48. Megill, N.D.; Pavičić, M. New Classes of Kochen-Specker Contextual Sets (Invited Talk). In Proceedings
of the 2017 40th International Convention on Information and Communication Technology, Electronics
and Microelectronics (MIPRO 2017), Opatija, Croatia, 22–26 May 2017; Biljanović, P., Ed.; Institute of
Electrical and Electronics Engineers (IEEE), Curran Associates, Inc.: Red Hook, NY, USA, 2017; pp. 195–200,
ISBN 9781509049691.

c 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

17

www.pdfgrip.com
 entropy
Article
Advanced Statistical Testing of Quantum Random
Number Generators
Aldo C. Martínez 1 , Aldo Solis 2 , Rafael Díaz Hernández Rojas 3 , Alfred B. U’Ren 2 ,
Jorge G. Hirsch 2 and Isaac Pérez Castillo 4,5, *
1 Department of Physics, Center for Research in Photonics, University of Ottawa, 25 Templeton St, Ottawa,
ON K1N 6N5, Canada; mbac@ciencias.unam.mx
2 Instituto de Ciencias Nucleares, Universidad Nacional Autónoma de México, Apdo. Postal 70-543, Cd. Mx.,
C.P. 04510 Mexico, Mexico; aldo.solis@correo.nucleares.unam.mx (A.S.);
alfred.uren@correo.nucleares.unam.mx (A.B.U.); hirsch@nucleares.unam.mx (J.G.H.)
3 Dipartimento di Fisica, Sapienza University of Rome, P.le Aldo Moro 5, I-00185 Rome, Italy;
rafael.diaz.hr@gmail.com
4 Departamento de Física Cuántica y Fótonica, Instituto de Física, Universidad Nacional Autónoma de
México, Apdo. Postal 20-364, Cd. Mx., C.P. 04510 Mexico, Mexico
5 London Mathematical Laboratory, 8 Margravine Gardens, London W6 8RH, UK
* Correspondence: isaacpc@fisica.unam.mx

Received: 20 October 2018; Accepted: 14 November 2018; Published: 17 November 2018

Abstract: Pseudo-random number generators are widely used in many branches of science, mainly in
applications related to Monte Carlo methods, although they are deterministic in design and, therefore,
unsuitable for tackling fundamental problems in security and cryptography. The natural laws of
the microscopic realm provide a fairly simple method to generate non-deterministic sequences of
random numbers, based on measurements of quantum states. In practice, however, the experimental
devices on which quantum random number generators are based are often unable to pass some tests
of randomness. In this review, we briefly discuss two such tests, point out the challenges that we
have encountered in experimental implementations and finally present a fairly simple method that
successfully generates non-deterministic maximally random sequences.

Keywords: Bell inequalities; algorithmic complexity; Borel normality; Bayesian inference; model
selection; random numbers

1. Introduction
Monte Carlo methods are one of the essential staples of the basic sciences in the modern age.
Although these gained prominence during the early 1940s, thanks to secret research projects carried
out in Los Alamos Scientific Laboratory by Ulam and von Neumann [1,2], their origins may be traced
back to the famous Buffon’s needle problem, posed by Georges-Louis Leclerc, Comte de Buffon, in the
18th century. In the present day, Monte Carlo “experiments” are seen as a broad class of computational
algorithms that use repeated random sampling to obtain numerical estimates of a given natural or
mathematical process. In order to use these methods efficiently, fully random sequences of numbers
are needed. Back in the 1940s, this was a tall order, and various methods to generate random sequences
were used (some of them literally using roulettes), until von Neumann pioneered the concept of
computer-based random number generators. During the following years, these became the standard
tool in Monte Carlo methods and are still generally well-suited for many applications. However, these
computer-based methods generate pseudo random numbers [3], which means that the generated
sequence can be determined given an algorithmic program and an initial seed, two ingredients which
are hardly random. Thus, in order to achieve a truly unpredictable source of random numbers, we must

Entropy 2018, 20, 886; doi:10.3390/e20110886 18 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2018, 20, 886

eliminate these two deterministic aspects. The former is easy to overcome using, for example, a pattern
of keystrokes typed on a computer keyboard as a random seed. On the other hand, the algorithmic
program could be replaced, for instance, by a classical chaotic system [4]. Examples of the latter abound
in the area of weather prediction and climate sciences.
In recent years, however, the community has been moving towards using the fundamental laws
dictating the behaviour of the quantum realm for the generation of sequences of truly random numbers.
This seems, at a first glance, to be at odds with the following rather naïve thought: if the natural laws of
the microscopic world are considered to be a computer program under which a system evolves from an
initial state (a seed), should not its corresponding generated sequence also be predictable? It turns out
that Quantum Mechanics, in its current standard view, related to the Copenhagen interpretation, has a
special ingredient that makes the random sequence inherently unpredictable for both the generator
and the observer. Such a strange behaviour has been eloquently recast over the years in various
forms, famously by the quote “spooky action at a distance” due to Einstein, or mathematically by the
celebrated work of Bell [5,6]. The application of quantum randomness in cryptography has given rise
to the concept of device independent randomness certification, which, in a nutshell, corresponds to
those processes that violate Bell’s inequalities [7,8]. However, there seems to be some confusion in the
literature regarding two different properties of a given sequence of random numbers. The first one,
rather important as we have argued above, is whether the sequence is truly random, meaning that it
is unpredictable. In contrast, the second one is related to the issue of assessing whether or not it is
biased. It is crucial to keep in mind that these two properties are independent, as evidenced by the
random number generator Quantis [9], which is based on a quantum system and is able to pass the
standard tests of randomness (NIST (National Institute of Standards and Technology) suite) [10] but
has difficulties with other tests [11].
Due to recent advances in quantum technologies, and since the NIST suite has been examined
in other works [12], together with a critical view on the use of p-values on which the NIST suite
relies [13], it becomes necessary to consider other criteria for measuring the performance of quantum
random number generators. Thus, we focus solely on two recently introduced approaches: the first
one is based on algorithmic complexity theory evaluating incompressibility and bias at the same time,
since an incompressible sequence is necessarily an unbiased one [14], while the second one relies on
Bayesian model selection. Both methods are based on solid structures which lead to a definition of
randomness that is very intuitive and which arises independently of the development of random
number generators. We apply them to analyze sequences of random bits generated in our laboratory
using quantum systems. We also address the issue about the origin of the biases observed when
utilizing these types of devices.

2. Tests of Randomness
A simple criterion for assessing the predictability of a sequence is the presence of patterns in it.
For example, for the sequence 01010101..., we can ask ourselves whether the next number is either 1
or 0. The natural answer is 0 based on the pattern observed in the previous bits. In general, we would
like to find any possible regularity that helps us to predict the next bit. Within the framework of
algorithmic information theory, it is possible to address this problem by noting that any sequence
which exhibits regularity can be compressed using a short algorithm which can produce as output
precisely such patterns. Thus, a sequence of this type could be reproduced using fewer bits than the
ones contained in its original form. Therefore, whenever a sequence lacks regularity, we refer to it as
“algorithmically” random.
We now introduce a remarkable result from algorithmic information theory: the Borel-normality
criterion due to Calude [14], which allows us to asymptotically check whether a sequence is not
“algorithmically” random. Assuming we are given a string  = {1001010110110 · · · } of || = n bits
(We will only consider binary sequences, but our results are easily generalizable to other alphabets),
the idea of the Borel-normality criterion consists primarily of dividing the original sequence  into

19

www.pdfgrip.com
Entropy 2018, 20, 886

consecutive substrings of length i and then computing the frequencies of occurrence of each of them.
For brevity and later use, let us define Ω(i) as the set of 2i substrings that can be formed with i characters,
j
let i be the sequence obtained after dividing it into substrings, and ||i ≡ [||/i ]. Additionally, let Ni ()
be the number of times the j-th substring of length i appears in . For example, when considering
substrings of length i = 1, we are looking at the frequencies of the symbols Ω(1) = {0, 1} that conform
the original string , while for i = 2, we have to consider the frequencies of four substrings, namely
Ω(2) = {00, 01, 10, 11}. According to Calude, a necessary condition for a sequence to be maximally
random is that the deviations of these frequencies with respect to the expected values in the ideal
random case should be bounded as follows [14,15]:
j
Ni () 1 log2 (n)
− i < , j = 0, . . . , 2i − 1 . (1)
||i 2 n

This condition must be satisfied for all substrings of lengths from i = 1 up to imax = log2 (log2 (n)).
Intuitively, this criterion “compresses” the original sequence by reading i bits at a time and tests
whether the substrings appear with a frequency that differs from what would be expected in the
random case, thus indicating the presence of some regularity. We emphasize that since Borel-normality
is not a sufficient criterion for randomness, it can only be used to assess whether a given sequence is
not random. In other words, even if a sequence satisfies Equation (1) for all substrings and allowed
values of i, the Borel-normality condition cannot guarantee that it is indeed random.
Recently, a Bayesian criterion has been introduced [16,17] by some of the authors of the present
article to test, from a purely probabilistic point of view, whether a sequence is maximally random
as understood within information theory [18]. The method works by exploiting the Borel-normality
compression scheme and then recasting the problem of finding possible biases in the sequence as an
inferential one in which Bayesian model selection can be applied. Specifically, for a fixed value of i,
(i )
we need to consider all the possible probabilistic models, henceforth denoted as {Mα }α , that could
have generated the sequence . Each such model determines a unique probability assignation to the
elements of Ω(i) , which depends on a set of prior parameters θ. For these parameters, the Jeffreys’
prior, PJeff (θ), turns out to be a convenient choice of prior parameter distribution, as it entails the
“Occam Razor principle” in which more complex models are penalized, as well as being mathematically
convenient for the case at hand; some other advantages are pointed out in [16,17].
Next, the question of finding all the generative models that can produce a sequence  is ultimately
solved by noticing that all the possible probabilities assignations are in a one-to-one correspondence
with all possible partitions of Ω(i) . Since obtaining the partitions of any set is a straightforward
combinatorial task [19], we are able to determine all the relevant models when searching for possible
biases in the generation of . For instance, when i = 1, there are two possible models: one in which
the two elements of Ω(1) are equiprobable, corresponding to the partition {{0, 1}} of Ω(1) into one
subset—i.e., the same set—and another model with probabilities p0 = θ, p1 = 1 − θ corresponding to
the partition {{0}, {1}} of Ω(1) into two subsets. Even though it might seem that the first model is
just a particular case of the second one (by letting θ = 1/2), we should keep in mind that the prior
distributions are different in both cases, δ(θ − 1/2) and √ 1 , respectively, thus yielding two
π θ (1− θ )
different models. Analogously, for i = 2, there is a single unbiased model, which corresponds to the
partition of Ω(2) into one subset (with probabilities p j = 1/4, for j = 00, 01, 10, 11), and 14 additional
models associated with the different ways of dividing Ω(2) into subsets. The latter are related to the
number of ways of distinguishing among the elements of Ω(2) during the assignation of probabilities,
and thus any of these models would entail some bias when generating a sequence. Note that, in general,
for any value of i, we will face a similar situation in which a single model can produce an unbiased,
and hence maximally random, sequence by means of a uniform distribution, while the rest of them
will be some type of categorical distribution.

20

www.pdfgrip.com
Entropy 2018, 20, 886

Once all the models have been identified, the remaining part is the computation of the posterior
(i )
distribution P Mα  , which from an inferential point of view is the most relevant distribution as
(i )
it gives the probability that the model Mα has indeed produced the given sequence . Note that,
(i )
since a generative approach was adopted, we have direct access to the distribution P  θ, Mα ,
(i )
which can be combined with the parameters’ prior PJeff (θ) to obtaining the distribution P  = Mα
 (i )
dθ P  Mα , θ PJeff (θ). One of the most important results of [16] is that this marginalization can
be accomplished exactly for all the models and any value of i. Therefore, we can obtain the posterior
distribution by a simple application of Bayes’ rule:

(i ) (i )
P  Mα P Mα
(i )
P Mα  = , (2)
(i ) (i )
∑γ P  Mγ P Mγ

(i )
with P Mγ being the prior distribution in the space of models that can generate sequences of
strings of length i. Therefore, the best model α that describes the dataset  is quite simply given by

(i )
α = arg maxα P Mα  . (3)

(i )
If the best model Mα turns out to be the unbiased one for all possible lengths i of the substrings,
then we can say that the process that generated that dataset was maximally random. However,
it remains to discuss how large the length i of substrings can be for a given dataset of n bits. To answer
this, we first note that, for any set containing N elements, the possible number of partitions is given by
the N-th Bell number BN [19]. Thus, for a given i, all possible partitions of the set Ω(i) will result in B2i
models to be tested, and, therefore, it is expected for them to be sampled at least once when observing
. This means that B2imax ∼ n, which, for sufficiently large n, yields imax = log2 (log2 (n)), precisely as
in the Borel-normality criterion.
Randomness characterization through Bayesian model selection has some clear and natural
advantages, as already pointed out in [16], but, unfortunately, it has an important drawback:
the number of all possible models for a given length i, given by B2i , grows supra-exponentially
with i: indeed, for i = 1, we have two possible models, for i = 2, we have 15 possible models, for i = 3,
we have instead 4140 possible models, while, for i = 4, we have 10,480,142,147 models. Thus, even
if we are able to acquire data for the evaluation of these many models, it becomes computationally
impractical to estimate the posterior for all of them using Equation (2). There is an elegant strategy
to overcome this difficulty: one can derive bounds similar to those provided by the Borel-normality
criterion, by comparing the log-likelihood ratio between the maximally random model and the
maximally biased one. This yields the following bound for the frequencies of occurrence [17]:
  ⎛ ⎞
    
 2i − 1 j
Ni ()
j
Ni ()  2− n Γ 2
i 1
Γ 1
+ n
 1 1  i2 2 21−i i
∑ − i − i < ln ⎝ ⎠, (4)
j ≤ j =1
||i 2 ||i 2 n2 ψ1 1
+ n
Γ 1
Γ2
i 1
+ n
2 i2i 21−i 2 i2i

where ψ1 is the polygamma function of order 1. Note that, unlike Calude’s bound given by Equation (1),
this new Borel-type bound couples all frequencies, and, moreover, results in highly restrictive bounds.

21

www.pdfgrip.com
Entropy 2018, 20, 886

3. Ideal Random Number Generation

While intuition dictates that quantum random number generators (QRNG) should be superior
to their classical counterparts, such a comparison was carried out in [11] and very recently in [20],
with rather disappointing results. For the classical case, the authors used three Pseudo-Random
Number Generators (PRNG): the generators included in the software packages Mathematica and
Maple, and the digits of π expressed in base 2. For QRNG, they used two devices: (i) Quantis,
developed by IDQ [9], a quantum random number generator interfaced with a common computer, and
(ii) an experiment from a quantum optics group in Vienna. The latter experiment consists of a very weak
light source, attenuated to the single photon level, a beam splitter, and two single photon detectors.
Leaving aside the question of which QRNG performs better, the real surprise was that the PRNGs come
out in this test with a superior performance, by far, as compared to their quantum counterparts. This
result appears to be at odds with the natural randomness associated with quantum phenomena. Why
is it that the inherent quantum randomness does not translate into better performance with respect to
classical systems? Does randomness, as discussed in this paper, have no impact on the performance of
the generators? Is this a fundamental or a technical problem?
In order to explore this apparent paradox, we will discuss the different technical and design
difficulties associated with quantum random number generation using light. These days, it is
straightforward to detect single photons using avalanche photo-diodes (APD), devices capable of
detecting up to a few million single photons per second with > 60% detection efficiencies employing
relatively simple electronics. With this simple design in mind, we only need a single photon source, a
beam splitter (BS), and a couple of single-photon detection devices in order to set up a QRNG device.
This minimalistic design is sketched in Figure 1.

Figure 1. Ideal experimental setup for a naïve QRNG (quantum random number generators) based
on an individual photon source and a beam splitter (BS). Neglecting the possible losses, a photon will
activate only one of the two detectors, therefore producing a random bit per photon.

4. Experimental Challenges in Random Number Generation

Suppose now that we have a single-photon source and we want to generate a sequence of bits
to be tested against the bounds given by Equation (1). Let us start by first focusing on the so-called
Borel level, the word length i, which can take a maximum value of imax = log2 (log2 (n)). In Table 1, we
report how imax grows with n, up to a value of imax = 6. In order to achieve a Borel level imax = 6, we
will require a dataset of length 1018 events. Assuming that our single-photon generator and detectors
can cope with around a million of events per second, we would then require on the order of 600 years
to generate a sequence of that length! It turns out that imax = 5 is a more realistic value, since it leads
to a required dataset  of size 4.3 × 109 events, which can be realistically produced in a couple of hours.

22

www.pdfgrip.com
Entropy 2018, 20, 886

Table 1. Necessary data lengths for maximum Borel level imax = log2 (log2 (n)). The double exponential
relation grows so quickly that it is not possible get to level 6.

Maximum Borel Level imax = log2 (log2 (n)) Data Length n

1 4
2 16
3 256
4 65,536
5 4,294,967,296
6 18,446,744,073,709,551,616

The bound on the right-hand-side of Equation (1) implies that the frequencies of occurrence
j
Ni ()
||i
for substrings of length i ≤ imax cannot deviate from the ideal random one, 1/2i , by more than
8.6 × 10−5 , which constitutes an extremely tight tolerance. Hence, in practice, any part of the naïve
experimental setup that gives rise to some bias will unfortunately make the dataset  unable to pass the
Calude criterion. The first component that we must be wary about is the BS. A regular BS usually has
an error figure in the region of 1%, which is very high with respect to the stringent tolerance  would
need in order fulfill Borel normality. Is it plausible to correct this using a Polarizing Beam Splitter
(PBS), instead of the BS, with an active control through feedback of the state of polarization so as to
compensate for any bias in the PBS? In what follows, we investigate this question through a simple
experiment. The state of polarization of a single photon entering the PBS can be written as

|ψ = a |V  + eiφ b | H  , (5)

where |V  and | H  refer to the vertical and horizontal polarization components, respectively. We can
approach the state in Equation (5) by transmitting the laser beam trough a half wave plate (HWP) so
as to achieve arbitrary rotation of the linear polarization.√
Assuming a perfect, unbiased BS, we would
need an incoming polarization state with a = b = 1/ 2 so that the resulting sequence of bits is
unbiased. If, on the other hand, the PBS exhibits biases (e.g., due to manufacturing error), we can
adjust the orientation of the above-mentioned half wave plate so as to adjust precisely the value of our
coefficients a and b to compensate for the PBS bias.
Our experimental setup, shown in Figure 2, can be regarded as the minimal realistic device
for the implementation of a QRNG. The main questions which we wish to address are: (i) how
good are the sequences of bits generated by such a device? In addition, (ii) do they pass the
Borel-normality criterion?

M
APD2 APD1

LD
NDF Motor
Controller AL2
P2

M
P1 AL1
HWP

Figure 2. Experimental setup. The relative angle of the half wave plate is controlled in order to
reduce bias.

23

www.pdfgrip.com
Entropy 2018, 20, 886

The input state is prepared using the beam from a laser diode (LD). The beam is transmitted
through a set of neutral density filters (NDF) with a combined optical density 7.3 for attenuation
to a level compatible with the maximum recommended count rate of our single photon detectors.
The beam is then transmitted through a half wave plate (HWP) mounted on a motorized rotation stage
so as to control its orientation angle relative to the PBS axes. The PBS splits the beam into two spatial
modes according to the H and V polarizations, each of which is coupled with the help of an aspheric
lens (AL1 and AL2) into a multimode fiber leading to an avalanche photodiode (APD1 and APD2).
We include a polariser (P1 and P2), with an extinction ratio, defined as the ratio of the maximum to the
minimum transmission of a linearly polarized input, of 100,000:1 prior to each of the aspheric lenses
(AL1 and AL2) for a reduction of the non-polarized intensity reaching the detectors.
Suppose now that we prepare our system so that the average relative power Pi /( P0 + P1 ) of each
APD detector i = 0, 1 starts ideally at 1/2. In Figure 3, we show how this average relative power
evolves with time (see curve labelled “without feedback”). Note that, even though the system starts
in a perfectly balanced state, it rapidly deviates from this condition. The slow change of this curve
can be attributed to thermal drift while the oscillatory component with a period of approximately
half an hour is related to the air conditioning system in the laboratory. These effects can be effectively
compensated by rotating the HWP. After some study of the response function of our experimental
setup, a correction every minute with a proportional controller was sufficient to correct for all these
effects leading to a steady response (see curve labelled “with feedback”) [21].

0.504
Without feedback
With feedback
0.503

0.502

0.501

0.5
P0 +P1

0.499
P0

0.498

0.497

0.496

0.495

0.494
1 2 3 4 5 6
Hours

Figure 3. Evolution in time of the normalized power. The device starts in a perfect balanced state but
quickly deviates from this condition. By using a feedback mechanism, we can obtain stability of the
normalized power within an error of 0.004.

5. First Battery of Results

We have used the experimental setup described in the previous section to generate a sequence of
4,294,967,296 bits, allowing us to test the Borel–Normality criterion up to level imax = 5. The results
of this analysis are depicted in Figure 4. In the plot, bars represent the deviations from the ideal
value for all the strings at each Borel level. For instance, in the first part of the analysis (purple bars),
there are only two bars corresponding to the frequency of occurrences of substrings “1” and “0”.
As our initial setup is very fine-tuned and stable, the bars have practically zero height, with value
5 × 10−6 . The green bars represent the second part of the analysis, or Borel level two, corresponding to

24

www.pdfgrip.com
Entropy 2018, 20, 886

frequencies of occurrences of symbols {00, 01, 10, 11}, and so on. In the same figure, the horizontal
lines represent the bound given by the right-hand-side of Equation (1). Our first battery of results are a
clear disappointment: only the first set for substrings of length one clearly passes the test, while, for
higher lengths, our QRNG fails miserably to pass Calude’s criterion.

0.006

0.004

0.002
2i
1
−

0
Nij (l)
|l|i

-0.002

-0.004

-0.006
0
1
00
01
10
11
000
001
010
011
100
101
110
111
0000
0001
0010
0011
0100
0101
0110
0111
1000
1001
1010
1011
1100
1101
1110
1111
00000
00001
00010
00011
00100
00101
00110
00111
01000
01001
01010
01011
01100
01101
01110
01111
10000
10001
10010
10011
10100
10101
10110
10111
11000
11001
11010
11011
11100
11101
11110
11111
Figure 4. Results from Borel analysis. The first two boxes correspond to the deviations from the mean
at the first Borel level; these exhibit the same height but opposite signs. The next four bars (green)
represent the deviations for level two, i.e., “00”, ”01”, ”10”, “11”. The blue and yellow boxes represent
the deviations for level three and four, respectively. The red lines correspond to Borel’s bound, which
turns out to be much smaller than the deviations. Only the first level passes the test.

Furthermore, a closer look at the green bars shows that events 00 and 11 appear more frequently
than expected, by about 0.005%, (while events 01 and 10 appear less frequently than expected by
the same margin). This effect reveals a correlation between “equal events”, that is, the same digit
appearing twice. In terms of our experiment, this means that it is more probable to observe an event
in a detector once a previous event has already been recorded. Other parts of our test validate this:
for Borel level three, the yellow bars indicate that events with alternate zeroes and ones (010 and 101)
appear less frequently than expected, also by about 0.005%. At Borel levels four and five, the larger
deviations appear for events 0101 and 1010, clearly in accordance with the previous results. This
indicates that certain parts of our experimental setup are introducing unwanted correlations between
bits, which results in the magnitude of some of the deviations to be 50 times larger than expected. Our
experimental effort clearly does not suffice for our sequences to pass the Borel–Normality criterion.
How is this possible?

6. APD Effects on Introducing Correlations

The two main effects in the behavior of our APDs which can introduce undesired correlations
in the resulting sequences of bits are called after-pulsing and dead time [22,23]. The first effect,
roughly speaking, corresponds to a false detection event due to the residual effects of an avalanche
triggered by a previous event, while the dead time is the time period after each event during which
the system is not able to record a subsequent incoming optical signal. In this case, we have a typical
dead time of 22 ns, a maximum after-pulsing probability of 1%, and a dark counts rate of 100 counts/s.

25

www.pdfgrip.com
Entropy 2018, 20, 886

While the device exhibits a linear behaviour up to 5 × 106 counts/s, the detection rates used in our
experiments are an order of magnitude lower.
The mechanism by which dead time introduces correlations in our data, particularly in
experimental arrangements with two or more APDs as in our case, is as follows: suppose that we
have an event in one of our detectors. During its dead time, it will have zero probability of recording
another incoming event in that detector, while the other detector still exhibits a non-zero probability of
recording an event. On the other hand, the way after-pulsing introduces correlations in our data is by
increasing the probability of observing consecutively the same event, resulting in the observation of an
excess of the events {00} and {11} in Figure 4.
These two effects can somewhat be corrected either by re-designing our experimental setup and/or
modifying the software. Instead of following this route to generate maximally random sequences of
bits, let us pursue a rather simple solution as discussed below.

7. Random Number Generation Using Time Measurements

We now follow a method introduced in [3]. Suppose that ρ( x ) is the probability density function
of a continuous random random variable X on an interval x ∈ ( a, b). Let us further assume that its
real value x is represented up to a given precision so that we assign a parity to x according to the
parity of its least significant digit. Next, we divide the interval ( a, b) into an even number 2L of bins
and introduce
i(b − a)
xi = a + , i = 0, . . . , 2L . (6)
2L
Suppose that 2L and the precision has been chosen so that xi is even for i even and odd for i odd.
It follows that
 b 2L−1  xi+1
1=
a
dxρ( x ) = ∑ xi
dxρ( x ) ≡ Neven + Nodd , (7)
i =0

with
L−1  x2i+1 L−1  x2i+2
Neven ≡ ∑ x2i
dxρ( x ) , Nodd ≡ ∑ x2i+1
dxρ( x ) . (8)
i =0 i =0

Approximating the integrals by the left sum rule, we can write that

L −1 L −1
b−a b−a
Neven ∼
2L ∑ ρ(x2i ) , Nodd ∼
2L ∑ ρ(x2i+1 ) , (9)
i =0 i =0

which implies that, roughly, the probability that the least significant digit is odd can be expressed as

L −1  2
1 1 b−a
Nodd ∼ +
2 2 ∑ ρ (x2i ) 2L
, (10)
i =0

where the bias term can be fine-tuned by increasing either the number of bins or through a smooth
density ρ( x ), or both.
This method can be very easily implemented in the lab as follows. Suppose that the random
variable X is the time difference between two consecutive photon arrivals to the detector. In our case,
these times are of the order of 500 ns to 10 μs. A typical sequence of these time differences look like:

26

www.pdfgrip.com
Entropy 2018, 20, 886

592 342ps,
595 634ps,
593 645ps,
592 342ps,
595 634ps.
..
.

We can then look at the parity of the least significant digit and assign, for instance, a 0-bit to
even parity and 1-bit to odd parity, thus generating a dataset  of n bits. In Figure 5, we show the
results of testing such a sequence using the Borel-normality bounds and Bayesian bounds, given by
Equations (1) and (4), respectively, up to Borel level imax = 5. The colored bins correspond to the
deviations from the ideal value of relative frequencies for all the possible subsequences. These are
ordered using its length i = 1 (purple bins), i = 2 (green bins), i = 3 (blue bins), i = 4 (orange bins),
and imax = 5 (yellow bins). The solid red line corresponds to the Borel bound, 8.6 × 10−5 . In the same
graph, the green line is the Bayesian bound given the right-hand-side of Equation (4) that depends
on i and therefore it is not a constant, as is the case for the Borel bound. Finally, the height of the
various background colored boxes correspond to the values given by the left-hand-side expression of
the Bayesian bound.

15×10−5

10×10−5


2i −1
j≤j  =1
5×10−5

N ()
||i
2i

i
1

j
−

0
− 1i
Nij (l)

2
|l|i


N
||i
i
j

−5×10−5
()
− 1i
2

−10×10−5

−15×10−5
0
1
00
01
10
11
000
001
010
011
100
101
110
111
0000
0001
0010
0011
0100
0101
0110
0111
1000
1001
1010
1011
1100
1101
1110
1111
00000
00001
00010
00011
00100
00101
00110
00111
01000
01001
01010
01011
01100
01101
01110
01111
10000
10001
10010
10011
10100
10101
10110
10111
11000
11001
11010
11011
11100
11101
11110
11111

Figure 5. Results for generation using the least significant bits of time tags. In this case, the deviations
are very small, so this generation scheme is excellent. The solid red lines represent the maximum
deviations allowed by the Calude test, while the solid green lines correspond to the bounds given the
Bayesian approach.

As we can see, this extremely simple QRNG passes the Borel-normality criterion up to i = 5,
and nearly passes the Bayesian criterion (passes it for i ≤ 4 and slightly exceeds the bound for i = 5).
Notice that, while the previous experimental setup required an accurate balance between zeroes and
ones, in the present case we already have very small deviations at Borel level i = 1, less than 10−5 ,
showing the convenience of this method. For i = 2 (green bins), the deviations are much larger, almost

27

www.pdfgrip.com
Entropy 2018, 20, 886

three times the value for i = 1, but nevertheless they pass the test again by a considerable margin.
These results show the lack of correlations between consecutive events, which is the main drawback
of the previous approach. It is important to note that, in the results at Borel level i = 4, there is a
substantial increase in the deviations compared with i = 1, 2, 3. While this increase may indicate the
presence of some as yet unidentified correlations, these are of an insufficient magnitude to reach the
bounds. On the other hand, this experimental setup fails to pass some of the requirements of the
Bayesian scheme. The deviations derived from the Bayesian criterion are shown in Table 2, and also
in Figure 5. As we can see, all Borel levels pass the Bayesian test, except for the last one, albeit by a
small margin.

Table 2. Comparison between the left-hand-side and the right-hand-side of the Borel-type bounds given
by Equation (4), applied to the sequence of bits generated in our lab. As we can observe, the Bayesian
bounds are satisfied for the first four Borel levels, but not the last one by a slight margin.

Borel Level LHS of Equation (4) RHS of Equation (4) LHS < RHS
1 5.719 × 10−6 3.62956 × 10−5 Yes
2 1.7129 × 10−5 6.08097 × 10−5 Yes
3 1.54974 × 10−5 7.82572 × 10−5 Yes
4 8.74186 × 10−5 9.11726 × 10−5 Yes
5 1.01138 × 10−4 1.01069 × 10−4 No

We can also look at the value of the posterior distribution given by Equation (2) for the maximally
random model. The value of the posterior for the four word lengths is reported in Table 3. For the
first three Borel levels, the posterior distribution of the maximally random model αsym is very close
to one, indicating that, given the dataset, this is the most likely model to have generated such data.
For Borel level i = 4, we are only able to analyse those models which are in the vicinity, in parameter
space, to the maximally random model. These models correspond to partitioning Ω(5) into two subsets,
resulting in 32,767 models, giving a total of 32,768, including the maximally random one. In this case,
it turns out that the most likely model is not the maximally random one. Actually, using the value of
the posterior probability, this model is ranked in the position 9240 out of all the explored models, and
therefore the sequence of bits fails to pass the Bayesian criterion already at Borel level i = 4. Note that
i = 5 was not included in Table 3 because we lack the computational power to address this Borel level.

(i )
Table 3. Value of the posterior distribution P Msym  given by Equation (2) for the maximally
random model. Note that the prior distribution for each Borel model is a flat distribution along all the
models tested. This means, for instance, that, at level i = 3, while we do not have any observational bias
(i )
to choose among any particular model, that is P(Mα ) = 1
4140 , after observing the data, the maximally
random model is the most plausible.

(i )
Borel Level i Number of Models Analysed P Msym 

1 B2i = 2 0.999984
2 B2i = 15 0.999634
3 B2i = 4140 0.995476
4 32,768 models considered out of B2i = 10,480,142,147 9.2179 × 10−42

8. Conclusions
A vast amount of literature exists which claims that QRNGs are superior when compared to
their classical counterparts, based on purely theoretical arguments. Indeed, randomness in Quantum
Mechanics is usually justified by the unpredictability of individual measurement outcomes given
some initial conditions. More concretely, quantum unpredictability is based on no-go theorems, such
as Bell’s theorem, that simply tells us that, given some initial conditions, it is impossible to predict

28

www.pdfgrip.com
Entropy 2018, 20, 886

the outcome of a single measurement. However, in the present review, we have shown that QRNGs
actually perform rather poorly in tests of randomness as compared to classical PRNGs. The reason is
fairly simple: unpredictability has nothing to do with bias, and while experimental devices based on
Quantum Mechanics may produce a truly unpredictable random signal, they also tend, more often
than not, to introduce correlations. In particular, for QNRGs based on optical devices, we have been
able to account for two, perhaps amongst the many, effects that introduce bias in our data. While in
our own experimental work involving a QNRG we have failed to obtain sequences which obey the
Borel and Bayesian criteria, we were able to show that extracting sequences from the least significant
digits of times of arrival represents a promising strategy.

Author Contributions: J.G.H. and A.B.U. conceived and designed the experiments; A.C.M. and A.S. performed
the experiments; I.P.C. and R.D.H.R. developed the method based on Bayesian Inference; A.C.M, A.S. and R.D.H.R.
analysed the data. All authors contributed to writing the paper.
Funding: Financial support of UNAM-DGAPA-PAPIIT IA103417 and IN109417 is acknowledged.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Metropolis, N.; Ulam, S. The Monte Carlo method. J. Am. Stat. Assoc. 1949, 44, 335–341. [CrossRef] [PubMed]
2. Von Neumann, J. Various techniques used in connection with random digits. J. Res. Nat. Bur. Stand. Appl.
Math. Ser. 1951, 12, 36–38.
3. Isida, M.; Ikeda, H. Random number generator. Ann. Inst. Stat. Math. 1956, 8, 119–126. [CrossRef]
4. Stojanovski, T.; Kocarev, L. Chaos-based random number generators-part I: Analysis [cryptography].
IEEE Trans. Circuits Syst. I Fundam. Theory Appl. 2001, 48, 281–288. [CrossRef]
5. Einstein, A.; Podolsky, B.; Rosen, N. Can quantum-mechanical description of physical reality be considered
complete? Phys. Rev. 1935, 47, 777–780. [CrossRef]
6. Bell, J.S. Speakable and Unspeakable in Quantum Mechanics: Collected Papers on Quantum Philosophy; Cambridge
University Press: Cambridge, UK, 2004.
7. Pironio, S.; Acín, A.; Massar, S.; de La Giroday, A.B.; Matsukevich, D.N.; Maunz, P.; Olmschenk, S.; Hayes, D.;
Luo, L.; Manning, T.A.; et al. Random numbers certified by Bell’s theorem. Nature 2010, 464, 1021–1024.
[CrossRef] [PubMed]
8. Acín, A.; Masanes, L. Certified randomness in quantum physics. Nature 2016, 540, 213–219. [CrossRef] [PubMed]
9. Quantis Random Number Generator. Available online: https://www.idquantique.com/random-number-
generation/products/quantis-random-number-generator (accessed on 9 October 2018).
10. Rukhin, A.; Soto, J.; Nechvatal, J.; Smid, M.; Barker, E. A Statistical Test Suite for Random and Pseudorandom
Number Generators for Cryptographic Applications; Technical Report; Booz-Allen and Hamilton Inc.: Mclean,
VA, USA, 2001.
11. Calude, C.S.; Dinneen, M.J.; Dumitrescu, M.; Svozil, K. Experimental evidence of quantum randomness
incomputability. Phys. Rev. A 2010, 82, 022102. [CrossRef]
12. Pareschi, F.; Rovatti, R.; Setti, G. Second-level NIST randomness tests for improving test reliability.
In Proceedings of the IEEE International Symposium on Circuits and Systems (ISCAS 2007), New Orleans,
LA, USA, 27–30 May 2007; pp. 1437–1440.
13. Wasserstein, R.L.; Lazar, N.A. The ASA’s statement on p-values: Context, process, and purpose. Am. Stat.
2016, 70, 129–133. [CrossRef]
14. Calude, C.S. Borel normality and algorithmic randomness. Dev. Lang. Theory 1993, 355, 113–129.
15. Calude, C.S. Information and Randomness: An Algorithmic Perspective, 2nd ed.; Springer Publishing Company:
New York, NY, USA, 2010.
16. Díaz Hernández Rojas, R.; Solís, A.; Angulo Martínez, A.M.; U’ren, A.B.; Hirsch, J.G.; Marsili, M.;
Pérez Castillo, I. Improving randomness characterization through Bayesian model selection. Sci. Rep. 2017,
7, 3096. [CrossRef] [PubMed]
17. Díaz Hernández Rojas, R. Mejora en la Caracterización de la Aleatoriedad Usando Selección Bayesiana de
Modelos. Master’s Thesis, Posgrado en Ciencias Físicas UNAM, Mexico, Mexico, 2017. (In Spanish)
18. Cover, T.M.; Thomas, J.A. Elements of Information Theory; John Wiley & Sons: Hoboken, NJ, USA, 2012.

29

www.pdfgrip.com
Entropy 2018, 20, 886

19. Pemmaraju, S.; Skiena, S.S. Computational Discrete Mathematics: Combinatorics and Graph Theory with Mathematica;
Cambridge University Press: Cambridge, UK, 2003.
20. Abbott, A.A.; Calude, C.S.; Dinneen, M.J.; Huang, N. Experimentally Probing the Incomputability of
Quantum Randomness. arXiv 2018, arXiv:1806.08762.
21. Martínez Becerril, A.C. Un Generador Cuántico de Números Al Azar: De un PBS a Etiquetas Temporales.
Bachelor’s Thesis, Facultad de Ciencias, UNAM, Mexico, Mexico, 2017. (In Spanish)
22. Horoshko, D.B.; Chizhevsky, V.N.; Kilin, S.Y. Afterpulsing model based on the quasi-continuous distribution
of deep levels in single-photon avalanche diodes. J. Mod. Opt. 2017, 64, 191–195. [CrossRef]
23. Kang, Y.; Lu, H.X.; Lo, Y.H.; Bethune, D.S.; Risk, W.P. Dark count probability and quantum efficiency of
avalanche photodiodes for single-photon detection. Appl. Phys. Lett. 2003, 83, 2955–2957. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

30

www.pdfgrip.com
 entropy
Article
Probabilities and Epistemic Operations in the Logics
of Quantum Computation
Maria Luisa Dalla Chiara 1, *, Hector Freytes 2 , Roberto Giuntini 2 , Roberto Leporini 3 and
Giuseppe Sergioli 2
1 Dipartimento di Lettere e Filosofia, Università di Firenze, Via Bolognese 52, I-50139 Firenze, Italy
2 Dipartimento di Pedagogia, Psicologia, Filosofia, Università di Cagliari, Via Is Mirrionis 1, I-09123 Cagliari,
Italy; hfreytes@gmail.com (H.F.); giuntini@unica.it (R.G.); giuseppe.sergioli@gmail.com (G.S.)
3 Dipartimento di Ingegneria Gestionale, dell’Informazione e della Produzione, Università di Bergamo,
viale Marconi 5, I-24044 Dalmine (BG), Italy; roberto.leporini@unibg.it
* Correspondence: dallachiara@unifi.it

Received: 28 August 2018; Accepted: 28 October 2018; Published: 31 October 2018

Abstract: Quantum computation theory has inspired new forms of quantum logic, called quantum
computational logics, where formulas are supposed to denote pieces of quantum information, while
logical connectives are interpreted as special examples of quantum logical gates. The most natural
semantics for these logics is a form of holistic semantics, where meanings behave in a contextual way.
In this framework, the concept of quantum probability can assume different forms. We distinguish
an absolute concept of probability, based on the idea of quantum truth, from a relative concept of
probability (a form of transition-probability, connected with the notion of fidelity between quantum
states). Quantum information has brought about some intriguing epistemic situations. A typical
example is represented by teleportation-experiments. In some previous works we have studied a
quantum version of the epistemic operations “to know”, “to believe”, “to understand”. In this article,
we investigate another epistemic operation (which is informally used in a number of interesting
quantum situations): the operation “being probabilistically informed”.

Keywords: quantum logics; quantum probability; holistic semantics; epistemic operations

1. Introduction
Quantum information and quantum computation have inspired new developments of some basic
concepts of the quantum theoretic formalism, which for a long time had been regarded as mysterious
and potentially paradoxical. In this framework the concept of quantum probability has been investigated
according to new perspectives, giving rise to possible applications to fields that are far apart from
microphysics (cognitive and social sciences, semantics of natural languages and of the languages of
art, see, for instance, [1–3]).
As is well known, the basic idea of quantum computation theory is that information can be stored
and transmitted by quantum physical objects. Accordingly, pieces of quantum information can be
identified with states of some special quantum systems that are storing the information in question.
In the simplest case a piece of quantum information corresponds to a pure state of a single particle:
a qubit (or qubit-state), the quantum counterpart of the classical concept of bit. Mathematically a
qubit can be represented as a quantum superposition (living in the two-dimensional Hilbert space C2 ),
whose form is
| ψ  = c0 |0 + c1 |1,

Entropy 2018, 20, 837; doi:10.3390/e20110837 31 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2018, 20, 837

where |0 and |1 are the two elements of the canonical orthonormal basis of the space, representing in
this framework the two classical truth-values 1 (Truth) and 0 (Falsity). From an intuitive point of view,
any qubit
c0 |0 + c1 |1

can be regarded as un uncertain information that might be true with probability |c1 |2 and might be
false with probability |c0 |2 . More generally, a piece of quantum information corresponds to a complex
knowledge that can be mathematically represented as a pure or mixed state of a composite quantum
system: a density operator ρ of a finite-dimensional Hilbert space whose standard form is

H (n) = C

2
⊗ .
. . ⊗ C2 (the n − fold tensor product of C2 ).

n−times

Quantum information is processed by (quantum logical) gates: special examples of unitary

quantum operations that transform pure and mixed states in a reversible way. Any finite sequence
of gates (defined on a space H (n) ) gives rise to a quantum circuit: when applied to a given input ρin ,
the circuit under consideration transforms ρin into an output ρout . This represents a mathematical
description for a physical process that might be performed by a quantum computer.
The theory of quantum circuits has inspired a natural logical abstraction, giving rise to the
development of new forms of quantum logic that have been termed quantum computational logics.
In these logics, formulas are supposed to denote pieces of quantum information, while logical
connectives are interpreted as special examples of gates. Consequently, all formulas of quantum
computational languages turn out to have a typical dynamic character, representing possible
computation-actions. The most natural semantics for quantum computational logics is a form of holistic
semantics, where quantum entanglement (often described as “mysterious”) can be used as a powerful
logical resource. Against the classical compositionality-principle, meanings of well-formed expressions
of a quantum computational language behave in a holistic and contextual way: the meaning of a
global expression determines the meanings of its well-formed parts, and not the other way around.
Furthermore, meanings are generally context-dependent: under one and the same interpretation of the
language an expression may receive different meanings in different contexts (as, in fact, happens in
our current use of natural languages and in many forms of informal reasoning).
An important “character” of the quantum computational semantics is the concept of quantum
probability, which can assume different forms. A basic notion of quantum probability (which plays an
important logical role) is connected with the idea of truth. In any quantum computational space H (n)
(n)
the concept of truth can be naturally represented as a special projection operator, indicated by P1 .
(1)
For instance, in the case of the space H (1) = C2 the truth-concept is identified with the projection
P1
P|1 that projects over the closed subspace determined by the bit |1 (corresponding to the classical
truth-value 1). In this way, truth is dealt with as a special example of a mathematical representative for
a possible physical event. Accordingly, one can naturally apply the basic probabilistic rule of quantum
theory, based on the concept of Born-probability. For any qubit |ψ = c0 |0 + c1 |1, the probability
that the quantum information |ψ satisfies the truth can be defined as follows:

(1)
p1 (|ψ) := P|1 |ψ 2
= |c1 |2 = tr( P|ψ P1 ),

where tr is the trace functional and P|1 |ψ is the length of the vector obtained by projecting |ψ over
the closed subspace determined by |1. In the next Section we will see how the probability-function p1
can be generalized to all pieces of quantum information, living in any space H (n) .
Interestingly enough, the contextual properties of the holistic quantum computational semantics
allow us to understand and to justify (at least to a certain extent) some strange uses of the concept
of probability that sometimes occur in the framework of intuitive ways of reasoning (for a general
discussion of this problem see, for instance, [1]). For instance, assigning to a conjunction α ∧ β a

32

www.pdfgrip.com
Entropy 2018, 20, 837

probability-value greater than the probabilities of both members (α, β) is not necessarily “irrational” or
“antiscientific”. For, it might happen that the meanings of the three sentences α ∧ β, α, β refer, in fact, to
different contexts.
(n)
Once fixed the truth-concept P1 (for any space H (n) ), the probability-function p1 represents a
kind of absolute concept of probability: any piece of quantum information has a well-determined
probability-value of being true. It is interesting to investigate another concept of probability that
represents a form of relative probability. Suppose that the pure state

| ψ  = ∑ ci | ϕi 
i

(where every | ϕi  is an element of the canonical orthonormal basis of the space H (n) ) represents
the information of an epistemic agent at a given time (say, at the beginning of an experiment or of
a computation). According to the quantum theoretic formalism, the information |ψ allows us to
assign probability-values to other pieces of information, that may represent possible outcomes of a
measurement or of a computation. For instance, the probability that the information |ψ assigns to the
outcome | ϕi  is |ci |2 . Thus, whenever we have the information |ψ we might have the information | ϕi 
with probability |ci |2 . Accordingly, we can write:

p|ψ (| ϕi ) =|ci |2 .

We will see how this concept of relative probability (which can be generalized to mixed states) is
strongly connected with the concept of fidelity between quantum states.
Quantum information has brought about some intriguing epistemic situations. A typical example
is represented by teleportation-experiments, where an agent (say, Alice) transmits qubit-states to a far
agent (say, Bob), by using some special quantum non-locality phenomena, which may appear prima
facie strange and mysterious. These puzzling (non-classical) situations have inspired new ideas in the
field of epistemic logics. See, for instance, [4].
As is well known, many standard approaches to epistemic logics have been developed in
the framework of a possible world semantics, where the basic epistemic operators (“to know”,
“to believe”) are dealt with as special examples of modal operators. Is it possible to represent epistemic
operators as particular examples of quantum operations in the framework of a quantum computational
semantics? This question admits a positive answer. In some previous works we have studied a
quantum version of the epistemic operators “to know”, “to believe”, “to understand”, whose semantic
properties depend on the notion of quantum truth and on the probability-function p1 . We have
seen, in particular, how quantum knowledge generally gives rise to forms of “reversibility-breaking”
that can be compared with what happens in the case of quantum measurements. In this article
we will investigate another epistemic concept: the operation “being probabilistically informed”,
which plays a significant role in a number of interesting quantum situations (for instance, in the case
of quantum teleportation-experiments).

2. Pieces of Quantum Information and Quantum Probabilities

We will first recall the basic “mathematical characters” that play an important role in quantum
computation. The “mathematical stage” where any piece of quantum information is usually supposed
to live is an n-fold tensor product of the Hilbert space C2 :

H (n) = C

2
⊗ .
. . ⊗ C2 .

n−times

Any piece of quantum information is a special mathematical object that lives in a particular
quantum computational space H (n) , representing a possible state of a physical system that is storing
the information in question. For some applications it may be useful to take, as a basic quantum

33

www.pdfgrip.com
Entropy 2018, 20, 837

computational space, a many-dimensional space Cd (with d > 2). In this way, qubits are generalized
to qudits. See, for instance, [5]. Important examples are represented by quregisters, qubits,
registers, bits and mixtures of quregisters. In the quantum computation-literature the terms “qubit”
(or “quantum bit”) and “quregister” (or “quantum register”) are sometimes used in an ambiguous
way. In many cases the expression “qubit” refers to a quantum system (say, an electron or a photon)
whose possible pure states live in the space C2 . In some other cases, instead, “qubit” simply means a
possible pure state of such a system. A similar ambiguity regards the term “quregister”. In this article,
we will always use the terms “qubit” and “quregister” in the sense of possible pure states of quantum
systems that can store pieces of quantum information.

Definition 1 (Quregisters and registers).

• A quregister (or quregister-state) is a unit vector of a space H (n) .

• A qubit (or qubit-state) is a quregister of the space C2 .
• A register (or register-state) is an element

| x1 , . . . , x n  = | x1  ⊗ . . . ⊗ | x n 

of the canonical orthonormal basis of the space H (n) (where xi ∈ {0, 1}).
• A bit is a register of the space C2 .

Any quregister |ψ of the space H (n) can be represented as a quantum superposition of registers
that belong to the canonical basis of the space:

| ψ  = ∑ c i | x i1 , . . . , x i n ,
i

where ci are complex numbers (called amplitudes) such that ∑i |ci |2 = 1.

Quregisters are pure states representing maximal pieces of information (about the quantum
systems under investigation) that cannot be consistently extended to a richer knowledge. More
generally, a piece of quantum information may correspond to a non-maximal knowledge: a mixed
state (or mixture of quregisters), that is mathematically represented as a density operator ρ of a space
H(n) . Of course, any quregister |ψ corresponds to a special case of a density operator: the projection
P|ψ that projects over the closed subspace determined by |ψ. We will indicate by D(H (n) ) the set of
  
all density operators of the space H (n) , while D = n D(H (n) ) will represent the set of all possible
pieces of quantum information.
A piece of quantum information is generally stored by a composite system S consisting of some
subsystems S1 , . . . Sr , where each part Si may be, in turn, a composite system. According to the
quantum theoretic formalism any possible (pure or mixed) state ρ of S determines the state ρi of each
subsystem Si ; this state is called the reduced state of ρ with respect to the i-th subsystem. Of course,
some composite systems S might be decomposed into parts in different ways; thus, the mathematical
formalism shall take into account all possible decomposition-choices.
Consider a Hilbert space H (n) that can be decomposed as

H ( n ) = H ( m1 ) ⊗ . . . ⊗ H ( mr ) ,

where m1 + . . . + mr = n. Accordingly, any density operator ρ of H (n) can be regarded as a possible

state of a composite system
S = S1 + . . . + S r ,

where H (mi ) is the Hilbert space associated to the subsystem Si . Consider now a particular subsystem
of S:
Si 1 + . . . + Si k

34

www.pdfgrip.com
Entropy 2018, 20, 837

(with 1 ≤ i1 , . . . , ik ≤ r). We will indicate by

(i ,...,ik )
Red[m1
1 ,...,mr ]

the reduced state of ρ with respect to the subsystem Si1 + . . . + Sik and with respect to the decomposition
H(n) = H(m1 ) ⊗ . . . ⊗ H(mr ) . By simplicity we will omit the subscript [m1 , . . . , mr ] in the case where
the decomposition of H (n) is obvious.
The mathematical representation of composite systems (via tensor products) has brought about
some deep changes in the relationships between parts and whole in the quantum world. As is well
known, in perfect harmony with the semantics of classical logic, classical physical systems satisfy a
physical compositionality-principle: the states of the subsystems of a given system S determine the
state of S and vice versa (from the parts to the whole and from the whole to the parts). In quantum
theory the compositionality-principle is strongly violated: the state of a composite system determines
the state of all its parts, but generally not the other way around. The mysterious quantum entanglement
(which has for a long time regarded as potentially paradoxical) is connected with the failure of the
compositionality-principle.
What exactly is entanglement? For the sake of simplicity, in this article we will only consider the
case of entangled pure states.

Definition 2 (Entangled pure state). Consider a composite quantum system

S = S1 + . . . + S r ,

and let HS , HS1 , . . . , HSr be the Hilbert spaces associated to the systems S, S1 , . . . , Sr (respectively). A pure
state |ψ of S is called entangled iff |ψ cannot be represented as a factorized state

|ψ1  ⊗ . . . ⊗ |ψr ,

where |ψ1 , . . . |ψr  belong to the spaces HS1 , . . . HS1 , respectively.

Important examples of entangled pure states are the so called Bell-states, that live in the space
H ( 2 ) = C2 ⊗ C 2 .

Example 1. A typical Bell-state is the following:

1 1
|ψ = √ |0, 0 + √ |1, 1.
2 2

Apparently, |ψ describes the state of a two-particle system (S = S1 + S2 ), assigning probability-value 1

2
to the two following possibilities:

• both subsystems are in the state |0;

• both subsystems are in the state |1.

One can show that the two reduced states of |ψ (with respect to the first subsystem S1 and with respect to
the second subsystem S2 ) are one and the same mixed state:

(1) (2) 1 (1)

Red[1,1] (|ψ) = Red[1,1] (|ψ) = I
2

(where I(1) is the identity operator of the space H (1) = C2 ).

Since |ψ is a pure state, while 12 I(1) is a proper mixture, we obtain:

(1) (2)
P|ψ = Red[1,1] (|ψ) ⊗ Red[1,1] (|ψ).

35

www.pdfgrip.com
Entropy 2018, 20, 837

Thus, the states of the two parts of S turn out to be indistinguishable and entangled in the context |ψ.

Let us now turn to the concept of quantum probability, a basic “character” of the quantum theoretic
formalism that can assume different forms. We will first consider the concept of truth-probability that
is naturally connected with the idea of quantum truth. In Section 4 we will see how this concept will
play an important role in the semantics of quantum computational logics. As noticed in the Introduction,
in any space H (n) the concept of truth can be identified with a special projection operator indicated by
(n)
P1 . In this way, truth is dealt with as a mathematical representative of a possible physical event.
(n)
In order to define the truth-concept P1 (of H (n) ), we will first distinguish the true registers from
the false registers of the space.

Definition 3 (True and false registers). Let | x1 , . . . , xn  be a register of H (n) .

• | x1 , . . . , xn  is called true iff its last bit | xn  is |1;

• | x1 , . . . , xn  is called false iff its last bit | xn  is |0.

Thus, the truth-value of a register is determined by its last bit. As we will see in the next Section,
this choice turns out to be natural and useful in the theory of some important quantum logical gates,
where the last bit of an input-register represents the target that is transformed by the gate in question
into the final truth-value of the output.
On this basis one can now define the concepts of truth and falsity of a space H (n) .

Definition 4 (Truth and falsity).

(n)
• The truth of the space H (n) is the projection P1 that projects over the closed subspace spanned by the set
of all true registers.
(n)
• The falsity of the space H (n) is the projection P0 that projects over the closed subspace spanned by the set
of all false registers.

(1)
As a particular case, we obtain that the truth P1 of the space C2 is the projection P|1 (which
projects over the closed subspace determined by the bit |1).
Now, we can naturally apply the basic probabilistic rule of quantum theory: the Born-rule, which
determines for any state ρ and for any physical event represented by a projection operator P (of a
Hilbert space H), the probability that a quantum system in state ρ verifies the event represented by P.
According to this rule we have:
Probρ ( P) = tr(ρP)

(where Probρ ( P) represents the probability that the state ρ assigns to the event P ).
(n)
By applying the Born-rule to the particular case of the truth-concept P1 we obtain for any state ρ
of H (n) :
(n) (n)
Probρ ( P1 ) = tr(ρP1 ).
(n)
From an intuitive point of view, Probρ ( P1 ) represents the probability that the quantum
(n)
information ρ is true. Since P1 is a constant projection operator (in the space H (n) ) we can briefly
(n) (n)
write p1 (ρ), instead of Probρ ( P1 ). Once chosen the truth-concept P1 (in any space H (n) ), every
piece of quantum information turns out to have a well-determined probability-value of being true.
An important property of the function p1 is asserted by the following Lemma.

Lemma 1. For any ρ ∈ D(H (n) ),

(2) (1)
p1 (ρ) = tr( Red[n−1,1] (ρ) P1 ).

36

www.pdfgrip.com
Entropy 2018, 20, 837

As observed in the Introduction, it is interesting to consider also a different notion of quantum

probability, that represents a kind of relative probability. Let us first refer to the case of quregisters
(living in a given space H (n) ) and suppose that |ψ = ∑i ci | xi1 , . . . , xin  represents the information that
an epistemic agent has at a given time. The quregister |ψ allows us to assign probability-values to all
registers | xi1 , . . . , xin  (which might represent possible outcomes of a measurement or of a computation).
For any | xi1 , . . . , xin , the probability that |ψ assigns to the outcome | xi1 , . . . , xin  is |ci |2 . Apparently,
one is dealing with a kind of relative probability: when we have the information |ψ, we might have the
information | xi1 , . . . , xin  with probability |ci |2 .
In the general case, this idea of relative probability turns out to be strongly connected with the
notion of fidelity: a concept (introduced by Uhlmann and by Jozsa), that represents a generalization of
the notion of of transition-probability for pure states (see [6,7]).

Definition 5 (Fidelity). Let |ψ and | ϕ be two quregisters of H (n) . The fidelity between |ψ and | ϕ is
defined as follows:
F (|ψ, | ϕ) := |ψ| ϕ|2

(where ψ| ϕ is the inner product of |ψ and | ϕ).

The following theorem sums up some basic properties of the notion of fidelity.

Theorem 1.

1. F (|ψ, | ϕ) ∈ [0, 1].

2. F (|ψ, | ϕ) = F (| ϕ, |ψ).
3. F (|ψ, | ϕ) = P|ψ | ϕ 2 = P| ϕ |ψ 2 .
4. F (|ψ, | ϕ) = tr( P|ψ P| ϕ ) = tr( P| ϕ P|ψ ).
5. F (|ψ, | ϕ) = 1 iff P|ψ = P| ϕ .
6. F (|ψ, | ϕ) = 0 iff |ψ ⊥ | ϕ.

From an intuitive point of view one can say that the real number F (|ψ, | ϕ) measures “how close”
are the two pure states |ψ and | ϕ in the Hilbert space under consideration.
Since F (|ψ, | ϕ) = tr( P|ψ P| ϕ ) = tr( P| ϕ P|ψ ), recalling the Born-rule, the number F (|ψ, | ϕ)
can be naturally interpreted as a relative probability: when we have the information |ψ, we might
have the information | ϕ with probability F (|ψ, | ϕ), or vice versa. Accordingly we can write:

p|ψ (| ϕ) = F (|ψ, | ϕ) = F (| ϕ, |ψ) = p| ϕ (|ψ),

stressing that F (|ψ, | ϕ) represents the probability of the information | ϕ under the condition |ψ,
or vice versa. In this sense one can say that the notion of fidelity represents a special concept of
conditional probability that, unlike the general case, does satisfy the symmetry-property. For a general
discussion of conditional probabilities in quantum theory see [8].
The concept of fidelity (which represents a form of transition-probability in the case of pure states)
can be generalized to density operators.

Definition 6 (Fidelity for density operators). Let ρ and σ be two density operators of H (n) .
√ √
F (ρ, σ) := tr( ρσ ρ)2 .

Interestingly enough, Jozsa has proved that this notion of fidelity between density operators
coincides with the notion of transition-probability between density operators (via purification of
mixtures), investigated by Uhlmann (see [7]).
The concept of fidelity for density operators represents a good generalization of the concept of
fidelity for pure states, as stated by the following Lemma.

37

www.pdfgrip.com
Entropy 2018, 20, 837

Lemma 2.
F ( P|ψ , P| ϕ ) = |ψ| ϕ|2 .

The following Theorem sums up the basic properties of the concept of fidelity for density operators.

Theorem 2.

1. F (ρ, σ ) ∈ [0, 1].

2. F (ρ, σ ) = 1 iff ρ = σ.
3. F (ρ, σ ) = F (σ, ρ).
4. F (UρU† , UσU† ) = F (ρ, σ), for any unitary operator U of H (n) (where U† is the adjoint of U). Thus, fidelity
is preserved by unitary operators.
5. F (ρ, σ ) = tr(ρσ), if either ρ or σ is a pure state.

On this basis, it seems reasonable to assume that the number F (ρ, σ ) represents a form of relative
probability for quantum states (which may be either pure or mixed). Accordingly, we will write (like in
the case of pure states):
pρ (σ ) = F (ρ, σ).

Is it possible to define fidelity for density operators that belong to different spaces? Suppose
that ρ ∈ D(H (n) ), σ ∈ D(H (m) ) and n > m. The intuitive idea is that ρ and σ can be compared in the
framework of the smaller space H (m) , by referring to a special reduced state of ρ that represents a part
of ρ living in H (m) . The choice of comparing pieces of information that live in different spaces in the
framework of the smaller space seems to be quite natural. In fact, both in the case of human and of
artificial intelligence it often happens that agents endowed with a richer knowledge communicate
with less informed agents by “reducing” their information-level to the level of the “more ignorant”
agents. Accordingly, the concept of generalized fidelity can be defined as follows.

Definition 7 (Generalized fidelity). Let ρ ∈ D(H (n) ) and σ ∈ D(H (m) ).

⎧
⎪
⎪ F (ρ, σ ), if n = m;
⎨
(2)
Fg (ρ, σ ) := F ( Red[n−m,m] (ρ), σ), if n > m;
⎪
⎪
⎩ F (ρ, Red(2) (σ)), if n < m.
[m−n,n]

The map Fg is clearly symmetric.

Thus, we can put:
pρ (σ) := Fg (ρ, σ ).

We obtain:

pρ (σ) := Fg (ρ, σ ) = Fg (σ, ρ) = pσ (ρ).

In this way, for any choice of ρ, the probability-function pρ turns out to be defined on the set D of
all possible pieces of quantum information.
It is interesting to compare the probability functions p1 and pρ . As we have seen, once chosen
the truth-concept in any space H (n) , the function p1 represents a kind of absolute concept of
probability: every piece of quantum information ρ (living in whatever space) has a well-determined
probability-value of being true. The probability-function pρ represents, instead, a relative concept
that depends on the choice of ρ. From an intuitive point of view, ρ can be regarded as an information
(available for a given epistemic agent) which describes a situation that is essentially characterized by
some uncertain and vague features. On the basis of his/her information our agent is able to valuate
probabilistically other possible alternative situations, using the pρ -function.
Interestingly enough, the truth-probability function p1 can be represented as a special case of the
relative probability pρ .

38

www.pdfgrip.com
Entropy 2018, 20, 837

Theorem 3. For any ρ ∈ D(H (n) ),

(1)
p1 (ρ) = pρ ( P1 ).

Proof. Let ρ ∈ D(H (n) ). By definition of p1 we have:

(n)
p1 (ρ) = tr(ρP1 ).

Suppose that n = 1. Then,

(1) (1) (1)

p1 (ρ) = tr(ρP1 ) = Fg (ρ, P1 ) = pρ ( P1 ).

Suppose that n > 1. Then, by Lemma 1,

(2) (1) (1) (1)

p1 (ρ) = tr( Red[n−1,1] (ρ) P1 ) = Fg (ρ, P1 ) = pρ ( P1 ).

Both the truth-probability function p1 and the relative probability-function pρ determine a preorder
relation on the set D of all possible pieces of quantum information.

Definition 8 (The truth-preorder). For any σ1 , σ2 ∈ D,

σ1  σ2 iff p1 (σ1 ) ≤ p1 (σ2 ).

Definition 9 (The relative preorder). Consider a density operator ρ. For any σ1 , σ2 ∈ D,

σ1 ρ σ2 iff pρ (σ1 ) ≤ pρ (σ2 ).

One can easily check that both relations  and ρ are reflexive, transitive and generally
non-antisymmetric.

3. Quantum Logical Circuits

The basic idea of the theory of quantum computers is that computations can be performed by
some quantum objects that evolve in time. Recalling that (according to Schrödinger’s equation) the
time-evolution of quantum systems is mathematically described by unitary operators, it is natural
to assume that quantum information is processed by quantum logical gates (briefly, gates): special
examples of unitary operators that transform (in a reversible way) the pure states of the quantum
systems that store the information in question. Any gate G(n) (defined on the space H (n) ) can be
canonically extended to a unitary operation D G(n) (defined on the set D(H (n) ) of all density operators
of H (n) ) according to the rule:

∀ρ ∈ D(H(n) ) : D (n)
ρ = G( n ) ρ G( n )
†
G

(where G(n) is the adjoint of G(n) ). For the sake of simplicity, we will call gate either a unitary operator
†

G(n) or the corresponding unitary operation D G(n) .

We will now recall the definitions of some basic gates that play an important role both from
the computational and from the logical point of view. We will first consider some gates, called
“semiclassical”, that cannot “create” superpositions from register-inputs. Two important examples are
represented by the negation-gate and by the Toffoli-gate.

39

www.pdfgrip.com
Entropy 2018, 20, 837

Definition 10 (The negation-gate on the space H (1) ). The negation-gate on H (1) is the linear operator
NOT(1) that satisfies the following condition for every element | x  of the canonical basis:

NOT(1) | x  := |1 − x .

The operator NOT(1) represents a natural quantum generalization of the classical negation.
We have:
NOT(1) |0 = |1; NOT(1) |1 = |0.

The negation-gate can be naturally generalized to higher-dimensional spaces. For any H (n) (with
n > 1), the operator NOT(n) is defined for every element | x1 , . . . , xn  of the canonical basis as follows:

NOT(n) | x1 , . . . , xn  := | x1 , . . . , xn−1  ⊗ NOT(1) | xn .

Apparently, NOT(n) always acts on the last bit of any register of H (n) .
The smallest space where the Toffoli-gate can be defined is the space H (3) .

Definition 11 (The Toffoli-gate on the space H (3) ). The Toffoli-gate on H (3) is the linear operator T(1,1,1)
that satisfies the following condition for every element | x, y, z of the canonical basis:
#
| x, y, x  y, if z = 0;
T(1,1,1,) | x, y, z :=
| x, y, ( x  y) , if z = 1
(where  and are the infimum and the complement of the two-valued Boolean algebra based on the set {0, 1}).

Thus, the Toffoli-gate leaves unchanged the first two bits | x  and |y (which represent the
control-bits); while the third bit |z (representing the target-bit) is transformed into

• the bit corresponding to the Boolean conjunction of the two control-bits, when z = 0;
• the bit corresponding to the Boolean negation of the conjunction of the two control-bits, when
z = 1.

Like the negation-gate, the Toffoli-gate also can be generalized to higher-dimensional spaces.
For any m, n ≥ 1, the Toffoli-gate T(m,n,1) on the space H (m+n+1) is defined as follows:

T(m,n,1) | x1 , . . . , xm , y1 , . . . , yn , z :=

| x1 , . . . , xm−1 , yn−1 , y1 , . . . , yn−2  ⊗ T(1,1,1) | xm , yn , z.

The Toffoli-gate has a special logical interest, since it allows us to define a quantum logical
conjunction that behaves as a reversible operation. For any choice of two natural numbers m, n (such
that m, n ≥ 1) the reversible conjunction AND(m,n) is dealt with as a holistic monadic operator that acts
on global pieces of quantum information represented by quregisters of the space H (m+n) . Accordingly,
any quregister of H (m+n) can be regarded as a holistic description of two possible members of the
conjunction AND(m,n) , which live in the space H (m) and H (n) , respectively.

Definition 12 (The conjunction on the space H (m+n) ). For any quregister |ψ of H (m+n) ,

AND(m,n) |ψ := T(m,n,1) (|ψ ⊗ |0)

(where the bit |0 plays the role of an ancilla).

In the particular case of mixed states ρ ∈ D(H (m+n) ) we will write:

(1)
D
AND(m,n) (ρ) for D (m,n,1)
T (ρ ⊗ P0 ),

40

www.pdfgrip.com
Entropy 2018, 20, 837

where D T(m,n,1) is the unitary quantum operation that corresponds to the unitary operator T(m,n,1) .
As a special case consider a register | x, y of the space H (2) . We obtain:
• AND(1,1) | x, y = T(1,1,1) | x, y, 0 = |1, 1, 1, if x = y = 1.
• AND(1,1) | x, y = T(1,1,1) | x, y, 0 = | x, y, 0, if x = 0 or y = 0.
Thus, AND(1,1) represents a “good” quantum generalization of classical conjunction. At the same
time, this particular form of quantum conjunction gives rise to a characteristic holistic behavior, which
is deeply rooted in the holistic features of the quantum formalism (as shown by the following example).

Example 2. Consider the quregister represented by the following (entangled) Bell-state:

1 1
|ψ = √ |1, 1 + √ |0, 0.
2 2
We have:
• AND(1,1) |ψ = T(1,1,1) (|ψ ⊗ |0) = √1 |0, 0, 0 + √1 |1, 1, 1;
2 2
D AND(1,1) ( P ) D T(1,1,1) ( P (1)
• |ψ = |ψ ⊗ P0 ) = P √1 |0,0,0+ √1 |1,1,1 .
2 2

Hence, AND(1,1) |ψ and D AND(1,1) ( P|ψ ) represent a pure state of the space H (3) . At the same time, the two
reduced states of P|ψ turn out to be one and the same proper mixture (of the space H (1) ):

1 (1)
Red(1) ( P|ψ ) = Red(2) ( P|ψ ) = I .
2
Consequently, the conjunction over the global state P|ψ cannot be represented as the conjunction of the
states of the two separate parts:

AND(1,1) ( P|ψ ) = AND(1,1) ( Red(1) ( P|ψ ) ⊗ Red(2) ( P|ψ )).

This gives rise to a clear violation of the compositionality-principle.

As semiclassical gates, the negation-gate and the Toffoli-gate are unable to “create” superpositions
from register-inputs. Of course, quantum computer theory cannot help using also “genuine quantum
gates” that can transform classical inputs (registers) into genuine superpositions (which are responsible
for the characteristic parallel structures of quantum computations). An important example is
represented by theHadamard-gate.

Definition 13 (The Hadamard-gate on the space H (1) ). The Hadamard-gate on H (1) is the linear operator
√ (1)
I that satisfies the following conditions:
√ (1) 1 1 √ (1) 1 1
I |0 = √ |0 + √ |1; I |1 = √ |0 − √ |1.
2 2 2 2

Thus, the Hadamard-gate transforms both bits into two distinct genuine superpositions that
might be either true or false with probability 12 .
Like the negation and the Toffoli-gate, the Hadamard-gate also can be generalized to
√ (n)
higher-dimensional spaces. For any H (n) (with n > 1), the operator I is defined for every
element | x1 , . . . , xn  of the canonical basis as follows:
√ (n) √ (1)
I | x1 , . . . , x n  : = | x1 , . . . , x n −1  ⊗ I | x n .

Quantum computations are performed by quantum circuits. Mathematically any quantum circuit
can be described as a finite sequence of gates, all defined on one and the same quantum computational
space H (n) . Since gates may be either unitary operators or unitary quantum operations, we will write:

41

www.pdfgrip.com
Entropy 2018, 20, 837

(n) (n) (n) D (n)

C = (G1 , . . . , Gt ) or C D = (D G1 , . . . , Gt ).

Example 3. An interesting example is represented by the following quantum circuit (also called
“Mach–Zehnder circuit”):
√ (1) √ (1)
C = ( I , NOT(1) , I ).
As is well known, this circuit can be implemented by a Mach–Zehnder interferometer (Figure 1), where
the Hadamard-gate is physically realized by a beam-splitter (BS), while the negation-gate is realized by a pair
of mirrors (M).

Figure 1. The Mach–Zehnder interferometer.

It is worth while recalling how the properties of the Mach–Zehnder circuit and of the Mach–Zehnder
interferometers have been an important object of discussion in many foundational debates about quantum theory.
The main intuitive “strangeness” is represented by the following mathematical result (which is confirmed by the
experimental evidence) :
√ (1) √ (1) √ (1) √ (1)
I NOT(1) I |0 = |0; I NOT(1) I |1 = −|1
(where the bit |0 is supposed to describe a photon-beam moving along the x-direction, while the bit |1 describes
a photon-beam moving along the y-direction). Such result seems to contradict any classical physical expectation,
according to which a beam that has entered into the interferometer along the x-direction, after having crossed
the second beam-splitter should be detected with probability-value 12 either along the x-direction or along
the y-direction.

4. A Logical Abstraction: Quantum Computational Logics

The theory of quantum circuits has inspired a natural logical abstraction, suggesting the
development of new forms of quantum logic, that have been called quantum computational logics.
As is well known, the prototypical example of quantum logic is the logic created by Birkhoff and
von Neumann in their celebrated article “The logic of quantum mechanics”, and usually called Birkhoff
and von Neumann’s quantum logic (see [9]). The basic aim of the original quantum logical approach
to quantum theory was the development of an abstract analysis of the relationships between the states

42

www.pdfgrip.com
Entropy 2018, 20, 837

of a quantum system S and the quantum events that may occur to S, which can be mathematically
represented as projections P of the Hilbert space HS associated to S. In the semantics of Birkhoff and
von Neumann’s quantum logic, the formulas of the quantum logical language are supposed to denote
quantum events: projections P (of a given space HS ) to which every state of S assigns a well-determined
probability-value. At the same time, the basic logical connectives (negation, conjunction, disjunction)
are interpreted as special (generally irreversible) algebraic operations that can be defined on the set of
all projections P of HS .
In the logical community, Birkhoff and von Neumann’s quantum logic has been often regarded
as a very peculiar and somewhat strange form of non-classical logic, for which some important
metalogical questions (like axiomatizability) are still open problems. An axiomatizable version of
quantum logic can be obtained by means of a convenient weakening of Birkhoff and von Neumann’s
quantum logic. This logic (often called abstract quantum logic) can be semantically characterized by
referring to the class of all orthomodular lattices (which contains, as special cases, the orthomodular
lattices based on the set of all projections of a Hilbert space, see, for instance, [10]).
The semantics of quantum computational logics has been inspired by quite different intuitive ideas,
which can be briefly sketched as follows:
• any formula α of a quantum computational language is supposed to denote a piece of quantum
information ρ, living in a Hilbert space whose dimension depends on the linguistic complexity of α;
• logical connectives are interpreted as particular gates (for a more detailed exposition see [11–13]).
Consequently, unlike the case of traditional quantum logics, the formulas of quantum computational
logics turn out to have a characteristic dynamic feature, representing possible computation-actions.
In this article we will refer to a minimal (quantum computational) sentential language L,
whose alphabet contains:
1. atomic formulas, including two special formulas t and f that denote the truth and the
falsity, respectively;
2. the following logical connectives:
• the negation ¬, corresponding to the gate negation;
• the Toffoli-connective , corresponding
√ to the Toffoli-gate;
• the Hadamard-connective id, corresponding to the Hadamard-gate.
These connectives simulate, at a syntactical level, the behavior of the corresponding gates. While
the negation and the Hadamard-connective are 1-ary connectives (which act on single formulas),
the Toffoli-connective is a ternary connective: if α, β are formulas and q is an atomic formula, then
(α, β, q) is a formula. On this basis, recalling the definition of the quantum computational conjunction
AND(m,n) , a binary conjunction-connective ∧ can be defined in terms of the Toffoli-connective:

α ∧ β := (α, β, f)

(where the false formula f plays the role of a syntactical ancilla).

We obtain, in this way, an appropriate logical language for the description of a class of quantum
circuits. For instance, the Mach–Zehnder circuit:
√ (1) √ (1)
C = ( I , NOT(1) , I )

can be naturally described by the following “Mach–Zehnder formula”:

√ √
id ¬ id q.

A syntactical notion that plays an important semantic role is the concept of atomic complexity of
a formula. As we will see, this concept provides a link between the language and the Hilbert-space
environment, where the meanings quantum computational formulas are supposed to live.

43

www.pdfgrip.com
Entropy 2018, 20, 837

Definition 14 (Atomic complexity). The atomic complexity At(α) of a formula α is the number of occurrences
of atomic formulas in α.

Example 4. Consider the (contradictory) formula

α = q ∧ ¬q = (q, ¬q, f).

We have: At(α) = 3.

For any formula α consider the space H ( At(α)) (which is determined by the atomic complexity of
α). This space is called the semantic space of α, where any piece of quantum information representing a
possible meaning of α shall live. We will briefly write: H α , instead of H ( At(α)) .
Any formula α can be decomposed into its parts, giving rise to a syntactical configuration called
the syntactical tree of α. Let us first consider a particular example.

Example 5. Consider again the formula

α = q ∧ ¬q = (q, ¬q, f).

The syntactical tree of α is the following sequence of levels, where each level is a particular sequence of
subformulas of α:

Level3α = (q, q, f)

Level2α = (q, ¬q, f)

Level1α = ((q, ¬q, f))

In the general case, the levels of the syntactical tree of a given formula α are determined in the
following way:

• the bottom level Level1α is (α);

• the top level Levelhα is the sequence of atomic formulas occurring in α;
• Leveli+1 (where 1 ≤ i < h) is obtained by dropping the principal connective in all molecular
formulas occurring at Leveli and by repeating all atomic formulas that occur at Leveli .

The syntactical tree of any formula α uniquely determines a quantum circuit: a sequence of gates
all defined on the semantic space of α. Such sequence is called the gate tree of α. For instance, the gate
tree of the formula α = (q, ¬q, f) is the following circuit:

(I(1) ⊗ NOT(1) ⊗ I(1) , T(1,1,1) ).

In fact, the second level of the syntactical tree of α has been obtained from the third level by
repeating the first occurrence of q, by negating the second occurrence of q and by repeating f. The first
level has been obtained from the second level by applying the Toffoli-connective to the three sentences
occurring at the second level.
While the atomic complexity of α corresponds to the width (i.e., the number of wires) of the circuit
described by α, the number of levels of the syntactical tree of α (called the height of α) corresponds to
the depth (i.e., the number of computational steps) of the circuit in question.
We will now briefly sum up the basic concepts of the holistic quantum computational semantics.
The notion of holistic model of the language L is based on the weaker notion of holistic map: a map

44

www.pdfgrip.com
Entropy 2018, 20, 837

Hol that assigns to each level of the syntactical tree of any formula α a global meaning, represented by
a density operator living in the semantic space of α. We have:

Hol : Leveliα → ρ ∈ D(H α ).

On this basis, the meaning that Hol assigns to α is identified with the meaning that Hol assigns to
the bottom level of the syntactical tree of α:

Hol(α) = Hol( Level1α ).

Given a formula α, any holistic map Hol determines the contextual meaning with respect to the
context Hol(α) of any occurrence of a subformula β in α. Suppose that

Leveliα = ( β i1 , . . . , β ir ).

The contextual meaning of β i j with respect to the context Hol(α) can be naturally defined using the
notion of reduced state:

Holα ( β i j ) := Red( j) (Hol( Leveli (α))).

Remark 1. Notice how our definition of contextual meaning of a quantum computational formula brings about
an interesting connection between the notion of contextuality in linguistic frameworks and the concept of physical
contextuality, that plays an important role in many quantum-theoretic problems. As an example, we might
recall the debates about the foundational consequences of Kochen and Specker’s theorems and about the logical
possibility of deterministic completions of quantum theory via a non-contextual hidden- variable theory.

The concept of holistic model of the language L can be now defined as a holistic map that satisfies
some natural logical restrictions.

Definition 15 (Holistic model). A holistic model of the language L is a holistic map Hol that satisfies the
following conditions:

1. Hol preserves the logical form of all formulas. Thus, the meaning of each Leveliα (different from the top
level) of the syntactical tree of α is obtained by applying the corresponding gate Giα (of the gate tree of α) to
the meaning of Leveliα+1 :
Hol( Leveliα ) = Giα ( Leveliα+1 ).
2. Hol assigns the same contextual meaning to different occurrences (in the syntactical tree of α) of one and
the same subformula of α.
(1)
3. The contextual meanings assigned by Hol to the true formula t and to the false formula f are: P1 (the
(1)
truth) and P0 (the falsity), respectively.

Suppose that the meaning assigned by a model Hol to a formula α is a pure state, whose form is:

c1 |ψ1  + . . . + cn |ψn  (where ci = 0).

From an intuitive point of view Hol(α) can be regarded as a vague, ambiguous idea that alludes
to other ideas (represented by the pieces of information |ψ1 , . . . , |ψn ) that are, in a sense, all co-existent.
Notice that any meaning Hol(α) represents a kind of autonomous semantic context that is not
necessarily correlated with the meanings of other formulas. Generally we have:

Holα (γ) = Holβ (γ).

45

www.pdfgrip.com
Entropy 2018, 20, 837

Hence, one and the same formula may receive different contextual meanings in different contexts.
As, in fact, happens in the case of our normal use of natural languages.
The following Lemma (which might appear prima facie obvious) asserts a highly non-trivial
property that plays an important role in the development of the holistic quantum computational
semantics (for a proof of Lemma 2 see [13]).

Lemma 3. Consider a formula γ and let η be a subformula of γ. For any model Hol and for any formula β there
exists a model ∗ Hol such that:

∗
Holγ∧ β (η ) = Holγ (η ).

The concepts of truth, validity and logical consequence (in the framework of the holistic quantum
computational semantics) can be now defined as follows.

Definition 16 (Truth, validity and logical consequence).

1. Hol α (α is true in a model Hol) iff p1 (Hol(α)) = 1.

2.  α (α is valid) iff for any model Hol, Hol α.
3. α  β (β is a logical consequence of α) iff for any formula γ such that α and β are subformulas of γ and for
any model Hol,
p1 (Holγ (α)) ≤ p1 (Holγ ( β)).

Apparently, both truth and logical consequence are, in this semantics, probabilistic concepts,
based on the probability-function p1 . In spite of the strong contextual features of the holistic quantum
computational semantics, one can prove that this holistic notion of logical consequence satisfies the
transitivity-property.

Theorem 4.
α  β and β  δ ⇒ α  δ.

Proof. Assume the hypothesis and suppose, by contradiction, that there exists a model Hol and a
formula γ, where α and δ occur as subformulas, such that: p1 (Holγ (α)) ≤ p1 (Holγ (δ)). Consider
the formula γ ∧ β. By Lemma 3 there exists a model ∗ Hol such that for any η that is a subformula of
γ: ∗ Holγ∧ β (η ) = Holγ (η ). Thus, we have: ∗ Holγ∧ β (α) = Holγ (α) and ∗ Holγ∧ β (δ) = Holγ (δ) . Since
we have assumed (by contradiction) that p1 (Holγ (α)) ≤ p1 (Holγ (δ)), we obtain: p1 (∗ Holγ∧ β (α)) ≤
p1 (∗ Holγ∧ β (δ)), against the hypothesis (and the transitivity of ≤), which imply: p1 (∗ Holγ∧ β (α)) ≤
p1 (∗ Holγ∧ β ( β)); p1 (∗ Holγ∧ β ( β)) ≤ p1 (∗ Holγ∧ β (δ)); ∗ Holγ∧ β (α) ≤ p1 (∗ Holγ∧ β (δ)).

The logic that is semantically characterized by the concept of logical consequence defined in
Definition 16 has been called holistic quantum computational logic (HQCL). One is dealing with a very
weak form of logic, where many important logical arguments (valid either in classical logic or in Birkhoff
and von Neumann’s quantum logic) may be violated.
The following two Theorems sum up some important logical arguments that are either valid or
possibly violated in the framework of the logic HQCL (proofs can be found in [13]).

Theorem 5.

1. αα
2. α ∧ β  α; α ∧ β  β
3. α  β =⇒ α ∧ δ  β
4. α  ¬¬α; ¬¬α  α
5. α  β =⇒ ¬ β  ¬α

46

www.pdfgrip.com
Entropy 2018, 20, 837

6. f  β;
√ √ βt √ √
7. √id idα  α; √
α  id√ idα √
8. id(α ∧ β)  idf; idf  id(α ∧ β)

Theorem 6.

1.  ¬(α ∧ ¬α)
2. α  α ∧ α
3. α ∧ β  β ∧ α
4. α ∧ ( β ∧ γ )  ( α ∧ β ) ∧ γ

Thus, conjunctions are generally non-idempotent, non-commutative and non-associative.

Such violations can be explained by recalling the contextual behavior of quantum meanings. It may
happen that:
p1 (Holδ (α)) = p1 (Holδ (α ∧ α))

p1 (Holδ (α ∧ β)) = p1 (Holδ ( β ∧ α))

p1 (Holδ (α ∧ ( β ∧ γ))) = p1 (Holδ ((α ∧ β) ∧ γ)).

All this seems to be strongly in agreement with a number of informal ways of reasoning where
conjunctions are frequently used as non-idempotent, non-commutative and non-associative logical
operations. As is well known, the semantics of natural languages is essentially holistic and contextual.
We need only think how children learn their mother-language, showing an extraordinary capacity
of understanding and using correctly the contextual meanings of expressions that occur in different
contexts. And it often happens that the meaning of a global expression is grasped and used in a clear
and correct way, while the meanings of its parts appear more vague and ambiguous.
Different forms of holistic quantum computational logics (also in a first-order version) can be applied
to investigate semantic phenomena where holism, contextuality and ambiguity play an important
role, as happens not only in the case of natural languages but also in the languages of art (say, poetry
or music, for instance, see [2]). Of course the holistic quantum computational semantics does not
forbid compositional semantic situations, which can be described as special cases of the holistic
semantics. Interestingly enough, conjunctions are always commutative and associative, but generally
non- idempotent in the framework of the compositional fragment of the holistic quantum semantics.

5. Quantum Epistemic Operations and Quantum Probabilities

Quantum information has brought about some intriguing epistemic situations, that have inspired
new ideas in the field of epistemic logics. Is it possible to represent epistemic operators as special
examples of operations in a Hilbert space environment? This question admits a positive answer.
A characteristic feature of the quantum computational approach to epistemic logic is the use of
the notion of truth-perspective: each epistemic agent (say, Alice, Bob, ...) is supposed to be associated
to a particular truth-perspective that represents his/her idea of truth. Truth-perspective changes
may give rise to some interesting relativistic-like epistemic effects: if Alice and Bob have different
truth-perspectives, Alice might see a kind of deformation in Bob’s logical behavior (see [14]).
From the mathematical point of view we assume that the choice of a given truth-perspective is
determined by the choice of a particular orthonormal basis of the Hilbert space C2 . In Section 2 we have
(n)
seen how the truth-concept P1 (of H (n) ) has been defined by referring to the canonical basis of the
space. But, of course, the choice of a particular basis of a given Hilbert space is a matter of convention.
Consider the space C2 . Any orthonormal basis of this space can be described as determined by the
application of a unitary operator T to the elements of the canonical basis {|0, |1}. We can think
that the operator T gives rise to a change of truth-perspective. While the classical truth-values Truth
and Falsity have been identified with the two bits |1 and |0, assuming a different basis corresponds
to a different idea of Truth and Falsity. Since any basis-change in C2 is determined by the choice

47

www.pdfgrip.com
Entropy 2018, 20, 837

of a particular unitary operator, we can identify a truth-perspective with a unitary operator T of C2 .

We will write:
|1T  = T|1; |0T  = T|0,
and we will assume that |1T  and |0T  represent, respectively, the truth-values Truth and Falsity of the
truth-perspective T. The canonical truth-perspective is, of course, determined by the identity operator
(1) (1)
I(1) . We will indicate by BT the orthonormal basis determined by T; while BI will represent the
canonical basis. From a physical point of view, we can suppose that each truth-perspective is associated
to an apparatus that allows one to measure a given observable.
Any unitary operator T of H (1) (representing a truth-perspective) can be naturally extended to a
unitary operator T(n) of H (n) (for any n > 1):

T( n ) | x 1 , . . . , x n  = T | x 1  ⊗ . . . ⊗ T | x n .

(n)
Accordingly, any choice of a unitary operator T of H (1) determines an orthonormal basis BT for
H (n) such that:  
(n) (n)
B T = T( n ) | x 1 , . . . , x n  : | x 1 , . . . , x n  ∈ B I .

(1)
Instead of T(n) | x1 , . . . , xn  we will also write: | x1T , . . . , xnT . The elements of BT will be called
(n)
the T-bits of H (1) ;
while the elements of will represent the T-registers of
BT H (n) .
The notions of truth, falsity and truth-probability can be now generalized to any truth-perspective T.

Definition 17 (T-true and T-false registers).

• | x1T , . . . , xnT  is a T-true register iff | xnT  = |1T ;

• | x1T , . . . , xnT  is a T-false register iff | xnT  = |0T .

Definition 18 (T-truth and T-falsity ).

(n)
• The T-truth of H (n) is the projection operator T P1 that projects over the closed subspace spanned by the
set of all T- true registers;
(n)
• the T-falsity of H (n) is the projection operator T P0 that projects over the closed subspace spanned by the
set of all T- false registers.

Definition 19 (T-probability). For any ρ ∈ D(H (n) ),

T (n)
pT
1 ( ρ ) : = tr( ρ P1 ).

It is worth while noticing that, unlike the probability function p1 , the relative probability function
pρ cannot be reasonably generalized to different truth-perspectives: as we have seen, the definition of
fidelity does not depend on the choice of a particular basis of the space.
One can show that all gates can be canonically transposed from the canonical truth-perspective
to any truth-perspective T. Hence, any quantum circuit C has a corresponding T-version C T (for any
truth-perspective T).
In some previous works we have studied a quantum representation of the epistemic operations
(n)
“to know”, “to believe”, “to understand”, whose properties depend on the notion of T-truth T P1 and
on the probability-function pT 1 (see [14,15]). We will now consider another epistemic operation, which
is informally used in a number of interesting quantum situations: the operation “being probabilistically
informed”.
Let us first recall the definition of a quantum version of the most important concept of epistemic
logics: the concept of knowledge-operation.

48

www.pdfgrip.com
 Entropy 2018, 20, 837

Definition 20 (Knowledge-operation). A knowledge-operation of a Hilbert space H (n) (with respect to

the truth-perspective T) is a map
(n)
KT : D(H (n) ) → D(H (n) ).

The following conditions are required:

(n)
(1) K(n) is associated with an epistemic domain EpD (KT ) that is a subset of D(H (n) );
(n) (n)
(2) pT T
1 (KT ρ ) ≤ p1 ( ρ ), for any ρ ∈ EpD (KT ).

(n)
As expected, the intuitive interpretation of KT ρ is the following: the piece of information ρ is
(n)
known by a given agent whose truth-perspective is T. The knowledge described by KT is limited
by a given epistemic domain, which is intended to represent the information accessible to our agent,
(n)
relatively to the space H (n) (the epistemic domain of KT should not be confused with the domain
(n) (n)
of KT , which coincides with the set of all density operators of the space: KT ρ is defined, even if ρ
(n) (n)
does not belong to the epistemic domain of KT ).Whenever ρ belongs to the epistemic domain of KT ,
(n)
it seems reasonable to assume that the probability-values of ρ and KT ρ
are correlated: the probability
of the quantum information asserting that “ρ is known” should always be less than or equal to the
probability of ρ. Hence, in particular, we have:

(n)
pT T
1 (KT ρ ) = 1 ⇒ p1 ( ρ ) = 1.

But generally, not the other way around. In other words, pieces of quantum information that are
certainly known are certainly true (with respect to the truth-perspective in question). This condition
is clearly in agreement with a general principle of standard epistemic logics, according to which
“knowledge implies truth, but generally not the other way around”.
(n)
A knowledge-operation KT is called non-trivial iff for at least one density operator ρ ∈
(n) (n)
EpD (KT ), pT
1 ( KT ρ )
< pT
1 ( ρ ).
Can knowledge-operations be always represented as (reversible) gates? This question has a
negative answer, as proved by the following theorem (for a proof of this theorem see [15]).

Theorem 7. Non-trivial knowledge-operations cannot be generally represented as unitary quantum operations.

Apparently, the “act of knowing” gives rise to a characteristic reversibility-breaking, which is

quite similar to what happens in the case of quantum measurements.
(n)
We will now introduce the operation “being probabilistically informed” (indicated by IT ),
which arises in some interesting situations when an epistemic agent (say, Alice) has a given
probabilistic information.

Example 6. As an example, we can refer to a teleportation-experiment, where, at the initial time, Alice has a
probabilistic information, represented by the genuine qubit

|ψ = c0 |0 + c1 |1 (with c0 , c1 = 1),

that shall be teleported to the “far” Bob. While Alice is probabilistically informed about the piece of quantum
information |ψ (with respect to the canonical truth-perspective), one cannot say that “Alice certainly knows
|ψ”. Since P|ψ is supposed to belong to Alice’s epistemic domain, we would obtain :

p1 ( P|ψ ) = 1

(against the hypothesis that |ψ is a genuine qubit).

49

www.pdfgrip.com
 Entropy 2018, 20, 837

(n)
The intuitive interpretation of IT ρ is the following: a given epistemic agent, with
truth-perspective T, is probabilistically informed about ρ (in the framework of the space H (n) ).
(n) (n)
As happens in the case of the knowledge-operation KT , the operation IT is associated to an
(n)
information-domain In f D (IT )
⊆ D(H (n) ),
which represents the set of pieces of information that the
agent under consideration is able to valuate probabilistically in the domain D(H (n) ). Unlike the case
of knowledge-operations (which shall satisfy the strong condition (2) of Definition 20) we admit the
following possibility (which occurs, for instance, in the case of the teleportation-example):
(n)
• ρ ∈ In f D (IT )
(n)
• pT
1 ( IT ρ ) = 1
• T
p1 (ρ) < 1

Definition 21 (Probabilistic information). A probabilistic information-operation of a Hilbert space H (n)

(with respect to the truth-perspective T) is a map

(n)
IT : D(H (n) ) → D(H (n) ).

The following conditions are required:

(n) (n)
(1) IT is associated with an information-domain In f D (IT ) that is a subset of D(H (n) );
(n) (n)
(2) ρ ∈ In f (IT ) ⇒ pT 1 (IT ρ ) = 1.

(n) (n)
It is worth while noticing that the domains EpD(KT ) and In f D(IT ) are not generally closed under
(n) (n)
the corresponding operations KT and IT .
Hence, the phenomenon of “epistemic self-consciousness” is
here avoided: Alice might know something (or might be probabilistically informed about something)
without knowing of knowing it (without being informed of being informed about it).
Every epistemic agent can be naturally associated to an epistemic situation, which is
characterized by the choice of a truth-perspective, of a knowledge-operation and of a probabilistic
information-operation (in any space H (n) ).

Definition 22 (Epistemic situation of an agent). Let i represent an epistemic agent. An epistemic situation
for i is a system
EpSiti = (Ti , EpDi , In f Di , Ki , Ii ),

where:

(1) Ti represents the truth-perspective of i.

(n)
(2) EpDi is a map that assigns to any n ≥ 1 a set EpDi ⊆ D(H (n) ) that represents the information
accessible to i in the information-environment D(H ). ( n )
(n) (n)
(3) In f Di is a map that assigns to any n ≥ 1 a set In f Di ⊆ EpDi that represents the information that i
is able to valuate probabilistically in the information-environment D(H (n) ).
(n)
(4) Ki is a map that assigns to any n ≥ 1 a knowledge-operation KTi (defined on H (n) ), which describes the
knowledge of i with respect to the information-environment D(H (n) ). The epistemic domain associated to
(n) (n)
the operation KTi is the set EpDi .
(n)
(5) Ii is a map that assigns to any n ≥ 1 an information-operation ITi (defined on H(n) ),
which describes the probabilistic information of i with respect to the information-environment D(H (n) ).
(n) (n)
The information-domain associated to the operation ITi is the set In f Di .
(n) (n)
(6) ∀ρ ∈ D(H(n) ) : pT Ti
1 (KTi ρ ) ≤ p1 (ITi ρ ).
i

The probability of knowing a given information is less than or equal to the probability of having a
probabilistic valuation about it. But, generally, not the other way around.

50

www.pdfgrip.com
Entropy 2018, 20, 837

On this basis one can develop a quantum epistemic semantics for a first-order language that can be
express formulas like:

• Kaα (a knows α);

• Iaα (a is probabilistically informed about α).

This semantics allows us to represent and to justify a number of significant features of our normal
use of epistemic notions in the framework of natural language. Interestingly enough, the unpleasant
phenomenon of logical omniscience is here avoided. Due to the limits of epistemic domains, Alice
might know a given sentence without knowing all its logical consequences. Furthermore, knowledge
and probabilistic information are not generally closed under logical conjunction, in accordance with
what happens in the case of concrete memories both of human and of artificial intelligence (see [14]).
In conclusion, we have seen how the holistic quantum computational semantics provides some
useful abstract tools that can be naturally applied to a formal analysis of concepts and problems in
fields that may be far apart from microphysics. Some interesting examples concern the use of crucial
epistemic concepts (like knowledge, information, belief, understanding) both in the case of rigorous
scientific arguments and in some informal ways of reasoning. Other examples regard the role of
ambiguity, vagueness, uncertainty and contextuality either in scientific theories or in our normal use
of natural languages or in the languages of art (say, poetry or music).
According to some traditional philosophical views, ambiguity and holism represent characteristic
features of human thought that cannot be adequately analyzed in the framework of scientific theories,
whose semantics is supposed to be essentially “sharp” and “analytical”. Interestingly enough, quantum
logics (in their different versions) have provided a significant bridge that might fill a gap between
humanistic and scientific disciplines.

Author Contributions: Funding acquisition, H.F., R.G. and G.S.; Writing—original draft, M.L.D.C., H.F., R.G., R.L.
and G.S.
Funding: This work has been partially supported by Regione Autonoma della Sardegna in the framework of the
project “Time-logical evolution of correlated microscopic systems” (CRP 55, L.R. 7/2007, 2015), by Fondazione
Banco di Sardegna & Regione Autonoma della Sardegna in the framework of the project “Science and its logics,
the representation’s dilemma”, cup: F72F16003220002, and by Fondazione Banco di Sardegna of the project
“Strategies and Technologies for Scientific Education and Dissemination”, cup: F71I17000330002.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Aerts, D.; Sassoli de Bianchi, M.; Sozzo, S.; Veloz, T. Modeling Human Decision-Making: An Overview of the
Brussels Quantum Approach. Found. Sci. 2018, 1–28. [CrossRef]
2. Dalla Chiara, M.L.; Giuntini, R.; Luciani, A.R.; Negri, E. From Quantum Information to Musical Semantics;
College Publications: London, UK, 2012.
3. Haven, E.; Khrennikov, A. Quantum Social Science; Cambridge University Press: Cambridge, UK, 2013.
4. Beltrametti, E.; Dalla Chiara, M.L.; Giuntini, R.; Sergioli, G. Quantum teleportation and quantum epistemic
semantics. Math. Slovaca 2012, 62, 1121–1144. [CrossRef]
5. Dalla Chiara, M.L.; Giuntini, R.; Leporini, R.; Sergioli, G. A many-valued approach to quantum computational
logics. Fuzzy Sets Syst. 2018, 335, 94–111. [CrossRef]
6. Jozsa, R. Fidelity for mixed quantum states. J. Mod. Opt. 1994, 41, 2315–2323. [CrossRef]
7. Uhlmann, A. The “transition probability” in the state space of a ∗-algebra. Rep. Math. Phys. 1976, 9, 273–279.
[CrossRef]
8. Beltrametti, E.; Cassinelli, G. The Logic of Quantum Mechanics; Encyclopedia of Mathematics and its
Applications 15; Addison-Wesley: Cambridge, MA, USA, 1981.
9. Birkhoff, G.; von Neumann, J. The logic of quantum mechanics. Ann. Math. 1936, 37, 823–843. [CrossRef]
10. Dalla Chiara, M.L.; Giuntini, R.; Greechie, R. Reasoning in Quantum Theory: Sharp and Unsharp Quantum Logics;
Kluwer: Dordrecht, The Netherlands, 2004.

51

www.pdfgrip.com
Entropy 2018, 20, 837

11. Dalla Chiara, M.L.; Giuntini, R.; Leporini, R. Logics from Quantum Computation. Int. J. Quantum Inf. 2005, 3,
293–337. [CrossRef]
12. Dalla Chiara, M.L.; Giuntini, R.; Ledda, A.; Leporini, R.; Sergioli, G. Entanglement as a Semantic Resource.
Found. Phys. 2011, 40, 1494–1518. [CrossRef]
13. Dalla Chiara, M.L.; Giuntini, R.; Leporini, R.; Sergioli, G. Holistic Logical Arguments in Quantum
Computation. Math. Slovaca 2016, 66, 313–334.
14. Dalla Chiara, M.L.; Giuntini, R.; Leporini, R.; Sergioli, G. A First-order Epistemic Quantum Computational
Semantics with Relativistic-like Epistemic Effects. Fuzzy Sets Syst. 2016, 298, 69–90. [CrossRef]
15. Beltrametti, E.; Dalla Chiara, M.L.; Giuntini, R.; Leporini, R.; Sergioli, G. Epistemic Quantum Computational
Structures in a Hilbert space Environment. Fundam. Inf. 2012, 115, 1–14.

c 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

52

www.pdfgrip.com
 entropy
Article
Enhancing Extractable Quantum Entropy in
Vacuum-Based Quantum Random Number Generator
Xiaomin Guo 1,2 , Ripeng Liu 1,2 , Pu Li 1,2 , Chen Cheng 1,2 , Mingchuan Wu 1,2 and
Yanqiang Guo 1,2, *
1 Key Laboratory of Advanced Transducers and Intelligent Control System, Ministry of Education,
Taiyuan 030024, China; guoxiaomin@tyut.edu.cn (X.G.); liuripeng0944@163.com (R.L.);
lipu8603@126.com (P.L.); chengchen248@163.com (C.C.); wumingchuan1@163.com (M.W.)
2 College of Physics and Optoelectronics, Taiyuan University of Technology, Taiyuan 030024, China
* Correspondence: guoyanqiang@tyut.edu.cn; Tel.: +86-351-6018-249

Received: 9 July 2018; Accepted: 22 October 2018; Published: 24 October 2018

Abstract: Information-theoretically provable unique true random numbers, which cannot be

correlated or controlled by an attacker, can be generated based on quantum measurement of vacuum
state and universal-hashing randomness extraction. Quantum entropy in the measurements decides
the quality and security of the random number generator (RNG). At the same time, it directly
determines the extraction ratio of true randomness from the raw data, in other words, it obviously
affects quantum random bits generating rate. In this work, we commit to enhancing quantum entropy
content in the vacuum noise based quantum RNG. We have taken into account main factors in
this proposal to establish the theoretical model of quantum entropy content, including the effects
of classical noise, the optimum dynamical analog-digital convertor (ADC) range, the local gain
and the electronic gain of the homodyne system. We demonstrate that by amplifying the vacuum
quantum noise, abundant quantum entropy is extractable in the step of post-processing even classical
noise excursion, which may be deliberately induced by an eavesdropper, is large. Based on the
discussion and the fact that the bandwidth of quantum vacuum noise is infinite, we propose large
dynamical range and moderate TIA gain to pursue higher local oscillator (LO) amplification of
vacuum quadrature and broader detection bandwidth in homodyne system. High true randomness
extraction ratio together with high sampling rate is attainable. Experimentally, an extraction ratio
of true randomness of 85.3% is achieved by finite enhancement of the laser power of the LO when
classical noise excursions of the raw data is obvious.

Keywords: quantum random number; vacuum state; maximization of quantum conditional

min-entropy

1. Introduction
Randomness is one vital ingredient in modern information science, in the regime of both classical
and quantum [1,2], since encryption is founded upon the trust in random numbers [3–5]. The demand
for true and unique randomness in these applications has triggered various proposals for producing
random numbers based on the measurements of quantum observables, which offer the verifiability
and ultimate in randomness. In the past two decades, there has been tremendous development for
various types of quantum RNG [6–15]. Among these proposals, random number generation based
on homodyne measurement of quantum vacuum state is especially appealing in practice since highly
efficient photodiodes working at room temperature can be applied [11]. Vacuum state is a pure
quantum state with the lowest energy and independent of any external physical quantities. It cannot be
correlated or controlled by an attacker, therefore unique random numbers can be yielded by measuring
the quadrature amplitude of the vacuum state [16,17]. All the components in this scheme, including

Entropy 2018, 20, 819; doi:10.3390/e20110819 53 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2018, 20, 819

laser source, beam splitter and photo detectors have been integrated on a single chip recently [18].
Meanwhile, bit conversion and post-processing are easy to be implemented in virtual “hardware”
inside the field-programmable gate array (FPGA). Chip-size integration of the QRNG is expectable.
Several dedicated researches have been developed to enhance the generation rate of random bits in this
proposal, such as schemes based on optimization of the digitization algorithm [19], implementation
of fast randomness extraction in the post-processing [20], application of squeezing vacuum state to
increase entropy in raw data [21] and optimization of ADC parameters to improve the quantum
entropy in the raw data [22]. In this paper, considering the effects of the classical noise, we discuss the
role of homodyne gain in enhancing quantum entropy in the vacuum-based quantum RNG working
in the optimum dynamical ADC range scenario. Conditional min-entropy is applied to critically assess
the quantum entropy in the quantum RNG. It is the key input parameter of randomness extractor and
determines the extraction ratio of true randomness from the raw random sequence, thereby affects the
generation speed of quantum RNG significantly.

2. Quantum Entropy Evaluation and Enhancing in Vacuum-Based Quantum RNG

Entropy is defined relative to one’s knowledge of an experiment’s output prior to observation.
The larger the amount of the entropy, the greater the uncertainty in predicting the value of an
observation. Among types of entropy, min-entropy is a very conservative measure. In cryptography,
the unpredictability of secret values is essential. The min-entropy measure the probability that a secret
is guessed correctly in the first trial. For mathematically determining min-entropy of a secret, the first
thing is to precisely identify the distribution that the secret was generated from [23].
Quadrature fluctuation of optical quantum vacuum state, the nature initial state of optical field at
room temperature, is the noise source for the random bits generation in this scheme. According to
Born’ s rule, the measurement outcome of a pure quantum state can be intrinsically random. A single
measurement of the quadrature of the vacuum state is completely random and multiple repeated
measurements satisfy the Gaussian distribution statistically, so we can extract random bits from the
measurement results. Based on homodyne measurement, the microscopic fluctuations of quadrature
of the vacuum state are detected, amplified and transferred into an electric signal
$ %
Vvac ∝ i− 2 − i− 2 ∝ 4α2 [δX (t)2 cos (θ )2 + δY (t)2 sin (θ )2 ] (1)

i− is the difference current from the two detectors. Measured quantum quadrature of vacuum state in
any local phase is amplified by the factor α2 = g TI A α L 2 , which includes the amplification effects from
LO gain and electronics gain in the system [24]. Without regard to classical noise, the electric signals
(voltage or current) obey a Gaussian distribution:

1 Vvac 2
P(Vvac ) = √ exp(− 2 ). (2)
πα α

The coefficient α has to be calibrated to rescale histogram of the associated marginal distribution
in optical homodyne tomography (OHT) [25]. In this scheme of quantum random numbers generation,
α is associated with the quantum entropy contained in the measured data and it is the critical parameter
for digitization of the measured analogue signal.
When classical noise is taken into account, such as electronic noise and local noise resulted from
imperfect balancing in balanced homodyne detection (BHD), the observed probability distribution of
the electric signal is in the form of a convolution of the scaled vacuum state marginal distribution and
the classical noise histogram

1 V
Pobs (V ) = P( ) P (V − V )dV . (3)
α α cl

54

www.pdfgrip.com
Entropy 2018, 20, 819

without loss of generality, the broadband electric noise and the LO noise distribution can be assumed
to be Gaussian:
1 V 2
Pcl (Vcl ) = √ exp(− cl ). (4)
πB B
The vacuum noise and the classical noise as two variables with normal distribution,
are independent with each other, thus their sum is also normally distributed with a total variance
equal to the sum of the two variances.
According to Equations (2)–(4), the homodyne measurement of the vacuum state yields a signal
distribution as follows
1 V2
Pobs (V ) = √ √ exp(− 2 ), (5)
π α +B
2 α +B
with the measurement variance of

2
σobs = σquan
2
+ σcl2 = (α2 + B /2, (6)

where factor 2 is added to renormalize the distribution. Then the quantum and classical noise ratio
(QCNR) in the homodyne measurement system is defined as

QCNR = 10Lg(σquan
2
/σcl2 ). (7)

The QCNR related to the signal-to-noise ratio of homodyne detection, is defined as the ratio
between the mean square noise of the measured vacuum state and the electronic noise, that is,
the quantity
S = (α2 + B)/B = σobs2
/σcl2 , (8)

or the clearance between the shot noise power spectrum and electronic noise power in dB units,
10Log10 (S) dB, reads on spectrum analyzer. In other words, when the homodyne detection system
works in linear region, the QCNR of the raw data can be indicated from the clearance shown by
spectrum analyzer.
In our proposal, as a continuous-variable, the measurement output consisting of scaled quadrature
of the vacuum state and the classical noise is discretized by an n-bit ADC with a dynamical range
[− R + δ/2, R − 3δ/2] . The sampled signals are binned over 2n bins with width of δ = R/2n−1 and are
assigned a corresponding bit combination with length of n. 3-bit ADC binning is shown in Figure 1a
as an example.

Figure 1. (a) Model of 3-bit analog-digital converter (ADC); (b) Numerical simulations of acquisition
conditions for a Gaussian signal when dynamical ADC range is chosen too small; (c) too big.

In order to design an entropy source that provides an adequate amount of entropy per output bit
string, the developer must be able to accurately estimate the amount of entropy that can be provided
by sampling its noise source. The behavior of the other components included in the entropy must also
be known clearly since the behavior of the other components may affect the assessment of the entropy.
In our system, the randomness or the entropy in the measurements could derive from multiple factors,

55

www.pdfgrip.com
Entropy 2018, 20, 819

such as the quantum fluctuation, classical influences on it and even malicious attack from the third
part [5]. Especially and strictly, quantum conditional min-entropy is used to evaluate the maximal
amount of randomness extractable from the total entropy of the system [26]. Firstly, the min-entropy
for the Gaussian distribution is defined as

Hmin ( X ) = −Log2 (maxV ∈{0,1}n Prob[ X = V ]). (9)

In this scheme, the min-entropy of the probability distribution of quadrature measurements can
be accurately predicted from the probability density function of the quantum signal. The maximum
probability in (9) can be acquired based on the probability distribution discretized by the bins
⎧  − R+δ/2
⎪
⎨ −∞ Pobs (V )dV, i = i L ,
Vi +δ/2
Pbin (Vi ) = − Pobs (V )dV, i L < i < iM , (10)
⎪  +i ∞
⎩
V δ/2

R−3δ/2 Pobs (V ) dV, i = iM .

Each bin is labelled by an integer i ∈ {iL , ..., i M }, with iL = −2n−1 , the least significant bits (LSB)
bin, iM = 2n−1 − 1, the most significant bit (MSB) bin and Vi = i × δ.
Secondly, some restrictions must be taken into account in analog-digital conversion process.
Those samples go off-scale, that is, points in saturation will be recorded as extrema values as depicted
in Figure 1b. So, underestimating the range will induce too many blocks of zeros and ones. Conversely,
overestimating the signal range will lead to undue unused bins (Figure 1c). In either situation, some bit
combinations are too frequent to be considered random. It is necessary to adjust the amplitude of the
analogue signal and the ADC dynamical range in order to employ the full n-bit sampling properly
whenever possible.
Further, considering the influence of classical noise on the measurement outcome, ADC dynamical
range should be optimized over the classical noise shifted quantum signal probability distribution.
In application scenario, inevitable classical noise excursion in the measurement system will result in
nonzero mean in the measured signal probability distribution. On the other hand, eavesdropper may
induce a deliberate offset over the sampling period. In a word, a noticeable classical noise excursion,
Δ, need to be considered in the optimization of the sampling dynamical range.
Taking into account all these factors offered above, we rewrite the discretized probability
distribution as, ⎧  − R+δ/2−Δ
⎪
⎨ −∞ P(Vi |Vcl )dV, i = i L ,
Vi +δ/2−Δ
Pbin (Vi |Vcl ) = V −δ/2−Δ (Vi |Vcl ) dV, i L < i < iM ,
P (11)
⎪
⎩  +i ∞
R−3δ/2−Δ P (Vi |Vcl ) dV, i = iM .

where,
1 (V − Vcl )2
P(Vi |Vcl ) = √ exp(− ) (12)
πα α2
is the probability density distribution of the quantum signal given full knowledge of the classical noise
Vcl , where Vcl ∈ [Vcl,min , Vcl,max ] with an excursion of Δ. Finally, the quantum conditional min-entropy
is expressed as
&  & '
−2(Vcl,min + R+Δ)+δ
Hmin (V |Vcl ) = −Log2 Max( 12 1 + Erf ,
 & '2α
(13)
δ 2(Vcl,max − R+Δ)+3δ
Erf( 2α ), 2 1 + Erf
1
2α )].

In the best-case scenario of ADC sampling range, the measurement outcome probability in the
center bin is equal to the higher one of the first and the last bins. In this way, the quantum conditional
min-entropy is information theoretically provably estimated and the amount of quantum-based
randomness in the total noise signal is rigorously evaluates. In applications with the requirement of

56

www.pdfgrip.com
Entropy 2018, 20, 819

information security, a random sequence is demanded to be truly unpredictable and have maximum
entropy [27].
At the same time, the conditional min-entropy sets the lower bound of extractable randomness
from the raw measurements and quantifies the least amount of randomness possessed by each sample
or P = Hmin ( X )/n bit per raw bit. Quantum randomness can be distilled from raw data by applying
information theoretically provable Toeplitz-hash extractor. As discussed above, the key point is to
find out the QCNR and derive the probability distribution of the quantum signal. The higher the
QCNR, the more true randomness can be extracted from the raw measurement. Only when QCNR
is high enough, both the quality and the security of the random number generator are guaranteed.
Fulfilling the condition of optimal dynamical sampling range R, minimum-entropy of the quantum
signal for growing clearance is theoretically analyzed. Proceeding from the directly measurable
quantity, homodyne clearance, corresponding QCNR is derived from Equation (8). Then quantum
noise variances are expressed as multiples of the σcl . For different clearance, probabilities of middle bin
and the LSB/MSB are compared and the optimal sampling range R is decided based on Equation (11).
Finally, based on Equation (13), the quantum conditional min-entropy in optimal sampling range
scenario as a function of different classical noise excursion is analyzed.
The classical noise excursions in our raw data have been collected from multiple measurements,
which range from almost 3 to 29 times of classical noise standard deviation σcl . In application scenario,
much larger DC offset may be induced deliberately by the eavesdropper. In Figure 2, we show the
quantum conditional min-entropy, Hmin (V |Vcl ), as a function of homodyne detection clearance for
three different classical noise excursions under the precondition of optimal sampling range. Δ = 3σcl is
the smallest classical noise excursion among our multiple measurements, Δ = 40σcl , a larger classical
noise excursion for comparison and Δ = 17.2σcl is the excursion in the raw data from which we extract
true random numbers. As shown in Figure 2, the extractable random bits are robust against the decline
of QCNR while the classical excursion is subtle. Whereas if classical noise excursion is evidence,
one can achieve high secure randomness only when clearance is high enough.

Figure 2. Optimized Hmin (V |Vcl ) as a function of homodyne detection clearance among different
classical excursions. The theoretical value circled in red corresponding to the highest extraction ratio of
true randomness in our experiment.

The clearance relies on the total gain in homodyne detection system (also α in Equation (1)),
including the LO amplification and the electrical gain. In quantum state measurements and
reconstructions, the clearance needed between shot noise and classical noise is dependent on the
amount of squeezing and entanglement one wishes to measure. Empirically, the homodyne system

57

www.pdfgrip.com
Entropy 2018, 20, 819

should satisfy the condition that the measured shot noise is 10 dB higher than the classical noise among
the analysis frequency range [28,29]. High TIA gain and moderate dynamical range are required so
that shot noise is the dominant spectral feature among the detection frequency range. In this scheme
of quantum RNG, however, high QCNR, but also large detection bandwidths, are pursued, since the
cut-off frequency of the homodyne detector upper bound the sampling frequency in random numbers
generation process [30].
On the other hand, the classical effects, which blur the distribution and cause classical entropy in
the raw bit sequence, include imperfect balancing of LO, non-unit quantum efficiency and electronic
noise of the detectors [31–34]. The non-unit detector efficiency can almost completely overcome by
using special fabricated diodes and the quantum efficiencies of more than 99% have been reported [35].
The detrimental electronic noise depends on numerous components in the circuit part as expressed by

VEL,noise = R (4KT/RPD + IPD,dk
2 + 4KT/Rr + ITIA,c
2 ) + (VTIA,v /R)2 (14)

One term is from the photodiode (PD) and comprise of thermal noise and dark current noise of PD,
both of which are usually negligible thanks to its big shunt resistance RPD and low dark current
IPD,dk [36]. The other term is from the TIA circuit including thermal noise 4KT/Rr , input noise current
ITIA,c and input noise voltage VTIA,v of the operational amplifier. The electrical gain of TIA amplifies
quantum fluctuations as well as the electronic fluctuations, so the electronic noise included in the
homodyne raw measurements comes mainly from the amplified TIA circuit noise. LO effectively acts
as a noise-less amplifier for the quantum fluctuations of the vacuum state and the electrical noise is
independent of the LO. In fact, the optical fluctuations seen by the detector can be made much larger
than the electronic fluctuations by increasing the laser intensity of LO beam to enhance the QCNR
signally [37].
At the same time, the gain of a typical op-amp is inversely proportional to frequency and
characterized by its gain–bandwidth product (GBWP). As a trade-off, lower electrical gain put up
with higher op-amp bandwidth. In fact, theoretically, vacuum quadrature fluctuates with unlimited
bandwidth in the frequency domain. The random number generation rate in this scheme is ultimately
limited by the bandwidth of the homodyne detector. Increased bandwidth of op-amp allows higher
sampling rate.

3. Experiment and Results

Experimentally, we dedicate to enhance quantum entropy in quantum RNG by enhancing the
laser power of LO beam to noise-independently amplify quadrature fluctuation of vacuum state on
the premise of optimizing ADC sampling range. An extraction ratio of true randomness of 85.3% is
achieved by finite enhancement of the LO power when classical noise excursions of the raw data is
obvious and the extracted random sequences passed the NIST (National Institute of Standards and
Technology), Diehard and the TestU01 tests.
The experimental setup is depicted in Figure 3. A 1550 nm laser diode (LD) is driven by constant
current with thermoelectric temperature control with a maximal out power of 15 mW. A half-wave
plate and a polarizing beamsplitter (PBS2) were combined to serve as accurate 50/50 beamsplitting.
Single-mode continuous-wave laser beam from the laser incident into one port of the beamsplitter
and acts as the LO, while the other port was blocked to ensure that only the vacuum state could
enter in. The vacuum field and the LO interfere on the symmetric beamsplitter to form two output
beams with balanced power. The outputs are simultaneously detected by balanced homodyne detector
(PDB480C, Thorlabs Inc., Newton, MA, USA) to cancel the excess noise in LO while amplify the
quadrature amplitude of the vacuum state, which fluctuates randomly and is independent of any
external physical quantities.

58

www.pdfgrip.com
Entropy 2018, 20, 819

Figure 3. Schematic of the experiment for the quantum random number generator based on homodyne
measurements of the quadrature amplitudes of the vacuum state.

Classical noise in the photocurrents is rejected effectively over the whole detection band while the
clearance has dependence on frequency as shown in Figure 4. We filtered out a part of the vacuum
spectrum, where the clearance is almost consistent, to extract true randomness based on a certain
quantum conditional min-entropy and analyze the effect of LO intensity on the conditional min-entropy.
The shot noise limited signal from the homodyne detector is mixed down with a 200 MHz carrier
(HP8648A) and then passes through a low-pass-filter (LPF) with 50 MHz cut-off frequency (BLP50+,
Mini-Circuits Corp., Brooklyn, NY, USA), that is, we actually use 100 MHz vacuum sideband frequency
spectrum centered at 200 MHz to act as the random noise resource.

Figure 4. Amplified vacuum noise power spectral when local oscillator (LO) power is 6 mW. 100 MHz
vacuum sideband centered at 200 MHz is filtered out as the entropy source of quantum RNG.

In OHT, BHD system is established and locked to every relative phase to measure the marginal
distributions of electromagnetic field quadrature for completely reconstruction of quantum states [25].
While the random numbers generation scheme discussed here focus on a marginal distribution of
vacuum state in any one phase thanks to the space rotational invariance of its distributions in the
phase space, that is no active modulation or phase (or polarization) stabilization is required.

59

www.pdfgrip.com
Entropy 2018, 20, 819

We present the QCNR as a function of the LO power arriving at the PD. The electrical noise
variance is relatively consistent for certain TIA gain. The clearance depends only on the LO power.
The noise power is given by
4e2 ( P/hν)ηBR2
PdBm = 10lg( ), (15)
Z × 1 mW
where e is the electron charge, η = 0.9 is the quantum efficiency of the photodiode (Hamamatsu
G8376), B = 100 KHz the resolution bandwidth, R = 16 × 103 V/A the transimpedance gain of the
photo detector and Z = 50 Ω the load impedance [38]. For each power value the distribution of the
random data was analyzed in time domain in the form of histogram to calculate the QCNR. QCNR as a
function of the LO power figured out from the measured clearance levels is plotted with open circles in
Figure 5. The LO power received by each PD is gradually increased from 300 μW to 6 mW by rotating
the HWP before PBS1. Here we interpolate between the experimental points to obtain the dependence
of QCNR on LO power. It is shown as the black dashed line in Figure 5. The experimental results are
given by red open circles and can be fitted well by the theoretical curve with a transimpedance gain of
13.1 × 103 V/A. The experimental results are about 2 dB lower than the theoretically excepted QCNR,
which is due to uncertainties in determining the transimpedance of the detector and the transmission
losses in the LPF.

Figure 5. QCNR as a function of the LO power. Inset: Resulting histograms of the vacuum (red) and
electronic (black) noise obtained at a LO power of 6 mW.

We increase the LO power up to 6 mW to achieve the largest QCNR of 17.8 dB in our system,
limited by the maximal output power of the laser. The signal is sampled with a rate of 100 MHz,
upper limit of twice the LPF band for the sampling rate to avoid temporal correlation between samples.
The resolution is 12 bits and the dynamical range is optimized according to the histogram of the time
series acquired with reasonably larger sampling range. The amplitude acquisition scale of oscilloscope
(SDA806Zi-A, LeCroy, New York, NY, USA) is continuously adjustable. By choosing the analog-digital
conversion range appropriately and tuning the LO intensity finely, the amount of off-scale points can
be controlled within allowed statistical deviation. The number of saturated points is easy to restrain
on-line from the oscilloscope. The distributions of the random data in time domain and in histogram
are shown as insets of Figure 5. The measured total variance of the raw data and electrical noise
variance are 154.43 mV2 and 5.89 mV2 , respectively. The classical noise excursions of the raw data are
about 17.2 times of the classical noise standard deviation σcl . Then the probability distribution of the
quantum signal is derived and the conditional min-entropy in the quantum signal is worked out to be
10.13 bit per sample, as circled in red in Figure 2.

60

www.pdfgrip.com
Entropy 2018, 20, 819

Finally, information-theoretically provable post-processing scheme, Toeplitz-hashing extractor,

is constructed on an FPGA to extract true randomness from the raw data and uniform the Gaussian
biased binary stream [39]. A binary Toeplitz matrix of m × n is constructed with a seed of m + n − 1
random bits (the seed can be reused since the Toeplitz-hashing extractor is a strong extractor).
m final random bits are extracted by multiplying the matrix and n raw bits, where m/n ≤ P and
P = Hmin ( X )/n. We employ 4096 × 3520 Toeplitz Hash extractor to distil 10.13 bits/sample.
The extraction ratio of 85.3% is the highest as ever reported. We recorded the data with the size
of 1 G bits to undergo random test. 1000 sequences with each one 1 M bits are applied to the NIST test
and significant level is set as α = 0.01. The NIST test is successful(if final P-values of all sequences
are larger than α with a proportion within the range of (1 − α) ± 3 (1 − α)α/n = 0.99 ± 0.00944 for
15 test suits [40]. P-value shown in the Figure 6 are the worst cases of our test outcomes.

Figure 6. Results of the NIST statistical test suite for a 109 -bit sequence.

Results of the Diehard statistical test suite for the same data file is shown in Figure 7.
Kolmogorov-Smirnov (KS) test is used to obtain a final p-value to measure the uniformity of the
multiple p-values. The test is considered successful if all the final p-values lies in the range from 0.01 to
0.99 [41].

Figure 7. Results of the Diehard statistical test suite for a 109 -bit sequence.

61

www.pdfgrip.com
Entropy 2018, 20, 819

Constrained by the computational power of crush of TestU01, small crush test is performed with a
data size of 8 G bits [42]. The random numbers can pass all the statistical tests successfully. The p-value
from a failing test converges to 0 or 1. Where the test has multiple p-values, the worst case is tabled in
Figure 8. All the test items are passed successfully.

Figure 8. Results of the TestU01 statistical test suite for a 5 × 109 -bit sequence.

On the other hand, we reduce the LO power in the homodyne system to 400 μW and
correspondingly, the clearance declines to 4.06 dB. The time series of the system outcomes are collected
and statistically analyzed. Classical noise excursion in the Gaussian distribution is about 19.3 times of
the classical noise standard deviation. Based on theoretical calculation, the min-entropy is worked out
to be 7.73 bits/sample. The hash extraction results with maximum extraction ratio of 0.63 can pass the
NIST, Diehard and TestU01 tests finally.

4. Conclusions
To summary, in this work, we discussed the role of LO power plays in random number generation
based on quantum detection of vacuum state. When classical noise excursion in the system is
trivial, LO power in the homodyne system affect the quantum entropy in raw data insignificantly.
Nevertheless, in realistic scenario, the mean of the measured signal distribution is normally nonzero,
even much larger noise excursion may be induced deliberately by the eavesdropper. In this case,
enough real randomness is attainable only when QCNR is high enough. With the LO power enhanced,
the vacuum quadrature fluctuations are amplified independent of the electrical noise and the quantum
entropy content in the raw data is enhanced effectively. Thus, we propose large dynamical range and
moderate TIA gain to pursue higher LO amplification of vacuum quadrature and larger detection
bandwidth in homodyne system for higher sampling rate in random numbers generation. Higher hash
extraction ratio along with higher sampling rate will enhance the real random number generation rate
effectively. More importantly, the quantum RNG system is more robust against to the third part attack.

5. Discussion
The central mathematical concept in true RNG is entropy, which is the assessment standard of the
security and quality of a RNG. There are many types of entropy. In recent years, min-entropy, a very
conservative evaluation, is applied to lower bound the entropy content in quantum RNG and as the
indicator for extraction ratio of universal hash extractor. In our work and some ever works [15,19],
quantum conditional min-entropy are deduced to impose stricter removal of side signal. Min-entropy

62

www.pdfgrip.com
Entropy 2018, 20, 819

is estimated by using the most common value estimate. However, the most common value estimate is
more appropriate for IID (independent identically distribution). For non-IID distribution, the estimate
may provide an overestimation. The NIST Special Publication 800-90 series of Recommendations
provides guidance on the construction and validation of random bit generators (RBGs) in the form of
deterministic random bit generators, in which pseudorandom bits are generated by using an unknown
seed, or in the form of non-deterministic random bit generators that can be used for cryptographic
applications. Entropy source validation is necessary in order to obtain assurance that all relevant
requirements of this Recommendation are met.
As discussed above, the raw noise-source output in our proposal is biased, Toeplitz hash extractor
(conditioning component) is used in the design to reduce that bias to an appropriately level before the
RNG exports any bits. For non-IID data, a list of estimators is proposed and the minimum of all the
estimates is taken as the entropy assessment of the entropy source for the entropy source validation
for the Recommendation. We apply our raw bit strings to the test suit on line [43]. The Test result is
shown in Figure 9. Because the size of the sample space in our work is 212 , we take the lower 8 bits
to meet the applicability of the test. The resulting min-entropy is taken from the minimum of all the
estimates as 5.818 per 8 bits. The restart tests are passed. Although the ratio of 72.7% is lower than the
evaluation of quantum conditional min-entropy, the quality of our entropy source is validated.

Figure 9. Entropy estimates NIST 800-90B for a 1.6 × 106 -bit sequence.

Author Contributions: X.G. and Y.G. designed the study and wrote the paper; X.G. carried out calculations and
analyzed the data; R.L., C.C. and M.W. contributed to the experiment; X.G. and P.L. performed the random number
tests. All authors discussed the results at all stages. All authors have read and approved the final manuscript.
Funding: This research was funded by the National Natural Science Foundation of China (NSFC) (Grants Nos.
61505136, 61875147, 61775158, 61671316), Shanxi Scholarship Council of China (SXSCC) (Grants No. 2017-040)
and Natural Science Foundation of Shanxi Province (Grants No. 201701D221116).
Acknowledgments: We would like to thank Tiancai Zhang for fruitful discussions.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Korzh, B.; Lim, C.C.W.; Houlmann, R.; Gisin, N.; Li, M.J.; Nolan, D. Provably secure and practical quantum
key distribution over 307 km of optical fibre. Nat. Photonics 2014, 9, 163–168. [CrossRef]
2. Ferguson, N.; Schneier, B.; Kohno, T. Cryptography Engineering: Design Principles and Practical Applications;
John Wiley & Sons: Hoboken, NJ, USA, 2010.
3. Stefanov, A.; Gisin, N.; Guinnard, O.; Guinnard, L.; Zbinden, H. Optical quantum random number generator.
J. Mod. Opt. 2000, 47, 595–598. [CrossRef]

63

www.pdfgrip.com
Entropy 2018, 20, 819

4. Gisin, N.; Ribordy, G.; Tittel, W.; Zbinden, H. Quantum cryptography. Rev. Mod. Phys. 2002, 74, 145–195.
[CrossRef]
5. Toffoli, T. Entropy? honest! Entropy 2016, 18, 247. [CrossRef]
6. Rarity, J.; Owens, P.; Tapster, P. Quantum random-number generation and key sharing. J. Mod. Opt. 1994, 41,
2435–2444. [CrossRef]
7. Guo, H.; Tang, W.Z.; Liu, Y.; Wei, W. Truly random number generation based on measurement of phase noise
of a laser. Phys. Rev. E 2010, 81, 051137. [CrossRef] [PubMed]
8. Ma, H.Q.; Xie, Y.; Wu, L.A. Random number generation based on the time of arrival of single photons.
Appl. Opt. 2005, 44, 7760–7763. [CrossRef] [PubMed]
9. Yan, Q.R.; Zhao, B.S.; Liao, Q.H.; Zhou, N.R. Multi-bit quantum random number generation by measuring
positions of arrival photons. Rev. Sci. Instrum. 2014, 85, 615–621. [CrossRef] [PubMed]
10. Ren, M.; Wu, E.; Liang, Y.; Jian, Y.; Wu, G.; Zeng, H.P. Quantum random-number generator based on a
photon-number-resolving detector. Phys. Rev. A 2011, 83, 1293–1304. [CrossRef]
11. Gabriel, C.; Wittmann, C.; Sych, D.; Dong, R.F.; Mauerer, W.; Andersen, U.L. A generator for unique quantum
random numbers based on vacuum states. Nat. Photonics 2010, 4, 711–715. [CrossRef]
12. Qi, B.; Chi, Y.M.; Lo, H.-K.; Qian, L. High-speed quantum random number generation by measuring phase
noise of a single-mode laser. Opt. Lett. 2010, 35, 312–314. [CrossRef] [PubMed]
13. Xu, F.H.; Qi, B.; Ma, X.F.; Xu, H.; Zheng, H.X.; Lo, H.K. Ultrafast quantum random number generation based
on quantum phase fluctuations. Opt. Express 2012, 20, 12366. [CrossRef] [PubMed]
14. Marangon, D.G.; Vallone, G.; Villoresi, P. Source-device-independent ultrafast quantum random number
generation. Phys. Rev. Lett. 2017, 118, 060503. [CrossRef] [PubMed]
15. Cao, Z.; Zhou, H.; Ma, X.F. Loss-tolerant measurement-device-independent quantum random number
generation. New J. Phys. 2015, 17, 125011. [CrossRef]
16. Sych, D.; Leuchs, G. Quantum uniqueness. Found. Phys. 2015, 45, 1613–1619. [CrossRef]
17. Fiorentino, M.; Santori, C.; Spillane, S.M.; Beausoleil, R.G.; Munro, W.J. Secure self-calibrating quantum
random-bit generator. Phys. Rev. A 2006, 75, 723–727. [CrossRef]
18. Abellan, C.; Amaya, W.; Domenech, D.; Muñoz, P.; Capmany, J.; Longhi, S. Quantum entropy source on an
InP photonic integrated circuit for random number generation. Optica 2016, 3, 989–994. [CrossRef]
19. Symul, T.; Assad, S. M.; Lam, P.K. Real time demonstration of high bitrate quantum random number
generation with coherent laser light. Appl. Phys. Lett. 2011, 98, 231103. [CrossRef]
20. Shi, Y.C.; Chng, B.; Kurtsiefer, C. Random numbers from vacuum fluctuations. Appl. Phys. Lett. 2016, 109,
041101. [CrossRef]
21. Zhu, Y.Y.; He, G.Q.; Zeng, G.H. Unbiased quantum random number generation based on squeezed vacuum
state. Int. J. Quantum Inf. 2012, 10, 1250012. [CrossRef]
22. Haw, J.Y.; Assad, S.M.; Lance, A.M.; Ng, N.H. Y.; Sharma, V.; Lam, P.K. Maximization of extractable
randomness in a quantum random-number generator. Phys. Rev. Appl. 2015, 3, 054004. [CrossRef]
23. Turan, M.S.; Barker, E.; Kelsey, J.; McKay, K.A.; Baish, M.L.; Boyle, M. NIST Draft Special Publication
800-90 B: Recommenda-tion for the Entropy Sources Used for Random Bit Generation. Available
online: https://csrc.nist.gov/csrc/media/publications/sp/800-90b/draft/documents/sp800-90b_second_
draft.pdf (accessed on January 2018).
24. Kumar, R.; Barrios, E.; MacRae, A.; Gairns, E.; Huntington, E.H.; Lvovsky, A.I. Versatile wideband balanced
detector for quantum optical homodyne tomography. Opt. Commun. 2012, 285, 5259–5267. [CrossRef]
25. Lvovsky, A.I.; Raymer, M.G. Continuous-variable optical quantum state tomography. Rev. Mod. Phys. 2005,
81, 299–332. [CrossRef]
26. Konig, R.; Renner, R.; Schaffner, C. The operational meaning of min- and max-entropy. IEEE Trans.
Inform. Theory 2009, 55, 4337–4347. [CrossRef]
27. Stipčević, M. Quantum random number generators and their applications in cryptography. Adv. Photon
Count. Tech. 2012, 837504.
28. Vahlbruch, H.; Mehmet, M.; Chelkowski, S.; Hage, B.; Franzen, A.; Lastzka, N. Observation of squeezed light
with 10-db quantum-noise reduction. Phys. Rev. Lett. 2008, 100, 033602. [CrossRef] [PubMed]
29. Olivares, S.; Paris, M.G.A. Bayesian estimation in homodyne interferometry. J. Phys. B At. Mol. Opt. Phys.
2012, 42, 55506–55512. [CrossRef]

64

www.pdfgrip.com
Entropy 2018, 20, 819

30. Shen, Y.; Tian, L.; Zou, H.X. Practical quantum random number generator based on measuring the shot noise
of vacuum states. Phys. Rev. A 2010, 81, 063814. [CrossRef]
31. Mcclelland, D.E.; Mckenzie, K.; Gray, M.B.; Ping, K.L. Technical limitations to homodyne detection at audio
frequencies. Appl. Opt. 2007, 46, 3389–3395.
32. Gramdi, S.; Zavatta, A.; Bellini, M.; Paris, M.G.A. Experimental quantum tomography of a homodyne
detector. New J. Phys. 2017, 19, 053051.
33. Combes, J.; Wiseman, H. Quantum feedback for rapid state preparation in the presence of control
imperfections. J. Phys. B At. Mol. Opt. Phys. 2011, 44, 154008. [CrossRef]
34. Chrzanowski, H.M.; Assad, S.M.; Bernu, J.; Hage, B.; Lund, A.P.; Ralph, T.C. Reconstruction of photon
number conditioned states using phase randomized homodyne measurements. J. Phys. B At. Mol. Opt. Phys.
2013, 46, 104009. [CrossRef]
35. Oshima, T.; Okuno, T.; Arai, N.; Suzuki, N.; Ohira, S.; Fujita, S. Vertical solar-blind deep-ultraviolet schottky
photodetectors based on beta-Ga2 O3 substrates. Appl. Phys. Express 2008, 1, 011202. [CrossRef]
36. Graeme, J. Photodiode Amplifiers: OP AMP Solutions; McGraw-Hill: New York, NY, USA, 1995.
37. Jin, X.L.; Su, J.; Zheng, Y.H.; Chen, C.; Wang, W.Z.; Peng, K.C. Balanced homodyne detection with high
common mode rejection ratio based on parameter compensation of two arbitrary photodiodes. Opt. Express
2015, 23, 23859. [CrossRef] [PubMed]
38. Gray, M.B.; Shaddock, D.A.; Harb, C.C.; Bachor, H.-A. Photodetector designs for low-noise, broadband and
high-power applications. Rev. Sci. Instrum. 1998, 69, 3755–3762. [CrossRef]
39. Carter, J.L.; Wegman, M.N. Universal classes of hash functions (Extended Abstract). J. Comput. Syst. Sci.
1977, 18, 106–112.
40. Rukhin, A.; Soto, J.; Nechvatal, J.; Miles, S.; Barker, E.; Leigh, S. A Statistical Test Suite for Random
and Pseudorandom Number Generators for Cryptographic Applications; National Institute of Standards and
Technology: Gaithersburg, MD, USA, 2001.
41. Marsaglia, G. DIEHARD Battery of Tests of Randomness. 1995.
42. L’Ecuyer, P.; Simard, R. TestU01: A C library for empirical testing of random number generators. ACM Trans.
Math. Softw. 2007, 33, 22. [CrossRef]
43. Turan, M.S.; Barker, E.; Kelsey, J.; McKay, K.A.; Baish, M.L.; Boyle, M. “The SP800-90B_EntropyAssessment
Python Package”. 2008. Available online: https://github.com/usnistgov/SP800-90B_EntropyAssessment
(accessed on 3 August 2018).

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

65

www.pdfgrip.com
 entropy
Article
Entanglement of Three-Qubit Random Pure States
Marco Enríquez 1 , Francisco Delgado 1, * and Karol Życzkowski 2,3
1 Escuela de Ingeniería y Ciencias, Tecnológico de Monterrey, Atizapán 52926, Mexico; menriquezf@itesm.mx
2 Faculty of Physics, Astronomy and Applied Computer Science, Jagiellonian University, ul. Łojasiewicza 11,
30-348 Kraków, Poland; karol.zyczkowski@uj.edu.pl
3 Center for Theoretical Physics, Polish Academy of Sciences, Al. Lotników 32/46, 02-668 Warsaw, Poland
* Correspondence: fdelgado@itesm.mx; Tel.: +52-55-5864-5670

Received: 29 August 2018; Accepted: 26 September 2018; Published: 29 September 2018

Abstract: We study entanglement properties of generic three-qubit pure states. First, we obtain the
distributions of both the coefficients and the only phase in the five-term decomposition of Acín et al.
for an ensemble of random pure states generated by the Haar measure on U (8). Furthermore,
we analyze the probability distributions of two sets of polynomial invariants. One of these sets allows
us to classify three-qubit pure states into four classes. Entanglement in each class is characterized
using the minimal Rényi-Ingarden-Urbanik entropy. Besides, the fidelity of a three-qubit random state
with the closest state in each entanglement class is investigated. We also present a characterization
of these classes in terms of the corresponding entanglement polytope. The entanglement classes
related to stochastic local operations and classical communication (SLOCC) are analyzed as well from
this geometric perspective. The numerical findings suggest some conjectures relating some of those
invariants with entanglement properties to be ground in future analytical work.

Keywords: quantum entanglement; three-qubit random states; entanglement classes; entanglement

polytope; anisotropic invariants

1. Introduction
Entanglement is possibly the most interesting and complex issue in Quantum Mechanics. Due to
this phenomenon it is not possible to describe properties of individual subsystems, even though the
entire system is known to be in a concrete pure quantum state. Quantification of entanglement is
still a challenge for any quantum system consisting of more than two parts [1,2]. The difficulty of the
problem grows quickly with the growing number of subsystems and it becomes intractable in the
asymptotic limit [3]. Several measures of quantum entanglement were proposed [4], but even in the
case of pure states of a multipartite quantum system, it is not possible to identify the single state which
can be called the most entangled, as the degree of entanglement depends on the measure used [5].
On the other hand, entanglement in bipartite systems is already well understood. In the case
of pure states, a key tool in describing entanglement properties is the Schmidt decomposition as
any entanglement measure is a function of the Schmidt coefficients [2]. Dealing with three-party
pure states, the problem becomes more intricate as the corresponding state is represented by a tensor
rather than a matrix, so one cannot rely on the Schmidt decomposition related to the singular value
decomposition of a matrix. Nevertheless, several decompositions for three-qubit states have been
studied in literature [6–8]. More recently, a canonical form for symmetric three-qubit states has been
proposed, showing that in this case the number of entanglement parameters can be reduced from five
to three [9].
Early studies on correlation in composite quantum systems revealed that for three or more parties
there exist quantum states with different forms of entanglement [8], as the states from one entanglement
class cannot be converted by local operations to any states of the other class. As the number of parties

Entropy 2018, 20, 745; doi:10.3390/e20100745 66 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2018, 20, 745

increases, the number of entanglement classes grows quickly [10]. Since local operations cannot
generate entanglement, one usually assumes that a faithful measure of quantum entanglement should
be invariant under local unitary operations and should not grow under arbitrary local operations.
For a given class of operations there exist invariants which are constant along every orbit of
equivalent states [11,12]. A full set of invariants determines a given orbit of locally equivalent states.
However, such sets of invariants are established only for systems consisting of few parties of a small
dimension including the simplest multipartite case of three-qubit systems [13–15].
An interesting question arises: To what extent single-particle properties can provide information
about the global entanglement [16]? The issue is related to the so-called quantum marginal problem:
Given a set of reduced density matrices one asks whether they might appear as partial trace of a
given state of a composed system [17]. Necessary conditions for such a “compatibility problem”
were provided in [18] for the two-qubit system and then developed by Klyachko [19] for the general
case. These conditions can be expressed as a set of linear inequalities concerning the eigenvalues of
the density matrix corresponding to the entire system and eigenvalues of the reduced matrices.
Interestingly, for multipartite systems the compatibility problem is related to the entanglement
characterization [20]. For instance, eigenvalues of three one-qubit reduced matrices of any three-qubit
pure state belong to the entanglement polytope and some of its parts correspond to certain classes of
quantum entanglement [21].
Not knowing a particular quantum state corresponding to a physical system it is interesting
to ask, what are properties of a typical state? More formally, one defines an ensemble of pure
quantum states induced by the unitary invariant Fubini-Study measure [2] and computes mean
values of various quantities averaging over the unitary group with respect to the Haar measure.
Such random quantum states are physically interesting as they arise during time-evolution of quantum
systems corresponding to classically chaotic systems [22,23] and are relevant for problems of quantum
information processing [24,25].
Research on non-local properties of generic multipartite states has been intensive in recent
years. This includes entanglement in two qudit systems [26–28], pairwise entanglement in
multi-qubit systems [29–31], entropic relations and entanglement [32], correlations and fidelities
in qutrits system [33], a characterization of entanglement through negativities and tangles in
several qubits systems and its relation to the emergence of the bulk geometry [34]. More recently,
genuine entanglement for typical states for a system composed out of three subsystems with d levels
each was studied with help of the geometric measure of entanglement [35], while for generic four-qubit
Alsina analyzed the distribution of the hyperdeterminant [36].
The aim of this work is to extend the analysis of entanglement properties of generic states of
three-qubit systems. We focus our attention on the five-term decomposition of an arbitrary pure
state [15] as it allows one to construct a set of polynomial invariants and to identify the classes of
entanglement. We generated an ensemble of pure quantum states induced by the Haar measure on
the unitary group U (8) corresponding to the system composed of three qubits and investigated the
distribution of various entanglement measures and local invariants.
The paper is organized as follows. In Section 2 we review the five-term decomposition of a
three-qubit stateand study statistical properties of the coefficients in such a representation of a generic
state. In Section 3, we investigate properties of the three qubits invariants, Ik and Jk [15] as well as two
newly discovered anisotropic invariants [37]. We obtain their probability distributions, either exact or
approximate, and compare them with accurate numerical approximations. The fourth section presents
an analysis for the entanglement classes defined in terms of the latter invariants. As a comparative
element, we use the Rényi and the minimal Rényi-Ingarden-Urbanik (RIU) entropies [35] to analyze
possible meanings for such classes. Another measure, the maximum overlap with respect to a selected
entanglement class, allows us to identify for an arbitrary three-qubit state the closest state in each
class resembling it. In Section 5 we discuss a characterization of quantum entanglement through the
corresponding entanglement polytope and we show how entanglement classes can be distinguished

67

www.pdfgrip.com
Entropy 2018, 20, 745

from a geometrical viewpoint. The last section presents concluding remarks, a list of open questions
with suggestions concerning the future work.

2. The Canonical Five-Term Decomposition

A three-qubit state in the Hilbert space H ⊗3 involves eight terms, thus, it can be written as

|ψ = tijk |ijk, tijk tijk = 1, tijk ∈ C, (1)

where we have used the repeated scripts notation. It is known [15] that through local unitaries,
the number of terms in |ψ can be reduced from eight to five. First, we define the two square matrices
T0 and T1 whose entries are given by ( Ti ) jk = tijk , with i, j, k = 0, 1. A local unitary transformation
U ⊗ 12 ⊗ 13 acting on the first qubit produces

T0 = u00 T0 + u01 T1 , T1 = −u01 T0 + u00 T1 . (2)

The matrix U is taken such that det( T0 ) = 0. On the other hand, the transformation 12 ⊗ V ⊗ W
changes the matrices Ti according to VTi W. We choose V and W so that T0 can be diagonalized via the
singular value decomposition (SVD). Explicitly, at the end of this procedure we arrive at
   
λ0 0 λ1 )
) λ2
T0 = , T1 = ) . (3)
0 0 λ3 ) λ4

In addition, the phase of the coefficients )

λ2 , )
λ3 and )
λ4 can be absorbed into )
λ1 to yield
the decomposition

|ψ = λ0 |000 + λ1 eiφ |100 + λ2 |101 + λ3 |110 + λ4 |111, (4)

where λi , φ ∈ R. Besides ∑ λ2i = 1. According to [15], the only phase φ should be restricted to
0 < φ < π to assure the uniqueness of the decomposition.

Distribution of the Coefficients

We take an ensemble of 106 random states in H ⊗3 distributed according to the unitary invariant
measure on the group U (8) and then first reduce them into the five-term representation (4), then we
track each coefficient λk to compute numerically its probability distributions as well as the distribution
of the phase φ. The result is shown in Figure 1 depicting the value of each component λk versus
their relative normalized density on H ⊗3 . Note that the state (1) depends on 14 real parameters,
say p = ( p1 , . . . , p14 ) where each pμ is the real or imaginary part of tijk . The unitary invariance implies
that after the action of the transformation U ⊗ V ⊗ W on the state |ψ the distribution of the coefficients
λi ’s and the phase φ fulfils P (p) = J × P (λ), where λ = (λ0 , λ1 , λ2 , λ3 , λ4 , φ) and J is the Jacobian of
the transformation. The evaluation of this 14 × 14 determinant becomes cumbersome and one has
to rely on numerical methods to compute the marginal distributions P(λk ) of the coefficients of the
state (4) as well as the phase φ. The data presented in Figure 1b suggest that the phase φ is distributed
uniformly on the entire range, P(φ) = 1/π for φ ∈ [0, π ]. As the beta distribution has been used to
model the behavior of random variables limited to finite length intervals in several contexts [24,35,38],
we propose the following distribution Pi (λi ) = c λia (1 − λi )b , to fit the distributions of the coefficients
λ j . The numerical fits are depicted as solid lines in Figure 1a and the values of the best fitting
parameters are reported in Table 1. Results presented suggest that the coefficients λ1 , λ2 and λ3 are
distributed according to the same probability distribution. Hence, we conjecture that out of the six
real parameters in Equation (4), only four are required to characterize entanglement in three-qubit
random states, say {λ0 , λ1 , λ4 , φ}. Interestingly, the coefficients λ0 and λ4 are related with the invariant
J4 connected with the three-qubit genuine entanglement (for the definition see subsequent section).

68

www.pdfgrip.com
Entropy 2018, 20, 745

As generic three-qubit states are typically strongly entangled [35], this analysis illustrates how each
coefficient λ j of a given state is linked with the degree of its entanglement. Note particularly how low
values of λ1 , λ2 and λ3 are more representative for entangled states in contrast to λ4 , the distribution
of which appears to be balanced. Furthermore, the higher values of the coefficient λ0 correspond to the
states with larger entanglement. This is particularly interesting as in the decomposition of Carteret et al.
this coefficient yields the maximum overlap with the closest separable state [7].

(a) (b)

Figure 1. Probability distribution of the Acín parameters in the state (4): (a) the coefficients λk , k =
0, 1, ..., 4 and (b) the phase φ for a set of 106 three-qubit random states on H2⊗3 . Solid lines represent the
best numerical fit in all the cases, the parameters of which are listed in Table 1.

Table 1. Best numerical fit parameters of the distributions Pi (λi ) = c λia (1 − λi )b for i = 0, 1, 2, 3, 4.

i a b c
0 3.74 6.05 1856.85
1 67.76 4.25 1.52
2 68.40 4.27 1.53
3 66.75 4.24 1.52
4 795.16 4.37 3.96

3. Three-Qubits Polynomial Invariants

Local unitary (LU) transformations performed on individual subsystems define orbits of locally
equivalent multipartite states. Local invariants can be understood as coordinates in the space of orbits
of locally equivalent states. Any complete set of local invariants allows one to distinguish between
different orbits of locally equivalent states and thus to describe the degree of quantum entanglement [7].
For pure states of a three-qubit system, the space of orbits has six dimensions and it is possible to find
six algebraically independent invariants [39].
In this section we will analyze the distributions P( Ik ) and P( Jk ) on H ⊗3 for the corresponding
three-qubit invariants (under local operations) Ik [13] and Jk [40], with k = 1, ..., 5. These polynomial
invariants set representative classes on H ⊗3 and cannot be directly used as the measures of
genuine entanglement.

Distribution of the Invariants

We first consider the set of five invariants used in [40]

I2 = tr(ρ2A ), I3 = tr(ρ2B ), I4 = tr(ρ2C ),

(5)
I5 = tr[(ρ A ⊗ ρ B )ρ AB ], I6 = |Hdet( T )|2

69

www.pdfgrip.com
Entropy 2018, 20, 745

where ρi stands for the reduced density matrix of the i-th system, ρij is the reduced density matrix
when the partial trace respect the system k is performed while i, j, k is a permutation of A, B, C. The last
invariant is related to the hyperdeterminant Hdet of the tensor coefficients T = (tijk ) representing
the state (1).
The invariants are labeled according to the notation used by Sudbery [13]. Note that the squared
norm of the state (1) is in itself a polynomial invariant usually denoted as I1 . In Figure 2a–c we show the
probability distribution of the above set of invariants over an ensemble of 106 random states. Moreover,
as for k = 2, 3, 4 the quantity Ik is related with the linear entropy, Sk = 1 − Ik , the corresponding
distributions show that the entanglement of each qubit with the other two is the same no matter which
partial trace is performed. On the other hand, the invariants Ik in terms of the coefficients tijk are
written as [13]:

(a) (b)

(c) (d)

Figure 2. (a–c) Probability distribution for the polynomial invariants Ii , i = 1, ..., 5 for a set of 106
three-qubit random states. Solid line in panel (a) stands for the distribution (9), while in panels (b,c)
the best numerical distributions are depicted by green curves. In panel (d) a dispersion plot comparing
I1 , I4 and I5 is shown. In addition, each dot has been colored as function of its S1 Rényi entropy [41]
calculated after the five terms reduction.

I2 = ti1 j1 k1 ti2 j1 k1 ti2 j2 k2 ti1 j2 k2 , I3 = ti1 j1 k1 ti1 j2 k1 ti2 j2 k2 ti2 j1 k2 , I4 = ti1 j1 k1 ti1 j1 k2 ti2 j2 k2 ti2 j2 k1 ,

I5 = ti1 j1 k1 ti1 j2 k2 ti2 j2 k2 ti2 j3 k1 ti3 j3 k3 ti3 j1 k3 (6)

I6 = 4|i1 ,j1 i2 j2 k1 1 k2 2 i3 k3  j3 3 ti1 i2 i3 t j1 j2 j3 tk1 k2 k3 t1 2 3 |2

70

www.pdfgrip.com
Entropy 2018, 20, 745

where the convention of summation over repeated indexes is used and k, stands for the Levi-Civita
tensor of order two. Since the coefficients can be regarded as a column of a random unitary
matrix, we can compute the average value of each invariant by evaluating integrals of polynomial
functions over the unitary group with respect to unique normalized Haar measure. Using symbolic
integration [42] we obtain  Ik  = 2/3 for k = 2, 3, 4. This result is consistent with the mean purity
of a single qubit traced out from a 2 × 4 system reported in [43]. Moreover,  I5  = 7/15 and
 I5 2  = 133/572. In order to compute the mean value of I6 , we use the second moment of the
three-tangle τ reported in [35] with the fact τ 2 = 16I6 to get  I6  = 1/110. On the other hand,
to compute the distributions of the invariants P( Ik ) for k = 2, 3, 4, we first note that the joint density
of eigenvalues ϑ1 and ϑ2 of a single qubit traced out of a system of a three-qubit system is given in
Equation (6) of [43] with N = 2 and K = 4. This reads

P (ϑ1 , ϑ2 ) = 210 δ(1 − ϑ1 − ϑ2 )(ϑ1 − ϑ2 )2 ϑ12 ϑ22 (7)

where δ stands for the Dirac delta. As each Ik is nothing other than the purity of a single qubit reduced
density matrix, we can compute the probability distribution by performing the following integral
 1 1
P( Ik ) = 210 dϑ1 dϑ2 P (ϑ1 , ϑ2 )δ( Ik − ϑ12 − ϑ22 ), (8)
0 0

this yields
105
P( Ik ) = (1 − Ik )2 (2Ik − 1)1/2 , 1/2 ≤ Ik ≤ 1, k = 2, 3, 4. (9)
2
This probability distribution is depicted in Figure 2. In addition, we approximate the distribution
P( I5 ) by the following beta distribution

Γ ( a + b + 2)
PF5 ( I5 ) = (1 − I5 ) a (4I5 − 1)b , (10)
3 a + b +1 4 a +1 Γ ( a + 1) Γ ( b + 1)

requiring the first two moments of this distribution coincide with the exact two moments of P( I5 )
reported above. We found a = 21, 989/5691 and b = 5554/5691. On the other hand, the distribution of
the square of the three tangle was approximated in [35] by a Beta distribution. Thus, making a variable
change in this result we may approximate P( I6 ) by

2 (
PF6 ( I6 ) = √ Beta(31/17, 62/17, 4 I6 ), 0 ≤ I6 ≤ 1/16. (11)
I6

As the distributions of the invariants I2 , I3 and I4 are the same, we only need three invariants to
characterize the entanglement in the set of three-qubit random states, say ( I2 , I5 , I6 ). In Figure 2d we
show a dispersion plot whose three axes correspond to such invariants and their colors correspond
to their S1 Rényi entropy calculated after of the five terms reduction [35] (which will be properly
presented in the next section) in agreement with the side color scale.
We also consider the set of invariants proposed by Acín et al. [15]. These invariants allow to
identify different entanglement classes (which will be discussed in the next section) and can be written
in terms of the six parameters of the five-term decomposition as

J1 = |λ1 λ4 eiϕ − λ2 λ3 |2 , J2 = μ0 μ2 , J3 = μ0 μ3 ,
(12)
J4 = μ0 μ4 , J5 = μ0 ( J1 + μ2 μ3 − μ1 μ4 ),

where μi = λ2i . For this analysis, the same set of 106 random states was considered but they are now
used to obtain the corresponding values of them through their expressions in terms of the five-term
coefficients [15]. All these invariants can be calculated departing from the set of λi . The outcomes are
shown in the Figures 2 and 3 in their respective ranges. Note in the Figure 3a–c how for J1 , J2 and J3

71

www.pdfgrip.com
Entropy 2018, 20, 745

the distribution is biased on low values of these invariants, denoting a possible relation with higher
entanglement. For the quantity J4 , related to the hyperdeterminant, the distribution peaks around of
⊗3
16 , denoting that separability as well as genuine entanglement are absent in the most of states in H .
1

A similar feature is observed for J5 but varying sharply for negative and positive values. On the other
hand, the invariants Jk ’s can be expressed in terms of the quantities Ik ’s [15]

1 ( 1 (
J1 = (1 + I2 − I3 − I4 − 2 I6 ), J2 = (1 − I2 + I3 − I4 − 2 I6 ),
4 ( 4 (
1
J3 = (1 − I2 − I3 + I4 − 2 I6 ), J4 = I6 ,
4 (
1
J5 = (3 − 3I2 − 3I3 − I4 + 4I5 − 2 I6 ).
4
√
Such expressions are useful to compute some averages. For instance, as  I5  = τ /4 it is
immediate to compute  J4  = 1/12. From the above definitions we can calculate directly  Jk  = 1/24,
for k = 1, 2, 3 and  J5  = 1/120. We approximate the probability distributions P( Jk ) with k = 1, 2, 3 by
a distribution PFk ( Jk ) ∼ Jka (1 − 4Jk )b , where the parameters in this case are determined numerically to
yield the best fit. In addition, making use of the approximation (11) for the distribution of the invariant
I6 . One can obtain the following approximation for the distribution of the variable J4

PF4 ( J4 ) = 4Beta(31/17, 62/17; 4J4 ), 0 ≤ J4 ≤ 1/4 (13)

On the other hand, as the distributions for J1 , J2 and J4 are uniform among them, we may
characterize the entanglement using only the invariants J1 , J4 and J5 . In Figure 3d we depict a scatter
plot using these invariants as coordinates, similarly as in Figure 2d for Ik .
Another interesting invariant is the one obtained by Kempe [44]

I5 = 3tr(ρ A ⊗ ρ B )ρ AB − trρ3A − trρ3B = ti1 j1 k1 ti2 j2 k2 ti3 j3 k3 ti1 j2 k3 ti2 j3 k1 ti3 j1 k2 , (14)

which distinguishes locally indistinguishable states. In terms of the Acín parameters, it reads

I5 = 1 − 3λ24 − 3λ23 + 3λ43 + 3λ44 + 3λ21 λ23 + 6λ23 λ24 (15)

+ λ21 3 − 6λ23 −3 λ23 −1 2λ23 + 2λ24 −1 λ22

+6λ1 λ3 λ4 λ21 + λ22 + λ23 + λ24 λ2 cos φ + 3 − 6λ23 λ42 .

Note that the form (14) of the Kempe Invariant I5 is manifestly permutation symmetric.
Although this quantity cannot be considered as a legitimate measure of entanglement, Osterloh has
pointed out [45] that different values of I5 allow to distinguish between different local orbits of three
qubit pure states. Integrating Equation (14) using symbolic integration on the Haar measure, we found
that  I5  = 2/5 and  I52  = 499/2860. In Figure 4a, we show the probability distribution of the
invariant I5 , which can be approximated by the distribution

9 a +1 Γ ( a + b + 2)
PFI (κ ) = (1 − κ ) a (9κ − 2)b , 2/9 ≤ κ ≤ 1 (16)
5 7 a + b +1 Γ ( a + 1) Γ ( b + 1)

where a = 90/23 and b = 283/621 are settled by the condition that the first two moments of PFI ( I5 )
5
correspond with the first two moments of P( I5 ) provided above. We remark that sextic invariant I5
can be written in terms of the Kempe invariant and the quadratic and quartic invariants [13].
Recently, an alternative set of invariants characterizing a three-qubit pure state |ψ was proposed
by Cheng and Hall [37]. To define them, consider a two-qubit reduced density matrix ρkl =

72

www.pdfgrip.com
Entropy 2018, 20, 745

Trm |ψklm ψklm | where indices k, l, m denote three subsystems A, B, C and m = k, l. Any such a
matrix of order four can be written in its Bloch representation,
 
3
1
ρkl =
4
1k ⊗ 1l + K ·σk ⊗ 1l + 1k ⊗ L ·σl + ∑ k,l k
Ti,j σi ⊗ σjl (17)
i,j=1

where σk = (σ1k , σ2k , σ3k ), while K and L denote the Bloch vectors for parts k and l respectively. Entries of
k,l
the correlation matrices of the reduced states read Tn,m = σnk ⊗ σml  = Tr(σnk ⊗ σml ρkl ), while the
superscripts denote two out of three subsystems A, B, C as required to determine a two-qubit partial
†
j denote the eigenvalues of the symmetric matrix S = T
trace. Let skl kl kl T kl and the average value
read, kl
siso = (s1kl + s2kl + s3kl )/3. The invariants are constructed in terms of the pairwise anisotropic
j with j = 1, 2, 3 which read δs j = s j − siso , with k, l = A, B, C. It was
δs jAB , δs jAC and δs BC
strengths kl kl kl

shown [37] that the pairwise anisotropic strengths fulfil the relations

(a) (b)

(c) (d)

Figure 3. (a–c) Probability distribution for the polynomial invariants Ji , i = 1, ..., 5 for a set of 106
three-qubit random states. In all graphics the numerical best fit distribution is depicted as the green
line. In Figure (d) we show a dispersion plot comparing J1 , J4 and J5 . In addition, each dot has been
colored as a function of its S1 Rényi entropy [41] calculated after the five term reduction in agreement
with the side color scale.

δs j = δs jAB = δs jAC = δs BC
j , j = 1, 2, 3 (18)

and they are also invariant under local transformations as well as any permutation of the parties.
Hence, the anisotropic strength and the anisotropic volume can be defined as

73

www.pdfgrip.com
Entropy 2018, 20, 745

s2ani = ∑(δsi )2 , Vani = ∏ δsi . (19)

i j

Note that for a given three-qubit pure state |ψ the above invariants can be related with parameters
entering the five-term form (4) —see Supplementary Material in [37].
In Figure 4b–d we show the probability distributions of the pairwise anisotropic strengths as well
as the probability distribution of the invariants sani and Vani for an ensemble of 106 three-qubit random
states. We approximate numerically the distribution of the quantities δs1 , δs2 and sani by respective
beta distributions:

PF (δs1 ) = c1 ( a1 − δs1 ) β1 (δs1 − a2 ) β2 , PF (δs2 ) = c2 ( a3 − δs2 ) β3 ( a4 + δs2 ) β4 ,

(20)
β
PF (sani ) = c3 ( a5 − sani ) β5 sani
6
,

while the positive part of the distribution of V can be approximated by an exponential distribution,

PF (Vani ) = c4 e−bVani . (21)

The fitting parameters read ci = (472.7, 1299, 135.6, 54.9) for i = 1, . . . , 4; a j =

(0.66, 0.01, 0.11, 0.33, 0.72) for j = 1, . . . , 5; β i = (2.5, 2.04, 1.88, 1.92, 2.26, 1.63) for i = 1, . . . , 6 and
b ≈ 61.6. Interestingly, the distribution of the negative quantity δs3 displays a singular peak, while
the distribution of Vani attains its maximum at anisotropic volume close to zero and exhibits an
exponential decay.

(a) (b)

(c) (d)

Figure 4. Probability distribution of: (a) the Kempe invariant I5 (the green line stands for the probability
distribution (16)). (b) The pairwise anisotropic strengths δs j with j = 1, 2, 3. (c) The invariant sani and
(d) The invariant Vani . Solid lines in all cases correspond to the best numerical fit.

4. Three-Qubits Entanglement Classes

A state classification has been presented in [15] based on the minimal number of product states
in (4). Acín et al. reported some entanglement classes which are presented in Table 2. The conditions

74

www.pdfgrip.com
Entropy 2018, 20, 745

for such class states are expressed in terms of the invariants Jk . Thus, in this section we consider
the invariant classes introduced there, departing from the coefficients of the five-term representation
in H ⊗3 . These classes barely describe some families around some characteristic states in this space.
The first aim is to analyze how those classes represent the entanglement of each state included there,
mainly based on the entanglement distribution knowledge on H ⊗3 [35]. Note that in some classes
the direct imposition of the conditions on the invariants leaves some product states that differ from
those reported by Acín, that is to say, to obtain such product states an additional LU transformation is
required. Such cases are remarked with  in Table 2.

Table 2. Acín entanglement classes introduced in [15]. Besides Δ J ≡ ( J4 + J5 )2 − 4( J1 + J4 )( J2 + J4 )( J3 + J4 ).

Basis elements marked with  are not directly obtained, instead they have additional relabellings.
Besides, the fourth column shows the identification of each class with subsets of the entanglement
polytope. The point G stands for (1/2, 1/2, 1/2). Details are presented in Section 5.

Class Conditions States Entanglement Polytope

1 Ji = 0 |000 point O = (0, 0, 0)

2a All Ji = 0 apart from J1 |000, |011 lines O A, O B and O C

2b All Ji = 0 apart from J4 |000, |111 line O G

3a J1 J2 + J1 J3 + J2 J3
√ = |000, |101, |110 2 O AB, 2 O AC, 2 O BC, 2 ABC
J1 J2 J3 = J5 /2, J4 = 0
3b J1 = J2 = J5 = 0 |000, |110, |111 2 ABG, 2 ACG, 2 BCG
√
4a J4 = 0, J1 J2 J3 = J5 /2 |000, |100, |101, |110 3 O ABC

4b J2 = J5 = 0 |000, |100, |110, |111

4c J1 J4 + J1 J2 + J1 J3 + J2 J3 =
√ |000, |101, |110, |111
= J5 /2
J1 J2 J3 √
4d Δ J = 0, J1 J2 J3 = | J5 |/2 |000, |010, |100, |111

4.1. The Minimal Decomposition Entropy

We characterize the entanglement degree of the classes in Table 2 using the minimal
Rényi-Ingarden-Urbanik (RIU) entropy, also known as minimal decomposition entropy [35]. For the
state (1) this is defined as
SqRIU (ψ) := min Sq [ p(Uloc |ψ)] , (22)
Uloc

where p(·) stands for the probability vector related to the state (1) and the minimum is taken on all local
transformations Uloc = U1 ⊗ U2 ⊗ U3 . Note that Sq is the q-order Rényi entropy [41]. Depending on
the parameter q the quantity (22) provides information about the state [35]. Thus, for

• q = 0: The decomposition entropy is related to the tensor rank of the state |ψ. As a direct
consequence of the decomposition (4) we have S0RIU (ψ) ≤ 5.
• q = 1: The minimal decomposition entropy S1RIU (|ψ) determines the minimal information gained
by the environment after performing a projective von-Neumann measurement of the pure state
|ψψ| in an arbitrary product basis [46].
• q → ∞: In such a limiting case, the minimal RIU entropy is associated with the maximal overlap
with the closest separable state Λmax = max |ψ|χsep |2 . Indeed, it can be shown that S∞
RIU (| ψ ) =

− log λmax . See [35] for details.

A direct computation shows that for a state in class 1, the minimal RIU entropy vanishes regardless
of the value of the parameter q. The corresponding calculation for the other entanglement classes is
presented below.

75

www.pdfgrip.com
Entropy 2018, 20, 745

4.1.1. Classes 2
A direct calculation shows that the decomposition of states in class 2b is optimal. That is to say,
if the state is given by
| ϕ2b  = cos α|000 + sin α|111, 0 < α < π/2, (23)

the minimal decomposition entropy reads

S1RIU ( ϕ2b ) = − cos2 α ln(cos2 α) − sin2 α ln(sin2 α). (24)

Our numeric calculations indicate that for the class 2a, the Acín decomposition is optimal as well.
The states with the largest minimal decomposition entropy in each class are

1 1 1 1
2a  = √ |000 + √ |111,
| ϕmax 2b  = √ |100 + √ |111,
| ϕmax (25)
2 2 2 2

note the reported basis for class 2b in Table 2 is different due to additional changes commonly reported
in the literature. A simple calculation shows the LU equivalence of the two local basis. Note that the
state | ϕmax
2b  is bi-separable and it attains the same minimal decomposition entropy as the GHZ state.

4.1.2. Classes 3
Any state belonging to class 3a can be parametrized as

| ϕ3a  = sin θ1 sin θ2 |000 + sin θ1 cos θ2 |101 + cos θ1 |110, 0 < θ1 , θ2 < π/2 (26)

note such state is LU-equivalent to the symmetric state

| ϕ)3a  = sin θ1 sin θ2 |100 + sin θ1 cos θ2 |001 + cos θ1 |010, (27)

hence, the minimal RIU entropy can √ be computed using √ the method described in [35] for symmetric
states. In particular, if cos θ1 = 1/ 3 and sin θ2 = 1/ 2 we obtain the well-known W-state for which
S1RIU (W ) = ln 3, which is the largest value of S1RIU for this class.
On the other hand, a state in class 3b can be written as

| ϕ3b  = sin θ1 sin θ2 |000 + sin θ1 cos θ2 |110 + cos θ1 |111, 0 < θ1 , θ2 < π/2. (28)

No state in class 3b has greater S1RIU than the W-state. For a general state in these classes, the minimal
decomposition entropy as a function of parameters θ1 and θ2 is depicted in Figure 5. Note that regions of
maximal S1RIU entropy are around the values θ1 , θ2 for the maximal entropy for such states.

(a) (b)

Figure 5. (a) The minimal decomposition entropy level curves as function of the parameters θ1 and θ2
for a state in class 3a; (b) Same as (a) for a state in class 3b.

76

www.pdfgrip.com
Entropy 2018, 20, 745

4.1.3. Classes 4
A general state in each one of the classes 4 can be written as

| ϕ4a  = β 1 |000 + eiϕ β 2 |100 + β 3 |101 + β 4 |110 (29)

| ϕ4b  = β 1 |000 + eiϕ β 2 |100 + β 3 |110 + β 4 |111 (30)
| ϕ4c  = β 1 |000 + β 2 |101 + β 3 |110 + β 4 |111 (31)
| ϕ4d  = β 1 |000 + β 2 |010 + β 3 |100 + β 4 |111 (32)

where β 1 = sin θ1 sin θ2 sin θ3 , β 2 = sin θ1 sin θ2 cos θ3 , β 3 = sin θ1 cos θ2 and β 4 = cos θ1 . As for
class 2b, the basis elements for class 4d reported in Table 2 are not those directly obtained from (4).
Class 4d corresponds to the real class (with all components real, thus eiϕ = ±1) which allows for
the performance of an additional reduction to only four terms. As in the previous case, we get
the surfaces of minimal decomposition entropy in terms of parameters θ1 , θ2 and θ3 in the Figure 6.
Those figures exhibit for each class the behavior for the entropy. There, the frontiers of the regions
shown θ1 , θ2 , θ3 = 0, π/2 correspond to separable states. In addition, our numerical calculations show
that the minimal decomposition entropy is independent of the phase φ. We also numerically found
that the the largest S1RIU (ψ4a
max ) = 1.213 is attained for a state in class 4a with θ = 3π/10, θ = 4π/15
1 2
and θ3 = 23π/60. Note that this value is smaller than the one reported earlier [35] as the maximal for
a random state with five components.

(a) (b)

(c) (d)

Figure 6. Surfaces of equal entanglement for classes 4 measured with respect the minimal
decomposition entropy as function of the parameters θ1 , θ2 and θ3 defining each class. Different
panels correspond to a state in: (a) Class 4a; (b) Class 4b; (c) Class 4c; (d) Class 4d.

77

www.pdfgrip.com
Entropy 2018, 20, 745

4.2. The Maximum Overlap with an Entanglement Class

Given an ensemble of random states, a natural question arises: how many states of such ensemble
belong to a particular Acín entanglement class? To tackle this question, observe first that numeric
calculations imply S0RIU (ψ) = log 5. Hence a generic three-qubit state has five non trivial components
in the decomposition (4). As each class has at most four components, we rather consider the
following quantity
Λi ( β) = max {| ϕ|Ulocal
†
| β|2 : | ϕ ∈ Class i }, (33)
| ϕ,Ulocal

where i = {1, 2a, 2b, 3a, 3b, 4a, 4b, 4c, 4d} and Ulocal = U1 ⊗ U2 ⊗ U3 . Such quantity provides an
information, how much a given state | β on H ⊗3 differs from the closest state | ϕ in the Acín
entanglement class i [15]. Note that the quantity Λi can be interpreted as the maximal fidelity of
a given state | β with respect to the closest state belonging to the class i. In particular, if i = 1 the
results are consistent with S∞ RIU ( β ) (see [35]) as this yields the maximum overlap with the closest

separable state.
By taking a set of 105 random states in H ⊗3 , we get their projection Λi on each Acín class,
tracking their hyperdeterminant Hdet(| ϕ), which is clearly invariant under local transformations.
Then we perform a numerical optimization on the three parameters depicting a local transformation
on each qubit (nine in total) together with the necessary coefficients depicting an arbitrary state in each
class [15]. Finally, we also track the hyperdeterminant of such a state, Hdet(| β). With this information,
we construct the corresponding distribution ρ(Λi ) of each projection i (33).
Numerical results are shown jointly in Figure 7. First, the line plot shows the value of ρ(Λi )
on the left axis versus the value of projection Λi on the horizontal axis. Superposed, a dispersion
plot of the entire set of states being analyzed is shown in color. Each dot represents a random
state located vertically on their projection value Λi and horizontally in its hyperdeterminant value
Hdet(| β), which remains invariant under the local optimization procedure. Additionally, each dot
is colored in agreement with the hyperdeterminant of the best class element | ϕ obtained in the
optimization. Colors are assigned from red for separable states to green for maximal genuine
entanglement. This structure of the plot allows one to compare the closeness between | β and | ϕ in
terms of genuine entanglement. Note the graph corresponding to class 4d has been omitted because it
is equivalent to that of class 4c: All coefficients in the class are real, then by exchanging 0 and 1 in all
qubits and swapping the qubits 1 and 3 we get the same state with local operations. Thus, the maximal
overlap and the hyperdeterminant statistics do not change.
Note particularly how in the Figure 7a the closest class states have Hdet(| ϕ) = 0 for some
random states which have Hdet(| β) near from the highest value 14 maintaining a closer distance
Λ4a ≈ 1. The opposite phenomenon is also observed in Figure 7b,c,e,g where some class states with
Hdet(| ϕ) ≈ 14 (in green) are close to some random states with lower Hdet(| β) values. On the other
hand, in Reference [24] the distribution of the fidelity between two random states has been computed
analytically. However, in our case the problem becomes more complicated due to the optimization of
the fidelity over all local unitaries.

78

www.pdfgrip.com
Entropy 2018, 20, 745

Figure 7. Dispersion graphs showing Hdet(| β) versus the maximum overlap Λi , colored from red
(separable) to green (maximal genuine entanglement). Each panel correspond to one of the Acín classes
(see Table 2) as follows: (a) Class 4a; (b) Class 4b; (c) Class 4c; (d) Class 3a; (e) Class 3b; (f) Class 2a;
(g) Class 2b; (h) Class 1. Besides, probability distributions of the maximum overlap (33) are shown in
the inset of each plot (vertical scale on the left). We have taken an ensemble of 105 three-qubit random
states. Graphs of classes 4c and 4d are equivalent so this last was omitted (see details in the core text).

5. The Entanglement Polytope of Three Qubits

Let λmin
k denote the smallest eigenvalue of the reduced density matrix of the subsystem of three
qubits, where k = A, B, C. The following set of compatibility conditions

A ≤ λ B + λC ,
λmin ≤ λmin
A + λC , ≤ λmin
A + λB .
min min
λmin
B
min
λmin
C
min
(34)

Form particular examples of polygon inequalities obtained by Higuchi et al. for systems of several
qubits [17]. The smaller eigenvalue of a one-qubit system is not larger then 1/2 so that 0 ≤ λmin
k ≤ 1/2.
Inequalities (34) determine jointly a convex polytope in the three-space (λmin A , λ B , λC ). Its five
min min

vertices represent distinguished three-qubit states: Fully separable states are identified by the point

79

www.pdfgrip.com
Entropy 2018, 20, 745

SEP = (0, 0, 0) whereas points A = (1/2, 1/2, 0), B = (1/2, 0, 1/2) and C = (0, 1/2, 1/2) stand for
bi-separable states. The GHZ-state is located at GHZ = (1/2, 1/2, 1/2). The convex hull of these
points is known as the Kirwan polytope [21,47,48]. In addition, the identification of a state belonging
to an entanglement classes reported in [20] is summarized in Table 2.
Consider now an ensemble of three-qubit random states. For such states, the probability
distribution of the minimal eigenvalue of a single-particle reduced density matrix fulfils P(λmin ) =
P(λmin
A ) = P ( λ B ) = P ( λC ). Using the following relation between the two eigenvalues ϑ1 and ϑ2 of
min min

a single qubit reduced density matrix

1 1
λmin = min(ϑ1 , ϑ2 ) = (ϑ + ϑ2 ) − |ϑ1 − ϑ2 |,
2 1 2
we can compute the probability distribution of the minimal eigenvalue λmin as
 1 1
P(λmin ) = dϑ1 dϑ2 P (ϑ1 , ϑ2 )δ[λmin − (ϑ1 + ϑ2 )/2 + |ϑ1 − ϑ2 |/2)], (35)
0 0

where P (ϑ1 , ϑ2 ) is the joint density (7) and δ stands for the Dirac delta function. Performing the
integral, we obtain

P(λmin ) = 420[λmin (2λmin − 1)(1 − λmin )]2 , 0 ≤ λmin ≤ 1/2. (36)

This distribution is depicted in Figure 8. Besides, a direct calculation yields the average value
λmin  = 29/128. In general, the k-the moment of λmin reads
* +
105 Γ(k + 3) Γ(k + 4) Γ ( k + 5)
λmin
k
= k − + . (37)
2 Γ(k + 6) Γ(k + 7) 4Γ(k + 8)

Note that a given pure state can be identified with a point in the entanglement polytope.
Its coordinates are (λmin
A , λ B , λC ). This is shown in Figure 8b for an ensemble of 10 three-qubit
min min 6

random states colored according to their joint probability distribution in the polytope. To compute
such probability distribution, the space containing the whole polytope [0, 12 ]×3 was divided into 803
cubic cells. Then, we state the statistics of random states falling in each cell to get the probability
density of those states (by volume unity). Note that the closer the points are to the faces, the lower the
value of the distribution. In Figure 8c we depict a transverse cut by the plane containing the vertices
S, C and GHZ to depict the distribution of the inner points. This shows that random states are more
concentrated near the line joining the vertices SEP and GHZ, which corresponds to class 2a.
On the other hand, two quantum pure states attain the same amount of entanglement if they
belong to the same class, that is to say if there is a finite probability of success that they can be converted
into each other using stochastic local operations and classical communication, referred to as SLOCC
by its acronyms. For the case of three qubits, there exist two SLOCC classes of entanglement: the one
containing the GHZ state, which exhibits genuine entanglement and the W class [8].These classes can
be distinguished from the entanglement polytope. Numerical calculation shows that around 6% of
the states are placed in the upper polytope, so that they belong to the GHZ SLOCC class [21]. As the
invariant I6 discriminates between such classes in panel Figure 8d we show the ensemble of random
states colored with respect to this invariant. For states placed near the bi-separable faces I6 goes to
zero, whereas the states landing in the GHZ simplex are characterized by a positive value of this
invariant. An equivalent approach can be done dealing with the maximum eigenvalues of the reduced
single qubit density matrices. For such a case, the joint probability distribution is known [49] and
hence the fraction of random states in the GHZ pyramid was computed in Reference [50] yielding
13/216 ≈ 6.02% which is consistent with our numerical calculation.

80

www.pdfgrip.com
Entropy 2018, 20, 745

(a) (b)

(c) (d)

Figure 8. (a) Probability distribution of the minimal eigenvalue of a single qubit reduced system (36).
(b) An ensemble 106 of three-qubit random states depicted in the entanglement polytope. The color
scale stands for the joint probability distribution. (c) Detail of (b): A transversal section by the plane
which contains the points S, C and GHZ. (d) The ensemble of three qubit random states labeled by
colour settled according to the value of the invariant I6 .

6. Conclusions and Future Work

We studied various quantities describing a three-qubit pure quantum state and analyzed their
probability distributions obtained for an ensemble of random pure states generated by the unitary
invariant Haar measure. In particular, we investigated the distribution of the six parameters
determining the five-terms decomposition (4) of a three-qubit state. The phase of the complex
coefficient occurs to be uniformly distributed. The distributions of the amplitudes λ0 and λ4 differ
from the distribution describing the remaining three coefficients. Interestingly, these two coefficients
can be related with the degree of entanglement as the invariant J4 depends only on them. In addition,
we have also analyzed the probability distributions of two sets of polynomial invariants. The invariants
I1 , I2 and I3 follow the same distribution. Thus, out of the five independent invariants, only three are
necessary to characterize entanglement in three-qubit states. This fact is consistent with the second
set of invariants reported by Acín et al. as the distributions of the invariants J1 , J2 and J3 do coincide.
For each invariant its mean value was computed using symbolic integration with respect to the
unitary invariant Haar measure. Moreover, we have also obtained the probability distribution of the
anisotropic strength sani and the anisotropic volume Vani introduced recently in [37]. These invariants
are useful in the study of strong monogamy relations, geometric discord and fidelity of remote state
preparation and studies of violation of the Bell inequality. In this last context, one could ask for
the probability that one of the three pairs violates a Bell inequality. However, these results will be
reported elsewhere.

81

www.pdfgrip.com
Entropy 2018, 20, 745

On the other hand, the set of invariants { Jk } allows us to identify certain entanglement
classes, whose entanglement was described through the minimal decomposition entropy. Moreover,
highly entangled states with respect to this measure were identified in each class. Our results imply
that the more terms in the decomposition (4) of a three-qubit state, the larger its degree of entanglement
measured by the minimal decomposition entropy.
The numerical outcomes provide us with several insights about possible meanings of the
entanglement invariants. First, there is an apparent underlying statistical equivalence between
coefficients λ1 , λ2 and λ3 (and their low values suggest a closer position of the states respecting
genuine entanglement states in terms of the RIU entropy statistics for the overall 3-qubits random
states). The same aspects seems true for I2 , I3 , I4 and J1 , J2 , J3 invariants. Together, larger values for λ0
and low values for I5 and J4 seem related with the presence of genuine entanglement (this affirmation
is based on the fact that larger values of RIU entropy are statistically more common for the overall
3-qubit random states).
Other outcomes relative to the type a in the Acín classes exhibit separable states. There,
the growing number of the class (1, 2, ..., 4) reflects the inclusion of most of the random states for
three qubits (see Figure 7a,d,f,h). In this sense, the use of RIU entropy as an exemplary measure of
quantum entanglement allows us to provide a classification of three qubit states and to describe their
hierarchy. The invariants with respect to local transformations are useful to identify certain types
of entangled structures in the entire system. As shown in Figure 7, the states displaying genuine
entanglement appear closer from other states in the classes with no genuine entanglement. Although
smooth measures of entanglement depend on the state in a continuous way, a small variation of a state
can lead to a considerable change of its entanglement. This feature was observed in larger systems [51].
In such a scenario, the current analysis in the quest of understanding the hierarchy of entanglement,
could set directions to transform states from maximally entangled into separable ones. By using
the SU (2) decomposition procedure, [52] has been clear about the existence of basic U (1) × SU (2)
operations among entangled pairs, showing how the entanglement phenomena can be generated in
a structured way form basic operations then transiting from separable to genuine entangled states.
This suggests that programmed local operations combined with entangling operations between two
previous entangled pairs can be realized in order to connect such state types. Thus, basic separable
states could be transformed into maximal entangled states as | GHZ  and |W  only with a series of
such operations. In a more ambitious task, those single types of operations could suggest they could
be responsible for the transit from certain classes to others among the hierarchies of entanglement.
In such a process, the track in the change of the invariants values could provide a strong road-map for
such transit.
Finally, we have analyzed the probability distribution of the maximal fidelity of a random state
with respect to the closest representative of each entanglement class. The highest maximal fidelity
is obtained for classes 4a–d listed in Table 2. This can be seen from the fact that the distributions
of five coefficients in the decomposition (4) are highly non-trivial, as these quantities carry some
information concerning the degree of entanglement. Our study comprises several ways to analyze
the entanglement in a three-qubit system showing the fact that entanglement can be characterized
from different approaches, each one providing different aspects of non-locality. Therefore, we hope the
results of this work will shed some light on the matter.

Author Contributions: M.E. and F.D. performed the numerical calculations. M.E. obtained the analytical results.
K.Ż. proposed the main idea of Section 5. All authors contributed equally to write the paper.
Funding: This research received no external funding.
Acknowledgments: The support of Escuela de Ingeniería y Ciencias of Tecnológico de Monterrey as well as
the support of CONACyT are gratefully acknowledged. K.Ż. acknowledges support by Narodowe Centrum
Nauki under the grant number DEC-2015/18/A/ST2/00274. We are obliged to Michael J.W. Hall for drawing our
attention to the reference [37].
Conflicts of Interest: The authors declare no conflict of interest.

82

www.pdfgrip.com
Entropy 2018, 20, 745

References
1. Walter, M.; Gross, D.; Eisert, J. Multi-partite entanglement. arXiv 2016, arXiv:1612.02437.
2. Bengtsson, I.; Życzkowski, K. Geometry of Quantum States: An Introduction to Quantum Entanglement, 2nd ed.;
Cambridge University Press: Cambridge, UK, 2017.
3. Gurvits, L. Classical complexity and quantum entanglement. J. Comput. Syst. Sci. 2004, 69, 448–484.
[CrossRef]
4. Horodecki, R.; Horodecki, P.; Horodecki, M.; Horodecki, K. Quantum entanglement. Rev. Mod. Phys. 2009,
81, 865–942. [CrossRef]
5. Enríquez, M.; Wintrowicz, I.; Życzkowski, K. Maximally entangled multipartite states: A brief survey. J. Phys.
Conf. Ser. 2016, 698, 012003. [CrossRef]
6. Higuchi, A.; Sudbery, A. How entangled can two couples get? Phys. Lett. A 2000, 273, 213–217. [CrossRef]
7. Carteret, H.A.; Higuchi, A.; Sudbery, A. Multipartite generalisation of the Schmidt decomposition.
J. Math. Phys. 2000, 41, 7932. [CrossRef]
8. Dür, W.; Vidal, G.; Cirac, J.I. Three qubits can be entangled in two inequivalent ways. Phys. Rev. A 2000,
62, 062314. [CrossRef]
9. Meill, A.; Meyer, D.A. Symmetric three-qubit-state invariants. Phys. Rev. A 2017, 96, 062310. [CrossRef]
10. Verstraete, F.; Dehaene, J.; De Moor, B.; Verschelde, H. Four qubits can be entangled in nine different ways.
Phys. Rev. A 2002, 65, 052112. [CrossRef]
11. Albeverio, S.; Fei, S. A note on invariants and entanglements. J. Opt. B 2011, 3, 223. [CrossRef]
12. Grassl, M.; Rötteler, M.; Beth, T. Computing local invariants of qubit systems. Phys. Rev. A 1998, 58, 1833.
[CrossRef]
13. Sudbery, A. On local invariants of pure three-qubit states. J. Phys. A. Math. Gen. 2001, 34, 643–652. [CrossRef]
14. Holweck, F.; Luque, J.; Thibon, J. Entanglement of four qubit systems: A geometric atlas with polynomial
compass I (the finite world). J. Math. Phys. 2014, 55, 012202. [CrossRef]
15. Acín, A.; Andrianov, A.; Jané, E.; Tarrach, R. Three-qubit pure-state canonical forms. J. Phys. A 2001, 34,
6725–6739. [CrossRef]
16. Sawicki, A.; Walter, M.; Kuś, M. When is a pure state of three qubits determined by its single-particle reduced
density matrices? J. Phys. A 2013, 46, 055304. [CrossRef]
17. Higuchi, A.; Sudbery, A.; Szulc, J. One-qubit reduced states of a pure many-qubit state: Polygon inequalities.
Phys. Rev. Lett. 2003, 90, 107902. [CrossRef] [PubMed]
18. Bravyi, S. Requirements for compatibility between local and multipartite quantum states. Quantum Inf. Comp.
2004, 4, 12–26.
19. Klyachko, A. Quantum marginal problem and representations of the symmetric group. arXiv 2004,
arXiv:quant-ph/0409113.
20. Han, Y.J.; Zhang, Y.S.; Guo, G.C. Compatible conditions, entanglement, and invariants. Phys. Rev. A 2004,
70, 042309. [CrossRef]
21. Walter, M.; Doran, B.; Gross, D.; Christandl, M. Entanglement polytopes: Multiparticle entanglement from
single-particle information. Science 2013, 340, 1205–1208. [CrossRef] [PubMed]
22. Kuś, M.; Mostowski, J.; Haake, F. Universality of eigenvector statistics of kicked tops of different symmetries.
J. Phys. A Math. Gen. 1988, 21, L1073–L1077. [CrossRef]
23. Haake, F. Quantum Signatures of Chaos, 2nd ed.; Springer Verlag: Berlin, Germany, 2001.
24. Życzkowski, K.; Sommers, H.-J. Average fidelity between random quantum states. Phys. Rev. A
2005, 71, 032313. [CrossRef]
25. Giraud, O.; Žnidarič, M.; Georgeot, B. Quantum circuit for three-qubit random states. Phys. Rev. A
2009, 80, 042309. [CrossRef]
26. Kendon, V.M.; Życzkowski, K.; Munro, W.J. Bounds on entanglement in qudit subsystems. Phys. Rev. A 2002,
66, 062310. [CrossRef]
27. Cappellini, V.; Sommers, H.-J.; Życzkowski, K. Distribution of G concurrence of random pure states.
Phys. Rev. A 2006, 74, 062322. [CrossRef]
28. Kumar, S.; Pandey, A. Entanglement in random pure states: Spectral density and average von Neumann
entropy. J. Phys. A 2011, 44, 445301. [CrossRef]

83

www.pdfgrip.com
Entropy 2018, 20, 745

29. Vivo, P.; Pato, M.P.; Oshanin, G. Random pure states: Quantifying bipartite entanglement beyond the linear
statistics. Phys. Rev. E 2016, 93, 052106. [CrossRef] [PubMed]
30. Kendon, V.; Nemoto, V.K.; Munro, W. Typical entanglement in multiple-qubit systems. J. Mod. Opt.
2002, 49, 1709–1716. [CrossRef]
31. Facchi, P.; Florio, G.; Pascazio, S. Probability-density-function characterization of multipartite entanglement.
Phys. Rev. A 2006, 74, 042331. [CrossRef]
32. Korzekwa, K.; Lostaglio, M.; Jennings, D.; Rudolph, T. Quantum and classical entropic uncertainty relations.
Phys. Rev. A 2014, 89, 042122. [CrossRef]
33. Fannes, M. Multi-state correlations and fidelities. Int. J. Geom. Methods Mod. Phys. 2012, 9, 1260021. [CrossRef]
34. Rangamani, M.; Rota, M. Entanglement structures in qubit systems. J. Phys. A 2015, 48, 385301. [CrossRef]
35. Enríquez, M.; Puchała, Z.; Życzkowski, K. Minimal Rényi-Ingarden-Urbanik entropy of multipartite
quantum states. Entropy 2015, 17, 5063–5084. [CrossRef]
36. Alsina, D. Multipartite Entanglement and Quantum Algorithms. Ph.D. Thesis, Universitat de Barcelona,
Barcelona, Spain, 2017.
37. Cheng, S.; Hall, M.J.W. Anisotropic invariance and the distribution of quantum correlations. Phys. Rev. Lett.
2017, 118, 010401. [CrossRef] [PubMed]
38. Grendar, M. Entropy and effective support size. Entropy 2006, 8, 169–174. [CrossRef]
39. Carteret, H.A.; Linden, N.; Popescu, S.; Sudbery, A. Multi-particle entanglement. Found. Phys. 1999, 29,
527–552. [CrossRef]
40. Acín, A.; Andrianov, A.; Costa, L.; Jané, E.; Latorre, J.; Tarrach, R. Generalized Schmidt decomposition and
classification of three-quantum-bit states. J. Phys. Lett. 2000, 85, 1560–1563. [CrossRef] [PubMed]
41. Rényi, A. On measures of information and entropy. In Proceedings of the Fourth Berkeley Symposium on
Mathematics, Statistics and Probability, Berkeley, CA, USA, 20 June–30 July 1960.
42. Puchała, Z.; Miszczak, J.A. Symbolic integration with respect to the Haar measure on the unitary groups.
Bull. Pol. Acad. Sci. Tech. Sci. 2017, 65, 21. [CrossRef]
43. Życzkowski, K.; Sommers, H.-J. Induced measures in the space of mixed quantum states. J. Phys. A 2001,
34, 7111–7125. [CrossRef]
44. Kempe, J. Multiparticle entanglement and its applications to cryptography. Phys. Rev. A 1999, 60, 910–916.
[CrossRef]
45. Osterloh, A. Classification of qubit entanglement: SL(2, C) versus SU (2) invariance. Appl. Phys. B 2010, 98,
609–616. [CrossRef]
46. Maziero, J. Understanding von Neumann entropy. Rev. Bras. Ensino Fís. 2015, 37, 1314. [CrossRef]
47. Sawicki, A.; Oszmaniec, M.; Kuś, M. Critical sets of the total variance can detect all stochastic local operations
and classical communication classes of multiparticle entanglement. Phys. Rev. A 2012, 86, 040304. [CrossRef]
48. Sawicki, A.; Oszmaniec, M.; Kuś, M. Convexity of momentum map, Morse index, and quantum entanglement.
Rev. Math. Phys. 2014, 26, 1450004. [CrossRef]
49. Christandl, M.; Doran, B.; Kousidis, S.; Walter, M. Eigenvalue distributions of reduced density matrices.
Commun. Math. Phys. 2014, 332, 1–52. [CrossRef]
50. Zhao, Y.Y.; Grassl, M.; Zeng, B.; Xiang, G.Y.; Zhang, C.; Li, C.F.; Guo, G.C. Experimental detection of
entanglement polytopes via local filters. Quantum Inf. 2017, 3, 11. [CrossRef]
51. Delgado, F. Assembling large entangled states in the Rényi-Ingarden-Urbanik entropy measure under the
SU (2)-dynamics decomposition for systems built from two-level subsystems. In Proceedings of the 6th
Annual International Conference on Physics, Athens, Greece, 23–26 July 2018.
52. Delgado, F. SU (2) decomposition for the Quantum Information Dynamics in 2d-Partite Two-Level Quantum
Systems. Entropy 2018, 20, 610. [CrossRef]

c 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

84

www.pdfgrip.com
 entropy
Article
New Entropic Inequalities and Hidden Correlations
in Quantum Suprematism Picture of Qudit States †
Margarita A. Man’ko 1,∗ and Vladimir I. Man’ko 1,2,3
1 Lebedev Physical Institute, Russian Academy of Sciences, Leninskii Prospect 53, Moscow 119991, Russia
2 Moscow Institute of Physics and Technology (State University), Institutskii per. 9, Dolgoprudnyi,
Moscow Region 141700, Russia; manko@sci.lebedev.ru
3 Department of Physics, Tomsk State University, Lenin Avenue 36, Tomsk 634050, Russia

* Correspondence: mmanko@sci.lebedev.ru
† We dedicate this article to the memory of Professor E. C. G. Sudarshan, the Great Scientist and our friend
and colleague with whom we were happy to collaborate during many years.

Received: 17 August 2018; Accepted: 7 September 2018; Published: 11 September 2018 

Abstract: We study an analog of Bayes’ formula and the nonnegativity property of mutual
information for systems with one random variable. For single-qudit states, we present new entropic
inequalities in the form of the subadditivity and condition corresponding to hidden correlations
in quantum systems. We present qubit states in the quantum suprematism picture, where these
states are identified with three probability distributions, describing the states of three classical
coins, and illustrate the states by Triada of Malevich’s squares with areas satisfying the quantum
constraints. We consider arbitrary quantum states belonging to N-dimensional Hilbert space as
( N 2 − 1) fair probability distributions describing the states of ( N 2 − 1) classical coins. We illustrate
the geometrical properties of the qudit states by a set of Triadas of Malevich’s squares. We obtain
new entropic inequalities for matrix elements of an arbitrary density N ×N-matrix of qudit systems
using the constructed maps of the density matrix on a set of the probability distributions. In addition,
to construct the bijective map of the qudit state onto the set of probabilities describing the positions
of classical coins, we show that there exists a bijective map of any quantum observable onto
the set of dihotomic classical random variables with statistics determined by the above classical
probabilities. Finally, we discuss the physical meaning and possibility to check derived inequalities
in the experiments with superconducting circuits based on Josephson junction devices.

Keywords: entropy; correlations; qubits; probability representation; Bayes’ formula

1. Introduction
Generic states of quantum systems are identified with the density matrices [1,2] or the density
operators ρ̂ acting in a Hilbert space. The pure states of quantum systems are identified with the
state vectors | ψ belonging to the Hilbert space [3] and complex wave functions [4,5] ψ( x ) =  x | ψ,
where x is an observable, e.g., the continuous position of a particle. The physical meaning of the
wave function ψ( x ) is related to measuring the observable x; in the state | ψ, the measurement of
the position of a particle yields the probability density | x | ψ|2 = |ψ( x )|2 , which does not contain
information on the phase of the complex wave function.
For spin-s systems with discrete observables like spin projections m = −s, −s + 1, . . . , s − 1, s;
s = 0, 1/2, 1, 3/2, 2, . . ., the state vectors belong to the Hilbert space of finite dimension N = 2s + 1,
and the complex wave function ψ(m) = m | ψ determines the probability distribution
|m | ψ|2 = |ψ(m)|2 associated with the state | ψ. The phase of the wave function is not determined
by the probability distribution; in view of this fact, information on the state | ψ, contained in the

Entropy 2018, 20, 692; doi:10.3390/e20090692 85 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2018, 20, 692

probability density |ψ( x )|2 or in the probability distribution |ψ(m)|2 , is not sufficient to describe the
particle’s pure state or the spin-s pure state.
The aim of this paper is to consider the old problem of looking for a such formulation of quantum
mechanics, where the system states can be identified with fair probability distributions of measurable
observables only. Such a possibility is based on quantum tomography methods of measuring [6]
quantum states, using the formalism of reconstructing [7,8] the state Wigner function by means of
Radon transform [9].
Wigner introduced the Wigner function [10] W (q, p) of the position q and momentum p that
is similar to the probability density f (q, p) describing the classical particle state in the presence of
fluctuations. The Wigner function can take negative values and, due to this circumstance, it is called
the quasidistribution function. The Wigner function is related to the density matrix ρ( x, x ) of the
quantum particle state by an invertible Fourier transform and contains the same information on the
state as the density matrix.
There exist other analogous quasidistributions like the Husimi–Kano Q-function [11,12] and
Glauber–Sudarshan function [13,14], which are functions on the phase space. The suggestion to
identify quantum states with fair probability densities was presented in [15], where the probability
density, called symplectic tomogram, was used. An analogous approach was elaborated for spin states
in [16,17], where the spin tomograms, being fair probability distributions of spin projections m on
an arbitrary direction in the space given by a unit vector n, were shown to determine the density
matrix ρ(m, m ).
In this paper, on the example of qubits, we show the bijective map of density operators of
spin-1/2 states onto the probability distributions. Since for probability distributions the notion of
Shannon entropy [18], relative entropy, and Tsallis entropy [19] is the standard tool to characterize the
statistical properties of the systems, we obtain, in view of the map introduced, some new relations like
entropic equalities and inequalities for quantum spin states. Other kinds of entropies also exist like
Rényi entropy [20], non-Shannonian and generalized (c, d) entropies; see, e.g., [21,22]. In this paper,
we consider new relations connected with Shannon and Tsallis entropies. In addition, we discuss new
geometric interpretation of spin-1/2 (qubit) states in terms of the Triada of Malevich’s squares [23,24]
and its relation to the Bloch sphere geometry of these states. Employing the identification of qubit
states with probability distributions, we present the construction of quantum observables (Hermitian
2 × 2-matrices) in terms of sets of classical-like variables and provide the bijective map of the qubit
states (density matrices) and observables onto classical-coin probability distributions and classical
observables associated with these coins.
We present the evolution equations for the density matrices of qubit states in the form of kinetic
equations for probability distributions determining the qubit states. We formulate the superposition
principle of qubit state vectors as a new addition rule for the probabilities determining the states.
In addition, we express the Born rule for calculating the probability |ψ1 | ψ2 |2 = w12 as a function of
probabilities determining the pure states | ψ1  and | ψ2 . Then, we extend the probability representation
of qubit states and express the matrix elements of an arbitrary density N ×N-matrix in terms of
classical-coin probability distributions. We consider in detail examples of qutrit (spin-1), identifying the
qutrit state with a set of Triadas of Malevich’s squares. We present new relations of areas of Malevich’s
squares and the possibility of checking these relations in the experiments with superconducting circuits.
The other goal of this work is to study within the probability representation of quantum
states [15–17,25–28] (reviewed in [29]) the triangle geometry of qudit states and discuss Bayes’ formula
for systems without subsystems and correlations (called the hidden correlations) in such systems.
It is worth noting that the classical probability distributions were discussed within the framework of
state vectors for spin-1/2 systems by Khrennikov [30–32] and the superposition principle for spin-1/2
states was expressed as the nonlinear superposition of classical probability distributions in [33,34].
Malevich’s squares and the approach called the suprematism in art are described in [35].
This paper is organized as follows.

86

www.pdfgrip.com
Entropy 2018, 20, 692

We present the notion of random variables in Section 2 and study Bayes’ formula for systems
with one random variable in Section 3. We discuss qubit states in Section 4 and consider classical-coin
random variables for qubit systems in Section 5. Then, we review the notion of quantum suprematism
in Section 6 and study qutrit states in the probability representation in Section 7. We devote Section 8 to
the superposition principle for the probabilities, demonstrating this principle on the example of qutrits.
Within the framework of the probability representation, we formulate the superposition principle for
qudit states in Section 9. Finally, in Section 10, we provide the conclusions and perspectives.

2. Random Variables and Probabilities

In probability theory, the notion of random variables and probability distributions were discussed
using rigorous approaches presented, for example, in [36–39]. We employ here the following empiric
approach. We define the relation of random variables to sets of integer numbers following [40,41].
Given a set of N different events, these events are associated with integers j = 1, 2, . . . , N. We call
relative frequencies P( j) of the realization of these random events in a series of experiments
“the probabilities of the events” where 0 ≤ P( j) ≤ 1. The function P( j) is the probability distribution;
it is normalized ∑ N j=1 P ( j ) = 1.
The properties of the events are characterized by some functions f ( j), which we call observables.
In this approach, random variables are mapped onto the integers j = 1, 2, . . . , N. The physical meaning
of the events can be different; for example, in the casino roulette, the event is the appearance of some
integer number j which is chosen from a set of integer numbers located between 1 and N. The event
may be also considered as positions “UP” and “DOWN” of two coins; in this case, the integer number
j is mapped onto a pair ( a, b) of integer numbers labeling the position of each coin. In both cases,
the relative frequencies of the events can be associated with the integer j, but the interpretation of this
random variable is different. In the case of casino roulette, we say about one random variable, and in
the case of two coins, we have two random variables associated with labeling positions of two coins
by other two integer numbers ( a, b). An analogous approach to random events can be employed in
quantum mechanics.
We extend the above approach to classical probabilities using in this case the identification of
random events with the integers 1 ≤ j, j ≤ N labeling the matrix elements of the density matrix ρ jj
determining the states, e.g., of qudit with spin s, where N = 2s + 1, or of the N-level atom. The physical
observables are given by the Hermitian matrices f jj , where indices of rows and columns are identified
with the random variables 1 ≤ j, j ≤ N. It is important that we can interpret the described above
association of integers j analogously to the case of classical casino roulette and the case of two
classical coins considering numerically the same density matrices ρ jj either as the density matrices of
noncomposite (nondivisible) systems (an analog of the casino roulette) or as the density matrices of
bipartite systems (an analog of the states of two coins).
In the next section, we consider Bayes’ formula, in view of the approach under discussion,
using it for one random variable and applying the map of integer numbers 1, 2, . . . , N onto pairs of
random numbers.

3. Bayes’ Formula for the Probability Distribution of One Random Variable

In this section, we discuss the application of Bayes’ formula available for probability distributions
of several random variables to the case of the probability distribution of one random variable.
First, we recall Bayes’ formula and the notion of conditional probability distribution for statistics
of two random variables. Given the function 1 ≥ P( j, k) ≥ 0, where j = 1, 2, . . . , n1 , k = 1, 2, . . . , n2 ,
and n1 n2 = N, with the normalization condition
n1 n2
∑ ∑ P( j, k) = 1. (1)
j =1 k =1

87

www.pdfgrip.com
Entropy 2018, 20, 692

This function is identified with the probability distribution of two random variables j and k.
The marginal probability distributions
n2 n1
P1 ( j) = ∑ P( j, k), P2 (k) = ∑ P( j, k) (2)
k =1 j =1

determine the statistical properties of each random variable.

The conditional probability distribution of the first random variable j for given k is presented by
the formula; see [36],
P( j, k)
P( j | k) = , (3)
P2 (k )
which means that
P( j, k ) = P2 (k ) P( j | k). (4)

For the case of joint probability distributions of two random variables describing the statistics
of the bipartite system, these relations correspond to Bayes’ formula connecting marginal probability
distributions and conditional probability distributions of these random variables. In Appendix A,
we present an example of application of the above formulas for a particular case N = 4. In view of the
example from Appendix A, we are in the position to formulate the rule for introducing Bayes’ formula
for the probability distribution P (n); N = n1 , n2 of one random variable. We apply the map of integers
n onto pairs of integers j and k, such that j = 1, 2, . . . , n1 and k = 1, 2, . . . , n2 . Then, for marginal
probability distributions and conditional probability distributions, we use the known expression
for joint probability distribution of two variables and define these distributions, in view of the
invertible map of integers 1, 2, . . . , N ↔ (1, 1), (2, 1), . . . , (n1 , 1)(1, 2), (2, 2), . . . , (n1 , 2), . . . , (n1 , n2 ),
where N = n1 n2 . This map can be described by the functions discussed in [42–44].
Following [44], we determine the functions y( x1 , x2 ), x1 (y), and x2 (y), where 1 ≤ x1 ≤ X1 ,
1 ≤ x2 ≤ X2 , and 1 ≤ y ≤ N = X1 X2 , as

y ( x 1 , x 2 ) = x 1 + ( x 2 − 1 ) X1 , (5)
x1 (y) = y mod X1 , 1 ≤ y ≤ N, (6)
y − x1 ( y )
x2 ( y ) − 1 = mod X2 , 1 ≤ y ≤ N. (7)
X1

We use these functions for representing the probability distribution of one random variable as a
joint probability distribution of two random variables. To do this, we introduce in Equations (5)–(7)
the following notation: y ≡ n, x1 ≡ j, x2 ≡ k, X1 ≡ n1 , X2 ≡ n2 , N = n1 n2 = X1 X2 , n = 1, 2, . . . , N,
P( j, k ) ≡ y( x1 , x2 ), and f (y) = P (n). In the case of N = 4 and n1 = n2 = 2, the map introduced
just provides the relations P (1) = P(1, 1), P (2) = P(2, 1), P (3) = P(1, 2), and P (4) = P(2, 2)
discussed above. Nevertheless, the functions introduced describe the invertible map of the probability
distribution of one random variable P (n); n = 1, 2, . . . , N onto the joint probability distribution P( j, k )
of two random variables j = 1, 2, . . . , n1 and k = 1, 2, . . . , n2 , with N = n1 n2 , for arbitrary integers n1
and n2 . In our new notation, n = n( j, k ), j = j(n), and k = k(n). Taking into account this discussion,
we introduce Bayes’ formula for the probability distribution P (n) of one random variable; it reads
, -
, - P n( j, k)
P j(n) | k(n) = , -, n = n1 n2 , (8)
∑nj=1 P n( j, k )

where functions n( j, k ), j(n), and k(n) are constructed in [44].

The relation of the joint probability distribution P( j, k ) to the marginal probability distributions
corresponds to the presence of correlations in the system with two random variables. Since we
introduced an analog of two random variables and their marginal and conditional probability

88

www.pdfgrip.com
Entropy 2018, 20, 692

distributions, the relation of these distributions reflect correlations, which we called [45] the
hidden correlations for systems without subsystems. Such correlations exist for both classical and
quantum systems.
Bayes’ formula can also be considered for the probability distribution of one random variable P (n),
if the integer n = 1, 2, . . . , N and N = n1 n2 n3 , where n1 , n2 , and n3 are integers of the joint
probability distribution of three random variables P( j, k, l ), with j = 1, 2, . . . , n1 , k = 1, 2, . . . , n2 ,
and l = 1, 2, . . . , n3 . To do this, we use an analogous invertible map [44] of functions y( x1 , x2 , x3 ), x1 (y),
x2 (y), and x3 (y), taking integer values y = 1, 2, . . . , N = X1 X2 X3 ; 1 ≤ xi ≤ Xi ; i = 1, 2, 3, defined by
the relations

y ( x 1 , x 2 , x 3 ) = x 1 + ( x 2 − 1 ) X1 + ( x 3 − 1 ) X1 X2 , (9)
x1 (y) = y mod X1 , (10)
y − x1 ( y )
x2 ( y ) − 1 = mod X2 , (11)
X1
y − x 1 ( y ) − x 2 ( y ) X1
x3 ( y ) − 1 = mod X3 . (12)
X1 X2

After substitution x1 ≡ j, x2 ≡ k, x3 ≡ l, and y ≡ n, we arrive at an analog of Bayes’ formula for

one random variable , -
, - P j ( n ), k ( n ), l ( n )
P j ( n ) | k ( n ), l ( n ) = n1 , - . (13)
∑ j=1 P n( j, k, l )
To illustrate this formula, we consider the example of N = 8 = 2 · 2 · 2, i.e., n1 = n2 = n3 = 2;
the map of integers reads 1 ↔ (1, 1, 1), 2 ↔ (2, 1, 1), 3 ↔ (1, 2, 1), 4 ↔ (2, 2, 1), 5 ↔ (1, 1, 2),
6 ↔ (2, 1, 2), 7 ↔ (1, 2, 2), 8 ↔ (2, 2, 2). This means that the probability distribution P (n)
takes the values P (1) ≡ P(1, 1, 1), P (2) ≡ P(2, 1, 1), P (3) ≡ P(1, 2, 1), P (4) ≡ P(2, 2, 1),
P (5) ≡ P(1, 1, 2), P (6) ≡ P(2, 1, 2), P (7) ≡ P(1, 2, 2), and P (8) ≡ P(2, 2, 2). The joint probability
distribution P( j, k, l ) has the values given by numbers P (n), and Bayes’ formula obtained provides,
e.g., the conditional probability

, - P(1, 1, 1) P (1)
P j(n) = 1 | k (n) = 1, l (n) = 1 = = .
P(1, 1, 1) + P(2, 1, 1) P (1) + P (2)

In the quantum case, the map of integers discussed provides a tool to consider the density matrix
of qudit state ρnn , where n, n = 1, 2, . . . , N, as the density matrix of a multipartite system. For example,
at N = 4, the ququart density matrix can be interpreted as the density matrix of two two-level atoms,
using the map discussed. In fact, if n, n = 1, 2, 3, 4, we consider the density matrix as ρnn ≡ ρ jk,j k ,
where j, j , k, k = 1, 2. Formally, we obtain the density matrix of the two-qubit system, which has the
same numerical matrix elements that the 4×4-matrix ρnn ; this means that all numerical properties of
the density matrix of two-qubit state and ququart state are identical.
This fact provides the possibility to consider formal entanglement properties of ququart system.
For example, if we consider the pure state, (ρ2 )nn = ρnn , then the properties of linear entropy
, - , -
S = 1 − Tr ρ2 (1) , with ρ(1) jj = ∑2k=1 ρ jk, j k , where the indices j, k, j are determined by the
numbers n, n = 1, 2, 3, 4 according to the discussed map, characterize the entanglement degree in the
bipartite system.
For the four-level atom, one has the same numerical characteristics. From the viewpoint of the
matrix properties, the ququart state with the density matrix, having only different from zero matrix
elements ρ11 = ρ14 = ρ41 = ρ44 = 1/2, provides the linear entropy S = 1/2 corresponding to
maximum entangled state of two qubits. The interpretation of this phenomenon for systems without
subsystems is the presence of hidden correlations in the degrees of freedom of such systems, formally
analogous to quantum correlations associated with the entanglement phenomenon, e.g., in bipartite
systems of two qubits.

89

www.pdfgrip.com
Entropy 2018, 20, 692

4. Probability Representation of Spin-1/2 States

We start our introduction of the probability representation of quantum system states with the
consideration of spin-1/2 systems. These systems realize qubits and their states, as well as they are
realized by two-level atom systems. In standard formulation of quantum mechanics, the spin-1/2
pure states are described
 by Pauli spinors, which are complex vectors | ψ with two components,
ϕ1
i.e., | ψ = . The vectors | ψ belong to the two-dimensional Hilbert space H with the
ϕ2
scalar product
(1)∗ (2) (1)∗ (2)
 ψ (1) | ψ (2)  = ϕ 1 ϕ1 + ϕ2 ϕ2 . (14)

The state vectors are normalized ψ | ψ = 1, and | ϕ1 |2 

+ | ϕ2 |2 = 1. Thedensity operators [1,2]
ϕ1∗ ϕ1 ϕ1∗ ϕ2
of the pure states ρψ =| ψψ | in matrix form read ρψ = . This matrix has the
ϕ2∗ ϕ1 ϕ2∗ ϕ2
properties of Hermiticity ρ†ψ = ρψ and nonnegativity ρψ ≥ 0, as well as it has the unit trace Tr ρψ = 1.
The physical meaning of the state-vector components ϕ1 and ϕ2 and matrix elements of the
density matrix ρ(ψ) is determined by the relation of these values to operators of physical observables
associated with spin projection operators h̄σx /2, h̄σy /2, and h̄σz /2 onto the axes x, y, and z, respectively.
Here, the Pauli matrices σx , σy , and σz are
     
1 1 0 −i 1 0
σx = , σy = , σz = , (15)
1 0 i 0 0 −1

and h̄ is the Planck constant. In this paper, we use dimensionless units and assume h̄ = 1. Three
normalized eigenstates of the matrices σx /2, σy /2, and σz /2 with eigenvalues +1/2 have the form
     
1 1 1 1 1
| ψx  = √ , | ψy  = √ , | ψz  = . (16)
2 1 2 i 0

The vectors are identified with spin-1/2 states, in which the spin projections on the axes x, y,
and z are equal to +1/2. The corresponding density matrices read
     
1/2 1/2 1/2 −i/2 1 0
| ψx ψx |= , | ψy ψy |= , | ψz ψz |= . (17)
1/2 1/2 i/2 1/2 0 0
 
ρ11 ρ12
Any density matrix describing mixed state of the spin-1/2 system ρ = , such that
ρ21 ρ22
ρ† = ρ, Tr ρ = 1, and ρ ≥ 0 (i.e., the matrix has nonnegative eigenvalues), is determined by three
real parameters. The physical meaning of these parameters can be clarified, if one considers the
probabilities to obtain in the state with the density matrix ρ the spin projections +1/2 on the axes x, y,
z, which we denote as p1 , p2 , and p3 , respectively. The probabilities p1 , p2 , and p3 play a fundamental
role in describing the spin-1/2 states and, as we show, they determine the density matrix of this system.
These probabilities are given by the Born rule as follows:
, -
p1 = Tr (ρ | ψx ψx |) , p2 = Tr ρ | ψy ψy | , p3 = Tr (ρ | ψz ψz |) . (18)

The spin tomogram w(m|n) introduced in [16,17], being equal to the conditional probability of
spin projection m = ±1/2 onto the direction given by the unit vector n, is expressed in terms of the
probability vector p = ( p1 , p2 , p3 ) [23], i.e.,
, -
w(m|n) = (1/2) + m(p − p0 )n, p0 = 1/2, 1/2, 1/2 . (19)

90

www.pdfgrip.com
Entropy 2018, 20, 692

In view of relations (18) employed as the equations for matrix elements of ρ, it is not difficult
to rewrite the density matrix ρ in the form where its matrix elements are expressed in terms of the
probabilities p1 , p2 , and p3 [23,24,46,47]; we have
 
p3 p1 − (1/2) − i ( p2 − 1/2)
ρ= . (20)
p1 − (1/2) + i ( p2 − 1/2) 1 − p3

The standard parameters of the Bloch sphere of qibit states x1 , x2 , and x3 are connected with the
probabilities through the bijective map xk = 2pk − 1; k = 1, 2, 3.
If the spin-1/2 state is the pure state, its density matrix satisfies the constraint ρ2 = ρ that provides
the condition for probabilities

( p1 − 1/2)2 + ( p2 − 1/2)2 = p3 (1 − p3 ). (21)

In this case, the Pauli spinor of the pure state | ψ can also be expressed in terms of the three probabilities
satisfying condition (21), i.e.,
⎛ √ ⎞
p3
⎜ ⎟
| ψ = ⎝ p1 − 1/2 p − 1/2 ⎠ . (22)
√ + i 2√
p3 p3

For mixed states, the nonnegativity condition of the density matrix (nonnegativity condition for
its eigenvalues) yields the inequality for the probabilities p1 , p2 , and p3 ; it reads

( p1 − 1/2)2 + ( p2 − 1/2)2 + ( p3 − 1/2)2 ≤ 1/4. (23)

As we see, all information on the spin-1/2 state density matrix (and its Pauli spinor describing the
pure state | ψ) is identified with three probabilities 0 ≤ p1 , p2 , p3 ≤ 1 satisfying inequality (23).
This observation provides the possibility to consider again very old problem of quantum
mechanics, namely: Is it possible to formulate the notion of quantum states employing only ingredients
of classical probability theory of systems with fluctuations, such as the probability distributions?
We observed that for spin-1/2
 systems  it is enough
 to have
 three probability
 distributions
 given
p 1 p 2 p 3
by the probability vectors P1 = P2 =
, P3 =
, and  . Inequality (23)
1 − p1 1 − p2 1 − p3
is the only one quantum condition which should be respected by the probabilities. Thus, instead of
vectors | ψ and density matrices ρ, we can introduce the notion of spin-1/2 states, ⎛ employing
⎞ the set
p1
of three probability distributions or identify the state with the vector P  = ⎜ ⎟
⎝ p2 ⎠. This means
p3
that all quantum phenomena like, e.g., quantum interference, can be described in terms of the
probabilities. Here, it worth noting that the interference of classical probabilities was discussed
in [30–32]. For example, the superposition principle of quantum states, expressed in terms of
normalized and orthogonal state vectors | ψ1  and | ψ2  by the equality
( (
| ψ = Π3 | ψ1  + 1 − Π3 eiξ | ψ2 , (24)

where | ψ is again the state vector, can be formulated as “superposition” of probabilities.

We present the result in the form of a nonlinear addition of two vectors

 (1) ⊕ P
P  (2) = P
 (3) , (25)

91

www.pdfgrip.com
Entropy 2018, 20, 692

(k) (k) (k)

where components of the vectors are the probabilities p1 , p2 , and p3 ; k = 1, 2, 3, satisfying
⎞ value of k. The notation of addition ⊕ is also associated with the probability
equality (21)⎛for each
Π1
 =⎜
vector Π
⎟
⎝ Π2 ⎠, where the components 0 ≤ Π1 , Π2 , Π3 ≤ 1 satisfy equality (21). These three
Π3
probabilities are related to the superposition parameters as follows:

Π1 − 1/2 Π2 − 1/2
cos ξ = ( , sin ξ = ( . (26)
Π3 (1 − Π3 ) Π3 (1 − Π3 )

 (3) are functions of the three probability vectors P

The three components of the vector P  1, P
 2,

and Π; they read [23,24,48]

(3) (1) (2) (2) (2)
p3 = Π3 p3 + (1 − Π3 ) p3 + 2 p3 p3 (Π1 − 1/2) , (27)
(3) (1) (2)
p1 − 1/2 = Π3 ( p1 − 1/2) + ( p1
− 1/2)(1 − Π3 )
& '
(1) (1)
(2) (1)
+ (Π1 − 1/2)( p1 − 1/2) + (Π2 − 1/2)( p2
− 1/2) p3 /p3
& ' 
(2) (2) (1) (2)
+ (Π1 − 1/2)( p1 − 1/2) − (Π2 − 1/2)( p2 − 1/2) p3 /p3 , (28)

(3)
and the third one p2 is determined in view of Equation (21).
   
1 1 1 1
As an example of the superposition of vectors √ =| ψ1  and √ =| ψ2 
2 1 2 −1
given by (24) and described by the probabilities (1, 1/2, 1/2) and (0, 1/2, 1/2), respectively, we obtain
(3) (3)
p3 = Π1 and p1 = Π3 .
(3) (3) (3)
One can check that the obtained numbers p1 , p2 , and p3 are nonnegative; they satisfy
(3) (3)
equality (21), which determines the probabilities and (1 − p2 p2 ).
The unitary evolution of the
probabilities p1 , p2 , and p3 is described by the following transform of the matrix ρ (20)

ρ → uρu† , (29)

where the unitary 2×2 matrix u is such that u† = u−1 and

 
−iHt u11 (t) u12 (t)
u(t) = e = , (30)
u21 (t) u22 (t)
 
H11 H12
with the Hamiltonian H = . From this evolution, which corresponds to the
H21 H22
von Neumann equation
∂ρ
+ i [ H, ρ] = 0, (31)
∂t
the evolution formula for the probabilities follows; it reads
⎛ ⎞ ⎛ ⎞
p3 ( t ) p3
⎜ p (t) − (1/2) − i ( p (t) − 1/2) ⎟ ⎜ p − (1/2) − i ( p − 1/2) ⎟
⎜ 1 ⎟ ⎜ ⎟
⎟ = [u(t) ⊗ u∗ (t)] ⎜
2 1 2
⎜ ⎟. (32)
⎝ p1 (t) − (1/2) + i ( p2 (t) − 1/2) ⎠ ⎝ p1 − (1/2) + i ( p2 − 1/2) ⎠
1 − p3 ( t ) 1 − p3

92

www.pdfgrip.com
Entropy 2018, 20, 692

This result is the solution of Equation (31) written as the kinetic equation for probabilities p1 (t),
p2 (t), and p3 (t). Equation (32) describes the temporal evolution of the initial probabilities p1 , p2 , and
p3 , which convert at time t to probabilities p1 (t), p2 (t), and p3 (t) satisfying relation (23).

5. Quantum Observables and Classical-Coin Random Variables for Qubit Systems

The
 quantum observable for spin-1/2 system is described by the Hermitian matrices
A11 A12
A= . It is possible [34] to consider quantum statistical properties of this observable in
A21 A22
the probability representation associating the matrix elements A jk with classical-like random variables.
Introducing the notation A12 = x − iy, A11 = z1 , and A22 = z2 , we can rewrite the mean value of the
observable  A in the form
 A = Tr ρA =  x cl + ycl + zcl , (33)

where

 x cl = p1 x + (1 − p1 )(− x ), ycl = p2 y + (1 − p2 )(−y), zcl = p3 z1 + (1 − p3 )z2 . (34)

This form shows that the mean values of the observable A calculated using the standard formalism
of quantum mechanics provide the connection with classical-like means of three dichotomic random
variables xcl , ycl , and zcl , employing the values ( x, − x ), (y, −y), and (z1 , z2 ), respectively.
The probability distributions for these values are given as P  1 = ( p1 , 1 − p1 ), P
 2 = ( p2 , 1 − p2 ),
and P3 = ( p3 , 1 − p3 ), which determine the density matrix ρ of the qubit state. The quantum observable

in the quantum suprematism representation has a classical analog.
We point out that the highest moments of the observable A given as  An  = Tr ρAn , n = 2, 3, . . .
are also expressed in terms of classical random variables xcl , ycl , and zcl , and the expressions reflect
quantum correlations of classical-like random observables due to the nonnegativity condition of the
density matrix (23). One can also associate the matrix elements of arbitrary qudit observables with
artificial-coin random variables.

6. Quantum Suprematism Representation

The relations described in the previous section can be illustrated in the quantum suprematism
picture [23,24,48], where the probabilities p1 , p2 , and p3 determine the Triada of Malevich’s squares.
We construct a triangle with vertices A1 , A2 , and A3 ,√which are located on three simplexes—sides
of equilateral triangle with the side length equal to 2, and three squares (black, red, and white)
determined by the sides of the triangle (Figure 1).
We call these squares the Triada of Malevich’s squares following [23,24,34,46,47]. The areas of the
squares S A1 ,A2 , S A2 ,A3 , and S A3 ,A1 are

S A1 ,A2 = 2 + 2p22 − 4p2 − 2p3 + 2p23 + 2p2 p3 ,

S A2 ,A3 = 2 + 2p23 − 4p3 − 2p1 + 2p21 + 2p3 p1 , (35)
S A3 ,A1 = 2 + 2p21 − 4p1 − 2p2 + 2p22 + 2p1 p2 .

The sum of the areas of Malevich’s squares, being the function of probabilities p1 , p2 , and p3 , reads
& '
S = 2 3 + 2 p21 + p22 + p23 − 3 ( p1 + p2 + p3 ) + p1 p2 + p2 p3 + p3 p1 . (36)

The map of the Bloch sphere parameters xk ; k = 1, 2, 3, onto the probabilities pk = ( xk + 1)/2 can
be used to express the area S in terms of the parameters xk .

93

www.pdfgrip.com
Entropy 2018, 20, 692

√
Figure 1. The equilateral triangle with a side length equal to 2. Each of three sides is simplex
corresponding to the coin probabilities pk and (1 − pk ) satisfying the relation pk + (1 − pk ) = 1;
k = 1, 2, 3. Here, the points on simplexes with probabilities p1 , p2 , and p3 determine the triangle
A1 A2 A3 (on the left). The three squares (black, red, and white) called Triada of Malevich’s squares,
which are constructed using the sides of the triangle A1 A2 A3 . The squares are in the one-to-one
correspondence with the density matrix of qubit (spin-1/2) states (on the right).

For quantum states, the sum satisfies the inequality S ≤ 3 [47]. For classical-coin states, the sum
can take maximum value Scl = 6. In view of this fact, the quantization condition (23) provides the
possibility to clarify the difference of classical and quantum properties of the systems, which states are
illustrated in the quantum suprematism representation by Triadas of Malevich’s squares.
The quantum states of spin-1/2 systems are described by spin tomogram w(m | n) [16], and the
spin tomogram determines the density matrix of the spin state [17]. The spin tomogram can be
expressed by the probabilities p1 , p2 , and p3 ; the tomogram is the probability to obtain the spin
projection m onto the direction in space determined by the unit vector n = (sin θ cos ϕ, sin θ sin ϕ, cos θ ).
For spin-1/2, the spin projection m takes two values⎛ ±
⎞1/2, and the ⎛ tomogram
⎞ (19) provides
p1 1/2
⎜ ⎟ ⎜ ⎟
w(m = +1/2 | n) = n(p − p0 )) + 1/2, where p = ⎝ p2 ⎠ and p0 = ⎝ 1/2 ⎠. The tomogram
p3 1/2
w(m | n) can be interpreted as the conditional probability distribution and, in view of this fact, one can
obtain a new entropic inequality associated with this distribution. For example, the relative Tsallis
entropy for two distributions w(m | n1 ) and w(m | n2 ) satisfies the inequality, which yields the new
condition for probabilities,

(1 − q)−1 [n1 (p − p0 ) + 1/2]q [n2 (p − p0 ) + 1/2]1−q
 (37)
+ [(1/2) − n1 (p − p0 )]q [(1/2) − n2 (p − p0 )]1−q − 1 ≥ 0.

In the limit q → 1, this inequality provides the nonnegativity condition for the von Neumann
relative entropy
 
[n1 (p − p0 ) + 1/2] ln [n1 (p − p0 ) + 1/2] [n2 (p − p0 ) + 1/2]−1
  (38)
+ [(1/2) − n1 (p − p0 )] ln [(1/2) − n1 (p − p0 )] [(1/2) − n2 (p − p0 )]−1 − 1 ≥ 0.

94

www.pdfgrip.com
Entropy 2018, 20, 692

The obtained new entropic inequalities for probabilities p1 , p2 , and p3 can be checked in the
experiments with superconducting qubits, as well as the maximum value S = 3 for the sum of areas of
Malevich’s squares.

7. Qutrit States in the Probability Representation

One can extend the consideration of suprematism ⎛ representation⎞to the case of any qudits,
ρ11 ρ12 ρ13
⎜ ⎟
for example, qutrit states with the density matrix ρ = ⎝ ρ21 ρ22 ρ23 ⎠. Using the tool [23,24,47]
ρ ρ32 ρ33
  31
ρ 0
of embedding this matrix into 4×4-matrix R = , one can obtain three qubit-state density
0 0
matrices applying the partial tracing procedure. These three-qubit density matrices are
     
ρ11 + ρ22 ρ13 ρ11 + ρ33 ρ12 ρ22 ρ23
ρ (1) = , ρ (2) = , ρ (3) = . (39)
ρ31 ρ33 ρ21 ρ22 ρ32 ρ11 + ρ33

Since these matrices can be expressed in terms of probabilities

 
(k) (k) (k)
p3 p1 − (1/2) − i ( p2 − 1/2)
ρ(k) = (k) (k) (k) ; k = 1, 2, 3, (40)
p1 − (1/2) + i ( p2 − 1/2) 1 − p3

(k)
the qutrit density matrix elements can also be expressed in terms of these probabilities p1,2,3 . In fact,
we arrive at
⎛ (1) (2) (2) (2) (1) (1) ⎞
p3 + p3 − 1 p1 − (1/2) − i ( p2 − 1/2) p1 − (1/2) − i ( p2 − 1/2)
⎜ ⎟
ρ = ⎝ p1(2) − (1/2) + i ( p2(2) − 1/2) 1−
(2)
p3
(3)
p1
(3)
− (1/2) − i ( p2 − 1/2) ⎠ ; (41)
(1) (1) (3) (3) (1)
p1 − (1/2) + i ( p2 − 1/2) p1 − (1/2) + i ( p2 − 1/2) 1 − p3

(k)
in other notation, Equation (41) is given in Appendix B. The probabilities p1,2,3 satisfy inequality (23)
for k = 1, 2, 3 and also the nonnegativity condition for det ρ ≥ 0. The expression of matrix elements of
the qutrit density matrix in terms of the probability distributions follows a new entropic inequality for
the matrix elements:
.& 'q & ' 1− q
2 ( ρ13 + ρ31 + 1) 2 ( ρ13 + ρ31 + 1)
1 1 1
1− q
& 'q & ' 1− q / (42)
+ 12 (−ρ13 − ρ31 + 1) 2 (1 − ρ23 − ρ32 )
1
− 1 ≥ 0.

This new inequality for the qutrit state comes from applying the nonnegativity condition of
the Tsallis entropy expressed in terms the qubit-state tomogram by Equations (37) and (38) to the
qutrit-state density matrix.
Now, we derive the other inequality for the probabilities determining the qutrit state (41).
First, we construct the qubit density matrices following [49]
 
(2) (2) (2)
p3 p1 − (1/2) − i ( p2 − 1/2)
ρ (2) = (2) (2) (2) , (43)
p1 − (1/2) + i ( p2 − 1/2) 1 − p3
 
(1) (1) (1)
p3 p1 − (1/2) − i ( p2 − 1/2)
ρ (2) = (1) (1) (1) . (44)
p1 − (1/2) + i ( p2 − 1/2) 1 − p3

95

www.pdfgrip.com
Entropy 2018, 20, 692

(k)
The subadditivity condition provides the inequality for the probabilities p1,2,3 , k = 1, 2, 3; it reads

− Tr ρ(1) ln ρ(1) − Tr ρ(2) ln ρ(2) ≥ −Tr (ρ ln ρ), (45)

where ρ, being given by (41), is determined by the probabilities.

In addition, the probabilities determining the qutrit states satisfy the Tsallis entropic inequality
for the distance between the states
1
Tr ρ(1)q ρ(2)1−q − 1 ≥ 0. (46)
1−q

Inequalities (42), (45), and (46) are compatible with the nonnegativity condition of the qutrit
density matrix.
Since the qubit density matrices (43) and (44) are obtained from the same qutrit density matrix,
the distance between these qubit states characterizes the hidden correlations between the artificial
qubits associated with the qutrit density matrix (41). The inequality can be checked in the experiments
where the tomography of qutrit states is performed, e.g., in the experiments with superconducting
circuits based on Josephson junction devices [50,51].
Since the qutrit states are described by the probabilities determining the states of three artificial
qubits, the density matrix (39) can be mapped onto the set of three Triadas of Malevich’s squares.

8. Pure Qutrit States and Their Superposition in the Probability Representation

In [52,53], the superposition principle of quantum states was formulated as a nonlinear addition
rule of the pure-state density operators. Namely, given two density operators ρ̂1 =| ψ1 ψ1 | and
ρ̂2 =| ψ2 ψ2 | satisfying the conditions ρ̂21 = ρ̂1 , ρ̂22 = ρ̂2 , Tr ρ̂1 = Tr ρ̂2 = 1, and ρ̂1 ρ̂2 = 0. Then,
for arbitrary real numbers 0 ≤ λ1 , λ2 ≤ 1; λ1 + λ2 = 1 and the density operator ρ̂0 =| ψ0 ψ0 |;
ρ̂20 = ρ̂0 , the state with the density operator ρ̂ψ of the form

( ρ̂1 ρ̂0 ρ̂2 + ρ̂2 ρ̂0 ρ̂1

ρ̂ψ = λ1 ρ̂1 + λ2 ρ̂2 + λ1 λ2 ( (47)
Tr (ρ̂1 ρ̂0 ρ̂2 ρ̂0 )

satisfies the conditions ρ̂†ψ = ρ̂ψ , ρ̂2ψ = ρ̂ψ , and Tr ρ̂ψ = 1.

The nonlinear addition rule (47) corresponds to the interference formula of two orthogonal pure
√ √
states | ψ1  and | ψ2  of the form | ψ = λ1 | ψ1  + eiϕ λ2 | ψ2 , where the phase ϕ is coded by an
artificial density operator ρ̂0 .
One can use generic Equation (47) to formulate the superposition rule of qudit states expressed
in terms of the probabilities; we obtain such a formula for two qutrit states. For this, we introduce
three probability vectors, i.e., three probability distributions Π  1 = ( Π1 , 1 − Π1 ), Π
 2 = ( Π2 , 1 − Π2 ),
and Π 3 = (Π3 , 1 − Π3 ), where 0 ≤ Π1 , Π2 , Π3 ≤ 1, and the phase ϕ is defined by the relations (26),

Π1 − 1/2 Π2 − 1/2
cos ϕ = ( , sin ϕ = ( . (48)
Π3 (1 − Π3 ) Π3 (1 − Π3 )

This means that we arrive at the condition (Π1 − 1/2)2 + (Π2 − 1/2)2 + (Π3 − 1/2)2 = 1/4
equivalent to the condition (21). To obtain the superposition rule for qutrit states in the probability
representation, we employ the expression of density matrix (41), introduce three 8-vectors P  1, P
 2,
and P
 0 of the form (see Appendix B):

 1 = p (I) , p (I) , p (I) , p (I) , p (I) , p (I) , p (I) , p (I) ,

P 1 2 3 4 5 6 7 8

 2 = p(II) , p(II) , p(II) , p(II) , p(II) , p(II) , p(II) , p(II) ,

P (49)
1 2 3 4 5 6 7 8

 0 = p(III) , p(III) , p(III) , p(III) , p(III) , p(III) , p(III) , p(III) ,

P 1 2 3 4 5 6 7 8

96

www.pdfgrip.com
Entropy 2018, 20, 692

and identify three qutrit states ρ̂1 , ρ̂2 , and ρ̂3 with these vectors P  1, P 2 , and P 0 , since the density
matrices are determined by their components. Formula (47) provides the dependence of the probability
vector P  ψ , which determines the state ρ̂ψ , on the probabilities P  1, P 2, P 0 , and Π
 1,2,3 . The density
matrix (41) of the qutrit pure state satisfies the condition ρ2 = ρ, which provides the formula for the
probabilities following from the equality of the matrix elements ∑k ρ jk ρkm = ρ jm written in terms
of probabilities.
Thus, we formulated the result for qutrits. We rewrite Equation (47) expressing the matrices ρ̂ψ ,
ρ̂1 , ρ̂2 , and ρ̂0 in terms of probabilities. The equality of matrix elements provides the expression of the
probability vectors P  ψ as functions of the probability vectors P
 1, P
 2, P
 0 , and Π1,2,3 . The approach can
be extended to other qudit states.

9. Probability Representation of the Density N ×N-Matrix of the Qudit State

We generalize Equation (41) and write the matrix element ρ jk for an arbitrary qudit state (N-level
atom state) in the form [48]:

( jk) ( jk)
ρ jk = p1 − (1/2) − i ( p2 − 1/2), k > j,
( jj)
ρ jj = 1 − p3 , j ≥ 2, (50)
N
ρ22 = 1 − ∑ ρ jj ,
j =2

( jk) ( jk)
where p1 and p2 are the probabilities for artificial qubits (spin-1/2 states) to have spin projections
m = +1/2 onto the x and y axes, respectively. The diagonal matrix elements ρ jj depend on
( jj)
probabilities p3 , which are probabilities of artificial spin projections m = +1/2 on the z-axis. Thus,
all the matrix elements of the qudit density matrix are expressed in terms of probabilities for ( N 2 − 1)
classical coins to have positions “UP” or “DOWN.” The new inequality (42) is also valid for the qudit
density matrix (50) as well as the entropic inequalities for the matrix elements where 1 → j and 3 → k.
The superposition principle of pure states given by relation (47) for the density matrices provides
the connection of the probabilities determining two orthogonal pure states ρ̂1 and ρ̂2 and the density
matrix ρ̂0 with the probabilities determining the pure superposition state ρ̂ψ . In spite of the fact
that the formulas are cumbersome, their existence demonstrates that such quantum phenomenon as
interference of quantum states for arbitrary N-level atoms can be formulated as nonlinear superposition
of classical probabilities (cf. [30–32]). Thus, for any qudit state, one has relationships connecting the
probability distributions determining the qudit states in view of generic formula (47).

10. Conclusions
To conclude, we point out the main results of our study.
We introduced the notion of hidden correlations for systems without subsystems, using the
explicitly written functions (5)–(7) and (9)–(12) providing the invertible map of the integers in both
classical and quantum domains. This approach provides the possibility to write Bayes’ formula and
introduce the conditional probability distribution for given probability distribution of one random
variable (22). Using the probability description of qubit states, we presented the solution of the von
Neumann equation for the two-level atom as the transform of probabilities determining the state
density matrix (35).
We obtained new inequality for the qubit-state probabilities, which can be checked experimentally
as the nonnegativity condition of classical relative Tsallis entropy—Equation (38). The properties
of generalized entropies discussed and employed in [21,22] can be also studied in view of the
probability distributions determining the matrix elements of the density matrix. We will do this
in future publications.

97

www.pdfgrip.com
Entropy 2018, 20, 692

The tomographic reconstruction of the density matrix, e.g., in the experiments with
superconductive circuits [50,51], provides the possibility to find the probabilities that should satisfy
the inequalities.
In view of a generic form of the superposition principle formulated in terms of density
matrices (47), we presented the general approach to get the addition rule for classical probabilities
determining the qutrit pure state. In this sense, we considered two related problems. One problem
is to reformulate the standard description of quantum mechanics by means of wave functions and
density matrices, using the Hilbert space formalism, in terms of classical probabilities; this can be
done within the framework of quantum tomography [15–17,29]. The inverse approach is to consider
classical probabilities associated with quantumlike objects in Hilbert spaces [30–32]. In this paper,
we concentrate on the presentation of the first problem and show that the quantum formalism of
Hilbert spaces can be mapped bijectively onto the classical-like formalism of the probability theory
and geometry of simplexes. New results obtained in this approach are the nonlinear addition rules of
probabilities, giving the probabilities. Such rules correspond either to Born rules of quantum mechanics
or the superposition principle of pure states of qudits.
In addition, we obtained an explicit form of the unitary evolution for probabilities determining
the probabilities in terms of their transform by means of unitary matrices for two-level atoms.
To illustrate the map of Bloch sphere parametrization of qubit states onto the probability
representation of the states, where the probabilities satisfy the quantum constraint inequalities,
we employed the geometric representation of the probabilities, in view of the suprematism picture of
Triadas of Malevich’ squares. The approach developed and its properties in the case of generic qudit
states will be elaborated in the future publication.
In addition, we point out that the tomographic-probability approach was applied in signal
analysis [54,55] due to the descriptions of the signals by an analog of the Wigner function proposed by
Ville [56]. The probability properties considered above can be used in the signal theory as well.

Author Contributions: Both authors contributed equally to the conception, design, and methodology of this study.
Both equally to the analysis of the results and the conclusions. M.A.M. wrote the first draft of the manuscript.
Both authors contributed equally to final writing of the manuscript.
Funding: Some results presented in this paper were reported at the 19th Växjö Conference on Quantum
Foundations “Towards Ultimate Quantum Theory” (Linnaeus University, Växjö, Sweden, June 11–14, 2018),
and we thank the Organizers and especially Professor Khrennikov for invitation and kind hospitality. The work
of V.I.M. was partially performed at the Moscow Institute of Physics and Technology, where V.I.M. was supported
by the Russian Science Foundation under Project No. 16-11-00084. Also, V.I.M. acknowledges the partial support
of the Tomsk State University Competitiveness Improvement Program.
Conflicts of Interest: The authors declare no conflict of interest.

Appendix A
We consider a simple example of N = 4 and n1 = n2 = 2. We have the distribution P( j, k) of
four numbers 1 ≥ P(1, 1), P(1, 2), P(2, 1), P(2, 2) ≥ 0. In view of Equations (2) and (3), we have the
conditional probabilities:

P(1, 1) P(2, 1)
P (1 | 1) = , P (2 | 1) = ,
P(1, 1) + P(2, 1) P(1, 1) + P(2, 1)
P(1, 2) P(2, 2)
P (1 | 2) = , P (2 | 2) = .
P(1, 2) + P(2, 2) P(1, 2) + P(2, 2)

Numbers P(1 | 1), P(2 | 1), P(1 | 2), and P(2 | 2) are nonnegative and normalized,
∑2j=1 P( j | k ) = 1.
Now, we consider the example of the same four nonnegative numbers, which we denote as P (n),
with n = 1, 2, 3, 4, and define the numbers as P (1) = P(1, 1), P (2) = P(2, 1), P (3) = P(1, 2),
and P (4) = P(2, 2). The set of numbers P (n) can be interpreted as the probability distribution

98

www.pdfgrip.com
Entropy 2018, 20, 692

of one random variable n. Since Bayes’ formula is determined through only the numbers P( j, k ),
we can apply it using the invertible map of integers 1 ↔ 1, 1, 2 ↔ 2, 1, 3 ↔ 1, 2, 4 ↔ 2, 2
and introducing two “marginal probability distributions” P1 (1) = P(1, 1) + P(1, 2) = P (1) + P (3),
P1 (2) = P(2, 1) + P(2, 2) = P (2) + P (4) and P2 (1) = P(1, 1) + P(2, 1) = P (1) + P (2),
P2 (2) = P(1, 2) + P(2, 2) = P (3) + P (4). After this, we can introduce the “conditional probability
distributions” in view of the definition given by the same relations (3) and (4) but rewritten in terms of
numbers P (n), with n = 1, 2, 3, 4. We define the conditional probabilities

P (1) P (2)
P (1 | 1) = , P (2 | 1) = ,
P (1) + P (3) P (1) + P (3)
P (3) P (4)
P (1 | 2) = , P (2 | 2) = .
P (2) + P (4) P (2) + P (4)

Appendix B
To express the qutrit density matrix ρ in the probability representation, we employ the 8-vector
P
 = ( p1 , p2 , p3 , p4 , p5 , px6 , p7 , p8 ) and obtain
⎛ ⎞
p3 + p6 − 1 p4 − (1/2) − i ( p5 − 1/2) p1 − (1/2) − i ( p2 − 1/2)
⎜ ⎟
ρ = ⎝ p4 − (1/2) + i ( p5 − (1/2) 1 − p6 p7 − (1/2) − i ( p8 − (1/2) ⎠ . (A1)
p1 − (1/2) + i ( p2 − 1/2) p7 − (1/2) + i ( p8 − 1/2) 1 − p3

We have the probability p9 = 1 − p6 , and nine probability distributions, corresponding to

classical-coin positions, read p j = ( p j , 1 − p j ); j = 1, 2, . . . , 8 and p9 = ( p9 , 1 − p9 ) = (1 − p6 , p6 ).
Thus, the density matrix of qutrit state formally corresponds to the probability distributions describing
positions "UP" and "DOWN" of nine classical coins. In our notation, the ninth coin probability
distribution is completely correlated with the sixth coin position, i.e., it is determined by the probability
distribution of the sixth coin. The other correlations are described by nonnegativity conditions for
the density-matrix eigenvalues. The density matrices ρ1 , ρ2 , and ρ0 have the form (A1) with the
(I) (II) (III)
replacement p j → p j , p j , p j ; j = 1, 2, . . . , 8.
 
The matrix ρ0 is determined by the probability vector P  0 = p(0) ; j = 1, 2, . . . , 8. All of the
j
8-vectors P
 0, P  1 , and P 2 satisfy the condition that the corresponding density matrices satisfy the
equality Tr ρ̂21 = Tr ρ̂22 = Tr ρ̂20 = 1 and the property Tr ρ̂2ψ = 1 follows relation (47).

References
1. Landau, L. Das Dämpfungsproblem in der Wellenmechanik. Z. Phys. 1927, 45, 430–441. (In German)
[CrossRef]
2. Von Neumann, J. Wahrscheinlichkeitstheoretischer Aufbau der Quantenmechanik. Gött. Nach. 1927, 1,
245–272. (In German)
3. Dirac, P.A.M. The Principles of Quantum Mechanics; Clarendon Press: Oxford, UK, 1981; ISBN 9780198520115.
4. Schrödinger, E. Quantisierung als Eigenwertproblem (Erste Mitteilung). Ann. Phys. 1926, 79, 361–376.
(In German) [CrossRef]
5. Schrödinger, E. Quantisierung als Eigenwertproblem (Zweite Mitteilung). Ann. Phys. 1926, 80, 489–527.
(In German) [CrossRef]
6. Smithey, D.T.; Beck, M.; Raymer, M.G.; Faridani, A. Measurement of the Wigner distribution and the density
matrix of a light mode using optical homodyne tomography: Application to squeezed states and the vacuum.
Phys. Rev. Lett. 1993, 70, 1244–1247. [CrossRef] [PubMed]
7. Bertrand, J.; Bertrand, P. A tomographic approach to Wigner’s function. Found. Phys. 1987, 17, 397–405.
[CrossRef]
8. Vogel, K.; Risken, H. Determination of quasiprobability distributions in terms of probability distributions for
the rotated quadrature phase. Phys. Rev. A 1989, 40, 2847–2849. [CrossRef]

99

www.pdfgrip.com
Entropy 2018, 20, 692

9. Radon, J. Ũber die Bestimmung von Funktionen durch ihre Integralwerte langs gewisser Mannigfaltigkeiten.
Berichte Sachsische Akademie der Wissenschaften Leipzig 1917, 29, 262–277. (In German)
10. Wigner, E. On the quantum correction for thermodynamic equilibrium. Phys. Rev. 1932, 40, 749–759.
[CrossRef]
11. Husimi, K. Some formal properties of the density matrix. Proc. Phys. Math. Soc. Jpn. 1940, 22, 264–314.
12. Kano, Y. A new phase-space distribution function in the statistical theory of the electromagnetic field.
J. Math. Phys. 1965, 6, 1913–1915. [CrossRef]
13. Glauber, R.J. Coherent and incoherent states of the radiation field. Phys. Rev. 1963, 131, 2766–2788. [CrossRef]
14. Sudarshan, E.C.G. Equivalence of semiclassical and quantum mechanical descriptions of statistical light
beams. Phys. Rev. Lett. 1963, 10, 277–279. [CrossRef]
15. Mancini, S.; Man’ko, V.I.; Tombesi, P. Symplectic tomography as classical approach to quantum systems.
Phys. Lett. A 1996, 213, 1–6. [CrossRef]
16. Dodonov, V.V.; Man’ko, V.I. Positive distribution description for spin states. Phys. Lett. A 1997, 229, 335–339.
[CrossRef]
17. Man’ko, V.I.; Man’ko, O.V. Spin state tomography. J. Exp. Theor. Phys. 1997, 85, 430–434. [CrossRef]
18. Shannon, C.E. A mathematical theory of communication. Bell Syst. Tech. J. 1948, 27, 379–423. [CrossRef]
19. Tsallis, C. Nonextensive statistical mechanics and thermodynamics: Historical background and present
status. In Nonextensive Statistical Mechanics and Its Applications; Abe, S., Okamoto, Y., Eds.; Lecture Notes in
Physics; Springer: Berlin, Germany, 2001; Volume 560, pp. 3–98.
20. Rényi, A. On measures of entropy and information. In Proceedings of the Fourth Berkeley Symposium on
Mathematical Statistics and Probability; University of California Press: Berkeley, CA, USA, 1961; Volume 1,
pp. 547–561.
21. Hanel, R.; Thumer, S. Generalized (c, d) entropy and aging random walks. Entropy 2013, 15, 5324–5337.
[CrossRef]
22. Jisba, P.; Korbel, J. Maximum entropy principle in statistical interference: Case for non-Shannonian entropies.
2018, arXiv:1808.01172 [cond-mat.stat-mech].
23. Chernega, V.N.; Man’ko, O.V.; Man’ko, V.I. Triangle geometry for qutrit states in the probability
representation. J. Russ. Laser Res. 2017, 38, 416–425. [CrossRef]
24. Chernega, V.N.; Man’ko, O.V.; Man’ko, V.I. Quantum suprematism picture of Malevich’s squares triada for
spin states and the parametric oscillator evolution in the probability representation of quantum mechanics.
J. Phys. Conf. Ser. 2018, in press, arXiv:1712.01927.
25. Man’ko, V.I.; Marmo, G.; Ventriglia, F.; Vitale, P. Metric on the space of quantum states from relative entropy.
Tomographic reconstruction. J. Phys. A 2017, 50, 335302. [CrossRef]
26. Filippov, S.N.; Man’ko, V.I. Symmetric informationally complete positive operator valued measure and
probability representation of quantum mechanics. J. Russ. Laser Res. 2010, 31, 211–231. [CrossRef]
27. Castaños, O.; López-Peña, R.; Man’ko, M.A.; Man’ko, V.I. Squeeze tomography of quantum states. J. Phys. A
2004, 37, 8529–8544. [CrossRef]
28. Man’ko, M.A.; Man’ko, V.I. Properties of nonnegative hermitian matrices and new entropic inequalities for
noncomposite quantum systems. Entropy 2015, 17, 2876–2894. [CrossRef]
29. Asorey, M.; Ibort, A.; Marmo, G.; Ventriglia, F. Quantum tomography twenty years later. Phys. Scr. 2015,
90, 074031. [CrossRef]
30. Khrennikov, A. Quantum-like representation algorithm: Transformation of probabilistic data into vectors on
Bloch’s sphere. arXiv 2008, arXiv:0803.1391v1.
31. Khrennikov, A. The principle of supplementarity: A contextual probabilistic viewpoint to complementarity,
the interference of probabilities and incompatibility of variables in quantum mechanics. Found. Phys. 2005,
35, 1655–1693. [CrossRef]
32. Khrennikov, A. Interference of probabilities and number field structure of quantum models. Ann. Phys. 2003,
12, 575–585. [CrossRef]
33. Chernega, V.N.; Man’ko, O.V.; Man’ko, V.I. God plays coins or superposition principle for classical
probabilities in quantum suprematism representation of qubit states. J. Russ. Laser Res. 2018, 39, 128–139.
[CrossRef]
34. Chernega, V.N.; Man’ko, O.V.; Man’ko, V.I. Probability representation of quantum observables and quantum
states. J. Russ. Laser Res. 2017, 38, 324–333. [CrossRef]

100

www.pdfgrip.com
Entropy 2018, 20, 692

35. Shatskikh, A. Black Square: Malevich and the Origin of Suprematism; Yale University Press: New Haven,
CT, USA, 2012.
36. Kolmogorov, A.N. Foundation of the Theory of Probability; Chelsea: New York, NY, USA, 1956.
37. Shiryaev Albert, N. Probability-1; Springer: Berlin, Germany, 2016.
38. Holevo, A.S. Probabilistic and Statistical Aspects of Quantum Theory; North Holland: Amsterdam,
The Netherlands, 1982.
39. Khrennikov, A. Probability and Randomness. Quantum versus Classical; World Scientific: Singapore, 2016.
40. Man’ko, M.A.; Man’ko, V.I. Inequalities for nonnegative numbers and information properties of qudit
tomograms. J. Russ. Laser Res. 2013, 34, 203–218. [CrossRef]
41. Man’ko, M.A.; Man’ko, V.I. Hidden quantum correlations in single qudit systems. J. Russ. Laser Res. 2015, 36,
301–311. [CrossRef]
42. De Pasquale, A. Bipartite entanglement of large quantum systems. arXiv 2012, arXiv:1206.6749.
43. De Pasquale, A.; Facchi, P.; Giovannetti, V.; Parisi, G.; Pascazio, S.; Scardicchio, A. Statistical distribution of
the local purity in a large quantum system. J. Phys. A Math. Theor. 2012, 45, 015308. [CrossRef]
44. Man’ko, V.I.; Seilov, Z. The partition formalism and new entropic-information inequalities for real numbers
on an example of Clebsch-Gordan coefficients. J. Russ. Laser Res. 2017, 38, 50–60. [CrossRef]
45. Man’ko, M.A.; Man’ko, V.I. Hidden correlations and entanglement in single-qudit states. J. Russ. Laser Res.
2018, 38, 301–311. [CrossRef]
46. Chernega, V.N.; Man’ko, O.V.; Man’ko, V.I. Triangle geometry of the qubit state in the probability
representation expressed in terms of the Triada of Malevich’s Squares. J. Russ. Laser Res. 2017, 38, 141–149.
[CrossRef]
47. Lopez-Saldivar, J.A.; Castaños, O.; Nahmad-Achar, E.; López-Peña, R.; Man’ko, M.A.; Man’ko, V.I. Geometry
and entanglement of two-qubit states in the quantum probabilistic representation. Entropy 2018, 20, 630.
[CrossRef]
48. Man’ko, M.A.; Man’ko, V.I. From quantum carpets to quantum suprematism—The probability representation
of qudit states and hidden correlations. Phys. Scr. 2018, 93, 084002. [CrossRef]
49. Chernega, V.N.; Man’ko, O.V.; Man’ko, V.I. New inequality for density matrices of single qudit states. J. Russ.
Laser Res. 2014, 35, 457–461. [CrossRef]
50. Devoret, M.H.; Schoelkopf, R.J. Superconducting circuits for quantum information: An outlook. Science 2013,
339, 1169–1174. [CrossRef] [PubMed]
51. Pashkin, Y.A.; Yamamoto, T.; Astafiev, O.; Nakamura, Y.; Averin, D.V.; Tsai, J.S. Quantum oscillations in two
coupled charge qubits. Nature 2003, 421, 823–826. [CrossRef] [PubMed]
52. Man’ko, V.I.; Marmo, G.; Sudarshan, E.C.G.; Zaccaria, F. Interference and entanglement: An intrinsic
approach. Phys. A Math. Gen. 2002, 35, 7137–7157.
53. Man’ko, V.I.; Marmo, G.; Sudarshan, E.C.G.; Zaccaria, F. Positive maps of density matrix and a tomographic
criterion of entanglement. Phys. Lett. A 2004, 327, 353–364. [CrossRef]
54. Man’ko, M.A.; Man’ko, V.I.; Mendes, R.V. Tomograms and other transforms: A unified view. J. Phys. A
Math. Gen. 2001, 34, 8321–8332. [CrossRef]
55. Mendes, R.V. Non-commutative tomography and signal processing. Phys. Scr. 2015, 90, 074022. [CrossRef]
56. Ville, J. Theorie et Applications de la Notion de Signal Analytique. Cables Transm. 1948, 24, 61–74. (In French)

c 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

101

www.pdfgrip.com
 entropy
Article
“The Heisenberg Method”: Geometry, Algebra,
and Probability in Quantum Theory
Arkady Plotnitsky
Literature, Theory, Cultural Studies Program, Purdue University, West Lafayette, IN 47907, USA;
plotnits@purdue.edu

Received: 2 June 2018; Accepted: 23 August 2018; Published: 30 August 2018

Abstract: The article reconsiders quantum theory in terms of the following principle, which can be
symbolically represented as QUANTUMNESS → PROBABILITY → ALGEBRA and will be referred
to as the QPA principle. The principle states that the quantumness of physical phenomena, that is,
the specific character of physical phenomena known as quantum, implies that our predictions
concerning them are irreducibly probabilistic, even in dealing with quantum phenomena resulting
from the elementary individual quantum behavior (such as that of elementary particles), which in turn
implies that our theories concerning these phenomena are fundamentally algebraic, in contrast to more
geometrical classical or relativistic theories, although these theories, too, have an algebraic component
to them. It follows that one needs to find an algebraic scheme able make these predictions in a given
quantum regime. Heisenberg was first to accomplish this in the case of quantum mechanics, as matrix
mechanics, whose matrix character testified to his algebraic method, as Einstein characterized it.
The article explores the implications of the Heisenberg method and of the QPA principle for quantum
theory, and for the relationships between mathematics and physics there, from a nonrealist or, in terms
of this article, “reality-without-realism” or RWR perspective, defining the RWR principle, thus joined
to the QPA principle.

Keywords: algebra; causality; geometry; probability; quantum information theory; realism; reality

Perhaps the success of the Heisenberg method points to a purely algebraic method of
description of nature, that is, to the elimination of continuous functions from physics. Then,
however, we must give up, in principle, the space–time continuum.
—Albert Einstein, “Physics and Reality” (1936)

1. Introduction
This article reconsiders quantum theory, from quantum mechanics to quantum field theory to
quantum information theory, primarily focusing on quantum mechanics, in terms of the following
principle, which can be symbolically represented as:

QUANTUMNESS → PROBABILITY → ALGEBRA

and will be referred to as the QPA principle. This principle states, first, defining the experimental
nature of my first implication, QUANTUMNESS → PROBABILITY, that the quantumness of physical
phenomena, that is, the specific character of physical phenomena known as quantum, implies that
our predictions concerning them are irreducibly probabilistic or statistical, even in dealing with
quantum phenomena resulting from the elementary individual quantum behavior (such as that
of elementary particles). This, in turn implies, defining the theoretical character on my second
implication, PROBABILITY → ALGEBRA, that our theories concerning these phenomena, quantum
theories, are fundamentally algebraic, in contrast to more geometrical classical or relativistic theories.

Entropy 2018, 20, 656; doi:10.3390/e20090656 102 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2018, 20, 656

Although this distinction is not unconditional, because quantum theories do have geometrical
aspects, while, conversely, geometrical theories have algebraic aspects, it is, I shall argue, irreducible;
and understanding this distinction, along with the shared aspects of both types of theories, helps us to
shed new light on the relationships between algebra and geometry in physics.
Physically, quantum phenomena are defined by the fact that, in considering them, Planck’s
constant, h, must be taken into account, which allows one to use the classical theory in describing
them, but not in predicting them. By “quantum physics” I shall refer to the overall assembly of the
available quantum phenomena and theoretical accounts of these phenomena. The terms “classical
physics” will be used along parallel lines for classical phenomena, which need not depend on h (or on c,
the role of which defines relativistic phenomena). While, however, the role of h is irreducible in
quantum phenomena, their specificity as quantum is defined by a broader set of physical features,
such as the uncertainty relations (which do contain h), complementarity, and quantum correlations,
some of which are not linked to h, at least not expressly. On the other hand, some of these features,
although not all of them, are also exhibited by classical phenomena or found in mathematical models
different from those of the standard quantum mechanics or quantum field theory. The ultimate
distinction between quantum and classical phenomena is the subject of ongoing investigations and
debates, on which subject I shall further comment below. From present perspective, h may not pertain
to quantum objects or behavior but only to our theories, and enters these theories via the interactions
between quantum objects and measuring instruments.
Quantum phenomena also obey the principle of discreteness or the QD principle, which,
necessarily coupled to the individuality of quantum phenomena, may indeed represent the “essence”
or “quantumness” of quantum theory, according to Bohr [1]. In other words, quantum phenomena are
individual and discrete in relation to each other, which, as emphasized by N. Bohr, is not the same as
the atomic, Democritean, discreteness of elementary quantum objects themselves, which was initially
(following Planck’s discovery of quantum physics in 1900) seen as defining quantum physics as
quantum [2]. By “elementary” I refer to those quantum objects, also known as “elementary particles,”
that cannot be considered as composite. Either character, elementary or composite, could be ascertained
on the basis of effects such objects have on measuring instruments, keeping in mind that some particles
considered elementary can reveal themselves to be composite, as it happened in the case of hadrons
that were found to be composed of quarks and gluons.
It follows that the difference between objects and phenomena is irreducible in quantum theory,
as opposed to classical theory, specifically classical mechanics, which deals with individual classical
objects or simple classical systems. Rigorously speaking, as Kant already realized, this difference exists
there as well, but it can be disregarded insofar as we can, at least ideally and in principle, consider
such objects by neglecting the interference of observation. This is not possible in quantum physics,
at least not explicitly. It is under debate whether it is possible indirectly, inferentially, to establish
the independent nature and behavior of quantum objects, similarly to the way in which we can
treat, by means of classical mechanics, the nature and behavior of the elemental constituents of
the systems considered in classical statistical physics, even though we do not directly observe this
behavior. (This approach helped the nineteenth-century physics to confirm the existence of atoms.)
The interpretation of quantum phenomena adopted in this article, following Bohr and Heisenberg
(in his early work, discussed below [3]), precludes this possibility of attributing any independent
properties to quantum objects, although it does not preclude alternative interpretations of quantum
phenomena. The QPA principle would still hold for most of these interpretations, possibly, in contrast
to the present view, under the assumption of a continuously connected underlying reality. There is,
thus, another implication: QUANTUMNESS → DISCRETENESS. I shall, however, subsume the
discreteness of quantum phenomena under quantumness.
I shall discuss the concepts of “geometry,” “algebra,” and “probability,” and the relationships
between them in more detail below. Briefly, I understand geometry as the mathematical formalization
of spatiality in terms of measurement (while topology as referring to the structure of spatiality apart

103

www.pdfgrip.com
Entropy 2018, 20, 656

from measurement), algebra as the mathematical formalization of the relationships between symbols,
arithmetic as dealing with numbers, and probability as the mathematical formalization the likelihood
of events and expectations concerning them (the corresponding mathematical fields are geometry,
topology, algebra, number theory, and probability theory). Geometrical and topological objects always
have algebraic components, while algebraic objects need not have a geometrical component.
Now, there is next to nothing geometrical about probability or probability theory. The origin
of probability theory coincides with the rise of algebra, in the works of Cardano, Fermat, Descartes,
and Pascal. Some form of algebra was necessary for probability theory, as Hacking persuasively
argued in explaining why the theory emerged in the seventeenth century rather than earlier [4].
Analytic geometry and calculus were introduced around the same time by, the first by Fermat and
Descartes, and the second by Newton and Leibniz (although Fermat was, again, an important precursor,
especially as concerns the algebraic aspects of calculus), and these fields, too, were the product of
the algebraization of mathematics, a defining feature of the mathematics and physics of modernity,
even though geometry continued to dominate both until the nineteenth century.
It is true that probability theory, uses spatialized mathematical concepts, such as that of
“probability space,” introduced by Kolmogorov as part of his axiomatization of probability theory,
just as quantum mechanics uses the concept of “Hilbert space.” Kolmogorov’s concept follows the
concepts of “space” developed in functional analysis and measure theory (which Kolmogorov used
to axiomatize probability theory [5]), that of Hilbert space, among them. As I shall argue, however,
these concepts are more algebraic than geometrical: They have algebraic structures that geometrical
objects possess but are not aimed at representing the physical space, the original and still continuing
task of geometry, even though it has, as a mathematical field, developed far beyond its concerns with
nature or its relations to physics.
Quantum mechanics reshaped the relationships between the algebra of probability and the
algebra of theoretical physics, as against previous uses of probability, for example, in classical
statistical physics. There the relationships between them is underlain by a geometrical picture of
the behavior of the individual constituents of the systems considered, assumed to follow the laws of
classical mechanics. By contrast, as became apparent beginning with Planck’s discovery of quantum
phenomena, even elementary individual quantum objects and the events they give rise to had to be
treated probabilistically. One needed, accordingly, to find a new theory to make correct probabilistic
or statistical predictions concerning them, a task that quantum theory pursued from its inception,
with mixed results. Heisenberg was able to accomplish this task with quantum mechanics as matrix
mechanics, which avoided the deficiencies of “the old quantum theory,” as it became called after the
introduction of quantum mechanics, and which only predicted the probabilities of what was observed
in measuring instruments, as quantum phenomena, without describing the behavior of quantum
objects [3]. Heisenberg’s use of his matrix variables as operators in linear vector spaces (essentially,
infinite-dimensional Hilbert spaces over C) defined the algebraic nature of “the Heisenberg method,”
as Einstein characterized it [6]. This was in contrast to Schrödinger’s more geometrical method in
his wave mechanics, accompanied by a geometrical conception of quantum-level reality in terms
of a continuous vibrational process, a conception never worked out by Schrödinger to accord with
the experimentally established discrete features of quantum phenomena. Indeed, the physical and
mathematical demands of accounting for these features led Schrödinger to a mathematically equivalent
scheme [7]. I shall explain below why Schrödinger’s equation may be seen in probabilistically predictive
terms, without (geometrically) representing either the propagation of wave or the motion of particles
in space and time.
Bohr, in his initial assessment of Heisenberg’s discovery in 1925, before Schrödinger’s wave
version, but after Born and Jordan’s paper, which gave Heisenberg’s initial scheme its proper matrix
form [8], explained Heisenberg’s approach as follows [9]:

In contrast to ordinary mechanics, the new quantum mechanics does not deal with a space–time
description of the motion of atomic particles. It operates with manifolds of quantities which

104

www.pdfgrip.com
Entropy 2018, 20, 656

replace the harmonic oscillating components of the motion and symbolize the possibilities
of transitions between stationary states in conformity with the correspondence principle
[which requires that quantum and classical predictions coincide in the classical limit].
These quantities satisfy certain relations which take the place of the mechanical equations of
motion and the quantization rules [of the old quantum theory]. (emphasis added)

Thus, in effect combining the QPA principle with another principle, the reality-without-realism,
RWR, principle (defined by the fact that “the new quantum mechanics does not deal with a space–time
description of the motion of atomic particles” and explained in detail below), Heisenberg’s quantum
mechanics was a radical departure from the preceding history of modern (mathematical-experimental)
physics, from Galileo’s mechanics to Einstein’s relativity and even to the previous quantum theory.
All these theories were based on such descriptions or representations, either phenomenally visualizable,
usually on geometrical lines (referring to the continuous motion of the objects considered), at least in
dealing with elementary individual processes, as in classical mechanics, or beyond visualization but
given a conceptual or mathematical representation, as in relativity when dealing with photons and
velocities close to c. Classical statistical physics still relied and even depended on the representational
treatment of the behavior of the elemental constituents of the multiplicities considered, constituents
that were viewed as behaving in accordance with the laws of classical mechanics. The description of
individual quantum objects and the corresponding mathematical representation became partial in
the so-called old quantum theory, such as Bohr’s atomic theory, introduced in 1913 and developed
by him and others over the following decade [10,11]. The old (semiclassical) quantum theory only
provided such a representation, in terms of orbits, for stationary states of electrons in atoms, but
not for the discrete transitions, “quantum jumps,” between stationary states. The relativistic law
v+u
of addition of velocities (defined by the Lorentz transformation) in special relativity, s = 2,
1+(vu/c)
for collinear motion (c is the speed of light in a vacuum), runs contrary to any intuitive (geometrical)
representation of motion that we can have. This concept of motion is, thus, no longer a mathematical
refinement of a daily concept of motion in the way the classical concept of motion is. Relativity was
the first physical theory that defeated our ability to form a phenomenal conception of an elementary
physical process, and it was a radical change in the history of physics. The reason that this aspect of
relativity worried Einstein less than quantum mechanics was that relativity still offered a conceptual,
as well as mathematical, representation of the behavior of individual systems, which could, moreover,
be handled deterministically rather than probabilistically. Besides, there was plenty of geometry in
the theory, all the more so in general relativity, the geometry grounded in the spacetime continuum,
which Einstein wanted to preserve as part of physical reality at all levels and which was threatened
by the Heisenberg method [6]. Ultimately, photons are quantum objects and are treated by quantum
electrodynamics, which, in the present view, no longer represents the behavior of quantum objects in
the corresponding (high-energy) regimes any more than quantum mechanics.
Heisenberg abandoned a geometrical representation of stationary states retained in Bohr’s 1913
theory and, thus, any description or representation, even a mathematical one, of the behavior of quantum
objects, in accordance with the RWR principle. Stationary states were only represented algebraically
by energy-level values, considered apart from representing the behavior of electrons themselves
to which these energy levels were assigned. One was no longer thinking, as in classical physics or
relativity, in terms of predicting, even in considering elementary individual processes, the spacetime
behavior of the objects considered due to (continuous) changes in their states, assumed to be definable
independently of the interactions between objects and measuring instruments, and representable
by the corresponding formalism. Instead, following Bohr’s thinking concerning quantum jumps in
his 1913 atomic theory, one was thinking in terms of discontinuous transitions between physical
states of quantum objects and the probabilities or statistics of such transitions, as only probabilistic
predictions were possible experimentally [12]. These states are, moreover, only manifested as effects
of the interactions between quantum objects and measuring instruments. While not part of Bohr’s
1913 theory, this understanding became central to Bohr’s thinking following Heisenberg’s discovery

105

www.pdfgrip.com
Entropy 2018, 20, 656

and came to define Bohr’s interpretation of quantum phenomena and quantum mechanics. (Although
quantum phenomena and quantum mechanics are commonly interpreted jointly, as they were by Bohr
or are here, quantum phenomena could be given an interpretation independent of a theory accounting
for them.)
Heisenberg’s theory may be seen in terms of transition from geometry to algebra in fundamental
physics, which was acutely sensed by Einstein, who was hardly welcoming this transition [6]:

[P]erhaps the success of the Heisenberg method points to a purely algebraic method of
description of nature, that is, to the elimination of continuous functions from physics. Then,
however, we must give up, in principle, the space–time continuum [at the ultimate level of
reality]. It is not unimaginable that human ingenuity will some day find methods which will
make it possible to proceed along such a path. At present however, such a program looks
like an attempt to breathe in empty space.

Einstein was equally unhappy with the recourse to probability in dealing with elementary
individual processes, and he thought it would be avoided by a kind of fundamental theory he
envisioned, a continuous geometrical field theory of the type general relativity was. Earlier, he referred
to Heisenberg’s scheme as a magical trick, “Jacob’s pillow,” of Göttingen: it was not “the real thing”
and “[did] not really bring us any closer to the secret of the ‘old one,’” who, Einstein added in his
famous pronouncement, “at any rate is . . . not playing at dice” [13]. Ultimately, Einstein was more
concerned with the absence of realism at the fundamental level. As, however, he must have realized,
randomness and the recourse to probability are automatic in this absence.
I shall also argue here, as a bridge to considering quantum information theory, that, while not,
technically, quantum-informational, Heisenberg’s thinking could be viewed as quantum-informational
in spirit, and conversely, quantum information theory as Heisenbergian in spirit, and thus both
as algebraic in spirit [14,15]. The reason for this view is that the quantum-mechanical situation,
as Heisenberg conceived of it, was, in retrospect, defined by:

(a) certain already obtained information, concerning the energy of an electron, derived from spectral
lines (due to the emission of radiation by the electron), observed in measuring instruments; and
(b) certain possible future information, concerning the energy of this electron, to be obtainable from
spectral lines to be observed in measuring instruments and predictable (on experimental grounds) in
probabilistic or statistical terms by the mathematical formalism of one or another quantum theory.

Heisenberg’s aim was to develop such a formalism without assuming that this formalism needed
to represent a spatiotemporal process connecting these two sets of information or how each set comes
about. Heisenberg’s quantum mechanics was about quantum information, albeit not only about it.
It was equally about the nature of quantum objects, even though and because this nature was beyond
human knowledge and even thought. But then, this is also true about much foundational thinking
in quantum information theory, which aims to understand the ultimate nature of reality through the
nature of quantum information.
The remainder of this article proceeds as follows: the next section outlines my main concepts.
Section 3 addresses algebra and geometry in fundamental physics. Section 4 revisits Heisenberg’s
discovery of quantum mechanics and Bohr’s interpretation of it. Section 5 considers some recent work
in quantum information theory.

2. Fundamentals of the QPA/RWR Approach to Quantum Theory

The currently standard version of quantum theory, the only one to be considered in this article,
is comprised of three theories, all discovered in quick succession between 1925 and 1928. The first
is quantum mechanics for continuous variables in infinite-dimensional Hilbert spaces (QM), the
second is quantum theory for discrete variables in finite-dimensional Hilbert spaces (QTFD), and
the third is quantum field theory in Hilbert spaces that are tensor products of finite and infinite

106

www.pdfgrip.com
Entropy 2018, 20, 656

dimensional Hilbert spaces (QFT), initially introduced in the form of quantum electrodynamics (QED).
All these theories are algebraic and probabilistic or statistical and are governed by the QPA principle.
QFT, which handles high-energy physics, is comprised of several theories, constituting the standard
model of particle physics: quantum electrodynamics (QED), the theory of weak forces, and the
theory of strong forces, quantum chromodynamics (QCD). While the first two are unified or (as some
prefer to see it) “merged” in the electroweak theory, the unification of all three, known as “grand
unification,” has not been achieved. More troubling is that QFT and general relativity are inconsistent
with each other. This inconsistency is one of the greatest outstanding problems of fundamental physics,
which motivated string and M-brane theories, and alternative approaches, including in quantum
information theory (e.g., [16]).
The interpretation of quantum phenomena and quantum theory adopted here is defined by
a nonrealist or “reality-without-realism” (RWR) view of quantum theory in any of its versions,
a view that follows “the Copenhagen spirit of quantum theory” [Kopenhagener Geist der Quantenheorie],
as Heisenberg called it [17]. This characterization, abbreviated here to “the spirit of Copenhagen,”
is preferable to the more common “Copenhagen interpretation,” because there is no single such
interpretation, even in the case of Bohr, who changed his views a few times [18] (here I shall be
primarily concerned with the ultimate version of his interpretation). This is an important point. First,
there is much confusion concerning this fact by both critics and advocates of Bohr and the spirit of
Copenhagen. Secondly, at stake are interpretations, those (again, several) in the spirit of Copenhagen
amidst still others, and not the ultimate truth of nature, which we do not know and may never know
or even imagine and concerning which this article makes no definitive claims. In most of this article,
I will be concerned with QM. I will give some attention to QTFD in the context of quantum information
theory, which has been primarily concerned with it. QFT will only be mentioned in passing, although
the QPA principle and the RWR principle apply there, and historically, QFT, beginning with QED,
has been used to support the spirit of Copenhagen all along. I shall now outline the key concepts
grounding my argument, in part in order to avoid misunderstandings concerning them, because these
concepts or, more accurately, concepts designated by these terms can be defined otherwise.
It is fitting to begin by addressing the concept of concept, first, because, it is rarely adequately
considered in physical or philosophical literature, and secondly and more importantly, because the
role of concepts is not sufficiently appreciated in the philosophy of physics, especially the analytic
philosophy of physics. If, as Wilczek, a leading elementary particle theorist and a Nobel Prize laureate,
argues, “the primary goal of fundamental physics is to discover profound concepts that illuminate
our understanding of nature,” then creative thinking in fundamental physics is defined by concepts
and is advanced by the discovery or invention of new concepts [19]. But what is a physical concept,
and what is a concept in the first place? Wilczek does not explain it, taking it for granted or assuming
some general sense of it presumably shared by his readers. One might safely assume, given the
specific concepts that Wilczek invokes, such as that of “elementary particle” associated with that of
“symmetry group,” that the concepts in question have mathematical components, the presence of which
has defined the concepts of all modern, post-Galilean, theoretical physics. I shall also understand a
physical theory as an organized assemblage of concepts in the sense about to be defined, an assemblage
that relate certain physical objects or phenomena, usually in terms of propositions that are considered
to be true, at least with a sufficient practical, even if not fully definitive, justification.
It is the latter aspect that tends to dominate the concepts of theory used in the analytic philosophy
of physics. This aspect is of course indispensable: no physical theory, or philosophical argument
concerning theoretical physics, can bypass it. I would, nevertheless, argue, following Borel’s 1907
critique of the logically based understanding of mathematics, which is, in my view, applicable to the
logically based understanding of theoretical physics as well [20]. For Borel, a truly fertile invention
in mathematics and theoretical physics alike consists of the discovery of new concepts that enable a
new point of view from which to interpret the facts, followed by a search for the necessary proofs by
plausible reasoning, and only then, necessarily, bringing logic in. According to Gray, “Borel’s criticisms

107

www.pdfgrip.com
Entropy 2018, 20, 656

point quite clearly toward a problem that has not gone away in philosophers’ treatment of mathematics:
a tendency to reduce it to some essence than not only deprives it of purpose but is false to mathematical
practice. The logical enterprise, even if it had succeeded, would only have been an account of part of
mathematics, its deductive skeleton” [21]. This, I would contend, is true about much of the analytic
philosophy of physics, again, indispensable as the propositional and logical aspects of theoretical
physics are, in physics also as concerns correspondence with the available experimental evidence.
Bohr clearly understood the significance of these aspects of a physical theory and specifically
QM, and used them in addressing Einstein’s criticism: “In my opinion, there could be no other way
to deem a logically consistent mathematical formalism as inadequate than by demonstrating the
departure of its consequences from experience or by proving that its predictions did not exhaust
the possibilities of observation, and Einstein’s argumentation could be directed to neither of these
ends” [22]. Bohr also understood, however, now in agreement with Einstein, that the invention of
concepts play a decisive creative role in theoretical physics. Einstein saw “conceptual construction”
[begrrifliche Konstruction] as essential to and irreducible in physics [23]. He also saw the practice of
theoretical physics as that of the invention of new concepts through which one can approach reality,
sometimes even to the point of overriding the experimental evidence [24]. Riemann, a major inspiration
for Einstein’s general relativity, including as concerns conceptual construction, observed already in
1854: “From Euclid to Lagrange this darkness [in our understanding of the nature of geometry] has
been dispelled neither by the mathematicians nor the philosophers who have concerned themselves
with it. The reason [Grund] for this is undoubtedly because the general concept of multiply extended
magnitudes, which includes spatial magnitudes, remains completely unexplored. I have therefore first
set myself the task of constructing the concept of a multiply extended magnitude from general notions of
magnitude” [25]. This led Riemann to his concept of manifold [Maningfaltigkeit], central to modern
geometry and topology [26]. This article, too, while recognizing the indispensability of logical and
propositional structures of physical theories, gives concepts and the invention of concepts the defining
role the creative practice of theoretical physics.
I shall adopt the following understanding of concepts, in part following Deleuze and Guattari,
whose thinking was inspired by Riemann and especially his concept of manifold [26,27]. In this
definition, a concept is not merely a generalization from particulars (which is commonly assumed to
define concepts) or a general or abstract idea, although a concept may contain such ideas, specifically
abstract mathematical ideas in physics or in mathematics itself, where these ideas may become
concepts in the present sense. A concept is a multicomponent entity, defined by the organization of
its components, and some of these components may be concepts in turn. The definition may be very
basic, but it reflects an essential character of concepts in any domain, from daily life to the stratosphere
of mathematics and science. What is crucial is how this basic architecture is specifically instantiated in
a given concept, which is defined by both the nature of the components and their organization, by how
they relate to each other in the structure of the concept.
Consider as, an example, the concept of motion, first, as it is used in daily life: it will involve
various components, such as a change of place, speed, acceleration, moving bodies, etc., which belong
to our phenomenal intuition and are not defined rigorously, especially mathematically, but are still
parts of the concept defined by the organization of these components. Now, one can, as both Bohr
and Heisenberg did, see classical mechanics as a physical and mathematical refinement of these daily
concepts by means of such mathematically defined concepts as coordinates, momentum, angular
momentum, energy, and so forth, and thus also that of motion. While the concepts of classical physics
are derived from the concepts of daily life, they are both mathematical and subject to an experimental
verification. A concept could also be borrowed from a preceding physical theory and modified (or left
intact). Every concept and every theory, no matter how innovative, has a history and depends on
it. In quantum theory (QM, QFT or QTFD), in RWR-type interpretations, classical concepts are no
longer applicable to quantum objects and their behavior, while they still have a limited applicability
at the level of quantum phenomena, defined by effects of the interaction between quantum objects

108

www.pdfgrip.com
Entropy 2018, 20, 656

and measuring instruments. The concepts of quantum theory still have their history in classical
physics, both physically (when applied to measuring instruments) and mathematically, for example,
by adopting the concept of the Hamiltonian, while changing the variables from those of functions of
real variables of classical mechanics to operator variables (in Hilbert spaces over C) in QM or QFT.
The standard conservations laws (those of momentum, energy, and angular momentum), too,
are preserved, although new conservation laws are added, such as the conservation of probability current
in QM and QFT, or the conservation of baryon or lepton number in QFT. According to Heisenberg [28]:

The concepts of velocity, energy, etc., have been developed from simple experiments with
common objects, in which the mechanical behavior of macroscopic bodies can be described
by use of such words. The same concepts have then been carried over to the electron, since in
certain fundamental experiments electrons show a mechanical behavior like that of the
objects of common experience [or classical mechanics]. Since it is known, however, that
this similarity exists only in a certain limited region of phenomena, the applicability of the
corpuscular theory must be limited in the corresponding way . . .
As a matter of fact, it is experimentally certain only that light [too] sometimes behaves as
if it possessed some of the attributes of a particle [as reflected in the uncertainty relations],
but there is no experiment which proves that it possesses all the properties of a particle;
similar statements hold for matter [e.g., electrons] and wave motion. The solution of the
difficulty is that the two mental pictures [derived from classical physics] which experiments
lead us to form—the one of particles, the other of waves—are both incomplete and have
only the validity of analogies which are accurate only in limited cases. It is a trite saying
that “analogies cannot be pushed too far,” but they may be justifiably used to describe things
for which our language has no words. Light and matter are both single entities, and the
apparent duality arises in the limitation of our language.

In the RWR-type view, quantum objects and behavior are beyond any representation, including
mathematical one (which need not depend on language or physical concepts) or, in the view adopted
here, are beyond conception. Either view transforms the wave-particle duality into the viewpoint
defined by the concept complementarity, moreover, in the way in which, contrary to a common
view of complementarity, there is no wave-particle complementarity. As Bohr noted, in referring to
complementarity, a word that does not appear to have been used as a noun before Bohr (as opposed
to the adjective “complementary”) and was introduced by Bohr to designate a new concept: “In the
last resort an artificial word like ‘complementarity’ which does not belong to our daily concepts
serves only briefly to remind us of the epistemological situation [found in quantum physics], which at
least in physics is of an entirely novel character” [29]. Both the epistemological situation in question,
essentially that of the RWR-type, and the architecture of the concept of complementarity, which does a
great more than merely serving as such a reminder, will be discussed below. My main point at the
moment is that complementarity is a new physical concept with several interrelated components.
As most innovative concepts, complementarity, when introduced, was not defined by generalization
from available entities: it was something entirely new, although it, too, had its history in physics and
beyond [30,31]. It then functioned, in part, by generalizing multiple specific entities, such as specific
complementary configurations, say, those of the position or the momentum measurement, always
mutually exclusive at any given moment of time in the case of quantum phenomena. All physical
concepts, including those of classical physics (which are closer to our daily concepts), are physical
concepts, with mathematical components, ultimately divorced from their daily meaning and should
be treated as such, which is not always the case, especially when it comes to complementarity and
other concepts introduced by Bohr (e.g., [31]).
These considerations extend to the concept of theory, again, as understood here, as other definitions
of this concept are possible: a theory is an organized (conceptually, logically, or otherwise) assemblage
of concepts, as just defined. Every theory, again, has its history in preceding theories and can change by

109

www.pdfgrip.com
Entropy 2018, 20, 656

modifying its concepts or the relationships among them. A viable physical theory must, however, relate,
by means of logically consistent and experimentally verifiable propositions (possibly probabilistic
or statistical in nature), to the multiplicity of phenomena or objects that are assumed to form the
reality considered by this theory. This relation, in modern physics usually by means of mathematical
models (defined below), might be representational, and derive its predictive capacity, essential for any
physical theory, from this representation, or be merely predictive, possibly only probabilistically or
statistically predictive. I refer to both phenomena and objects, because, as Kant realized, they are not
the same even in classical mechanics, which deals with individual classical objects or sufficiently small
classical systems. However, classical objects, say, planets moving around the Sun, and our phenomenal
representation of them could be treated as the same for all practical purposes. This is because our
observational interference could, in principle, be neglected or compensated for, thus allowing us to
consider this behavior independently, a circumstance that does not appear to be expressly noted by Kant
but that is crucial to Bohr, because this is no longer the case in quantum physics [32,33]. Doing so was
assumed to be possible, at least in principle, in the case of all classical physical objects, even when they
were not or even could not be observed, as in the case of atoms or molecules in the kinetic theory of gases.
Quantum phenomena put this assumption into question. Defined by the effects of the interactions
between quantum objects and measuring instruments, quantum phenomena are observable in the
same way as are classical physical objects and could be treated as classical objects. By contrast, the
“uncontrollable” (quantum) nature of these interactions precludes any observation and, in RWR-type
interpretations, an inferential reconstitution of the independent behavior of quantum objects [34,35].
Nobody has ever observed a moving electron or photon as such, independently, to the degree that
the concept of motion, as opposed to a change of a state, ultimately applies to them, or any kind of
quantum object. It is only possible to observe traces, such as spots on photographic plates, left by their
interactions with measuring instruments. This still allows for a spectrum of assumptions concerning
quantum objects and their behavior, beginning with the assumption of the existence of such objects
(or what is so idealized), inferred from these traces.
The present interpretation, while assuming this existence on the basis of these effects and their
particular character (not found in classical physics), places quantum objects and behavior beyond
conception, which I shall term “the strong RWR view,” rather than only representation, which I shall
term “the weak RWR view.” The strong RWR view is a radical position. Not all interpretations in the
spirit of Copenhagen go that far. Thus, while there are indications that Bohr, especially in the ultimate
version of his interpretation (my primary focus here), might have agreed with this view (e.g., [35]),
he never expressly stated so. One could assume the possibility of a mathematical representation of
quantum objects and behavior in the absence of a physical conception of them. While Bohr’s and the
present view exclude this possibility, Heisenberg was open to it in his later thinking (e.g., [36]).
The history of a theory is accompanied by the history of its interpretations. The history of
QM, in particular, has been shaped by a seemingly uncontainable proliferation of, sometimes
conflicting, interpretations. It is not possible to survey these interpretations here. Each rubric on
by now a long list (e.g., the Copenhagen, the many-worlds, consistent-histories, modal, relational,
transcendental-pragmatist, and so forth) contains different versions. The literature dealing with each
interpretation is immense. Standard reference sources would list and summarize most common rubrics.
Although often implicit, an interpretation is essential for establishing the relationships between a
theory and the phenomena or objects it considers, essential to any theory. This is customarily done by
means of mathematical models.
I define a mathematical model in physics as a mathematical structure or a set of mathematical
structures that enables such relationships (the concept of models in mathematics or mathematical
logic is a separate subject, put aside here). As that of theory and other major concepts discussed here,
the concept of a mathematical model or model, in the first place, has a long history, which is also a
history of diverse definitions, and literature on the subject is extensive as well. It is not my aim to
discuss the subject as such or engage with this literature, which would be difficult within the scope

110

www.pdfgrip.com
Entropy 2018, 20, 656

of this article. The present concept of a mathematical model, while relatively open, is sufficient to
accommodate those models that I shall consider. A more detailed discussion of the present view of
mathematical models is given [37,38] and of modeling in general, on the lines of analytic philosophy
of science, in [39,40]. One must also keep in mind the difference between a mathematical model of a
theory (a concept especially important in mathematical logic or the philosophy of mathematics) and a
mathematical model used by a theory, with which I am concerned here. Mathematical models used
in physics may be geometrical (as in general relativity, for example) or algebraic, as in QM and QFT,
although geometrical models contain algebraic elements. (The geometrical aspects of algebraic models
are a more complex matter considered below.) The relationships between a model and the objects or
phenomena considered may be representational. In this case the elements of a model and relations
among them would correspond or map the elements of reality and the relations among them and
relate the theory to reality by means of this mathematical representation. The predictive capacity of
the theory, essential for any theory, would then derive this representation. The mathematical models
used in classical mechanics or relativity are examples of such models. Models may, however, also be
strictly predictive, without being representational, as are the mathematical models used in QM or QFT,
in RWR-type interpretations, the predictions of which are, moreover, probabilistic or statistical, again,
even in the case of elementary individual quantum objects and behavior. An interpretation of a given
theory is, thus, always an interpretation of how the mathematical model or models used by it relate
to the phenomena or objects considered. A theory may, however, involve other interpretive aspects,
defined by its concepts. For example, part of Bohr’s and, following Bohr, the present interpretation is a
particular (interpretive) concept of measuring instruments used in quantum physics. According to this
concept, the observable parts of measuring instruments are described my means of classical physics,
while these instruments also have quantum parts, through which they interact with quantum objects,
an interaction “irreversibly amplified” to what is observed in measuring instruments [41]. Placing
quantum objects and behavior beyond representation or even conception is another interpretative
feature, in the second case without giving quantum objects and behavior concepts.
Rigorously, a different interpretation of a given mathematical model defines a different theory in
the present definition of a theory, because this interpretation may involve specific concepts, such as the
ones just mentioned in the case of the interpretation adopted here, which may not be shared by other
interpretations, even those in the spirit of Copenhagen. For simplicity, however, I shall speak of the
corresponding interpretation of the theory itself containing a given mathematical model, interpreted by
this theory, say, of one or another interpretation of QM. Thus, initially, Heisenberg’s and Schrödinger’s
versions of the formalism appeared as two different mathematical models, giving the same predictions.
They were also accompanied by two different theories, initially designated quantum mechanics and wave
mechanics, the first strictly algebraic and the second geometrical, by virtue of conceiving quantum-level
reality as a continuous wave-like process, each theory, moreover, given different interpretations at the
time (actually, interpreting either theory posed major difficulties then). These two models were quickly
proven to be mathematically equivalent (there are several proofs, all of which involve additional
assumptions and complexities, the most general one given by the Stone-von Neumann theorem),
which allowed one to unify the mathematical model of QM, ultimately in terms of its Hilbert-space
formalism, with some yet more abstract versions added later. By contrast, the two theories—quantum
mechanics (underlying the model of matrix mechanics) and wave mechanics—which were based in
two different sets of concepts remained different. While Schrödinger’s theory, based in the idea of a
wave-like ultimate reality, had receded by the late 1920s, as Schrödinger’s wave function received an
interpretation as a tool for predicting probabilities, rather than representing any physical process, this
theory has never been entirely abandoned.
I now turn to the concept of reality, which I shall approach via Bohr’s elaboration, partially cited
above, concerning the epistemological situation that the concept of complementarity reflects [29]:

The renunciation of the ideal of causality in atomic physics which has been forced on us is
founded logically only on our not being any longer in a position to speak of the autonomous

111

www.pdfgrip.com
Entropy 2018, 20, 656

behavior of a physical object, due to the unavoidable interaction between the object and
the measuring instruments which [interaction] in principle cannot be taken into account,
if these instruments according to their purpose shall allow the unambiguous use of the
concepts necessary for the description of experience. In the last resort an artificial word
like “complementarity” which does not belong to our daily concepts serves only briefly to
remind us of the epistemological situation here encountered, which at least in physics is of
an entirely novel character.

It follows that “our not being any longer in a position to speak of the autonomous behavior of a
physical object,” demands “a radical revision of our attitude toward the problem of physical reality,”
ultimately depriving us of realism [42]. “The renunciation of the [classical] ideal of causality,” invoked
by Bohr, is automatic, because, as I explain below, causality requires realism [42].
I shall now introduce a concept of reality that permits this revision. This concept itself is very
general and is, arguably, in accord with most, even if not all (which would be impossible), currently
available concepts of reality in realism and nonrealism, which would, respectively, assume this reality
to be representable or at least conceivable and to be beyond representation or even conception. By reality
I refer to that which exists or is assumed to exist, without making any claim concerning the nature of
this existence, which thus may be placed beyond representation or even conception. I understand
existence as a capacity to have effects on the world with which we interact and that, because it exists,
has such effects upon itself. To ascertain such effects entails representation of these effects, but not
necessarily of how they come about, which implies that a given theory might assume different levels
of reality, some allowing for a representation or at least conceptions and others not.
In physics, the primary reality considered is that of matter, including radiation, generally governed
by the concept of field, classical or quantum. The idea of matter is still a product of thought, which,
however, is customarily assumed to be a product of the material processes in the brain, and thus
of matter. Matter is commonly, but not always (although exceptions are rare), assumed to exist
independently, and to have existed when we did not exist and to continue to exist when we will no
longer exist, which may be seen as defining the independent existence of matter. This view is upheld
in the RWR-type interpretations of QM, but in the absence of a representation or even conception
of the character of this existence, for example, as either discrete or continuous. Discreteness only
pertains to quantum phenomena, observed in measuring instruments, while continuity has no physical
significance at all. It is only a feature of the formalism of QM, which, while mathematically continuous,
relates to discrete phenomena by predicting the probabilities or statistics of their occurrence.
Physical theories prior to quantum theory have been realist theories, usually representational
realist theories. Such theories aim to represent the corresponding objects and their behavior by
mathematical models, assumed to idealize how nature works, an assumption sometimes referred to as
“scientific realism.” More exactly, as noted earlier, such a theory is a representation that is then realized
by a mathematical model, which mathematically represents the reality considered. Thus, classical
mechanics (used in dealing with elemental individual objects and small classical systems), classical
statistical mechanics (used in dealing, statistically, with large classical systems), or chaos theory (used
in dealing with classical systems that exhibit a highly nonlinear behavior) are all realist theories,
as concerns the ultimate reality they consider. While classical statistical mechanics does not represent
the overall behavior of the systems considered because their great mechanical complexity prevents
such a representation, it assumes that the individual constituents of these systems are represented by
classical mechanics. As indicated earlier, the status of these theories as realist could be questioned,
on Kantian lines, even in the case of classical mechanics, where the representational idealizations
used are more in accord with our phenomenal experience, which is only partially the case in relativity.
However, all these cases still allow for viable idealized realist and (classically) causal models.
One could also define another type of realism, which is not representational. This realism
encompasses theories that presuppose an independent structure of reality governing the behavior of
the ultimate objects these theories consider, while allowing that this architecture cannot be represented,

112

www.pdfgrip.com
Entropy 2018, 20, 656

even ideally, either at a given moment in history or perhaps ever, but if so, only due to practical
epistemological limitations. In the first eventuality, a theory that is merely predictive may be accepted
for lack of a realist alternative, but under the assumption that a future theory will do better, in particular
as a representational realist theory. Einstein adopted this view toward QM, which he expected to be
eventually replaced by such a theory.
The assumption of realism of either type is abandoned or even precluded in reality-without-
realism (RWR) type interpretations of quantum phenomena and QM, beginning with that of Bohr.
In such interpretations, the mathematical model of QM, defined by its mathematical formalism,
becomes a strictly probabilistically or statistically predictive, rather than deterministic, model, even in
considering elementary individual quantum objects and processes, which form quantum-level reality,
while suspending or even precluding a representation and possibly a conception of this reality, and an
assumption that this reality is causal, or classically causal. I distinguish “causality” and “determinism.”
By classical causality I refer, ontologically, to the conception that the state of the system considered is
determined at all future moments of time, once it is determined at a given moment of time, and by
determinism, epistemologically, to the possibility of predicting the outcomes of such processes ideally
exactly. This conception of causality has defined modern classical physics since Descartes, Galileo,
and Newton, and philosophy beginning at least with Plato. As will be seen, causality may be defined
differently, first, in a relativistic or local sense and, second, in a quantum-theoretical probabilistic
sense. The probabilistic or statistical character of quantum predictions must, however, be maintained
by realist interpretations of QM or alternative theories (such as Bohmian mechanics), to accord with
quantum experiments, where only probabilistic or statistical predictions are possible. This is because
the repetition of identically prepared experiments in general leads to different outcomes, and unlike in
classical physics, this difference cannot be diminished beyond the limit defined by Planck’s constant, h,
by improving the capacity of our measuring instruments, as manifested in the uncertainty relations,
which would remain valid even if we had perfect instruments.
RWR-type interpretations do assume the concept of reality, defined as that which is assumed to
exist, without, in contrast to realist theories, making any claims concerning the character of this existence,
which is what makes this concept of reality that of “reality without realism” [43,44]. The existence of
quantum objects or something that leads to this idealization (it is still an idealization) is inferred from
the totality of effects they have on world we observed, specifically on experimental technology, without
making claims concerning their independent behavior. Such interpretations place quantum objects and
processes either beyond representation, the weak RWR view, or more radically, beyond conception,
the strong RWR view, which I adopt here. As I said, Heisenberg at the time of his discovery and Bohr
at nearly all stages of his thinking held at least a weak RWR view, with Bohr eventually moving closer
to the strong RWR view, while Heisenberg eventually adopted a form of mathematical realism. In 1927,
Bohr briefly and ambivalently entertained the idea that independent quantum behavior and thus the
ultimate nature of quantum reality could be represented, moreover, causally, by the mathematical
formalism of QM, while indeterminism was introduced by measurement [45]. Bohr’s ambivalence
was due to the fact that one deals with the formalism over C, which is difficult to associate with
physical representation and the fact that Schrödinger’s wave equation in fact applied to the coordinate
and not a real space: “The symbolic character of Schrödinger’s method appears not only from the
circumstance that its simplicity, similarly to that of the matrix theory, depends essentially upon the use
of imaginary arithmetic quantities. But above all there can be no question of an immediate connection
with our ordinary conceptions because the ‘geometrical’ problem represented by the wave equation
is associated with the so-called co-ordinate space, the number of dimensions of which is equal to
the number of degrees of freedom of the system, and, hence, in general greater than the number of
dimensions of ordinary space” [46]. In any event, Bohr quickly abandoned the view that independent
quantum behavior is represented by QM, under the impact of his exchanges with Einstein. However,
championed by both Dirac’s and von Neumann’s influential books [47,48], this view has persisted and
remains common [49].

113

www.pdfgrip.com
Entropy 2018, 20, 656

Although Kant’s philosophy may be seen as an important precursor the RWR view, the strong
RWR view is manifestly more radical than Kant’s view of noumena or things-in-themselves
vis-à-vis phenomena or appearances formed in our minds. According to Kant, while noumena are
unknowable, they are still in principle conceivable, especially when one’s thinking is helped by what
he calls “Reason” [Vernunft], a higher faculty than “Understanding” [Verstand], which only concerns
phenomena, although there is no guarantee, even for Reason, that this conception is correct [50].
Even the weak RWR view is still more radical than that of Kant, because, while a conception of
quantum objects and behavior is in principle possible, it cannot be unambiguously used in considering
quantum phenomena, at least as things stand now. I am not saying that the strong RWR view is
physically necessary, but only that it is interpretively possible. There does not appear to be any
experimental data that would compel one to prefer either the strong or the weak RWR view, or to
definitively claim for either anything beyond its consistency or effectiveness. These views are, however,
different philosophically because they reflect different limits that nature allows our thought in reaching
its ultimate constitution.
Two qualifications are in order. First, one could, in principle, see the claim concerning merely
the existence or reality of something to which a theory can relate without representing it as a form of
realism. This use of the term realism is sometimes found in advocating interpretations of QM that are
nonrealist in the present sense (e.g., [51–53]), although none of these works adopted the strong RWR
view. However, the present definition of realism or, similarly, ontology is more in accord with most
understandings of realism, representational or nonrepresentational, including in considering quantum
theory in physics and the philosophy of physics. Secondly, the present argument does not aim to deny
realism even in the present sense, still a generally preferred view.
It could be assumed that something “happens” between observations, as manifested in changes
that we observed in the instruments used, such as a discrete change of the energy, a “quantum jump,”
of an electron in an atom (statistical as any claim concerning such changes may be), if one keeps in
mind the provisional nature of such concepts as “happens.” According to Heisenberg [54]:
There is no description of what happens to the system between the initial observation and the
next measurement. . . . The demand to “describe what happens” in the quantum-theoretical
process between two successive observations is a contradiction in adjecto, since the word
“describe” [or “represent”] refers to the use of classical concepts, while these concepts cannot
be applied in the space between the observations; they can only be applied at the points
of observation.
The same, it follows, must apply to the word “happen” or any word we use, and we must use
words and concepts associated to them, even when we try to restrict ourselves to mathematics as much
as possible. There can be no physics without language, but quantum physics imposes new limitations
on using it. Heisenberg adds later in the same book: “But the problems of language are really serious.
We wish to speak in some way about the structure of the atoms and not only about ‘facts’—the latter
being, for instance, the black spots on a photographic plate or the water droplets in a cloud chamber.
But we cannot speak about the atoms in ordinary language” [55].
On the other hand, as Heisenberg noted on an earlier occasion, mathematics is, “fortunately,”
free from the limitations of daily language and concepts, fortunately because one could take
advantage of this freedom in creating QM, doing which might not even have been possible
otherwise [56]. Mathematics, especially algebra (geometry is more connected to our phenomenal
intuition of spatiality), also allows to circumvent the limits our phenomenal, representational
intuition, also involving visualization, sometimes used, including by Bohr, to translate the German
word for intuition, Anschaulichkeit. Bohr often spoke of quantum objects and behavior as beyond
altogether beyond visualization, although ultimately for him both were beyond any representation,
including mathematical one (e.g., [57,58]). As free from these limitations of language and ordinary,
or philosophical or even physical concepts, mathematics could be assumed to represent quantum-level
reality, as Heisenberg eventually came to believe. Physics and Philosophy and his other later writings

114

www.pdfgrip.com
Entropy 2018, 20, 656

do give mathematics at least some capacity to do so, still in algebraic terms, to the point of defining,
following Wigner [59], elementary particles themselves as representations of symmetry groups.
However, while crucial in QM or, even more so, in QFT, the role of symmetry need not depend
on realism, physical or mathematical, because symmetry groups can be viewed as part of the
probabilistically or statistically predictive machinery of QM and QFT. There are ontological and,
specifically, geometrical symmetries, for example, those embedded in conservation laws by Noether’s
theorems, which apply in quantum theory, where, however, they are manifested at the macro level of
measuring instruments, described classically. The concept of group is, however, algebraic, even when
used, on realist lines, in classical physics or relativity, or of course in geometry or topology. In any
event, a form of mathematical (algebraic) realism advocated by Heisenberg in his later works appears
to exclude the application of representational language or concepts apart from mathematical ones to
the ultimate constitution of reality [60]. On the other hand, as his argument in his paper introducing
QM [3] (discussed in Section 4) suggests, at the time of his discovery Heisenberg appears to have seen
mathematics’ freedom from these limitations, while crucial for QM and even making its invention
possible, in terms of its probabilistically predictive rather than representational capacity, a view held
by Bohr, a least from 1928 on. For Bohr, again, a mathematical representation, even if not a conception
of quantum objects and behavior was “in principle excluded,” along with a physical one, at least as
things stands now [35].
“As things stand now” is an important qualification, equally applicable to the strong RWR view,
even though it might appear otherwise, given that this view precludes any conception of the ultimate
reality not only now but also ever, by placing it beyond thought altogether. The qualification “as things
stand now” still applies because a return to realism is possible, either on experimental or theoretical
grounds even for those who hold this view. This return may take place because quantum theory,
as currently constituted (QM, QFT, and QTFD), may be replaced by an alternative theory that allows
for or requires a realist interpretation, or because RWR-type interpretations, either of the weak or
the strong type, may become obsolete, even for those who hold this view, with quantum theory in
place in its present form. As things stand now, either RWR view is interpretively possible. It is also
possible, however, that the RWR view, in either weak or strong version, will remain part of our future
fundamental theories, as the development of QFT theory appears to indicate. QFT has been open to the
RWR-type view and the corresponding interpretation from its inception with Dirac until now and was
used, specifically by Bohr, in support of this view (e.g., [61,62]). It also conforms to the QPA principle,
even though it may appear and in some respects is more geometrical than QM, given the role of
certain geometrical concepts, such as symmetries, in quantum theory, especially in QED, such as gauge
symmetry, introduced by Weyl, initially in his (failed) attempt to (geometrically) unify general relativity
and electromagnetism. In QED or QFT in general, however, the “geometry” of gauge symmetries is
symbolic, and their real significance is the invariance under the corresponding gauge groups, say,
as applies to the “phases” of electrons (“phase” being a symbolic concept, which physically relates to
probabilities), is just part of the algebra of QFT theory which ultimately relates to the probabilities or
statistics of experiments. As I said, there are spatial (or temporal) geometrical symmetries, such as those
involved, by Noether’s theorems, in conservation laws, or still others, which are used in QFT, or QM,
but these are only manifested at macro-levels. Feynman’s path integrals, which suggest trajectories
and thus geometry, can be seen along the same algebraic lines (as part of the algebraic probabilistic
machinery of QFT), especially given that these “paths” do not refer to the actual motion of particles,
although the subject, admittedly, needs more discussion, which cannot be pursued here.
It is also true that we use visual tools, such as Feynman’s diagrams. Enormously helpful as they
are, however, Feynman’s diagrams are only diagrams, heuristic devices: they do not represent the
quantum processes to which they refer, even if one holds that these processes are representable. The role
of Feynman’s diagrams may be said to represent the predictive workings of the formalism of QED or
QFT and thus to help one to work with this formalism in order to make probabilistic or statistical
predictions concerning the outcomes of the experiments these diagrams are connected to, predictions

115

www.pdfgrip.com
Entropy 2018, 20, 656

only possible rigorously, numerically, by means of the algebra of QED or QFT. This, however, is quite
different from representing quantum behavior itself. In sum, while there are additional nuances as
concerns their difference, the formalism of QFT is algebraic in the same way as is that of QM, and the
QPA and (interpretively) RWR principles combine analogously.
I now turn to the question of causality. As noted, RWR-type interpretations make the absence of
classical causality nearly automatic. This absence is strictly automatic if one adopts the strong RWR
view, which places the ultimate nature of reality beyond conception, because the assumption that this
nature is classically causal would imply at least a partial conception of this reality. However, even if
one adopts the weak RWR view, which only precludes a representation of this reality, classical causality
is still difficult to maintain in considering quantum phenomena. This is because to do so one requires
a degree of representation, analogous to that found in classical physics, that appears to be prevented,
in particular, by the uncertainty relations (which are independent of QM). Schrödinger expressed this
difficulty, while disparaging QM, or at least the spirit of Copenhagen, as “the doctrine born of distress,”
in his cat-paradox paper: “if a classical state does not exist at any moment, it can hardly change
causally,” where a classical state is defined by the (ideally) exact position and momentum of an object
at any moment of time [63]. According to Bohr, who did not share Schrödinger’s reservations [64]:
It is most important to realize that the recourse to probability laws under such circumstances
is essentially different in aim from the familiar application of statistical considerations as
practical means of accounting for the properties of mechanical systems of great structural
complexity. In fact, in quantum physics we are presented not with intricacies of this kind,
but with the inability of the classical frame of concepts to comprise the peculiar feature[s] of
the elementary [quantum] processes.
While “the classical frame of concepts” might refer to those of classical physics, Bohr might have
been here closer to the strong RWR view, because at this stage of his thinking, he argues that all our
representational concepts (“object” and “process,” among them) are classical, possibly apart from
purely mathematical concepts, considered as entirely divorced from any phenomenal representation.
Indeed, according to Wittgenstein, we may be unable to conceive of a process that is not causal [65].
Complementarity is different: While it has representational aspects, referring to phenomena observed
in measuring instruments and thus to our experience, it does not represent the independent properties
and behavior of quantum objects, but is instead designed to deal with a lack of this representation.
The question of causality is, however, a subtle matter, especially given that one can define concepts
of causality that are not classical, and it merits a further discussion. First, I shall consider the concepts
of indeterminacy, randomness, chance, and probability, again, as I understand them, because they, too,
can be defined otherwise. In the present definition, indeterminacy or chance is a more general category,
while randomness will refer to a most radical form of indeterminacy, when even a probability is not and
cannot be assigned to a possible future event. Indeterminacy (including randomness) and chance may
be understood as different from each other as well. These differences are, however, not germane in the
present context, and I shall for convenience only refer to indeterminacy. An indeterminate, including
random, event may or may not result from some underlying classical causal processes, whether this
process is accessible to us or not. The first eventuality defines classical indeterminacy or randomness,
conceived as ultimately underlain by a hidden classically causal architecture; the second irreducible
indeterminacy and randomness. The ontological validity of an application of the latter cannot be
guaranteed: it is impossible to ascertain that an apparently indeterminate or random sequence is in fact
indeterminate or random, and there is no mathematical proof that any sequence is [66]. This concept is
an assumption that may only be practically justified insofar as an effective theory or interpretation
is developed.
As explained, factually, quantum phenomena only preclude determinism, because identically
prepared quantum experiments, as concerns the state of measuring instruments, in general lead to
different outcomes. Only the statistics of multiple identically prepared experiments are repeatable.
It would be difficult, if not impossible, to do science without being able to reproduce at least the

116

www.pdfgrip.com
Entropy 2018, 20, 656

statistical data. The lack of classical causality or of realism in the RWR-type interpretations of quantum
phenomena and QM are interpretive inferences from this situation and additional features such as
correlations, the uncertainty relations, or complementarity. Such interpretations, again, do not exclude
the possibility of causal or realist interpretations of QM, or alternative causal or realist quantum
theories, such as Bohmian mechanics (which is, however, nonlocal), or theories defined by deeper
underlying causal dynamics, which makes QM an “emergent” theory, such as A. Khrennikov’s
“pre-quantum classical statistical field theory” [44,67].
Although sometimes glossed over, the difference between probability and statistics is important
in quantum theory. I would like to briefly comment on this difference and on the role of probability
and statistics in quantum theory more generally from the RWR-type perspective. My remarks cannot
do justice to the subject, extensively considered in literature (e.g., [5,68]). They are only aimed to
address those points that are especially relevant for my argument. “Probabilistic” commonly refers
to our estimates of the probabilities of either individual or collective events, such as that of a coin
toss or of finding a quantum object in a given region of space. “Statistical” refers to our estimates
concerning the outcomes of identical or similar experiments, such as that of multiple coin-tosses or
repeated identically prepared experiments with quantum objects, or to the average behavior of certain
objects or systems. The standard use of the term “quantum statistics” refers to the behavior of large
multiplicities of identical quantum objects, such as electrons and photons, which behave differently,
in accordance with, respectively, the Fermi-Dirac and the Bose-Einstein statistics, for identical particles
with, respectively, half-integer and integer spin. The Bayesian understanding defines probability as a
degree of belief concerning a possible occurrence of an individual event on the basis of the relevant
information we possess (e.g., [69] or, in a different version [70]). This makes the probabilistic estimates,
generally, subjective, although there may be agreement (possibly among a large number of individuals)
concerning such estimates. The frequentist understanding, also referred to as “frequentist statistics,”
defines probability in terms of sample data by emphasis on the frequency or proportion of these
data, which is considered more objective. In quantum physics, as noted, exact predictions are in
general impossible even in dealing with elemental individual processes and events. This situation
could, however, be interpreted either on Bayesian lines, under the assumption that a probability could
be assigned to individual quantum events, or on frequentist lines, under the assumption that each
individual effect is strictly random. A prominent recent example of a nonrealist Bayesian approach is
Quantum Bayesianism, QBism, which, however, contains other philosophical dimensions (e.g., [52]).
Although most of its argument would apply if one adopts a Bayesian view, this article adopts the
frequentist, RWR-type, view, considered in detail in [43,44]. Bohr and Heisenberg appear to have been
inclined to a statistical view of the type adopted here [71].
A brief qualification might be in order concerning two different uses of the statistical just
mentioned, concerning, respectively, multiple repeated experiment and the average behavior of
large system. One can make a probabilistic estimate for an event of finding an electron gas occupying
less than a given volume, similarly to that of finding a quantum object in a given region of space.
(In fact, this is true for events pertaining to classical statistical systems.) My point here, however,
is that, unlike in classical mechanics, in QM we are dealing with randomness or probabilities even in
considering events associated with elemental individual objects, such as electrons, rather than with
large (“statistical”) multiplicities, and that such individual events can be interpreted on either Bayesian
or statistical lines, in the latter case, under the assumption that such individual events are strictly
random. These two interpretations would also distinguish individual events, such as those of finding
an electron gas occupying less than a given volume. In the statistical interpretation in the present sense
we would need to repeat this experiment many times to establish these statistics, while no probability
is, in general, assigned to a given event.
Finally, probability introduces an element of order into situations defined by the role of
randomness in them and enables us to handle such situations better. In other words, probability
or statistics is about the interplay of indeterminacy or randomness and order. This interplay takes

117

www.pdfgrip.com
Entropy 2018, 20, 656

on a unique significance in quantum physics, because of the existence of quantum correlations,

such as the EPR or (as they are also known) EPR-Bell correlations, found in the experiments of the
Einstein-Podolsky-Rosen (EPR) type and considered, in the case of discrete variables, in Bell’s and
the Kochen-Specker theorems, and related findings. These correlations are a form of statistical order.
They are properly predicted by QM, which is, thus, along with and responding to quantum phenomena
themselves, as much about order as about indeterminacy or randomness, and, most crucially, about
their unique combination in quantum physics. The correlations themselves are collective, statistical,
and as such they would not depend on either interpretation, Bayesian or frequentist, of our predictions
concerning the individual events involved. That, in certain circumstances, indeterminate or random
individual events form statistically correlated and thus ordered multiplicities is, however, one of the
greatest mysteries of quantum physics, which makes it as much about order as about indeterminacy
and randomness, a statistically correlated order without an ontologically underlying classical order
that merely cannot be accessed epistemologically.
I shall now consider two alternative conceptions of causality important for my argument.
Thus, the term “causality” is often used in accordance with the requirements of special relativity,
which restricts (classical) causes to those occurring in the backward (past) light cone of the event that is
seen as an effect of this cause, while no event can be a cause of any event outside the forward (future)
light cone of that event. In other words, no physical causes can propagate faster that the speed of
light in a vacuum, c, which requirement also implies temporal locality. Technically, this requirement
only restricts classical causality by a relativistic antecedence postulate, rather than precludes it,
and relativity theory itself, special or general, is (locally) a classically causal and indeed deterministic
theory. By contrast, while, as a probabilistic or statistical theory of quantum phenomena, QM, at least
in RWR-type interpretations, lacks classical causality, its probabilistic or statistical predictions are
consistent with both temporal and spatial locality, and hence the relativistic antecedence. The same
is true in the case of QTFD or QFT, and QFT in its standard form conforms to special relativity
(although there are nonrelativistic versions of QFT). Thus, the compatibility with relativistic or, more
generally, locality requirements would be maintained insofar as an already performed experiment
determines, probabilistically or (if repeated many times) statistically, a possible outcome of a future
experiment, without assuming classical causality. Determinism is, again, precluded on experimental
grounds. Whatever actually happens is defined by spatially and temporally local factors, although
the probabilistic or statistical predictions, could concern distant events, sometimes, as in the EPR-type
experiments, without previously performing a measurement on the object concerning which one
makes a prediction [34,72].
Relativistic causality is, thus, a manifestation of a more general concept or principle, that of
locality. This principle states that no instantaneous transmission of physical influences between
spatially separated physical systems (“action at a distance”) is allowed or that physical systems can
only be physically influenced by their immediate environment. It is true that locality is a spatial or
spatiotemporal concept (there is a temporal locality, which precludes, for example, retroaction in time
and backward in time causality), which makes it geometrical. However, although it is an effect of the
ultimate reality (which is, in RWR-type interpretations, beyond representation or even conception),
locality manifests itself only classically, in what is observed in measuring instruments, where the
geometrical considerations are fully applicable. Events, as we observe them, do happen in space and
time: otherwise they could not be observed. In general, some geometrical considerations remain
unavoidable in algebraic physical theories, such as QM, in part, again, because these theories must
relate, at least in term predictions to phenomena observed in space and time.
Locality of quantum phenomena and QM was at stake in the Bohr-Einstein debate from its
inception in the late 1920s, but especially following EPR’s paper [34,72]. As Bohr argued in his reply
to EPR’s paper (which argued that QM is either incomplete or else nonlocal), standard QM avoids
nonlocality, at least in the RWR-type interpretations of the theory and quantum phenomena themselves,
even though, as I said, under certain circumstances, such as those of the EPR-type experiments, QM can

118

www.pdfgrip.com
Entropy 2018, 20, 656

make predictions concerning the state of spatially separated systems, while, crucially to Bohr’s argument,
the physical circumstances of making these predictions and verifying them are local [73]. The question
of the locality of QM or quantum phenomena is, however, a matter of much debate and controversy,
especially in the wake of the Bell and Kochen-Specker theorems and related findings, as well as
numerous experiments dealing with correlations, beginning with, most famously, those by Aspect [74],
based on Bohm’s version of the EPR experiments. These debates cannot be addressed within the
scope of this article, and the literature dealing with these subjects is nearly as extensive as that on
interpretations of QM (e.g., [75–78]). As in Bohr’s exchange with EPR, the question of the relationships
between locality and realism, or a lack thereof, figures centrally in these debates and the findings
just mentioned.
Finally, I would like to propose the concept of quantum causality. I shall do so via Bohr’s concept
of complementarity, which Bohr saw as a generalization of causality. Complementarity is defined by:

(a) a mutual exclusivity of certain phenomena, entities, or conceptions; and yet

(b) the possibility of considering each one of them separately at any given point; and
(c) the necessity of considering all of them at different moments for a comprehensive account of the
totality of phenomena that one must consider in quantum physics.

Complementarity may be seen as a reflection of the fact that, in a radical departure from classical
physics or relativity, the behavior of quantum objects of the same type, say, electrons, is not governed,
individually or collectively, by the same physical law, in all possible contexts, specifically in complementary
contexts. Speaking of “physical law” in this connection requires caution, because, in Bohr’s interpretation,
there is no physical law representing this behavior, not even a probabilistic law if one adopts a statistical,
rather than a Bayesian, view of the individual quantum behavior. The behavior of quantum objects
leads to mutually incompatible observable physical effects in complementary contexts. On the other
hand, the mathematical formalism of QM offers correct probabilistic or statistical predictions (no other
predictions are possible) of quantum phenomena in all contexts.
It follows, especially if one adopts an RWR-type of interpretation, that the nature of both
experimental and theoretical physics changes. Experimentally we no longer track, as we do in classical
physics or relativity, the independent behavior of the systems considered, track, in effect geometrically,
what happens in any event. Instead we define what will happen in the experiments we perform,
by how we experiment with nature by means of our experimental technology, even though and
because we can only predict what will happen probabilistically or statistically. Thus, in the double-slit
experiment, the two alternative setups of the experiment, whether we, respectively, can or cannot know,
even in principle, through which slit each particle, say, an electron, passes, we obtain two different
outcomes of the statistical distributions of the traces on the screen (with which each particle collides).
Although sometimes associated with the “wave” behavior of quantum objects, the “interference
pattern” observed in the second setup is a statistically ordered pattern of discrete traces left by the
collisions between the particles and the screen. This is one of the reasons why Bohr avoided speaking
of wave-particle complementarity, even though the latter is commonly used to illustrate Bohr’s concept.
In Bohr’s view, quantum objects could not be represented either in terms of particles or in terms of
waves, and the pattern of traces in questions were, again, only effects on the interactions between
quantum objects and measuring instruments. Or, in effect equivalently to the double-slit experiments,
we can set up our apparatus so as to measure and correspondingly predict, again, probabilistically or
statistically, either the position or the momentum of a given quantum object, but never both together.
Either case requires a separate experiment, incompatible with the other, rather than representing “the
arbitrarily picking out of different elements of physical reality at the cost of other such elements [all
pertaining to the same quantum object]” within the same physical situation, by tracking either one of
its aspects or the other, as we do in classical mechanics [79]. There, this is possible because we can,
at least in principle, assign simultaneously both quantities within the same experimental arrangement.
In quantum physics, we cannot. Quantum physics, again, changes what experiments do: they define

119

www.pdfgrip.com
Entropy 2018, 20, 656

what will happen, rather than follow what is bound to happen in accordance with classical causality.
It is true that we can sometimes define by an experiment what will happen in classical physics. In this
case, however, we can then observe the resulting process without affecting it by observation. This is
not the case in quantum physics, because any new observation defines a new course of events. Only some
observations do in classical physics. By the same token, at least in RWR-type interpretations, quantum
theory only tells us possible things about the future, never about the past, which is only determined
by measurements.
It is this probabilistic or statistical determination (which precludes classical causality but respects
locality) of what can happen as a result of our conscious decision concerning which experiment to
perform at a given moment in time, that defines what I call “quantum causality” [80]. Whatever is
registered as a quantum event (providing the initial data) defines a possible set of, probabilistically
or statistically, predictable future events, outcomes of possible future experiments. This definition is
in accord with recent views causality in quantum information theory (e.g., [81–83]), except that it is
linked to our conscious decision concerning experiments we perform, which is rarely considered. It is,
however, this aspect of the situation that brings complementarity into play, because, in complementary
situations, such a decision irrevocably rules out the possibility of making any predictions concerning
certain other, complementary, events.
With these considerations in mind one can understand Bohr’s view of complementarity as a
generalization of causality [84]. On the one hand, “our freedom of handling the measuring instruments,
characteristic of the very idea of experiment” in all physics, our “free choice” concerning what kind of
experiment we want to perform is essential to complementarity [79]. On the other hand, as against
classical physics or relativity, implementing our decision concerning what we want to do will allow us
to make only certain types of predictions and will exclude the possibility of certain other, complementary,
types of predictions. Complementarity generalizes causality in the absence of classical causality and,
in the first place, realism, because it defines which reality can and cannot be brought about by our
decision concerning what experiment to perform.

3. Geometry and Algebra in Physics and Beyond

However one assesses Einstein’s skeptical attitude toward the Heisenberg method, his view of
it as algebraic was, I argue here, correct, and it helps one to better understand and contextualize
Heisenberg’s discovery. Before I consider this discovery itself, however, a more proper examination
of this characterization is necessary. First, I briefly revisit the basic understanding of geometry and
algebra, starting with algebra, which is more straightforward, because geometry involves further
complexities, in part by virtue of always containing algebraic components, while algebra can be free
from geometry. Most generally, algebra is, as said, the mathematical formalization of the relationships
between symbols, which makes it part of all mathematics, at least all modern mathematics (as the
ancient Greek geometry only contained arithmetic). Thus, for example, both mathematical logic and
calculus are forms of algebra in this general sense, even though, as fields, each contains its specificity.
Another, narrower or field specific sense of algebra, that referring to algebraic structures such as groups
or associative algebras, is equally crucial to quantum theory and its algebraic character. (As noted,
symmetry groups are also crucial to geometry and geometrical physical theories, such as relativity,
but there, too, they form part of the algebraic structures associated to geometrical and topological
ones.) While the role of symmetry groups in quantum theory only became apparent a few years after
Heisenberg’s discovery, it was well in place by the time of Einstein’s comment on the algebraic nature
of the Heisenberg method in 1936. Third, algebra also refers to, and was born from, the study of
algebraic equations, which still define a large part of the mathematical discipline of algebra. Forth,
finally, the algebra of probability is central, at a fundamental, rather than, as in classical physics, merely
practical, level in quantum theory. Heisenberg’s algebraic method and, following his work, quantum
theory encompasses all these senses and aspects of algebra, some of which are shared with classical
physics or relativity.

120

www.pdfgrip.com
Entropy 2018, 20, 656

Geometry, generally defined here as the mathematical formalization of spatiality, especially

(although not only) in terms of measurement is a more complex matter, because, on the one hand,
this formalized spatiality still connects to our general phenomenal intuition, including visualization,
of spatiality, and on the other, the role of mathematical formalization in geometry connects it to algebra.
This connection allows one to generalize geometrical or topological objects far beyond anything our
phenomenal intuition can access, especially by means of visualization. I would like now to address
some of these complexities, as they pertain to my argument concerning the algebraic character of
quantum theory. It would not be possible to treat this complex subject more generally.
Thus, we speak of and can rigorously define Hilbert spaces and their geometry, and in this sense,
one could speak, as Dirac was reportedly fond of doing, of geometrical thinking in QM or QFT.
I would argue, however, that these spaces and geometries are such primarily by extrapolation or,
as it were, metaphorically, while, rigorously, they are essentially forms of algebra, as against the
more standard forms of geometry, such as Euclidean or even non-Euclidean geometry, or differential
geometry, used in general relativity, which are still more closely connected to our phenomenal intuition
of spatiality. I would contend that this distinction is warranted and useful even though it is not
absolute and there are gray areas. The geometries of these more conventional spaces are defined
by certain algebraic properties and relations (beginning with metric), some of which can be used to
define more abstract objects. Thus, these properties and relations define the structure of Hilbert spaces
or, similarly, other mathematical “spaces,” such as those of projective geometries, abstract algebraic
varieties, the spaces of noncommutative geometry, geometric groups, and so forth, in the absence of
certain other, more conventionally spatial elements and structures, geometrical or topological, found
in more conventional spatial objects, such as and in particular R3 . The latter is the mathematical space
that is the closest to our phenomenal spatiality, even though some of its mathematical (topological
and geometrical) properties, beginning with continuity are far beyond our phenomenal intuition.
As Weyl argued: “the conceptual world of mathematics is so foreign to what the intuitive continuum
presents to us that the demand for coincidence between the two must be dismissed as absurd” [85].
“Coincidence” is, of course, not the same as “relation,” which might be unavoidable, at least in that
it is difficult to think of continuity or spatially apart from one or other phenomenal intuitions of it.
On the other hand, it is entirely possible to mathematically define continuous mathematical objects,
such as R3 , algebraically. This is why I speak of the extrapolated or metaphorical character of “space”
and “geometry” of such objects as Hilbert spaces. Technically, R3 is a Hilbert space too, but it need not
be considered as such, while Hilbert spaces considered in quantum theory do. This extrapolated or
algebraic character arises because of the infinite-dimensional nature of some of those spaces or because
they are defined over C, which appear irreducible in QM. While the Hilbert spaces involved are finite
in the case of discrete variables in QTFD, they can have higher dimensions and are over C.
It is difficult to be certain, especially from reported statements, what Dirac exactly had in mind in
his appeals to geometrical thinking in quantum theory. If, however, one is to judge by his writings,
they appear to suggest that at stake were algebraic properties and relations modeled on those found
in geometrical objects, as just explained. Indicatively, notwithstanding his insistence on the role of
geometrical thinking in Dirac, Darigold’s analysis of this thinking shows precisely the significance
of this type of algebra there. Thus, he says: “roughly, Dirac’s quantum mechanics could be said to
be to ordinary mechanics what noncommutative geometry is to intuitive geometry” [86]. However,
noncommutative geometry, the invention of which was in part inspired by the mathematics of QM,
is a form of this kind of algebra [87,88].
In what sense, then, apart from being defined by such algebraic structures, may such spaces be
seen as spaces, in particular, as relates to our phenomenal intuition, including visualization, from
which QM departs, although, in Bohr’s or the present view it breaks with any representation of
quantum objects and behavior, including a mathematical one? The subject is complex and it is far
from sufficiently explored in cognitive psychology and related fields, an extensive research during
recent decades notwithstanding, including as concerns cultural or technological (digital technology

121

www.pdfgrip.com
Entropy 2018, 20, 656

in particular) factors affecting our spatial thinking. Accordingly, it would be difficult to make any
definitive claims. It does appear, however, that, these factors notwithstanding, our three-dimensional
phenomenal intuition is shared by us cognitively and even neurologically in shaping our sense of
spatiality. Part of this sense appear to be Euclidean, insofar as it corresponds to what is embodied
in R3 (again, a mathematical concept), keeping in mind that the idea of empty space, apart from
bodies of one kind or another defining or faming it, is an extrapolation, because we cannot have
such a conception phenomenally or, as Leibniz argued against Newton, physically. We can have
a mathematical conception of space itself. To what degree our phenomenal spatiality is Euclidean
remains an open question, for example, in dealing with the visual perception of extent and perspective
(e.g., [89–91]).
It is nearly certain, however, that, when we visualize such algebraically defined spatial objects
as Hilbert spaces or even more conventional geometrical spaces or geometries, once the number of
dimensions is more than three, or for spaces of any dimensions defined over fields other than R, such as
C or algebraic fields of finite characteristics, we visualize only three- (and even two-) dimensional
configurations and supplement them by algebraic structures and intuitions. Feynman instructively
explained this process in describing visual intuition in thinking about quantum objects [92]. Obviously,
such anecdotal evidence hardly suffices for any definitive claim, but it appears to be in accord
with some of the current neurological and cognitive-psychological research, as just mentioned,
which suggests the dependence of our spatial intuition, including visualization, on two- and three-
dimensional phenomenality.
As noted earlier, according to Bohr and Heisenberg, classical mechanics, a mathematical science of
bodies and motion in space and time, is a mathematical refinement of our daily phenomenal thinking
concerning space, time, and motion, no longer workable in quantum theory or, again, even in relativity
in considering photons or motions with speed close to the speed of light in a vacuum, in which cases,
however, the behavior of the objects considered can be properly handled algebraically, in quantum
theory in probabilistic or statistical terms [56,93]. Accordingly, one could maintain the difference that I
argue for between geometrical character of classical mechanics or (with qualifications) relativity and
the algebraic character of quantum theory, especially in RWR-type representations of the latter.
This is not to say that the geometrical or topological character of such objects disappears; quite
the contrary, this character remains crucial on at least two counts. First, algebra has a special form
that may be called “spatial algebra.” Selecting a good term poses some difficulties because such
suitable terms as “geometric algebra” and “algebraic geometry” are already in use for designating,
respectively, the Clifford algebra over a vector space with a quadratic form, and the study of algebraic
varieties, defined as the solutions of the systems of polynomial equations. “Spatial algebra” arises
from algebraic structures that mathematically define what are conventionally known as geometrical
or topological objects, such as projective spaces or topological manifolds, and reflect their proximity
to R3 and mathematical spatial objects that are close to our phenomenal intuition, and the geometry
and topology associated with it. This proximity may be left behind in the rigorous mathematical
treatment of such objects, beginning with R3 . The same type of spatial algebra also defines objects,
from projective and finite (topologically discrete) spaces to the infinite-dimensional spaces, where these
connections to phenomenally visualizable spatial objects are only spatial algebraic, unless used by
an extrapolation or metaphorically, as in the case of a projective space (a set of lines through the
origin of a vector space, such as R2 in the case of projective plane, with projective curves defined
algebraically, as algebraic varieties) or an infinite-dimensional Hilbert space (the points of which are
typically square-integrable functions or infinite series). Spatial algebra is the algebraization of spatiality
that makes it rigorously mathematical.
At the same time, however, and this is the second count on which mathematical objects defined
by spatial algebra retain their connections to geometrical and topological thinking, analogies with
R3 continue to remain useful and even indispensable. Such analogies may be rigorous (and hence
algebraic) or phenomenally intuitive or metaphorical. Thus, the analogues of the Pythagorean theorem

122

www.pdfgrip.com
Entropy 2018, 20, 656

or parallelogram law in Euclidean geometry, which holds in infinite-dimensional Hilbert spaces

over either R or C, are important, including in applications to physics, such as QM. More generally,
our thinking concerning geometrical and topological objects is not entirely translatable into algebra.
This was well understood by Hilbert in his axiomatization of Euclidean geometry, even though this
axiomatization had a spatial algebraic character, including in establishing an algebraic model (the field
of complex numbers) for his system of axioms in order to prove its consistency [94].
One can gain a further insight into this situation by considering a related principle, “Think
Geometrically, Prove Algebraically,” advanced by Tate, whose thinking bridged number theory and
algebraic geometry in highly original and profound ways. The principle was introduced in the book
(with Silverman), on “the rational points of elliptic curves.” The title-phrase itself combines algebra
(“rational points”) and “geometry” (curves), and implies that geometry, at least beyond that of R3 and
even there, requires algebra to be mathematically rigorous. According to Silverman and Tate [95]:
It is also possible to look at polynomial equations and their solutions in rings and fields
other than Z or Q or R or C. For example, one might look at polynomial with coefficients
in the finite field Fp with p elements and ask for solutions whose coordinates are also in the
field Fp. You may worry about your geometric intuitions in situations like this. How can
one visualize points and curves and directions in A2 when the points of A2 are pairs (x, y)
with x, y ∈ Fp ? There are two answers to this question. The first and most reassuring is that
you can continue to think of the usual Euclidean plane, i.e., R2 , and most of your geometric
intuitions concerning points and curves will still be true when you switch to coordinates
in Fp . The second and more practical answer is that the affine and projective planes and
affine and projective curves are defined algebraically in terms of ordered pairs (r, s) or
homogeneous triples [a, b, c] without any reference to geometry. So in proving things one can
work algebraically using coordinates, without worrying at all about geometrical intuitions.
We might summarize this general philosophy as: Think Geometrically, Prove Algebraically”.
The affine and projective planes and curves can, in principle, be defined without any reference to
ordinary language and concepts, which are more difficult and perhaps impossible to avoid in geometry,
that is, in the kind of intuitive geometry they refer to, rather than spatial algebra that ultimately defines
all geometry rigorously. Rigorously, in these cases, we think algebraically, too, by using spatial algebra,
even if with the help of geometrical intuitions. It is true that a mathematician like Tate can develop
and use intuition in dealing with discrete geometries as such, say, that of the Fano plane of order 2,
which has the smallest number of points and lines (seven each). However, beyond the fact that they
occur in the two-dimensional regular plane, the diagrammatic representations of even the Fano plane
are still difficult to think of as other than spatial algebra, in this case, combinatorial in character. For the
moment, while useful and even indispensable, our Euclidean intuitions are limited even when we deal
with algebraic curves in the usual Euclidean plane, let alone in considering something like a Riemann
surface as a curve over C, or curves in finite geometries, abstract algebraic varieties, Hilbert spaces,
the spaces of noncommutative geometry, or geometric groups, a great example of the extension, by a
reversal, of spatial algebra to conventionally algebraic objects as it is. This also means, as Tate must
have been aware, that mathematical thinking concerning geometrical and topological objects cannot
be reduced to those naïve intuitions. Silverman and Tate’s next example from differential calculus,
that of finding a tangent line to a curve, confirms this point. The invention of calculus, an essentially
algebraic form of mathematics, was not about proving things algebraically, as the standard of proof
then was geometry, which compelled Newton to present his mechanics in terms of geometry in his
1687 Principia, in order, as he said, to assure a geometrical demonstration of his findings, also in the
direct sense of showing something by means of visualization, rather than in terms of calculus [96].
Calculus was about thinking algebraically, as was especially manifested in Leibniz’s version, rather
than about rigorous proofs.
Einstein was, then, quite correct in characterizing the Heisenberg method as “algebraic.”
The “geometry” of the Hilbert spaces of quantum theory only confirms the algebraic character of this

123

www.pdfgrip.com
Entropy 2018, 20, 656

method, especially manifested in his matrix or operator algebra. Still, a few additional qualifications
are necessary for rigorously maintaining Einstein’s and the present view of the Heisenberg method as
algebraic. First of all, in saying that “we must give up, in principle, the space-time continuum,” Einstein
must have had in mind the spacetime continuum in representing, by means of the corresponding
theory, the ultimate reality considered, and possibly in attributing the space-continuum to this reality.
The idea that this reality may ultimately be discrete had been around for quite a while by then: it was,
for example, proposed by Riemann as early as 1854, by way of a remarkable phrase “the reality
underlying space,” thus potentially divorcing on proto-RWR lines reality and realism, at least a form
of realism according to which our phenomenal or even mathematical representation of space was
continuous [97]. The idea of the discrete nature of the ultimate reality has acquired new currency in
view of QM and QFT, advocated by, among others, Heisenberg in the 1930s, and is still around.
One must also keep in mind the complexity of the algebra of QM or QFT, which involves objects
that are not, in general, discontinuous, although certain key elements involved are no longer continuous
functions, such as those in classical physics. Some continuous functions are retained, because the
Hilbert spaces involved are those of such functions, considered as infinite-dimensional vectors in
the case of continuous variables such as position and momentum, which variables themselves are
represented by operators. These functions are those of complex (rather than, as in classical physics,
real) variables and the vector spaces that they comprise or associated objects, such as operator algebras,
have special properties, such as noncommutativity. Indeed, given that it deals with Hilbert spaces,
QM or QFT involves mathematical objects whose continuity is denser than that of regular continua
such as the (real number) spacetime continuum of classical physics or relativity. In contrast to these
theories, however, the continuous and differential mathematics used in quantum theory, along with the
discontinuous algebraic one, relates, in terms of probabilistic predictions, to the physical discontinuity
defining quantum phenomena, which are discrete in relation to the observable spacetime continuum
and to each other, while, at least in non-realist, RWR-type, interpretations, quantum objects and their
behavior are not given any physical or mathematical representation—continuous or discontinuous
(or mixed). Born and Jordan developed a differential calculus, “symbolic differentiation,” as they
called it, for matrices used in quantum mechanics [8,98]. So did Dirac in his first paper on quantum
mechanics [99]. In the style of Leibniz, this differentiation was defined algebraically by using the
noncommutation rules. This quantum differentiation enables one to retain the differential equations of
classical mechanics and their accompanying machinery, such as and in particular the Poisson bracket,
while using new quantum variables. The quantum-mechanical analogue of the Poisson bracket is
the expression 2πi h ( pq − qp ), as Dirac was first to realize, with far-reaching implications for quantum
mechanics. Dirac’s starting point, again, in the style of Leibniz, was the quantum-mechanical analogue
of the rule for the differential of the product of two functions, which may be seen as a linear operator
and which may be suitably algebraically quantized [99].

4. From Geometry to Algebra, with Heisenberg

For nearly a century since the publication of von Neumann’s seminal The Mathematical Foundation
of Quantum Mechanics [48], the mathematical models of quantum theory—QM, QTFD, and QFT—have
been commonly defined in terms of the Hilbert-space formalism, which remains dominant, even though
there are other versions, such as C*-algebras and, more recently, the one based in category theory.
Von Neumann’s aim was to give a proper mathematical grounding to the quantum-mechanical
formalism, already developed by Heisenberg, Born, Jordan, Dirac, and others. By contrast, Heisenberg’s
aim was to find a successful theory accounting for the behavior of the electrons in the atoms, which he
accomplished with his new mathematical model, admittedly preliminary, but quickly developed into
a more rigorous model, matrix mechanics, by Born and Jordan [8]. Dirac offered the most general
version of the formalism before von Neumann, who thought that Dirac’s version lacked a proper
mathematical rigor, in part because of Dirac’s use of delta function, not considered mathematically

124

www.pdfgrip.com
Entropy 2018, 20, 656

legitimate. Eventually, in the 1940s, delta function was given a proper definition, as the so-called
“distribution,” a functional, by Schwartz, which legitimatized Dirac’s formalism mathematically.
As stated from the outset, Heisenberg abandoned the project of representing the behavior of
electrons in atoms. It is true that he only thought that such a representation was unlikely to be
achieved at the time, as opposed to arguing, as Bohr did in the 1930s, that such a representation
and even an analysis, if not conception, of quantum behavior was “in principle excluded” [35]. Still,
Heisenberg’s was an audacious and radical move, which decisively shaped Bohr’s subsequent thinking,
although Heisenberg’s thinking was in turn influenced by Bohr’s 1913 atomic theory. Bohr’s theory
only partially abandoned the geometrical representation of quantum behavior in the case of “quantum
jumps”—transitions between the stationary states of electrons, conceived in terms of electrons orbiting
the nuclei. While Bohr’s theory, as developed by him and others had major successes, by the early
1920s it proved to be ultimately unsustainable. This failure compelled Heisenberg to renounce a
geometrical representation of any quantum behavior in space and time, including that of stationary
states in terms of orbits. This renunciation led him to his discovery of QM. There was no longer any
algebraic representation of quantum objects and behavior either.
Heisenberg’s approach was grounded in a set of fundamental principles, in part stemming from
Bohr’s 1913 theory. As in the case of concepts, although one could sometimes surmise the meaning of
the term “principle” from its use, it is, similarly to the term “concept,” rarely defined or explained in
physical or even philosophical literature. His title notwithstanding, Heisenberg did not do so in his
first book, The Physical Principles of the Quantum Theory [100] nor did Dirac, in his famous Principles of
Quantum Mechanics, published in the same year [47]. Terms like “principle,” “postulate,” and “axiom,”
are often used in physics somewhat indiscriminately, and it is difficult to entirely avoid overlapping
between the concepts designated by these terms, or those designated as “laws,” especially because
physical principles often derive from (or give rise to) postulates or laws. It may also be a matter of the
functioning of these concepts. Thus, conservation laws are sometimes seen as conservation principles.
For present purposes, I shall adopt the concept of principle from Einstein’s concept of a “principle
theory,” which he introduced by way of juxtaposing this concept to that of a “constructive theory.”
This concept corresponds to the use of principles by Bohr and Heisenberg, and by quantum-information
theorists discussed in the next section. According to Einstein, constructive theories aim “to build
up a picture of the more complex phenomena out of the materials of a relatively simple formal
scheme from which they start out,” which, it follows, also make such theories realist. By contrast,
principle theories “employ the analytic, not the synthetic, method. The elements which form their
basis and starting point are not hypothetically constructed but empirically discovered ones, general
characteristics of natural processes, principles that give rise to mathematically formulated criteria
which the separate processes or the theoretical representations of them have to satisfy” [101]. I would
add the following qualification, which is likely to have been accepted by Einstein: Principles are not
empirically discovered but formulated on the basis of empirically established evidence. A principle
theory may also be a constructive theory, but it need not be, and Heisenberg’s mechanics was not.
Heisenberg’s approach and then Bohr’s interpretation of QM were grounded in the following three
principles (with Bohr’s principle or at least concept of complementarity added in 1927), which fit and
even embody the “equation” QUANTUMNESS → PROBABILITY → ALGEBRA and the QPA principle:

(1) the principle of discreteness, the QD principle, according to which all observable quantum
phenomena are individual and discrete in relation to each other, which is different from the
discreteness of quantum objects;
(2) the principle of the probabilistic or statistical nature of quantum predictions, the QP/QS principle,
which is maintained, in contrast to classical statistical physics, even in considering elemental
individual quantum processes, and is accompanied by a special, nonadditive, character of
quantum probabilities and rules, such as Born’s rule, for deriving them; and
(3) the correspondence principle, which, as initially understood by Bohr, required that the predictions
of quantum theory must coincide with those of classical mechanics in the classical limit, but which

125

www.pdfgrip.com
Entropy 2018, 20, 656

was given by Heisenberg a form of “the mathematical correspondence principle,” requiring that
the equations and variables of QM convert into those of classical mechanics in the classical limit.

To connect his formalism (defined over C) to the probabilities of outcomes of quantum experiments
(which probabilities are real numbers), Heisenberg used a version of the Born rule in the special
case of the transitions between stationary states, and not, as Born did, as universally applicable in
QM. Referring to stationary states requires caution, because stationary only means that the electrons
remained in their orbits with the same energy, were in the same “energy-state,” but would continuously
change their position or their “position-state” along each orbit. On the other hand, the electrons would
discontinuously, by quantum jumps, change their energy states, or their other states, by moving from
one orbit to another. In Heisenberg, there were no longer orbits but only states and discontinuous
transitions between states. As noted from the outset, one was no longer thinking, as in classical
mechanics, in terms of predictions, even probabilistic predictions, concerning a moving object, say,
an electron, free or orbiting the nucleus of an atom, but instead in terms of the probabilities of transitions
between the states of an electron, transitions that were always discrete. This type of thinking emerged
in Bohr’s 1913 theory in considering an electron’s transitions from one energy level to another, but,
following Heisenberg, came to define quantum physics in general as a physics of predicting discrete
transitions between states [12]. As Heisenberg said in his letter to Kronig (5 June 1925): “What I really
like in this scheme is that one can really reduce all interactions between atoms and the external world
. . . to transition probabilities” (cited in [102]).
Heisenberg’s scheme, thus, extended Bohr’s 1913 concept of discrete transitions or quantum jumps,
which, unlike the orbital behavior of electrons in stationary states, had no mechanical or geometrical
model, to all quantum behavior. The “concept of orbit,” as Heisenberg noted later, “had been somewhat
doubtful from the beginning,” because of “the discrepancy between the calculated orbital frequency of
the electrons and the frequency of the emitted radiation,” which had to be interpreted as a limitation to
the concept of the electronic orbit” [103]. Accepting this discrepancy and, thus, dissociating these two
types of frequencies was a revolutionary move on Bohr’s part, emphasized by Bohr himself: “How
much the above considerations differ from an interpretation based on the ordinary electrodynamics is
perhaps most clearly shown by the fact that we have been forced to assume that a system of electrons
will absorb radiation of a frequency different from the frequency of vibration of electrons calculated in
the ordinary way” [104]. Heisenberg rethought stationary states as just energy states, permitting no
mechanical model or geometrical representation. There were, again, only transitions between quantum
states (using the term “state” physically, rather than mathematically, as “a state vector”).
By speaking of the “interactions between atoms and the external world,” Heisenberg’s statement
in his letter to Kronig also suggests that QM, as he saw it, was about (predicting) these interactions,
specifically with the measuring instruments involved, a view manifested in Heisenberg’s paper and
adopted by Bohr. All that one could say about quantum objects and behavior could only concern their
effects on measuring instruments, probabilistically or statistically predictable by QM.
The mathematical correspondence principle motivated Heisenberg’s decision to retain the
equations of classical mechanics, while, necessarily, introducing different variables, both, however,
being now equally parts of his algebra. The correspondence with classical theory could still be
maintained because new variables could be substituted for conventional classical variables (such
as those of position and momentum) in the classical limit, as for large quantum numbers, when the
electrons were far away from the nuclei and when classical concepts, such as orbits, could apply,
thus also restoring the geometry of classical mechanics. The electrons’ behavior itself was still quantum
and certain quantum effects, not observed in dealing with classical objects, could be observed, effects
predictable only by the algebra of QM. The old quantum theory was defined by the strategy of
retaining the variables of classical mechanics while adjusting the equations to achieve better predictions.
Heisenberg’s reversal of this strategy was, thus, unexpected, and it required a radical change in the
role these equations were to play. They no longer represented the motion of electrons, but served

126

www.pdfgrip.com
Entropy 2018, 20, 656

as the mathematical means enabling probabilistic or statistical predictions concerning effects of the
interaction between electrons and measuring instruments.
Heisenberg’s discovery was a remarkable achievement, ranked among the greatest in the history
of physics. A detailed discussion of his derivation of QM is beyond my scope [105]. Several key
features of his thinking are, however, worth commenting on, following [105], to further illustrate
the algebraic nature of his thinking. Heisenberg’s new quantum variables were infinite unbounded
matrices with complex elements. Their multiplication, which Heisenberg, who was famously unaware
of the existence of matrix algebra and reinvented it, had to define, is in general not commutative.
Essentially, these variables are operators in Hilbert spaces over C. Such mathematical objects had never
been used in physics previously, and their noncommutative nature was, initially, questionable and
even off-putting for some, including Heisenberg himself and Pauli [106]. In fact, while matrix algebra,
in finite and infinite dimensions, was developed in mathematics by then, unbounded infinite matrices
were not previously studied. As became apparent later, such matrices are necessary to derive the
uncertainty relations for continuous variables. There are further details: for example, as unbounded
self-adjoint operators, defined on infinite dimensional Hilbert spaces, these matrices do not form an
algebra with respect to the composition as a noncommutative product, although some of them satisfy
the canonical commutation relation. These details are, however, secondary. Most crucial was that
the concept was used in a fundamentally new way. Heisenberg’s variables were algebraic entities
enabling probabilistic or statistical predictions concerning quantum phenomena, observed in measuring
instruments, without providing a mathematically idealized representation, geometrical or other, of the
spacetime behavior of quantum objects responsible for these phenomena.
In this regard, although understandable historically, the term “observables” is misleading and is
especially inadequate if one adopts a RWR-type view. As noted, Bohr saw the quantum-mechanical
formalism as “symbolic” in the following, essentially algebraic, sense. While the mathematical symbols
used in it appear, as variables, in the same equations as those used in classical mechanics, these
symbols did not represent physical quantities pertaining to quantum objects themselves and their
behavior, in the way such symbols do in classical mechanics. By the same token, the equations of QM,
Schrödinger’s equation included, no longer functioned as equations of motion. Instead they are part of
the probabilistic algebra of QM, enabling us to compile, in Schrödinger’s terms, “expectation-catalogs”
concerning events observed in quantum phenomena, which, in RWR-type interpretations, gives
Schrödinger’s equation an algebraic character along with a probabilistic one [63]. Schrödinger’s
waves were symbolic waves, symbolizing these expectation-catalogs. If Schrödinger’s equation may
be seen as “deterministic,” as it is sometimes, it is only in the sense that it strictly determines such
expectation-catalogs, which are, however, catalogs of predictions that are not deterministic even in
realist interpretations.
In his 1925 paper, introducing QM, Heisenberg began his derivation with an observation that
reflects a radical departure from the classical ideal of continuous mathematical representation of
individual physical processes, which would connect discrete quantum events. He says: “in quantum
theory it has not been possible to associate the electron with a point in space, considered as a function of
time, by means of observable quantities. However, even in quantum theory it is possible to ascribe to
an electron the emission of radiation” [the effect of which is observed in a measuring instrument] [107].
Technically, a measurement could associate an electron with a point in space, but not by linking
this association to a function of time representing the continuous motion of this electron, in the
way it is possible in classical mechanics. Matrix mechanics did not offer a treatment of stationary
states, when and only when one could in principle speak of the position of an electron in an atom,
although while there are stationary energy states, with the same energy-levels, an electron itself is
never stationary. Only an instantly repeated measurement can give the same value of its position,
which instant repetition is an idealization. Heisenberg described his next task as follows: “In order to

127

www.pdfgrip.com
Entropy 2018, 20, 656

characterize this radiation we first need the frequencies which appear as functions of two variables.
In quantum theory these functions are in the form” [107]:

v(n, n − α) = 1/h {W(n) − W (n − α)}

and in classical theory in the form

v(n, α) = αv(n) = α/h(dW/dn)

This difference leads to a difference between classical and quantum theories as regards the
combination relations for frequencies, which, in the quantum case, correspond to the Rydberg-Ritz
combination rules, reflecting, to return to Heisenberg’s locution, “the discrepancy between the
calculated orbital frequency of the electrons and the frequency of the emitted radiation.” However,
“in order to complete the description of radiation [in correspondence, by the correspondence principle,
with the classical Fourier representation of motion] it is necessary to have not only frequencies but
also the amplitudes” [107]. On the one hand, then, the new, quantum-mechanical equations must
formally contain amplitudes, as well as frequencies. On the other hand, these amplitudes could no
longer serve their classical physical function (as part of a continuous representation of motion) and are
instead related to discrete transitions between stationary states. In Heisenberg’s theory and in QM
since then, these “amplitudes” are no longer amplitudes of physical motions, which makes the name
“amplitude” a symbolic term. In commenting on linear superposition in quantum mechanics in his
classic book, Dirac emphasized this difference: “the superposition that occurs in quantum mechanics is of
an essentially different nature from any occurring in the classical theory” [108]. In RWR-type interpretations,
this superposition is not even physical: it is only mathematical. In classical physics the mathematics of
(wave) superpositions represents physical processes; in QM, at least in the nonrealist view, it does not.
Amplitudes are instead linked to the probabilities of transitions between stationary states: they are
what became known as probability amplitudes. The corresponding probabilities are derived by a form
of Born’s rule for this limited case (technically, one needs to use the probability density functions).
The standard rule for adding the probabilities of alternative outcomes is changed to adding the
corresponding amplitudes and deriving the final probability by squaring the modulus of the sum.
The mathematical structure thus emerging is in effect that of vectors and (in general,
noncommuting) Hermitian operators in Hilbert spaces over C, which are infinite-dimensional, given
that one deals with continuous variables. Heisenberg explains the situation in these terms in [100].
In his original paper, which reflect his thinking more directly, he argues as follows [107]:

The amplitudes may be treated as complex vectors, each determined by six independent
components, and they determine both the polarization and the phase. As the amplitudes are
also functions of the two variables n and α, the corresponding part of the radiation is given
by the following expressions:
Quantum-theoretical:
Re{A(n, n − α)eiω(n, n − α)t }

Classical
Re{Aα (n)eiω(n)αt }

The problem—a difficult and, “at first sight,” even insurmountable problem—is that “the phase
contained in A would seem to be devoid of physical significance in quantum theory, since in
this theory frequencies are in general not commensurable with their harmonics” [109]. As noted,
this incommensurability, which is in an irreconcilable conflict with classical electrodynamics, was one
of the most radical features of Bohr’s 1913 atomic theory, on which Heisenberg builds here. His strategy
is still based on the shift from calculating the probability of finding a moving electron in a given state
to calculating the probability of an electron’s transition from one state to another, without describing

128

www.pdfgrip.com
Entropy 2018, 20, 656

the physical mechanism responsible for this transition. Heisenberg’s theory is more in harmony with
this approach because there are no longer orbits, where the classical approach would still apply.
Heisenberg says next: “However, we shall see presently that also in quantum theory the phase
has a definitive significance which is analogous to its significance in classical theory” [109]. “Analogous”
could only mean here that, rather than being analogous physically, the way the phase enters
mathematically is analogous to the way the classical phase enters mathematically in classical theory,
in accordance with the mathematical form of the correspondence principle, insofar as quantum- mechanical
equations are formally the same as those of classical physics. Heisenberg only considered a toy model
of an aharmonic quantum oscillator, and thus needed only a Newtonian equation for it, rather than
the Hamiltonian equations required for a full-fledged theory, developed by Born and Jordan [8,110].
As Heisenberg explains, if one considers “a given quantity x(t) [a coordinate as a function of time] in
classical theory, this can be regarded as represented by a set of quantities of the form” [109]:

Aα (n)eiω(n)αt ,

which, depending on whether the motion is periodic or not, can be combined into a sum or integral
which represents x(t) [60]:
α=+∞
x (n, t) = ∑ Aα (n)eiω (n)αt
α=−∞
or:  +∞
x (n, t) = Aα (n)eiω (n)αt dα
−∞
Heisenberg next makes his most decisive and most extraordinary move. He notes that “a similar
combination of the corresponding quantum-theoretical quantities seems to be impossible in a unique
manner and therefore not meaningful, in view of the equal weight of the variables n and n − α.
However, he says, “one might readily regard the ensemble of quantities A(n, n − α)eiω(n, n − α)t
[an infinite square matrix] as a representation of the quantity x(t)” [109]. The arrangement of the
data into these ensembles, in effect square tables, was a remarkable way to handle the transitions
between stationary states, although in retrospect it is also a natural way to do so, but only in retrospect.
However, it does not by itself establish an algebra of these arrangements, for which one needs to find
the rigorous rules for adding and multiplying these elements. Otherwise Heisenberg cannot use these
variables in the equations of his new mechanics. To produce a quantum-theoretical version of the
classical equation of motion considered, which would apply (no longer as an equation of motion) to
these variables, Heisenberg needs to be able to construct the powers of such quantities, beginning with
x(t)2 , which is actually all that he needs for his equation. The answer in classical theory is obvious
and, for the reasons just explained, obviously unworkable in quantum theory, where, Heisenberg
proposes, “it seems that the simplest and most natural assumption would be to replace classical
[Fourier] equations . . . by” [111]:

α=+∞
B(n, n − β)eiω (n,n− β)t = ∑ A(n, n − α) A(n − α, n − β)eiω (n,n− β)t
α=−∞

or:  +∞
= A(n, n − α) A(n − α, n − β)eiω (n,n− β)t dα
−∞
This is the main mathematical postulate, the (matrix) multiplication postulate, of Heisenberg’s
theory, “an almost necessary consequence of the frequency combination rules” [111].
Although it is commutative in the case of x2 , this multiplication is in general noncommutative,
expressly for position and momentum variables, and Heisenberg, without quite realizing, used this
noncommutativity in solving his equation, as Dirac was the first to notice. Heisenberg spoke of his
new algebra of matrices as the “new kinematics.” This was not the best choice of term because his new

129

www.pdfgrip.com
Entropy 2018, 20, 656

variables no longer described or were even related to motion as the term kinematic would suggest,
one of many, historically understandable, but potentially confusing terms. Planck’s constant, h, which is
a dimensional, dynamic entity, has played no role thus far. Technically, in Einstein’s view, the theory
wasn’t even a mechanics: it did not offer a representation of individual quantum processes, but only
predicted, probabilistically or statistically, what is observed in measuring instruments. To make these
predictions, one will need Planck’s constant, h.
That in general his new variables did not commute, PQ − QP = 0, was, again, an especially novel
feature of Heisenberg’s theory, confirming its essentially algebraic nature. This feature, which was,
again, an automatic consequence of his choice of variables, proved to be momentous physically.
Most famously, it came to represent the uncertainty relations constraining certain simultaneous
measurements, such as those of the momentum (P) and the coordinate (Q), associated with a
given quantum object in the mathematical formalism of quantum mechanics and (correlatively)
the complementary character of such measurements. Given, however, the nature of the situation to
which Heisenberg’s new mechanics responded, the noncommutative character of quantum variables
should not be surprising. Schwinger instructively commented on the subject in his unpublished lecture,
cited at length in [112]. He notes the most commonly stated physical feature corresponding to this
character, namely, that if one measures two physical properties in one order, and then in the other,
the outcome would in general be different. But he goes further in explaining why this is the case and
its implications for the formalism [112]:

If we once recognize that the act of measurement introduces in the [microscopic] object
of measurement changes which are not arbitrarily small, and which cannot be precisely
controlled . . . then every time we make a measurement, we introduced a new physical
situation and we can no longer be sure that the new physical situation corresponds to the
same physical properties which we had obtained by an earlier measurement. In other words,
if you measure two physical properties in one order, and then the other, which classically
would absolutely make no difference, these in the microscopic realm are simply two different
experiments . . .
So, therefore, the mathematical scheme can certainly not be the assignment, the association,
or the representation of physical properties by numbers because numbers do not have this
property of depending upon the order in which the measurements are carried out. . . . We
must instead look for a new mathematical scheme in which the order of performance of
physical operations is represented by an order of performance of mathematical operations.

This is not how Heisenberg discovered QM, in particular given that the noncommutativity of some
among the operators representing quantum observables was not his starting point but a consequence
of the multiplication rule for his matrices. The type of thinking described by Schwinger is more in
accord with quantum-informational approaches to deriving quantum theory, primarily QTFD, from the
(formalized) structure of quantum measurements, or as Schwinger revealingly put it the “measurement
algebra.” [113,114]). The difficulty here is that any such scheme requires a mathematics that does not
appear to be naturally connected to this measurement algebra, for one thing, because of the use of
complex, rather than real, variables, and rules, such as Born’s rule, by means of which this scheme
would be related to the probabilities or statistics of quantum predictions, which are real numbers.
There is no homomorphic (let alone isomorphic) mapping from measurement algebra to the algebra of
QM, “the new mathematical scheme.” One needs additional pieces of structure to arrive at this scheme.
In Heisenberg, these additional pieces we partly borrowed from classical physics, formally defining
his equations, and partly invented by Heisenberg in finding the variables needed.
As explained earlier and as Schwinger stresses in his lecture, no identical assignment of the
single quantity is ever possible, or in any event ever guaranteed, in two “identically” prepared
experiments in the way it can be in classical physics [112]. This is because quantum experiments cannot
be controlled so as to identically prepare quantum objects but only so as to identically prepare the

130

www.pdfgrip.com
Entropy 2018, 20, 656

measuring instruments involved, because this behavior can be considered classical. The quantum
strata of measuring instruments, through which they interact with quantum objects, do not affect these
preparations but only the outcomes of measurements. This interaction is uncontrollable. This fact is
central to Bohr’s argument, which invokes this “finite and incontrollable interaction” at key junctures
of his reply to EPR’s paper [34]. Hence, as noted earlier, the outcomes of repeated identically prepared
experiments, including those involving sequences of measurements, cannot be controlled even ideally
(as in classical physics), and these outcomes will, in general, be different. This circumstance makes
statistical considerations unavoidable, as reflected, among other things, in the statistical character of
the uncertainty relations, inherent in Heisenberg’s formula, ΔqΔp ∼ = h. The noncommutative nature of
the corresponding variables responds to this character, along with the uncertainty relations themselves.
The quantum-mechanical situation that emerged with Heisenberg’s discovery of quantum
mechanics and then Bohr’s interpretation of it was (sometime in the late 1930s) recast by Bohr in terms
of his concept of “phenomenon,” defined by what is observed in measuring instruments under the
impact of quantum objects, in contradistinction to quantum objects themselves, which could not be
observed or represented, or in the present view, even conceived of. According to Bohr [115]:
I advocated the application of the word phenomenon exclusively to refer to the observations
obtained under specified circumstances, including an account of the whole experimental
arrangement. In such terminology, the observational problem is free of any special intricacy
since, in actual experiments, all observations are expressed by unambiguous statements
referring, for instance, to the registration of the point at which an electron arrives at a
photographic plate. Moreover, speaking in such a way is just suited to emphasize that
the appropriate physical interpretation of the symbolic quantum-mechanical formalism
amounts only to predictions, of determinate or statistical character, pertaining to individual
phenomena appearing under conditions defined by classical physical concepts [describing
the observable parts of measuring instruments].
Phenomena are discrete in relation to each other, and, in Bohr’s scheme, one cannot assume
that there are continuous processes that connect them, especially classically causally, even in dealing
with elemental individual processes and events. Part of Bohr’s concept of phenomenon and the main
reason for its introduction was that this concept “in principle exclude[s]” any representation or analysis,
even if not a possible conception, of quantum objects and their behavior, at least, by means of QM [35].
The concept is, thus, correlative to the RWR-type view, reached by Bohr at this stage of his thinking,
at least the weak RWR view. Physical quantities obtained in quantum measurements and defining
the physical behavior of certain (classically described) parts of measuring instruments are effects of
the interactions between quantum objects and these instruments, and do not pertain to quantum
objects themselves. It is often forgotten by those who comment of Bohr’s insistence on the role of
classical concepts in quantum theory that Bohr clearly realized and relied on the fact that the measuring
instruments used in quantum experiments also have quantum parts through which they interact with
quantum objects. Otherwise quantum measurements and the effects defining phenomena would not
be possible. These effects are manifested by classical states of these parts of measuring instruments,
to which these quantum interactions are “irreversibly amplified” [41]. The language of effects (in the
absence of classical causes), found throughout Bohr’s writings on quantum mechanics, becomes
especially prominent in his later articles, presenting his ultimate interpretation (e.g., [116]).
These effects are no longer assumed to correspond to any properties of quantum objects, even to
single such properties, rather than only certain joint properties, in accordance with the uncertainty
relations. An attribution of even of a single property, such as that of “position,” “moment in time,”
“momentum,” or “energy,” or even invariant properties, such as the rest mass and charge of a particle,
which are defined by the fact that they are the same in all measurements, to any quantum object is never
possible—before, during, or after measurement. One could only rigorously specify measurable quantities
physically pertaining to measuring instruments. Even when we do not want to know the momentum
or energy of a given quantum object and thus need not worry about the uncertainty relations, neither

131

www.pdfgrip.com
Entropy 2018, 20, 656

the exact position of this object itself nor the actual time at which this “position” is established is ever
available and, hence, in any way verifiable. Any possible information concerning quantum objects as
independent entities is lost in “the finite [quantum] and uncontrollable interaction” between them and
measuring instruments [34]. However, this interaction can leave a mark in measuring instruments,
a mark, a bit of information, that can be treated as part of a permanent, objective record, which can be
discussed, communicated, and so forth. The uncertainty relations, too, now apply to the corresponding
(classical) variables of suitably prepared measuring instruments, impacted by quantum objects. We can
either prepare our instruments so as to measure or predict a change of momentum of certain parts of
those instruments or so as to locate the spot that registers an impact by a quantum object, but never do
both in the same experiment. The uncertainty relations are correlative to the complementary nature of
these arrangements.
Wheeler spoke of “law without law” in quantum theory [117]. One might see this concept, via the
combination of the QPA and the RWR principles, as the algebraic probabilistic law of QM, without any
law that would be assumed to govern the independent behavior of quantum objects. It is not surprising
either that Wheeler eventually linked this “law without law” to quantum information theory, which he
helped to usher, along with Feynman, his one-time student. Quantum objects, in their interactions with
measuring instruments, create specifically organized collections of information (composed of classical
bits) and make possible certain calculations, by using mathematical models, but we cannot know and
possibly cannot conceive how quantum processes do this. The ultimate (quantum) constitution of
matter is, according to Wheeler, “it from bit,” “it” inferred from “bit” [118]. In the present view, this “it,”
while real, is beyond thought, and as such, cannot ultimately be called “it,” any more than anything
else. Wheeler’s visionary manifesto was inspired by Bohr, whom Wheeler invoked on the same page
that announced “it from bit:” “The overarching principle of 20th-century physics, the quantum—and
the principle of complementarity that is the central idea of the quantum—leaves us no escape, Niels
Bohr tells us, from ‘a radical revision of our attitude [towards the problem of] physical reality’” [118]
(I correct Wheeler’s slight misquotation of Bohr).
Bohr’s argument for the necessity of this revision originated in Heisenberg’s algebraic method, of
which Einstein, by contrast, remained ever skeptical, not the least because the concept of reality that
was the product of this revision remained unpalatable to Einstein, or in his words, “while logically
possible without contradiction, it [was] so contrary to [his] scientific instinct that [he] could not forego
a search for a more complete conception” [119]. From Bohr’s perspective, QM is only incomplete when
compared with the kind of realist knowledge possible in classical physics or relativity, which may
be called the Einstein completeness. Otherwise, it is as complete as possible, as things stand now,
which may be called the Bohr completeness. The question is whether nature would allow us to do
better. While Einstein thought that it should, Bohr thought that it might not, which is not the same
as it never will. As we haven’t heard nature’s last word on this matter, that is to say, nature’s next
word (the only last word that nature gives us), the debate concerning this question continues with
undiminished intensity.

5. From the Algebra of Circuits to the Algebra of Categories in Quantum Information Theory
Although Heisenberg’s creativity and inventiveness were remarkable and although it would
be difficult to challenge him on the outcome, his derivation of QM may not have been as rigorous
as one could ideally wish. While borrowing the form of equations from classical mechanics by the
mathematical correspondence principle was a logical deduction concerning part of the mathematical
structure of QM, Heisenberg virtually “guessed” the variables he needed. The mathematical expression
of the principles in question was only partially worked out and sometimes more intuited than properly
developed, which was in part remedied in the later work of Born, Jordan, and Heisenberg himself,
but only in part. Even the derivation offered, following this more rigorous treatment, by Heisenberg in
his Chicago lectures [100] might still be seen as falling short of a rigorous derivation from first principles,
because it relied on intuitive moves of the type found in Heisenberg’s original derivation, especially as

132

www.pdfgrip.com
Entropy 2018, 20, 656

concerns his matrix variables, still essentially a guess and, arguably, the main difficulty for any rigorous
(re)construction, especially for continuous variables. One could accordingly envision a more rigorous
derivation. Most of the recent work in this direction has been in quantum information theory in dealing
with discrete variables and finite-dimensional Hilbert spaces (QTFD). Some of these efforts, however,
have affinities with that of Heisenberg, which, as I argue, exhibits a spirit of quantum-informational
thinking. I shall now comment on two such cases, by D’Ariano and coworkers and by Hardy.
D’Ariano, Chiribella, and Perinotti’s (DACP’s) program, developed over the last decade and
presented comprehensively in their book [82], belongs to a particular trend in quantum information
theory, and as most of the work there, it deals with discrete variables and the corresponding,
finite-dimensional, Hilbert spaces (e.g., [120–122]). This is in part because, as DACP note (a view
shared by others in this field), “the study of finite-dimensional systems allows one to decouple the
conceptual difficulties in our understanding of quantum theory from the technical difficulties of
infinite-dimensional systems” [123]. A rigorous (or at least more rigorous than that of founding figures)
derivation of QM, let alone QFT, from fundamental principles remains an open and difficult task, to
which I return below.
DACP’s project is motivated by “a need for a deeper understanding of quantum theory [QTFD]
in terms of fundamental principles,” and by the aim of deriving QTFD from such principles, which,
the authors contend, has never been quite achieved by their predecessors. As indicated earlier,
the fluctuations of terms such as principles, axioms, postulates, commonly used in these reconstructive
projects may be confusing and obscure this common aim. DACP use the term “axioms” as well. On the
other hand, while one can surmise their understanding of the term “principle” from their use of it,
they do not define the concept of “principle” either. As earlier, I adopt the concept of principle defined
above, via Einstein. This concept is, I would argue, in accord with DACP’s use of principles, and their
derivation of QTFD, or that of Hardy, may be seen as that of a principle theory in Einstein’s sense.
I put aside the question to what degree this (or Hardy’s) derivation amounts to a fully rigorous
derivation, which, or in the first place, the question of what could be considered as a fully rigorous
derivation, would require a separate analysis. One might even question the necessity of a “fully rigorous
definition.” After all, Heisenberg, at least as his scheme was developed by Born and Jordan, and then
differently by Dirac, did establish a correct theory, and then Dirac similarly invented QED, and various
parts of QFT were created similarly. Heisenberg in effect posed this question in commenting on his own
derivation of QM: “It should be distinctly understood, however, this [deduction of the fundamental
equation of quantum mechanics] cannot be a deduction in the mathematical sense of the word, since the
equations to be obtained form themselves the postulates of the theory. Although made highly plausible
by the following considerations [of the type that led him to his discovery of QM], their ultimate
justification lies in the agreement of their predictions with the experiment” [124]. While a derivation of
QM, or QTFD, might be made more rigorous than that offered by Heisenberg even there, it is doubtful
that any such derivation, from (physical) first principles, could ever be as rigorous as “a deduction
in the mathematical sense of the word.” As Hardy suggested, even by his title, in [120], it may be
more a matter what are “reasonable” initial axioms or postulates, although such a “reasonableness”
is not a simple or unconditional matter either. It goes without saying that these qualifications in
no way diminish the significance of DACP’s or Hardy’s work, or that of others pursuing this line
research. Besides, as I shall explain, more than merely re-deriving QTFD, or QM and QFT is at stake in
these programs.
The main new feature of DACP’s approach is adding to the view of QTFD as an extension of
probability theory (a view found in the works of their predecessors and applicable to QM as well)
“the crucial ingredient of connectivity among events” by using the operational framework of “circuits”
and giving it an algebra [125]. The framework of circuits has been similarly used by others, specifically
by Hardy. This addition allows DACP “to derive key results of quantum information theory and
general features of quantum theory [QTFD]” without first assuming Hilbert spaces. Unlike von
Neumann, to whom DACP refer for a contrast, Heisenberg, as we have seen, did not begin with a

133

www.pdfgrip.com
Entropy 2018, 20, 656

formalism either, although he used the equation of classical mechanics by the correspondence principle.
He arrived at his formalism from fundamental principles, even if, again, not fully rigorously derived it
from these principles.
Among the principles adopted by DACP, the purification principle plays a unique role as
an essentially quantum principle, because conforming to it distinguishes QTFD from classical
probabilistic information theories. According to them: “The purification of mixed states is specifically
quantum” [126] (it may be a question whether it uniquely quantum, on which I shall comment presently).
It is the single principle necessary to do so, which may not be surprising given the history of quantum
theory and attempts at its axiomatic derivations. Hardy’s pioneering derivation also needed only
one axiom, the continuity axiom, to do so [120]. On the other hand, that Hardy’s continuity axiom
is different from DACP’s purification postulate suggests that there may not be a single system of
postulates or principles from which QTFD could be derived, even though all such systems should
capture something that pertains uniquely to quantum phenomena, and thus to quantum objects, even
if one assumes that they are beyond representation or conception. Their effects are representable and
enable one to distinguish classical and quantum phenomena, and infer from them the existence of
quantum objects.
Whether one can do so definitively remains a question, noted from the outset in connection
with Planck’s constant, h. While this question and literature concerning it are beyond my scope,
I would like to comment on Spekkens’s recent work, which is especially relevant here because it
proceeds along the lines of quantum information theory. Spekkens introduced several toy models or
theories, “epirestricted theories” (so-called because of epistemic restrictions on the classical theory,
assumed as a starting point), that reproduce many quantum phenomena and features of QM or QTFD,
such as the presence of h, the uncertainty relations, noncommuting operator observables, entanglement,
or the purification of mixed states [127–129]. Many but not all! These models expressly fail to
reproduce some among the crucial features of quantum theory, specifically some of those dealing
with correlations and entanglement, such as violations of Bell inequalities and the existence of the
Kochen-Specker theorem, which is, however, predictable given the nature of Spekkens’s models,
as Spekkens explained [130]. This is crucial because these theorems reflect the essential features of
quantum phenomena (independently of any theory, which, however, must, if correct, satisfy them),
and are also crucial to the question of realism and locality, all of which is noted by Spekkens [130].
Accordingly, whether actual quantum phenomena found in nature can be captured by models of this
type remains an open question. According to Spekkens himself, this is unlikely [131]:
The investigation of epirestricted theories, therefore, need not—and indeed should not-be
considered as the first step in a research program that seeks to find a ψ-epistemic ontological
model [a realist model that assumes quantum states in the formalism to be epistemic, states of
knowledge] of the full quantum theory. Even though such a model could always circumvent
any no-go theorems by violating their assumptions, it would be just as unsatisfying as a
ψ-ontic model [a realist model that assumes quantum states to be ontic, states of reality]
insofar as it would need to be explicitly nonlocal and contextual. Rather, the investigation of
epistricted theories is best considered as a first step in a larger research program wherein the
framework of ontological models—in particular the use of classical probability theory for
describing an agent’s incomplete knowledge—is ultimately rejected, but where one holds
fast to the notion that a quantum state is epistemic.
Although common in quantum information theory, viewing quantum states (state vectors
in the formalism) as states of knowledge could be misleading, unless one strictly refers to this
knowledge as probabilistic or statistical. (Even then quantum states are only part of the corresponding
predictive machinery, as one needs Born’s or related rule, such as Lüders’ postulate, to have these
probabilities or statistics.) As noted earlier, from the present, RWR-type perspective, in which quantum
states in this mathematical sense never refer to any actual, determined knowledge (which is only
obtained in measurements), our knowledge, actual (obtained in measurement) or probabilistic is

134

www.pdfgrip.com
Entropy 2018, 20, 656

incomplete only when compared with the kind of realist knowledge possible in classical physics
or relativity—the Einstein completeness. Otherwise, it is as complete as possible, as things stand
now—the Bohr completeness.
I am not sure whether, his skepticism concerning ontological models of quantum theory
notwithstanding, Spekkens would be willing to go that far, especially to the strong RWR view.
It appears, however, at least to the present reader, that his epirestricted theories and his argument
suggest that the difference between the classical and the quantum may ultimate be irreducible,
even though finding and rigorously grounding reconstructive programs remains difficult. According
to Spekkens, “it may be that there are particularly elegant axiomatic schemes that are not currently
in our reach and the road to program involves temporarily setting one’s sight a bit lower,” moving
to partial reconstructive models of the kind he proposes [132]. That may be, especially in the case of
continuous variables, to which Spekkens’s epirestricted theories apply. DACP’s or Hardy’s approach
may suggest otherwise, even while dealing only with QTFD. There is also a question whether such
elegance should be the main criterion here, especially given that one’s goals may ultimately be beyond
QTFD or QM? While Dirac would have thought so, neither Heisenberg (in his derivation) nor Bohr
was much worried about elegance. It is difficult to be certain which trajectories will lead us there,
especially if this “there” is beyond QTFD, as it ultimately must. There may not be one such trajectory,
just as in the discovery of QM, there were two—Heisenberg’s and Schrödinger’s. I leave this complex
set of subjects on the following note, concerning the role of h.
It is true that there are classical phenomena, in considering which h must be taken into account,
even trivially true, although Spekkens’s model considered in [129] is nontrivial because it involves
other quantum-like features. On the other hand, most classical phenomena do not involve h at all,
while all quantum phenomena known thus far require taking h into account, just as all (special)
relativistic account require taking c into account. This does, I think, tell us something about the nature
of the quantum or our technological interactions with that which we call the quantum, even if does not
tell us the whole story, assuming that such a story can ever be told. If one adopts an RWR-type view,
there is no story to be told about how quantum phenomena come about, and h, too, may only appear
in our interaction with nature and not be a property of nature itself. That, however, does not mean
that more cannot be said, perhaps even definitively, about what distinguishes quantum from classical
phenomena. I now return to DACP’s derivation of QTFD and the purification principle, which they
see as “specifically quantum”.
In nontechnical terms, the purification principle states that “every random preparation of a system
can be achieved by a pure preparation of the system with an environment, in a way that is essentially
unique” [133]. The principle originates in Schrödinger’s insight in his response, in several papers,
to the EPR paper and his concept of entanglement [72]. According to DACP [134]:

The purification principle stipulates that, whenever you are ignorant about the state of a
system A, you can always claim that your ignorance comes from the fact that A is part of a
large [composite] system AB, of which you have full knowledge. When you do this, the pure
state that you have to assign to the composite system AB is determined by the state of A in
an essentially unique way.
The purification of mixed states is a peculiar feature—surely, not one that we experience in
our everyday life. How can you claim that you know A and B if you don’t have A alone?
This counterintuitive feature has been noted in the early days of quantum theory, when Erwin
Schrödinger famously wrote: “Another way of expressing the peculiar situation is: the best
possible knowledge of a whole does not necessarily include the best possible knowledge of all its parts.”
And, in the same paper: “I would not call that one but rather the characteristic trait of quantum
mechanics, the one that enforced its entire departure from classical lines of thought . . . ” [135]

The purification of mixed states is specifically quantum. But why should we assume it as a
fundamental principle of Nature? At first, it looks like a weird feature—and it must look so,

135

www.pdfgrip.com
Entropy 2018, 20, 656

because quantum theory itself is weird and if you squeeze it inside a principle, it is likely that
the principle looks weird too. However, on second thought one realizes that purification is a
fundamental requirement: essentially, it is the link between physics and information theory.
Information theory would not make sense without the notions of probability and mixed state,
for the whole point about information is that there are things that we do not know in advance.
But in the world of classical physics of Newton and Laplace, every event is determined and
there is no space for information at the fundamental level. In principle, it does not make
sense to toss a coin or to play a game of chance, for the outcome is already determined and,
with sufficient technology and computational power, can always be predicted. In contrast,
purification tells us that “ignorance is physical.” Every mixed state can be generated in a
single shot by a reliable procedure, which consists of putting two systems in a pure state
and discarding one of them. As a result of this procedure, the remaining system will be a
physical token of our ignorance. This discussion suggests that, only if purification holds,
information can aspire to a fundamental role in physics.

Technically, the purification of mixed states is the fundamental principle arising in our interactions
with nature by means of our experimental technology, rather than of nature itself, except insofar as
we and our technologies are also nature. But then, in the present view, there is no other principles of
nature than those defined by us in our interactions with it. DACP’s formulation of the purification
principle and their derivation of QTFD allows for an RWR-type interpretation, and I shall interpret it
in this way, without claiming that this necessarily corresponds to DACP’s own view.
Thinking in terms of “circuits” is close to Bohr’s thinking concerning the role of measuring
instruments in the constitution of quantum phenomena, as distinguished from quantum objects,
which give rise to quantum phenomena by interacting with measuring instruments but which are
never observable. Circuits and their arrangements, too, embody those of measuring instruments
capable of detecting quantum events, and thus enabling the probabilistic predictions of future events.
Their arrangements and operations, defining their “measurement algebra,” are enabled by rules that
should ideally be derived from certain sufficiently natural assumptions. They are described classically,
and thus embody the structure of quantum information as a particular form of organization of classical
information, which can be used, as by DACP (or Hardy), to derive the mathematical formalism of QTFD.
While indispensable for the authors’ derivation of QTFD, the purification principle is not sufficient
to do so. They need five additional postulates (termed “axioms”): causality (essentially locality),
local discriminability, perfect distinguishability, ideal compression, and atomicity of composition [133].
These postulates define a large class of classical probabilistic informational theories, while the
purification postulate, giving rise to the purification principle, distinguishes QTFD. The appearance
of these additional postulates or principles is not surprising. Heisenberg’s grounding principles,
the quantum discreteness (QD) principle and the quantum probability or statistics (QP/QS) principle,
were not sufficient for him to derive QM either. To do so, he needed the correspondence principle,
which gave him half of the mathematical architecture of quantum theory. The other half was supplied
by his matrix variables.
There are instructive parallels between DACP’s and Heisenberg’s approaches. Both the QD and
QP/QS principles are present in both cases. As they say in their earlier article: “The operational-
probabilistic framework combines the operational language of circuits with the toolbox of probability
theory: on the one hand experiments are described by circuits resulting from the connection of
physical devices, on the other hand each device in the circuit can have classical outcomes and the
theory provides the probability distribution of outcomes when the devices are connected to form
closed circuits (that is, circuits that start with a preparation and end with a measurement)” [136].
This is similar to Heisenberg’s thinking in his paper introducing QM, as the classical outcomes are
discrete in both cases as well. The concept of “circuit” is not found in Heisenberg and is, again, closer
to Bohr’s view of the role of measuring instruments and his concept of phenomenon, defined by
this role. As I explained, however, the idea that in quantum theory we only deal with transition

136

www.pdfgrip.com
Entropy 2018, 20, 656

probabilities between the outcomes of the interactions between quantum objects and measuring
instruments was introduced by Heisenberg, as part of his approach to QM, and was then adopted by
Bohr. Heisenberg discovered that Bohr’s frequencies rules are satisfied by, in general, non-commuting
matrix variables with complex coefficients, from which one derives, by means of a Born-type rule,
probabilities or statistics for transitions between stationary states, manifested in spectra observed in
measuring devices. Thus, Heisenberg’s derivation depended on measuring instruments as devices
with classically describable observable parts, which are akin to “operational circuits,” in his case
dealing with continuous variables.
DACP aim to arrive at the mathematical structure of QTFD in a more first-principle-like way,
for example, independently of classical physics, which, because of the correspondence principle,
was central to Heisenberg (classical physics, to begin with, does not have discrete variables, so there is
no correspondence principle). The rules governing the structure of operational devices, circuits, should,
they argue, allow them to do so, because these rules are empirical. They are, however, not completely
empirical, because circuits are given a mathematical structure, in effect algebra, by human agents,
even though this algebra may be partially defined by the organization, required by experiments, of the
experimental arrangements in which circuits appear. The mathematized structure or algebra of circuits
become a necessary condition for establishing the mathematical structure or algebra of QTFD, but it is
not sufficient to do so. As noted above in considering Schwinger’s argument, these two structures are
not isomorphic or even homomorphic, and they do not appear to be in DACP’s derivation. This means
that additional pieces of structure, provided by additional postulates or axioms, are required for one
thing, to get to the (Hilbert-space or other) formalism over C. Then, one needs a Born-type rule for
the probabilities of predictions. DACP’s derivation requires enormous technical work. It is next to
impossible to do it justice here. I shall instead close the article by considering the algebra of circuits in
considering Hardy’s work.
Hardy has at a different set of main assumptions necessary to derive QTFD than those of DACP,
but the main strategy is the same: establishing the architecture, algebra, of circuits that, with additional
axioms, would allow one to derive the mathematical formalism of QTFD. According to Hardy [137]:
Circuits have:

• A setting, s(H), given by specifying the setting on each operation.

• An outcome set, o(H), given by specifying the outcome set at each operation (equals o(A) × o(B)
× o(C) × o(D) × o(E) in this case). We say the fragment “happened” if the outcome is in the
outcome set.
• A wiring, w(E), given by specifying which input/output pairs are wired together.

With this algebraic definition in hand, I shall comment on some of Hardy’s fundamental
assumptions discussed by him in a different paper. Hardy says [138]:

We will make two assumptions to set up the framework in this paper . . .

Assumption 1. The probability, Prob (A), for any circuit, A (this has no open inputs or outputs), is well
conditioned. It is therefore determined by the operations and the wiring of the circuit alone and is independent of
settings and outcomes elsewhere.
This is a physical postulate, essentially that of locality, combined with probability or statistics,
along the lines of the QP/QS principle. The task now becomes how to derive a QTFD that could
correctly predict these probabilities. One needs another assumption:
d4e5... f 6
Assumption 2. Operations are fully decomposable . . . We assume that any operation A a1b2...c3 can be
d4e5... f 6
written, . . . in a symbolic notation, Aa1b2...c3 ≡d4e5... f 6 Aa1b2...c3 Xa1
a1 b2 c3
Xb2 . . . Xc3 d4 e5 f6
d4 X e5 X... f 6 X . In words we
will say that any operation is equivalent to a linear combination of operations each of which consists of an effect for
each input and a preparation for each output . . . We allow the possibility that the entries in d4e5 . . . f6 Aa1b2 . . . c3

137

www.pdfgrip.com
Entropy 2018, 20, 656

are negative (and this will, indeed, be the case in quantum theory). Hence, in general, this cannot be thought of
as physical mixing . . . (emphasis added).
I only need the italicized sentence for my conceptual point, insofar as it means that one can
construct a suitable algebra. I cite the passage at a greater length to illustrate a manifested algebraic
view of circuits and operations in Hardy’s scheme. Hardy then says [138]:

Assumption 2 introduces a subtly different attitude than the usual one concerning how we
think about what an operation is. Usually we think of operations as effecting a transformation
on systems as they pass through. Here we think of an operation as corresponding to a bunch
of separate effects and preparations. We need not think of systems as things that preserve
their identity as they pass through—we do not use the same labels for wires coming out
as going in. This is certainly a more natural attitude when there can be different numbers
of input and output systems and when they can be of different types. Both classical and
quantum transformations satisfy this assumption. In spite of the different attitude just
mentioned, we can implement arbitrary transformations, such as unitary transformations in
quantum theory, by taking an appropriate sum over such effect and preparation operations.

This rethinking of the concept of operation is important, especially if one adopts an RWR-type
view. An “operation” is now defined in terms of observable “effects” of the interactions between
quantum objects and measuring instruments, and not in terms of what happens, even in the course
of these interactions (let alone apart from them), to the quantum objects or systems, considered as
independent systems. It is useful that we can treat classical systems in this way as well. In the classical
case, however, we can, equivalently, use a more conventional concept of operation mentioned here,
which is not the case in quantum theory. After a technical discussion of “duotensors,” which I put
aside, Hardy suggests a principle [139]:

Physics to mathematics correspondence principle. For any physical theory, there [exist] a small
number of simple hybrid statements that enable us to translate from the physical description
to the corresponding mathematical calculation such that the mathematical calculation
(in appropriate notation) looks the same as the physical description (in appropriate notation).
Such a principle might be useful in obtaining new physical theories (such as a theory of
quantum gravity). Related ideas to this have been considered by category theorists. A category
of physical processes can be defined corresponding to the physical description. A category
corresponding to the mathematical calculation can also be given. The mapping from the first
category to the second is given by a functor (this takes us from one category to another).

The language of correspondence should not mislead one into relating this principle to Bohr’s
correspondence principle, even in Heisenberg’s mathematical form. Bohr’s correspondence principle
deals with the correspondence between different physical theories (such as classical mechanics
and quantum mechanics), insofar as their predictions would coincide in the regions when both
theories could be used. By contrast, apart from the fact that, as explained, there is no correspondence
principle in QFDT, Hardy’s “hybrid” construction implies that category of physical processes could be
functorially “translated” into a proper formalism of QTFD, which can then, through mathematical
calculations enabled by this formalism, be related to what is observed. Hardy’s principle may be better
called “physics to mathematics functoriality principle.” Hardy’s suggestion, inviting but somewhat
speculative and not really worked out, to begin with, would require a separate discussion, which cannot
be undertaken here. It is worthwhile, however, to offer a few brief comments, via the role of category
theory in the algebraization of physics, without definitively claiming that these comments are fully in
accord with Hardy’s view of the situation.
Category theory originated in algebraic topology and then was extensively used, especially thanks
to Grothendieck’s work, in algebraic geometry, in order to study certain algebraic invariants, such as
cohomology or homotopy groups, associated with topological spaces [140]. It was then extended,

138

www.pdfgrip.com
Entropy 2018, 20, 656

via topos theory, introduced by Grothendieck as well, to mathematical logic, a subject I shall put
aside here [141]. Roughly, category theory considers multiplicities (categories) of mathematical objects
conforming to a given concept, such as the category of topological spaces or geometrical spaces (say,
Riemannian manifolds), and the morphisms, also called arrows ( X → Y ), which are the mappings
between these objects that preserve this structure. Studying morphisms allows one to learn about the
individual objects involved more than we would by considering them individually. Thus, in geometry
one does not have to start with a Euclidean space. Instead the latter is just one specifiable object of a
large categorical multiplicity, such as the category of Riemannian manifolds, an object marked by a
particularly simple way we can measure the distance between points. Categories themselves may be
viewed as such objects, and in this case one speaks of “functors” rather than “morphisms.”
Now, one can more easily think of properly defining mathematically, the second category in
Hardy’s suggestion, say, as that of Hilbert spaces [142] or some more directly categorical equivalent
algebra by means of which the mathematical calculations in question would be performed. On the
other hand, the first category, that is, the structure of its objects and the morphisms between them,
and, thus, the nature of the functor in question between these two categories is a more complex
matter. Hardy’s formulations above “for any physical theory, there [exist] a small number of simple
hybrid statements that enable us to translate from the physical description to the corresponding
mathematical calculation such that the mathematical calculation (in appropriate notation) looks the same
as the physical description (in appropriate notation)” (emphasis added) might require qualification as
concerns the meaning of the expression “looks the same” and the relationships between “calculation”
and “description,” because this formulation allows for different interpretations. The same may be said
about his characterization of the functor in question as “virtually direct” (emphasis added) and then the
statement “a category of physical processes can be defined corresponding to the physical description.”
I shall sketch the reasons for these qualifications, beginning with a possible meaning of “physical
processes” in “a category of physical processes can be defined corresponding to the physical description.”
“Physical processes” may refer either to quantum processes, or to circuits, which appears more likely
because “the physical description” in Hardy’s scheme (or that of DACP) is given only at the level of
circuits. This view would also be suggested by the category theorists who work on using categories in
quantum theory, specifically QTFD, such as Abramsky, Coecke, and others, to whom Hardy appears to
refer here. Most of this work is primarily concerned with recasting the Hilbert-space language of the
QTFD formalism in a categorical framework by replacing Hilbert spaces, belonging to the category
of the finite-dimensional Hilbert spaces over C (e.g., [142]), with objects of monoidal categories
and the morphism between Hilbert spaces with morphism between these objects, rather than with
deriving the formalism from the first (physical) principles. There are some moves in this direction, say,
by starting with a suitable simple monoidal category and then consider which additional pieces of
structure one needs to arrive at QTFD (e.g., [143]). This work further testifies to the dominant role of
algebra in quantum theory, even though it uses a lot of diagrammatic operations, somewhat akin to
Feynman diagrams in QFT. As Coecke and Kissinger’s title, “picturing quantum processes,” indicates,
they appear to hold a realist view in assuming that their diagrammatic picturalism represents the
actual quantum behavior, which assumption, I argue here, has complexities, in part because of the role
of complex quantities, complexities not addressed by Coecke and Kissinger [143]. In RWR-type views,
quantum behavior is beyond representation or even conception, and thus pieces of this behavior (say,
between measurements) cannot be assumed to form a category. Hardy’s position on this point is not
entirely clear. If quantum processes themselves are given some representation, the corresponding
category may need to incorporate this representation in one way or another, by combing it with the
effects of quantum processes manifested in circuits and their arrangements.
If one adopts an RWR-type view (either weak or strong), one only deals with the physical
description of circuits, similarly to dealing only with the description of measuring instruments,
according to Bohr. Circuits and their organizations, their algebra, again, embody the arrangements
of measuring instruments capable of detecting quantum events and enabling the probabilistic or

139

www.pdfgrip.com
Entropy 2018, 20, 656

statistical predictions of future events, in other words, a structure that may be mathematizable
and as such categorially translatable into the mathematical formalism of QTFD (which enables the
necessary mathematical calculations). As discussed earlier in considering both Schwinger’s argument
and DACP’s derivation of QTFD, one need not and, in the RWR-type view, should not expect a
representational correspondence between the structures or algebra of circuits and the mathematical
structures or the algebra of the formalism of QTFD. For one thing, one needs, again, a formalism,
Hilbert-space one, C*-algebra, categorical, or other, over C. As classical, the data manifested in circuits
is over R (in fact, all measurements are rational numbers, while probabilities need not be). Then,
one needs a Born-type rule to get to probabilities. It follows, then, as, again, considered earlier, that
one needs additional pieces of structure to those defined by circuits.
One could contemplate two approaches. The one, closer to QTFD category theorists, is to use
these additional pieces of structure (algebraic relations, axioms, etc.) to build a category of algebraic
mathematical objects, which need not be Hilbert spaces, but would, again, have to be over C, and rules,
which enable proper probabilistic or statistical predictions of quantum phenomena, again, observed in
circuits [143]. Alternatively, closer to what Hardy appears to suggest and to Bohr’s way of thinking,
one could attempt to establish a category of circuits, each defined by an algebraic structure of units
and operations, and morphisms between them, and then, again, by using some reasonable axioms,
a category of objects defining the formalism, such as Hilbert spaces or some monoidal categories,
and morphism between them (along, again, with a Born-like rule for probabilities), and then connect
these two categories by means of a well-defined functor. The advantage of the second approach is
that it preserves the role of the algebraic structure of circuits, which are observable. The difficulty
is, again, that there is no natural categorical structure for the algebra of circuits, morphisms, etc.,
all of which need to be defined, which, however, is also true in the first approach. To do so requires
additional pieces of structure, and thus additional more or less reasonable principles or postulates,
rather than intuitive guesses, assuming, again, that the latter could be entirely avoided. In the second
approach (which I prefer), it is the question of a functorial relationship between categories, which,
again, still needs to be established, rather than morphism between objects, such as the algebras of
circuits and the algebras of the formalism of QTFD, or by implication, QM or QFT, a more complex
project. There are morphisms in each category, and there functors between categories, and the view just
considered is about the latter. From this perspective, Hardy’s “physics to mathematics correspondence
principle” would, if rigorously established, be the “physics to mathematics functoriality principle”:
it would be realized in terms of the functorial relationship between a category based on circuits (the
objects and morphisms of which need, again, to be defined) and a category of the algebraic objects
defining the formalism, such as that of Hilbert spaces over C, finite-dimensional ones in QTFD, or some
other category of objects and morphisms.
One of the most essential aspects of category theory, even its raison d’être, at least in its use in
algebraic topology and algebraic geometry (mathematical logic is, again, a different matter), consists
in establishing the relationships between multiplicities, “categories,” of objects of different types,
for example, geometrical or topological objects and algebraic objects, or between algebraic objects of
different types (such as Lie groups and Lie algebras), rather than, apart from trivial cases, directly
mapping the objects of the first category on those of the second. Indeed, the concept of category was
introduced in the field of algebraic topology and then extended to algebraic geometry to help study
certain algebraic invariants, such as groups, associated with topological spaces. In contradistinction
to geometry, defined, as a mathematical discipline, by the concept of measurement (geo-metry),
topology, as a mathematical discipline, is defined by associating an algebraic structure or a set of
structures, most especially groups, such as homotopy or cohomology groups (which came to play
an important role in QFT) to a topological space. The structure of a topological space is defined
by its continuities and discontinuities, and not by its geometry, even if it had a geometry, and not
all topological spaces do. Insofar as one deforms a given figure continuously (i.e., insofar as one
does not separate points previously connected and, conversely, does not connect points previously

140

www.pdfgrip.com
Entropy 2018, 20, 656

separated), the resulting figure or space is treated as mathematically equivalent. Thus, no matter how
much you expand or continuously (without separating connected points or joining separated points)
deform the two-dimensional surface of a sphere the resulting spaces are topologically equivalent
(homeomorphic), while some of these objects are no longer spherical geometrically speaking. Such
spaces are, however, topologically distinct from those of topologically deformed two-dimensional
surfaces of tori because spheres and tori cannot be converted into each other without disjoining their
connected points or joining the separated ones: the holes in tori make this impossible. By contrast,
the spaces of spheres (or any spaces) of different dimensions are not homeomorphic. In algebraic
topology, mathematical objects of each type are arranged in categories and are related to one another
by morphisms, while categories are related by functors. The category of topological spaces (or a
subcategory, such as that of Riemannian manifolds) and their morphisms becomes related to a
category of algebraic objects, such as groups and their morphisms. The relationships between these
two categories, topological and algebraic, allow one to extract an enormous amount of information
concerning topological spaces and, conversely, groups (through the structure of topological spaces
to which these groups are associated). A good categorical approach to, and a possible derivation of,
quantum theory would similarly establish the functorial relationships between two, now algebraic,
categories, one based in circuits and the other defining the formalism.
An important and difficult question, which need not depend on a categorical formalization,
but may be helped by it, is that of the relationship between the structure of the circuits, defined
by the corresponding experimental arrangements, and the infinite-dimensional mathematical
architecture of QM, or the same relationship in QFT. Consider the double-slit experiment, say, in the
interference pattern setup. It is a circuit, which embodies preparations, measurements, and predictions,
all manifested in the emergence of the interference pattern. I would not presume to be able to
mathematize it. But it is a circuit nevertheless, a complex one, albeit child’s play in comparison
to the circuitry found in high-energy quantum physics, such as that of the Large Hadron Collider
(LHC), which led to the detection of the Higgs boson.
Such questions will need be addressed if one is to extend the programs to derivations of QTFD to
QM or to QED and QFT, or beyond. Indeed, it is hardly sufficient to merely derive already established
theories. The ultimate value of these programs lies in what they can do for the future of fundamental
physics. Thus, Hardy aims to rethink general relativity in operational terms, analogous to those of
QTFD, and then to reach, in principle, quantum gravity, bypassing QFT or even QM, which would
then be merely special cases of the ultimate theory [16,83]. D’Ariano and co-workers, by contrast,
appear first to move from QTFD to QFT. In their more recent work, they aim to develop a new
approach to QFT, which, as based on the concept of quantum cellular automata, is different from
the operational framework discussed thus far, but shares with it certain key informational features
and, most especially, the aim of developing QFT from fundamental (first) principles. D’Ariano and
Perinotti’s derivation of Dirac’s equation is a step in this direction, for now in the absence of an external
field, essential to the proper QED [144,145]. Unlike Dirac’s own or other previous derivations of the
equation, their derivation only uses, along with other principles (homogeneity, isotropy, and unitarity),
the principle of locality, rather than special relativity. The approach may, they hope, offer new
possibilities for fundamental physics on Planck’s scale, suggesting a potential extension of quantum
information theory as far as one can envision it now.
Whether the QPA principle (which is in part experimental, empirical) and the RWR principle
(which is interpretive) will remain viable or will be defeated, fulfilling the hope of Einstein and those
who followed him, is an open question. On the other hand, it appears likely, as no currently known
attempts to move beyond QFT (such as string and M-brane theory, or loop quantum gravity) would
indicated otherwise, that we will see the emergence of new algebraic and spatial-algebraic structures.
We very much need them for quantum gravity, for example. Because algebraic topology and algebraic
geometry are likely to play a role in quantum gravity, the age of algebra—the age of Fermat, Descartes,
and Leibniz—is likely to continue for the foreseeable future in mathematics and physics alike.

141

www.pdfgrip.com
Entropy 2018, 20, 656

Funding: This research received no external funding.

Acknowledgments: I would like to thank Mauro D’Ariano, Laurent Freidel, Christopher A. Fuchs, Lucien Hardy,
Gregg Jaeger, Andrei Khrennikov, and Paolo Perinotti for valuable discussions concerning the subjects addressed
in this article. The suggestions of the anonymous reviewers were invaluable in helping to clarify and sharpen the
article’s argument.
Conflicts of Interest: The author declares no conflict of interest.

References
1. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 2, p. 53.
2. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 2, p. 33.
3. Heisenberg, W. Quantum-theoretical re-interpretation of kinematical and mechanical relations (1925).
In Sources of Quantum Mechanics; Van der Waerden, B.L., Ed.; Dover: New York, NY, USA, 1968; pp. 261–277.
4. Hacking, I. The Emergence of Probability: A Philosophical Study of Early Ideas about Probability, Induction and
Statistical Inference, 2nd ed.; Cambridge University Press: Cambridge, UK, 2006.
5. Hájek, A. Interpretation of Probability, Stanford Encyclopedia of Philosophy. Zalta, E.N., Ed.; 2012. Available online:
http://plato.stanford.edu/archives/win2012/entries/probability-interpret/ (accessed on 26 August 2018).
6. Einstein, A. Physics and reality. J. Frankl. Inst. 1936, 221, 378. [CrossRef]
7. Plotnitsky, A. The Principles of Quantum Theory, from Planck’s Quanta to the Higgs Boson: The Nature of Quantum
Reality and the Spirit of Copenhagen; Springer: New York, NY, USA, 2016; pp. 84–98.
8. Born, M.; Jordan, P. Zur Quantenmechanik. Z. Phys. 1925, 34, 858–888. [CrossRef]
9. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 1, p. 48.
10. Kragh, H. Niels Bohr and the Quantum Atom: The Bohr Model of Atomic Structure 1913–1925; Oxford University
Press: Oxford, UK, 2012.
11. Plotnitsky, A. The Principles of Quantum Theory, from Planck’s Quanta to the Higgs Boson: The Nature of Quantum
Reality and the Spirit of Copenhagen; Springer: New York, NY, USA, 2016; pp. 54–67.
12. Freidel, L. On the discovery of quantum mechanics by Heisenberg, Born, and Jordan. Unpublished
work, 2016.
13. Born, M. The Einstein-Born Letters; Walker: New York, NY, USA, 2005; p. 78.
14. Plotnitsky, A. Quantum Atomicity and Quantum Information: Bohr, Heisenberg, and Quantum Mechanics
as an Information Theory. In Quantum Theory: Reexamination of Foundations 2; Khrennikov, A., Ed.;
Växjö University Press: Växjö, Sweden, 2002; pp. 309–343.
15. Plotnitsky, A. The Principles of Quantum Theory, from Planck’s Quanta to the Higgs Boson: The Nature of Quantum
Reality and the Spirit of Copenhagen; Springer: New York, NY, USA, 2016; pp. 72–73.
16. Hardy, L. Towards quantum gravity: A framework for probabilistic theories with non-fixed causal structure.
J. Phys. A 2007, 40, 3081–3099. [CrossRef]
17. Heisenberg, W. The Physical Principles of the Quantum Theory; Courier Corporation: Dover, UK; New York,
NY, USA, 1930.
18. Plotnitsky, A. Niels Bohr and Complementarity: An Introduction; Springer: New York, NY, USA, 2012.
19. Wilczek, F. In search of symmetry lost. Nature 2005, 432, 239–247. [CrossRef] [PubMed]
20. Borel, E. La logique et l’intuition en mathématique. Revue de Métaphysique et de Morale 1907, 15, 273–283.
21. Gray, J. Plato’s Ghost: The Modernist Transformation of Mathematics; Princeton University Press: Princeton, NJ,
USA, 2008; p. 202.
22. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 2, p. 57.
23. Einstein, A. Autobiographical Notes; Open Court: La Salle, IL, USA, 1949; p. 47.
24. Van Dongen, J. Einstein’s Unification; Cambridge University Press: Cambridge, UK, 2010; pp. 89–95.
25. Riemann, B. On the Hypotheses That Lie at the Foundations of Geometry. In Beyond Geometry: Classic Papers
from Riemann to Einstein; Pesic, P., Ed.; Dover: Mineola, NY, USA, 1854; pp. 23–40.
26. Plotnitsky, A. Comprehending the Connection of Things: Bernhard Riemann and the Architecture
of Mathematical Concepts. In From Riemann to Differential Geometry and Relativity; Ji, L., Yamada, S.,
Papadopoulos, A., Eds.; Springer: Berlin, Germany, 2017; pp. 329–363.
27. Deleuze, G.; Guattari, F. What Is Philosophy? Columbia University Press: New York, NY, USA, 1994.

142

www.pdfgrip.com
Entropy 2018, 20, 656

28. Heisenberg, W. The Physical Principles of the Quantum Theory; Courier Corporation: Dover, UK; New York,
NY, USA, 1930; pp. 10, 13.
29. Bohr, N. Causality and complementarity. In The Philosophical Writings of Niels Bohr, Volume 4: Causality and
Complementarity, Supplementary Papers; Faye, J., Folse, H.J., Eds.; Ox Bow Press: Woodbridge, CT, USA, 1999;
pp. 83–91.
30. Plotnitsky, A. Niels Bohr and Complementarity: An Introduction; Springer: New York, NY, USA, 2012;
pp. 173–179.
31. Plotnitsky, A. The Principles of Quantum Theory, from Planck’s Quanta to the Higgs Boson: The Nature of Quantum
Reality and the Spirit of Copenhagen; Springer: New York, NY, USA, 2016; pp. 107–120.
32. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 1, p. 53.
33. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 2, p. 73.
34. Bohr, N. Can Quantum-Mechanical Description of Physical Reality Be Considered Complete? Phys. Rev.
1935, 48, 696–703. [CrossRef]
35. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 2,
pp. 61–62.
36. Heisenberg, W. Physics and Philosophy: The Revolution in Modern Science; Harper & Row: New York, NY,
USA, 1962.
37. Plotnitsky, A. The Principles of Quantum Theory, from Planck’s Quanta to the Higgs Boson: The Nature of Quantum
Reality and the Spirit of Copenhagen; Springer: New York, NY, USA, 2016; pp. 4–23.
38. Plotnitsky, A. The Visualizable, the Representable, and the Inconceivable: Realist and Non-Realist
Mathematical Models in Physics and Beyond. Philos. Trans. R. Soc. A 2016, 374, 20150101. [CrossRef]
[PubMed]
39. Frigg, R.; Hartmann, S. Models in Science. The Stanford Encyclopedia of Philosophy; Zalta, E.N., Ed.;
2012. Available online: http://plato.stanford.edu/archives/fall2012/entries/models-science/ (accessed on
26 August 2018).
40. Frigg, R. Theories and Models; Acumen: Slough, UK, 2014.
41. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 2, p. 51.
42. Bohr, N. Can Quantum-Mechanical Description of Physical Reality Be Considered Complete? Phys. Rev.
1935, 48, 697. [CrossRef]
43. Plotnitsky, A. The Principles of Quantum Theory, from Planck’s Quanta to the Higgs Boson: The Nature of Quantum
Reality and the Spirit of Copenhagen; Springer: New York, NY, USA, 2016.
44. Plotnitsky, A.; Khrennikov, A. Reality without Realism: On the Ontological and Epistemological Architecture
of Quantum Mechanics. Found. Phys. 2015, 25, 1269–1300. [CrossRef]
45. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 1,
pp. 54–57.
46. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 1,
pp. 76–77.
47. Dirac, P.A.M. The Principles of Quantum Mechanics; Oxford University Press: Oxford, UK, 1930.
48. Von Neumann, J. Mathematical Foundations of Quantum Mechanics; Princeton University Press: Princeton, NJ,
USA, 1932.
49. Plotnitsky, A. The Principles of Quantum Theory, from Planck’s Quanta to the Higgs Boson: The Nature of Quantum
Reality and the Spirit of Copenhagen; Springer: New York, NY, USA, 2016; pp. 66–68, 124–131, 159.
50. Kant, I. Critique of Pure Reason; Cambridge University Press: Cambridge, UK, 1997; p. 115.
51. Cabello, A. Interpretations of quantum theory: A map of madness. In What Is Quantum Information?
Lombardi, O., Fortin, S., Holik, F., López, C., Eds.; Cambridge University Press: Cambridge, UK, 2017;
pp. 138–144.
52. Fuchs, C.A.; Mermin, N.D.; Schack, R. An introduction to QBism with an application to the locality of
quantum mechanics. Am. J. Phys. 2014, 82, 749. [CrossRef]
53. Werner, R.F. Comment on ‘What Bell did’. J. Phys. A 2014, 47, 424011. [CrossRef]
54. Heisenberg, W. Physics and Philosophy: The Revolution in Modern Science; Harper & Row: New York, NY, USA,
1962; pp. 47, 145.

143

www.pdfgrip.com
Entropy 2018, 20, 656

55. Heisenberg, W. Physics and Philosophy: The Revolution in Modern Science; Harper & Row: New York, NY, USA,
1962; pp. 178–179.
56. Heisenberg, W. The Physical Principles of the Quantum Theory; Courier Corporation: Dover, UK; New York,
NY, USA, 1930; p. 11.
57. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 1, pp. 51,
98–100, 108.
58. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 2, p. 59.
59. Wigner, E.P. On unitary representations of the inhomogeneous Lorentz group. Ann. Math. 1939, 40, 149–204.
[CrossRef]
60. Heisenberg, W. Physics and Philosophy: The Revolution in Modern Science; Harper & Row: New York, NY, USA,
1962; pp. 145, 167–186.
61. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 2, p. 64.
62. Plotnitsky, A. The Principles of Quantum Theory, from Planck’s Quanta to the Higgs Boson: The Nature of Quantum
Reality and the Spirit of Copenhagen; Springer: New York, NY, USA, 2016; pp. 207–246.
63. Schrödinger, E. The present situation in quantum mechanics (1935). In Quantum Theory and Measurement;
Wheeler, J.A., Zurek, W.H., Eds.; Princeton University Press: Princeton, NJ, USA, 1983; pp. 152–167.
64. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 2, p. 34.
65. Wittgenstein, L. Tractatus Logico-Philosophicus; Routledge: London, UK, 1924; p. 175.
66. Aaronson, S. Quantum Computing Since Democritus; Cambridge University Press: Cambridge, UK, 2013.
67. Khrennikov, A. Quantum probabilities and violation of CHSH-inequality from classical random signals and
threshold type detection scheme. Prog. Theor. Phys. 2012, 128, 31–58. [CrossRef]
68. Khrennikov, A. Interpretations of Probability; de Gruyter: Berlin, Germany, 2009.
69. De Finetti, B. Philosophical Lectures on Probability; Springer: Berlin, Germany, 2008.
70. Jaynes, E.T. Probability Theory: The Logic of Science; Cambridge University Press: Cambridge, UK, 2003.
71. Plotnitsky, A. The Principles of Quantum Theory, from Planck’s Quanta to the Higgs Boson: The Nature of Quantum
Reality and the Spirit of Copenhagen; Springer: New York, NY, USA, 2016; pp. 180–184.
72. Einstein, A.; Podolsky, B.; Rosen, N. Can Quantum-Mechanical Description of Physical Reality be Considered
Complete? In Quantum Theory and Measurement; Wheeler, J.A., Zurek, W.H., Eds.; Princeton University Press:
Princeton, NJ, USA, 1935; pp. 138–141.
73. Plotnitsky, A. The Principles of Quantum Theory, from Planck’s Quanta to the Higgs Boson: The Nature of Quantum
Reality and the Spirit of Copenhagen; Springer: New York, NY, USA, 2016; pp. 136–154.
74. Aspect, A.; Dalibard, J.; Roger, G. Experimental test of Bell’s inequalities using time varying analyzers.
Phys. Rev. Lett. 1982, 49, 1804–1807. [CrossRef]
75. Bell, J.S. Speakable and Unspeakable in Quantum Mechanics; Cambridge University Press: Cambridge, UK, 2004.
76. Cushing, J.T.; McMullin, E. Philosophical Consequences of Quantum Theory: Reflections on Bell’s Theorem;
Notre Dame University Press: Notre Dame, IN, USA, 1989.
77. Ellis, J.; Amati, D. Quantum Reflections; Cambridge University Press: Cambridge, UK, 2000.
78. Brunner, N.; Gühne, O.; Huber, M. Special issue on 50 years of Bell’s theorem. J. Phys. A 2014, 42, 424024.
79. Bohr, N. Can Quantum-Mechanical Description of Physical Reality Be Considered Complete? Phys. Rev.
1935, 48, 699. [CrossRef]
80. Plotnitsky, A. The Principles of Quantum Theory, from Planck’s Quanta to the Higgs Boson: The Nature of Quantum
Reality and the Spirit of Copenhagen; Springer: New York, NY, USA, 2016; pp. 203–206.
81. Brukner, C. Quantum Causality. Nat. Phys. 2014, 10, 259–263. [CrossRef]
82. D’Ariano, G.M.; Chiribella, G.; Perinotti, P. Quantum Theory from First Principles: An Informational Approach;
Cambridge University Press: Cambridge, UK, 2017.
83. Hardy, L. A formalism-local framework for general probabilistic theories, including quantum theory. arXiv
2010, arXiv:1005.5164.
84. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 2, p. 41.
85. Weyl, H. The Continuum: A Critical Examination of the Foundation of Analysis; Dover: Mineola, NY, USA, 1928;
p. 108.
86. Darigold, O. From c-Numbers to q-Numbers: The Classical Analogy in the History of Quantum Theory; University
of California Press: Berkeley, CA, USA, 1993; p. 307.

144

www.pdfgrip.com
Entropy 2018, 20, 656

87. Connes, A. Noncommutative Geometry; Academic Press: San Diego, CA, USA, 1994.
88. Plotnitsky, A. Epistemology and Probability: Bohr, Heisenberg, Schrödinger and the Nature of Quantum-Theoretical
Thinking; Springer: New York, NY, USA, 2009; pp. 111–112.
89. Suppes, P. Is Visual Space Euclidean. Synthese 1977, 35, 397–421. [CrossRef]
90. Foley, J.; Ribeiro-Filho, N.; Da Silva, J. Visual Perception of extend and the geometry of visual space. Vis. Res.
2004, 44, 147–156. [CrossRef] [PubMed]
91. Erkelens, C. The Perceptual Structure of Visual Space. i-Perception 2015, 6. [CrossRef] [PubMed]
92. Schweber, S.S. QED and the Men Who Made It: Dyson, Feynman, Schwinger, and Tomonaga; Princeton University
Press: Princeton, NJ, USA, 1994; pp. 465–466.
93. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 2, p. 72.
94. Hilbert, D. Foundations of Geometry; Open Court: La Salle, IL, USA, 1999.
95. Silverman, J.; Tate, J. Rational Points on Elliptic Curves; Springer: Heidelberg, Germany; New York, NY, USA,
2015; p. 277.
96. Newton, I. The Principia: Mathematical Principles of Natural Philosophy; University of California Press: Berkeley,
CA, USA, 1999.
97. Riemann, B. On the Hypotheses That Lie at the Foundations of Geometry. In Beyond Geometry: Classic Papers
from Riemann to Einstein; Pesic, P., Ed.; Dover: Mineola, NY, USA, 1854; pp. 23–40.
98. Mehra, J.; Rechenberg, H. The Historical Development of Quantum Theory; Springer: Berlin, Germany, 2001;
Volume 3, p. 69.
99. Dirac, P.A.M. The fundamental equations of quantum mechanics. In Sources of Quantum Mechanics;
van der Warden, B.L., Ed.; Dover: New York, NY, USA, 1925; pp. 307–320.
100. Heisenberg, W. The Physical Principles of the Quantum Theory; Dover: New York, NY, USA, 1930.
101. Einstein, A. What is the Theory of Relativity? In Ideas and Opinions; Einstein, A., Ed.; Bonanza Books:
New York, NY, USA, 1954; p. 228.
102. Mehra, J.; Rechenberg, H. The Historical Development of Quantum Theory; Springer: Berlin, Germany, 2001;
Volume 2, p. 242.
103. Heisenberg, W. Physics and Philosophy: The Revolution in Modern Science; Harper & Row: New York, NY, USA,
1962; p. 41.
104. Bohr, N. On the constitution of atoms and Molecules (Part 1). Philos. Mag. 1913, 26, 1–25. [CrossRef]
105. Plotnitsky, A. The Principles of Quantum Theory, from Planck’s Quanta to the Higgs Boson: The Nature of Quantum
Reality and the Spirit of Copenhagen; Springer: New York, NY, USA, 2016; pp. 68–83.
106. Plotnitsky, A. Epistemology and Probability: Bohr, Heisenberg, Schrödinger and the Nature of Quantum-Theoretical
Thinking; Springer: New York, NY, USA, 2009; pp. 90, 111, 116.
107. Heisenberg, W. Quantum-theoretical re-interpretation of kinematical and mechanical relations (1925).
In Sources of Quantum Mechanics; Van der Waerden, B.L., Ed.; Dover: New York, NY, USA, 1968; p. 3.
108. Dirac, P.A.M. The Principles of Quantum Mechanics, 4th ed.; Oxford University Press: Oxford, UK, 1958; p. 14.
109. Heisenberg, W. Quantum-theoretical re-interpretation of kinematical and mechanical relations (1925).
In Sources of Quantum Mechanics; Van der Waerden, B.L., Ed.; Dover: New York, NY, USA, 1968; p. 264.
110. Born, M.; Heisenberg, W.; Jordan, P. On quantum mechanics. In Sources of Quantum Mechanics;
Van der Waerden, B.L., Ed.; Dover: New York, NY, USA, 1926; pp. 321–385.
111. Heisenberg, W. Quantum-theoretical re-interpretation of kinematical and mechanical relations (1925).
In Sources of Quantum Mechanics; Van der Waerden, B.L., Ed.; Dover: New York, NY, USA, 1968; p. 265.
112. Schweber, S.S. QED and the Men Who Made It: Dyson, Feynman, Schwinger, and Tomonaga; Princeton University
Press: Princeton, NJ, USA, 1994; pp. 360–361.
113. Schwinger, J. Quantum Mechanics: Symbolism of Atomic Measurement; Springer: New York, NY, USA, 2001.
114. Jaeger, G. Grounding the randomness of quantum measurement. Philos. Trans. R. Soc. A 2016. [CrossRef]
[PubMed]
115. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 2, p. 63.
116. Bohr, N. The Philosophical Writings of Niels Bohr; Ox Bow Press: Woodbridge, CT, USA, 1987; Volume 2,
pp. 46–47, 56, 62.
117. Wheeler, J.A. Law without law. In Quantum Theory and Measurement; Wheeler, J.A., Zurek, W.H., Eds.;
Princeton University Press: Princeton, NJ, USA, 1983; pp. 182–216.

145

www.pdfgrip.com
Entropy 2018, 20, 656

118. Wheeler, J.A. Information, physics, quantum: The search for links. In Complexity, Entropy, and the Physics of
Information; Zurek, W.H., Ed.; Addison-Wesley: Redwood City, CA, USA, 1990; pp. 309–336.
119. Einstein, A. Physics and reality. J. Frankl. Inst. 1936, 221, 375. [CrossRef]
120. Hardy, L. Quantum mechanics from five reasonable axioms. arXiv 2001, arXiv:quant-ph/0101012v4.
121. Fuchs, C.A. Quantum mechanics as quantum information, mostly. J. Mod. Opt. 2003, 50, 987–1023. [CrossRef]
122. Chiribella, G.; Spekkens, R. Quantum Theory: Informational Foundations and Foils; Springer: New York, NY,
USA, 2016.
123. D’Ariano, G.M.; Chiribella, G.; Perinotti, P. Quantum Theory from First Principles: An Informational Approach;
Cambridge University Press: Cambridge, UK, 2017; p. 4.
124. Heisenberg, W. The Physical Principles of the Quantum Theory; Dove: New York, NY, USA, 1930; p. 108.
125. D’Ariano, G.M.; Chiribella, G.; Perinotti, P. Quantum Theory from First Principles: An Informational Approach;
Cambridge University Press: Cambridge, UK, 2017; pp. 4–5.
126. D’Ariano, G.M.; Chiribella, G.; Perinotti, P. Quantum Theory from First Principles: An Informational Approach;
Cambridge University Press: Cambridge, UK, 2017; p. 169.
127. Spekkens, R.W. Evidence for the epistemic view of quantum states: A toy theory. Phys. Rev. A 2007, 75,
032110. [CrossRef]
128. Spekkens, R.W. Quasi-quantization: Classical statistical theories with an epistemic restriction. In Quantum
Theory: Informational Foundations and Foils; Chiribella, J., Spekkens, R.W., Eds.; Springer: New York, NY, USA,
2016; pp. 83–136.
129. Bartlett, S.D.; Rudolph, T.; Spekkens, R.W. Reconstruction of Gaussian quantum mechanics from Liouville
mechanics with an epistemic restriction. Phys. Rev. A 2012, 86, 012103. [CrossRef]
130. Spekkens, R.W. Quasi-quantization: Classical statistical theories with an epistemic restriction. In Quantum
Theory: Informational Foundations and Foils; Chiribella, J., Spekkens, R.W., Eds.; Springer: New York, NY, USA,
2016; p. 91.
131. Spekkens, R.W. Quasi-quantization: Classical statistical theories with an epistemic restriction. In Quantum
Theory: Informational Foundations and Foils; Chiribella, J., Spekkens, R.W., Eds.; Springer: New York, NY, USA,
2016; p. 95.
132. Spekkens, R.W. Quasi-quantization: Classical statistical theories with an epistemic restriction. In Quantum
Theory: Informational Foundations and Foils; Chiribella, J., Spekkens, R.W., Eds.; Springer: New York, NY, USA,
2016; p. 96.
133. D’Ariano, G.M.; Chiribella, G.; Perinotti, P. Quantum Theory from First Principles: An Informational Approach;
Cambridge University Press: Cambridge, UK, 2017; p. 6.
134. D’Ariano, G.M.; Chiribella, G.; Perinotti, P. Quantum Theory from First Principles: An Informational Approach;
Cambridge University Press: Cambridge, UK, 2017; pp. 168–169.
135. Schrödinger, E. Discussion of probability relations between separated systems. Proc. Camb. Philos. Soc. 1953,
31, 555–563. [CrossRef]
136. Chiribella, G.; D’Ariano, G.M.; Perinotti, P. Informational derivation of quantum theory. Phys. Rev. A 2011,
84, 012311-1–012311-39. [CrossRef]
137. Hardy, L. Foliable operational structures for general probabilistic theory. In Deep beauty: Understanding the
Quantum World through Mathematical Innovation; Halvorson, H., Ed.; Cambridge University Press: Cambridge,
UK, 2011; pp. 409–442.
138. Hardy, L. A formalism-local framework for general probabilistic theories, including quantum theory. arXiv
2010, arXiv:1005.5164; pp. 19–20.
139. Hardy, L. A formalism-local framework for general probabilistic theories, including quantum theory. arXiv
2010, arXiv:1005.5164; p. 39.
140. MacLane, S. Categories for the Working Mathematician; Springer: Berlin, Germany, 2013.
141. MacLane, S.; Moerdijk, I. Sheaves in Geometry and Logic: A First Introduction to Topos Theory; Springer: Berlin,
Germany, 1994.
142. Hasegawa, M.; Hofmann, M.; Plotkin, G. Finite dimensional vector spaces are complete for traced symmetric
monoidal categories. Lect. Notes Comput. Sci. 2008, 4800, 367–385.
143. Coecke, R.; Kissinger, A. Picturing Quantum Processes: A First Course in Quantum Theory and Diagrammatic
Reasoning; Cambridge University Press: Cambridge, UK, 2017.

146

www.pdfgrip.com
Entropy 2018, 20, 656

144. D’Ariano, G.M.; Perinotti, P. Derivation of the Dirac equation from principles of information processing.
Phys. Rev. A 2014, 90, 062106. [CrossRef]
145. D’Ariano, G.M. Physics without Physics. Int. J. Theor. Phys. 2017, 56, 97–128. [CrossRef]

© 2018 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

147

www.pdfgrip.com
 entropy
Article
SU (2) Decomposition for the Quantum Information
Dynamics in 2d-Partite Two-Level Quantum Systems
Francisco Delgado †
Escuela de Ingeniería y Ciencias, Tecnológico de Monterrey, Atizapán 52926, Mexico; fdelgado@itesm.mx;
Tel.: +52-55-5864-5670
† Current address: Departamento de Física y Matemáticas, Tecnológico de Monterrey,
Campus Estado de México, Atizapán, Estado de México, Mexico.

Received: 1 June 2018; Accepted: 2 August 2018; Published: 17 August 2018 

Abstract: The gate array version of quantum computation uses logical gates adopting convenient
forms for computational algorithms based on the algorithms classical computation. Two-level
quantum systems are the basic elements connecting the binary nature of classical computation with
the settlement of quantum processing. Despite this, their design depends on specific quantum systems
and the physical interactions involved, thus complicating the dynamics analysis. Predictable and
controllable manipulation should be addressed in order to control the quantum states in terms of
the physical control parameters. Resources are restricted to limitations imposed by the physical
settlement. This work presents a formalism to decompose the quantum information dynamics in
SU (22d ) for 2d-partite two-level systems into 22d−1 SU (2) quantum subsystems. It generates an
easier and more direct physical implementation of quantum processing developments for qubits.
Easy and traditional operations proposed by quantum computation are recovered for larger and
more complex systems. Alternating the parameters of local and non-local interactions, the procedure
states a universal exchange semantics on the basis of generalized Bell states. Although the main
procedure could still be settled on other interaction architectures by the proper selection of the basis
as natural grammar, the procedure can be understood as a momentary splitting of the 2d information
channels into 22d−1 pairs of 2 level quantum information subsystems. Additionally, it is a settlement
of the quantum information manipulation that is free of the restrictions imposed by the underlying
physical system. Thus, the motivation of decomposition is to set control procedures easily in order to
generate large entangled states and to design specialized dedicated quantum gates. They are potential
applications that properly bypass the general induced superposition generated by physical dynamics.

Keywords: quantum information; quantum dynamics; entanglement

1. Introduction
Quantum information is generating new applications and tentative future technologies such
as quantum computation [1–3] and quantum cryptography, based on disruptive phenomena in its
main trends: quantum key distribution [4,5], quantum secret sharing [6], and quantum secure direct
communication [7,8]. All these trends highlight the importance of entangled states—a basic aspect
involved in the current work in order to achieve quantum information processing tasks. In this
arena, the understanding of quantum information dynamics and the control of quantum systems
is a compulsory development to manage the quantum resources involved. Applications require a
tight control of resources and interactions—especially those related with coherence and entanglement.
They are fundamental in most applications. Quantum control has developed the fine management
of physical variables to prepare, maintain, and transform quantum states in order to exploit them

Entropy 2018, 20, 610; doi:10.3390/e20080610 148 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2018, 20, 610

for concrete purposes. The outstanding high-tech commercial appliances D-Wave and IBM-Q use
qubits in the form of two-level systems, either with superconducting circuits or ions as well as several
approaches to their interconnecting architecture.
For multipartite systems, research in control is oriented to achieve different goals in quantum
applications. Most of them are numerical approaches rather than analytical due to the inherited
complexity in the quantum information dynamics when the number of parts grows. For a single system
with a two-level spectrum, the control problem has been extensively studied in terms of exact optimal
control for energy or time cost [9,10]. Recently, research of the anisotropic Heisenberg–Ising model
for bipartite systems in SU (4) [11] has shown how this model exhibits SU (2) block decomposition
when it is written in the non-local basis of Bell states instead of the traditional computational basis.
This means that H2 becomes a direct sum of two subspaces, each one generated by a pair of Bell
states, while U underlies in the direct product U (1) × SU (2)2 . Thus, control can be reduced to SU (2)
control problems, each one in each block. Then, exact solutions for some control procedures can
be found [12,13]. There, controlled blocks can be configured by the direction of external driven
interactions introduced. That scheme allows controlled transformations between Bell states on demand
and therefore on a general state. Thus, the procedure sets a method of control to manipulate quantum
information on magnetic systems, where the computational grammar is based on Bell states instead
of the traditional computational basis. It allows an easier programmed transformation among any
pair of elements in that basis. This result provided the inspiration to reproduce similar decomposition
schemes for larger systems in terms of simpler problems based on quasi-isolated two-level subsystems,
developing easier and universal (not necessarily optimal) controlled manipulation procedures for
quantum information. Technology to set up the possible architecture of these generic systems is
currently being achieved through trapped-ion qubits [14] and superconducting qubits [15].
Thus, the generalization of SU (2) block decomposition is a convenient formalism to express
dynamics, revealing certain quantum information states algebraically free of the complexity
introduced by the entangling operations (doing few convenient the use of the computational basis).
Nevertheless, they still conserve their entangled properties. This reveals how the probability exchange
happens together with the structure of entanglement behind the randomness introduced by the
complexity of large quantum information systems. Still, as for their predecessor, those bases maintain
a certain degree of universality, including several alternative local and non-local interactions. As for
their SU (4) predecessor, when they are combined, it states a series of punctual operations that can
be set: (a) fine control based on well-known SU (2) control procedures; (b) the construction of universal
gates for the entire process based on two-channel like operations; and (c) the design of more complex
dedicated multi-channel gates by factorization.
The general aim of this paper is to show that such decomposition and reduction is achievable for
large qubit systems, not only for those in [11,13]. The second section states the general Hamiltonian to
be analyzed. The third section shows how the SU (2) decomposition procedure can be generalized
on general n-partite two-level systems (not only for the driven Heisenberg–Ising interactions),
reducing them to 2n−1 selectable transformations between pairs of specific quantum energy states.
Then, these transformations can be based on known control schemes for SU (2) systems such as those
in [9,10]. The selection of these 2n−1 pairs of states can be based on the convenience of the quantum
process being considered and the resources involved. Thus, the basis on which the decomposition
can be established works as a computational grammar for the quantum procedures being attained.
These bases are not completely arbitrary, and thus the fourth section shows how a kind of generalized
Bell basis is able to generate the SU (2) decomposition for an even number of parts, n = 2d. The fifth
section is devoted to analyzing the restrictions on the Hamiltonian to get the SU (2) decomposition,
the inherited states, and the block properties. This analysis includes a classification of interactions
able to generate the SU (2) decomposition. Because the presented procedure can reproduce complex
quantum gates, generate large entangled states, and introduce control procedures in SU (22d ) if the
grammar is based on the proposed basis, the sixth section analyzes potential applications in these

149

www.pdfgrip.com
Entropy 2018, 20, 610

trends. The final section concludes, summarizing the findings and settling the related future work to
be developed. Because several aspects in the work may be complex for the reader, an appendix with
some critical concepts has been included to clarify the contents.

2. Generalized Hamiltonian
The problem can be established for a general Hamiltonian for n coupled two-level systems on
U (2n ) forming a closed system. It can be written as a general combination of tensor products of
Pauli matrices for each subsystem (for a more detailed discussion of this Hamiltonian, please see
Appendix A.1):

0
n 4n −1 0
n
H̃ = ∑ h {ik } σik = ∑ hI4n σI n ,
4,k
(1)
{ik } k =1 I=0 k =1

where {ik } = {i1 , i2 , . . . , in }, ik = 0, 1, 2, 3, and h{ik } is a general set of time-dependent real

functions. Sometimes, as in the second expression in (1), {ik } will be represented as the number
I ∈ {0, 1, . . . , 4n − 1}, as it is expressed in base-4 with n digits, I4n . Then, I4,k
n = i for k = 1, 2, . . . , n.
k
σik for ik = 0, 1, 2, 3 are the traditional Pauli matrices in the computational basis |0 , |1 ∈ H2 for each
part k. Note that due to the SU (2) algebra of Pauli matrices, this Hamiltonian comprises all analytical
Hamiltonians based on two-level systems with n parts. The Hamiltonian obeys the Schrödinger
equation for its associated evolution operator Ũ:

∂Ũ
H̃ Ũ = ih̄ . (2)
∂t

Although h{0,...,0} is not necessarily zero, if { Ẽj | j = 1, . . . , 2n } are the eigenvalues of H̃ and
n
E ≡ ∑2j=1 Ẽj = 2n h{0,...,0} , then defining

0
n
i
H ≡ H̃ − h{0,...,0} σ0 , U ≡ Ũe h̄ h{0,...,0} t , (3)
k =1

these operators become the equivalent traceless Hamiltonian and its corresponding evolution operator
with eigenvalues Ej = Ẽj − h{0,...,0} , both fulfilling (2) as well. H and H̃ have the same set of
1 n
eigenvectors { b j ∈ H2 | j = 1, . . . , 2n }. Thus, the Hamiltonian can be written alternatively as
2n
1 2
H = ∑ j=1 Ej b j b j . Thus, in the following, the Hamiltonian can be assumed traceless without loss
of generality. Note that while Ũ ∈ U (2n ), then U ∈ SU (2n ). In the following, only H and U symbols
will be used as equivalent to H̃ and Ũ. H can be split in two parts—the local Hl and the non-local
Hnl interactions:

n 3 0
n
Hl = ∑ ∑ h(i4k−1 )4n σ(i4k−1 )n → H̃ = H̃nl + Hl ,
4,s
(4)
k =1 i =1 s =1

where (i4k−1 )4n is the number i4k−1 represented in base-4 with n digits and (i4k−1 )4,s
n is its sth term in

that basis.

3. SU (2) Decomposition Generalities

In order to support the understanding of some aspects in the further discussion, Appendix A.2
contains a brief of group theory that is relevant for this work as well as some critical points in the
decomposition procedure being presented here. Delgado [11] found that the SU (2) decomposition
procedure can be induced by considering a set of 2n orthogonal states: {|αi } and 2n−1 pairs

150

www.pdfgrip.com
Entropy 2018, 20, 610

{ j(i ), k(i )}, i = 1, 2, ..., 2n−1 , with k(i ) = j(i ) + 1 ∈ {2, 4, ..., 2n } related with the eigenvalues through a
mixing matrix, in such way that they fulfill:
% % %
2 |b2i−1  = Ai α j(i) + Bi αk(i) → α j(i) = Ai∗ |b2i−1  − Bi e−iφ |b2i  ,
% % % (5)
|b2i  = − Bi∗ eiφ α j(i) + Ai∗ eiφ αk(i) → αk(i) = Bi∗ |b2i−1  + Ai e−iφ |b2i  ,

with | Ai |2 + | Bi |2 = 1, where last relations are clearly true because of orthogonality (note that energies
1
Ej become ordered as the states are paired). States { α j } are then defined by the selection of Ai , Bi .
Each pair sets one of the orthogonal subspaces:

% % n
n −1
23
Hi2 = span({|b2i−1  , |b2i }) = span({ α j(i) , αk(i) }) → H2 = Hi2 . (6)
i =1

There are many possibilities for this selection, but not necessarily all practical bases fit
in this construction. In particular, separability or entanglement properties are not
necessarily assured
$ % {|αi }
for $ as in [11].
%∗ Clearly,
$ because%∗these $states are% assumed unitary,
then Ai = b2i |αk(i) eiφ = b2i−1 |α j(i) , Bi = b2i−1 |αk(i) = − b2i |α j(i) eiφ . By applying H
on (5) and considering that |bi  has the eigenvalue Ei , it is possible to arrive at the following expressions:
% % %
H α j(i) = (| Ai |2 E2i−1 + | Bi |2 E2i ) α j(i) + Ai∗ Bi ( E2i−1 − E2i ) αk(i) ,
% % % (7)
H αk(i) = Ai Bi∗ ( E2i−1 − E2i ) α j(i) + (| Ai |2 E2i + | Bi |2 E2i−1 ) αk(i) ,

giving the Hamiltonian components in this basis:

$ %
α j(i) | H |α j(i) = | Ai |2 E2i−1 + | Bi |2 E2i ,
$ %
αk(i) | H |αk(i) = | Ai |2 E2i + | Bi |2 E2i−1 , (8)
$ %
α j(i) | H |αk(i) = Ai Bi∗ ( E2i−1 − E2i ),

which can be alternatively obtained from (5). Note that the phase φ is non-physical. This basis
transformation changes the diagonal structure for the basis {|bi } into a 2 × 2 diagonal block structure
1
for the basis { α j }. For simplicity, we define the following quantities:

Ai = r Ai eiγ Ai , Bi = r Bi eiγBi ,
1
Δi± = ( E ± E2i−1 ), (9)
2h̄ 2i
Γi = γ Ai − γBi .

Then, each 2 × 2 block in H (labeled as S H i ) can be written in matrix form as (see Appendix A.2):
 
Δi+ − (r Ai 2 − r Bi 2 )Δi− −2r Ai r Bi Δi− eiΓi
SH i =
−2r Ai r Bi Δi− e−iΓi Δi+ + (r Ai 2 − r Bi 2 )Δi−
= Δi+ Ii − 2r Ai r Bi Δi− cos Γi Xi + 2r Ai r Bi Δi− sin Γi Yi − (r Ai 2 − r Bi 2 )Δi− Zi (10)
≡ Δi+ Ii + S H 0i ,

where although r Ai 2 + r Bi 2 = 1, we use both terms r Ai and r Bi for the symmetry. In addition, Ii , Xi , Yi ,
and Zi are respectively the 2 × 2 unitary matrix and the Pauli matrices settled as basis for the block S H i .

151

www.pdfgrip.com
Entropy 2018, 20, 610

Thus, H can be written as a sum of operators acting on the different subspaces Hi2 or as the following
direct sum structure of 2n−1 2 × 2 block-diagonal matrices:
⎛ ⎞
SH 1 0 ... 0
2n −1 ⎜ SH 2 ⎟
3 ⎜ 0 ... 0 ⎟
H = SH i =⎜
⎜ .. .. .. .. ⎟,
⎟ (11)
i =1 ⎝ . . . . ⎠
0 0 ... S H 2n −1

with 0, the 2 × 2 zero matrix. Because this structure is preserved under matrix products, it is
inherited by the evolution matrix U. In particular, if the Hamiltonian (1) is not time-dependent,
12
then U = ∑2j=1 e− h̄ Ej t b j b j . Thus, when the basis is changed (see Appendix A.2):
n i

SU i = e− h̄ E2i−1 t |b2i−1  b2i−1 | + e− h̄ E2i t |b2i  b2i |

i i


+ − 12
= e−iΔi t (eiΔi t − 2ir Bi 2 sin Δi− t) α j (i ) α j (i ) +
1 2 (12)
2ir Ai r Bi sin Δi− t(eiΓi α j (i ) αk (i )| + e−iΓi |αk (i ) α j (i ) )+

−
(e−iΔi t + 2ir Bi 2 sin Δi− t) |αk (i ) αk (i )| .

Similarly, in matrix form or in terms of Ii , Xi , Yi , and Zi :

 
+ cos Δi− t + i (r Ai 2 − r Bi 2 ) sin Δi− t 2ir Ai r Bi eiΓi sin Δi− t
SU i = e−iΔi t
2ir Ai r Bi e−iΓi sin Δi− t cos Δi− t − i (r Ai 2 − r Bi 2 ) sin Δi− t

+
= e−iΔi t
cos Δi− tIi + 2ir Ai r Bi cos Γi sin Δi− tXi − (13)

+
2ir Ai r Bi sin Γi sin Δi− tYi + i (r Ai 2 − r Bi 2 ) sin Δi− tZi ≡ e−iΔi t SU 0i .

Note that the election of Γi lets us simplify the last expression to contain only one operator
between Xi and Yi (as in [11,12]). This property is useful to set the optimal control in [9] in each block.
Then, similar to H:
⎛ ⎞
SU 1 0 ... 0
n − 1 ⎜ SU 2 . . . ⎟
23 ⎜ 0 0 ⎟
U = SU i = ⎜
⎜ .. . .. . ⎟,
⎟ (14)
⎝ . .
. . .
. ⎠
i =1
0 0 . . . SU 2n−1

where in general for the time-dependent Hamiltonian:

t +
t +
S H 0i
= τ {e− h̄ SH i
dt } = e−iΔi t τ {e− h̄ dt } ≡ e−iΔi t SU 0i ,
i i
SU i 0 0 (15)

where τ is the time-ordered integral. This implies that U is an element in the direct product
n −1 n −1 +
U (1)2 −1 × SU (2)2 ⊂ SU (2n ) (because any factor phase e−iΔi t depends on the remaining
phase factors through E , see Appendix A.2). In the following, we will informally call this
factorization the SU (2) decomposition (in reality, each block has the form U (1) × SU (2)) due to
the block structure. Consequently,
% the%Hilbert space H n becomes the direct sum of 2n−1 subspaces
generated by each pair { α j(i) , αk(i) }, i = 1, 2, . . . , 2n−1 . In each subspace, dynamics mixes the
probabilities, but probabilities among subspaces remain unmixed if there is no rearrangement in
1 1
the pairing between {|bi } and { α j } (clearly, in this rearrangement, the basis { α j } could change).

152

www.pdfgrip.com
Entropy 2018, 20, 610

n −1
% %
Thus, if |ψ0  = ∑2i=1 |ψ0 i  is the initial state with |ψ0 i  = ψ0 i,j(i) α j(i) + ψ0 i,k(i) αk(i) , then each
component is evolved in the subspace i = 1, 2, ..., 2n−1 , fulfilling |ψt i  = SU 0i |ψ0 i  = |ψ0 i  .
Finally, note that the SU (2) decomposition is not the only one available, although it is the most
valuable for the binary inheritance from the classical computation. In fact, other decompositions
involving bigger subgroups are possible, whether using bigger systems than two-level ones and/or
simply involving more than two eigenvectors in (5). Inclusively, a mixed-sized block structure can
be realized.
1
4. GBS: A Non-Local Basis Fitting in { α j }
Non-local bases are used as a theoretical resource to explicitly show how evolution [16] and
measurement [17] can generate entangled states. In [11], it was shown that the Heisenberg–Ising model
including driven magnetic fields in a fixed direction allows the generation of the block structure in the
traditional Bell basis. Thus, the Bell basis for two-level bipartite systems has been shown to fit in the
U (1) × SU (2)2 decomposition of SU (4). Despite the added complexity to manage non-local states,
recent work has moved towards the control of entangled states [18]. This basis works as a universal
basis for the Heisenberg–Ising interaction, including an external magnetic field in any specific direction
on a couple of qubits [11–13]. This model includes other interaction models, such as XXX [19], XY [20],
and XXZ [21]. In the current development, the most obvious guess is the generalized Bell states (GBS)
basis for n = 2d presented in [22] as tensor products of Bell states. In the next sections, some further
useful formulas are obtained to show how the GBS basis fits in the SU (2) decomposition for larger
systems than bipartite ones.

4.1. GBS Basis and Hamiltonian Components

For n = 2d, the GBS basis [22] forms an orthogonal basis of partial entangled states for 2d
particles. A more extended treatment for this basis is given in Appendix A.3 in order to ease further
understanding in the current context, particularly related with the underlying single Bell states in their
construction together with their index notation—a key aspect in the remaining development. Each
element in this basis can be written briefly as:

% 0
d
1 1
ΨI d
4
= √ ∑
2 s ,δs =0
(σ̃is )s ,δs |s δs 
s =1
1
= √ ∑ (σ̃i1 ⊗ . . . ⊗ σ̃id )1 ...d ,δ1 ...δd |1 . . . d  ⊗ |δ1 . . . δd  (16)
2d { j },{δk }

1 2d − 1 % %
= √ ∑ (σ̃i1 ⊗ . . . ⊗ σ̃id )E d ,D d E2d ⊗ D2d ,
2d E ,D=0
2 2

where { j } = {1 , . . . , d }, {δk } = {δ1 , . . . , δd };  j , δk = 0, 1. At this point, σ̃i can be considered as

proportionally unitary to the traditional Pauli matrices [22]. In addition, I4d is a brief expression
of {i1 , i2 , . . . , id } as the digits set of I ∈ {0, 1, . . . , 4d − 1} when it is written in base-4 with d digits.
Similarly, E2d , D2d are numbers written in base-2 with d digits (E , D ∈ {0, 1, . . . , 2d − 1}) representing
{1 , . . . , d }, {δ1 , . . . , δd }, respectively (note that digits are used inverted, as they commonly appear in
E2d or I4d expressions). In the following, for simplicity, we use Ibd and I interchangeably because the
base b can normally be inferred from the context. Each element in this basis is not maximally entangled.
Instead, they have maximally entangled bipartite subsystems (see Appendix A.3), which are separable
from the remaining system. Separable pairs contain the parts [s, s + d], s = 1, 2, ..., d (in the following,
square brackets will be%used to point out a subsystem of parts in the whole system).
1
In order for { ΨI d } (I ∈ {0, 1, . . . , 4d − 1}) to reach the kind of sets { α j } stated in the previous
4
section where H and U achieve the SU (2) block structure, H should fulfill some restrictions. We are

153

www.pdfgrip.com
Entropy 2018, 20, 610

interested in setting these in the current subsections. Combining expressions (1) and (16), we can
express the components of H in the GBS basis. First, we note [23]:

$ % d
1 1
1 1 2 1$ %
ΨI d |σj1 ⊗ . . . ⊗ σj2d |ΨKd = ∏ √2 ∑ (σ̃i∗s )s ,δs √ ∑ (σ̃ks )γs ,φs s |σjs |γs δs |σjs+d |φs
4 4
s =1 s ,δs =0 2 γs ,φs =0
1
=
2d
∑ (σ̃i∗1 ⊗ . . . ⊗ σ̃i∗d )E2d ,D2d (σj1 ⊗ . . . ⊗ σj2d )E2d D2d ,F2d G2d (σ̃k1 ⊗ . . . ⊗ σ̃kd )F2d ,G2d (17)
E ,D
F ,G
d
1
=
2d
∏ Tr(σ̃i∗s σjd+s σ̃kTs σjTs ),
s =1

where combined subscripts as E2d D2d represent the set of subscripts obtained by merging {1 . . . d } and
{δ1 . . . δd }. Therefore, the final and notable expression for the Hamiltonian components becomes [23]:

$ % 1 42d −1 d
ΨI d | H |ΨKd
4 4
=
2d
∑ hJ 2d
4
∏ Tr(σ̃i∗s σjd+s σ̃kTs σjTs ), (18)
J =0 s =1

where J ∈ {0, 1, . . . , 42d − 1} (here, J = 0 can be removed in spite of the discussion in the first section).
In the last expressions, the product σ̃i∗s σjd+s σ̃kTs σjTs has some properties inherited from Pauli matrices.
Because σ1 , σ2 , σ3 are traceless and σiT = ±σi (negative sign only if i = 2), then Tr(σ̃i∗s σjd+s σ̃kTs σjTs ) is
non-zero only if is , jd+s , k s , js are: (a) completely different between them; or (b) equal by pairs.
A remark is convenient at this point. In some works (e.g., [22]), GBS are preferred to be defined
using σ̃i = σi for i = 0, 1, 3 and σ̃2 = iσ2 because it allows real coefficients when they are expressed in
the computational basis |0 , |1 (alternative definitions introduce specific phase factors in σ̃i ). We adopt
the last definition in the following, which does not produce changes in the previous discussion. Note
that σ̃i ∗ = σiT = σi . The last expression should be fitted to (11), in particular with the non-diagonal
block entries. In the following sections, we will show that the GBS basis naturally generates the SU (2)
decomposition if the Hamiltonian fulfills certain restrictions. The use of the GBS basis allows the
management of this analysis because it is based on Pauli matrices.

4.2. Case d = 1
For d = 1 there are three possibilities to arrange the pairs in the corresponding GBS basis (reduced
in this case to the traditional Bell states: {| β 00  , | β 01  , | β 10  , | β 11 }). A direct but large analysis shows
that by fitting (18) to (11), the Hamiltonian should be reduced to the forms shown in Table 1 (assuming
always h02d = 0 and H0 = ∑3j=1 h jj σj ⊗ σj ). The first column shows the pairs arrangement to construct
4
the blocks. These results generalize those found in [11,12] for the anisotropic Heisenberg–Ising model
reached if the crossed interaction terms such as hij σi ⊗ σj with i, j = 1, 2, 3; i = j are not present.
These terms are similar to the Dzyaloshinskii–Moriya model [24,25], opening additional possibilities
for control in the pair exchange. Case d = 1 is special in the current context because for d > 1 crossed
terms can be present only for a unique pair in order to keep the SU (2) decomposition.

Table 1. Basis pairs and Hamiltonian required to get the SU (2) block decomposition for case d = 1.

Basis Arrangement Hamiltonian

{{| β 00  , | β 01 }, {| β 11  , | β 10 }} H = H0 + h01 σ0 ⊗ σ1 + h10 σ1 ⊗ σ0 + h23 σ2 ⊗ σ3 + h32 σ3 ⊗ σ2
{{| β 00  , | β 11 }, {| β 01  , | β 10 }} H = H0 + h02 σ0 ⊗ σ2 + h20 σ2 ⊗ σ0 + h13 σ1 ⊗ σ3 + h31 σ3 ⊗ σ1
{{| β 00  , | β 10 }, {| β 01  , | β 11 }} H = H0 + h03 σ0 ⊗ σ3 + h30 σ3 ⊗ σ0 + h12 σ1 ⊗ σ2 + h21 σ2 ⊗ σ1

Although the eigenvalues { Ej } do not follow a specific order, expressions in (18) can be arranged in
% %
several orders as functions of the pairs selected { α j(i) , αk(i) }, being related with the decomposition

154

www.pdfgrip.com
Entropy 2018, 20, 610

(22d )!
process. In general, there are combinations for these pairs, which grow very quickly with
(22d−1 )!222d−1
d (3 for d = 1; 2, 027, 025 for d = 2, etc.), making the cases for d > 1 unmanageable in an analogous
direct analysis.

4.3. Case d > 1

The exponential growth of the problem with d makes an exhaustive analysis for d > 1 based on a
large algebraic equation system impossible, as in the previous case. The previous case and the results
in [11,12] suggest some possible Hamiltonians for more complex cases. Thus, some of the following
forms (see Appendix A.1) could allow the SU (2) decomposition for the basis (16):

3
( j) ( j)
H0 = ∑ H0 , H0 = h 2d −1 2d σj⊗2d , (19)
(j 43 )4
j =1
2d 2d
0
(k,k ) (k,k )
Hnli = ∑ Hnl
i
, Hnl
i
= h(i(4k−1 +4k −1 ))2d
4
σ(i(4k−1 +4k −1 ))2d ,
4,s
(20)
k > k =1 s =1
2d 1 2d
0
(k,k ) (k,k )
Hcnli = ∑ Hcnl ,
i
Hcnl
i
= ∑ h( jp 4k−1 +k p 4k −1 )2d4 σ( j
p4
k −1 + k
p4
k −1 )2d
4,s
, (21)
k > k =1 p =0 s =1
2d 2d
0
(k) (k)
Hli = ∑ Hli , Hl
i
= h(i4k−1 )2d
4
σ(i4k−1 )2d ,
4,s
(22)
k =1 s =1

where (i4k−1 )2d

4 is the base-4 representation with 2d digits of i4
k −1 , a number with only one i in position
2d −1
k and zero in the other; (i4k−1 )2d
4,s is its element s; and ( j
4
3 )2d
4 is the base-4 representation with 2d
42d −1
digits of j 3 , a number with j in each digit position (by using the geometric partial sums properties).
Note that i ∈ {1, 2, 3} is fixed in all expressions. Physically, H0 represents a full simultaneous
interaction between all particles (as in the bipartite Heisenberg–Ising anisotropic interaction). Although
this kind of interaction is non-physical for d > 1, it is included here for reference. Hnli represents the
interaction between the component i of the spin for pairs of particles as in the Heisenberg–Ising model.
Hcnli is the crossed non-local interactions by pairs in the direction i (as those for d = 1 in Table 1), a
label used to characterize these interactions (as in the Dzyaloshinskii–Moriya model). Note that i, j p , k p
is a permutation of 1, 2, 3 with parity p = 0, 1, even and odd, respectively. Finally, Hli is the component
i of the local interactions with h(i4k−1 )2d as strengths (e.g., magnetic fields in the direction i for magnetic
4
systems). These cases generalize the bipartite models presented in [11,12] and those found for d = 1.
Some observations are useful at this point: (a) σ̃i = αi σi , αi ∈ {1, i }; (b) σiT = β i σi ,
β i ∈ {−1, 1}; (c) σi σj = γi,j σj σi , γi,j ∈ {−1, 1}. Thus, 2cijss ,k ,jd+s ≡ Tr( σ̃is σjd+s σ̃k s σjs ) =
s T T

αis αks β js β ks γks js γks is Tr(σis σks σjd+s σjs ) ∈ {0, ±2, ±2i }. We do not provide extensive formulas for
the coefficients αi , β i , γi,j , cijss ,k s
,jd+s , but they are trivially constructed departing from the Pauli
matrices properties.
At this point, a convenient definition is introduced for the following cases. We will say that
two particles or parts, i, j, are correspondents if j = i + d, with i, j − d ∈ {1, 2, ..., d}. This means
simply that one is in the same position of the first group of the Hamiltonian $ subscripts
% $1, 2, ..., d as the
%
other is in the second group d + 1, d + 2, ..., 2d. Then, the analysis of ΨI d | H0 |ΨKd , ΨI d | Hli |ΨKd ,
$ % $ % 4 4 4 4

ΨI d | Hcnli |ΨKd and ΨI d | Hnli |ΨKd is conducted with the following results.
4 4 4 4

$ %
4.3.1. Analysis of ΨI d | H0 |ΨKd
4 4

j 4 3−1
2d
Because J = in (18), then jd+s = js = j ∀s = 1, 2, ..., d, implying cijss ,k
,jd+s  = 0 only if
s

is = k s ∀s = 1, 2, . . . , d. Thus, H0 is diagonal in the GBS basis representation and each entry will

155

www.pdfgrip.com
Entropy 2018, 20, 610

contain the same three terms h 2d −1 2d for j = 1, 2, 3, but each with diverse signs. Despite the similitude
(j 4 3 )4
of H0 with the bipartite case (d = 1), for multipartite cases this interaction is non-physical, but it allows
the main idea to be introduced and understood in the remaining analysis.
$ %
4.3.2. Analysis of ΨI d | Hli |ΨKd
4 4

The treatment for the remaining cases is compressed in the explanation of the current case. By first
(k)
considering only an isolated term Hl (in this case J = i4k−1 for some i ∈ {1, 2, 3} and k = 1, 2, . . . , 2d
i
in (18)), then J in the base-4 representation contains only one i (in the position k) while other digits are
zero. Thus, there are only two meaningful possibilities for each correspondent part: (1) jd+s = js = 0
in most cases, so is = k s is the only case with cijss ,k
,j
s
= 0; or (2) one and only one position s = k or
d+s
d + s = k in J4d has jk = i, either for js or jd+s , while the other is zero. This last case implies only two
possibilities for is , k s : Case (A) one of is , k s is i and other is zero (both possibilities are possible); or Case
(B) i, is , k s are different among them and from zero, thus they are a permutation i, i , i of 1, 2, 3. In this
case, there are two possibilities, is = i , k s = i or is = i , k s = i . $ %
Case A is depicted in Figure 1 for indexes I , K being considered in ΨI d | Hli |ΨKd . There
4 4
is a pair of entries whose labels for rows and columns have 0 or i in the position s = k:
((i1 , i2 , . . . , i, . . . , id ), (i1 , i2 , . . . , 0, . . . , id )) and ((i1 , i2 , . . . , 0, . . . , id ), (i1 , i2 , . . . , i, . . . , id )). This will be
named the 0 ↔ i association (or index exchange) rule.

$ %
Figure 1. First case for a pair of entries in which ΨI d | Hli |ΨKd is non-zero. In them, for a fixed
4 4
position s = k in the row and the column labels appears i or 0, while the other corresponding positions
in the row and in the column have the same values.

Case B is depicted in Figure 2. Here, there is a pair of entries whose labels for rows
and columns have i or i in the position s = k: ((i1 , i2 , . . . , i , . . . , id ), (i1 , i2 , . . . , i , . . . , id )) and
((i1 , i2 , . . . , i , . . . , id ), (i1 , i2 , . . . , i , . . . , id )), i, i , i being a permutation of 1, 2, 3. This will be named the
i ↔ i association (or index exchange) rule.
Clearly, in each case (A or B), for each pair of correspondent interaction terms with i and k fixed
(k ≤ d and k + d positions), there are only two pairs on non-zero entries in rows (i1 , i2 , . . . , i, . . . , id ),
(i1 , i2 , . . . , 0, . . . , id ) for case A and in rows (i1 , i2 , . . . , i , . . . , id ), ((i1 , i2 , . . . , i , . . . , id )) for case B (with
the corresponding column labels exchanged in both cases). Together with the diagonal entries
generated by other adequate Hamiltonians (e.g., H0 or Hnli as it will be seen), they will form 2 × 2
blocks. In fact, each non-zero entry for Hli will have only two hJ terms corresponding with h0,0...0,i,0,...,0
with i in positions s or d + s (meaning local interaction with each element of the pair of correspondent
parts in position k). Noting that labels in the position s = k in I (row) and K (column) for the non-zero
entries are 0, i; i, 0; i , i ; or i , i , they cover all possibilities ik = 0, 1, 2, 3. Thus, for a fixed column and

156

www.pdfgrip.com
Entropy 2018, 20, 610

$ %
defined i, k values in ΨI d | Hli |ΨKd , there is exactly one non-zero row. Still, if two correspondent k
4 4
elements are considered (local interactions on each element of a correspondent pair), they still generate
only one non-zero row (each one with the two terms explained before).

$ %
Figure 2. Second case for a pair of entries in which ΨI d | Hli |ΨKd is non-zero. In them, for a fixed
4 4
position s = k in the row and the column labels appears i or i alternatively (i, i , i being a permutation
of 1, 2, 3), while other corresponding positions in the row and in the column have the same values.

Although there are 2d possibilities to select the position s = k in (22), they do not count as separate
blocks because they appear in other entries in (18). Instead, each term in non-correspondent terms
will appear in a different non-zero row, giving d non-zero rows as total. For each i direction of the
interaction being included, additional non-zero rows will appear. This implies that 3d rows could
appear when all parts have local interactions in the three spatial directions at time, destroying in this
case the 2 × 2 block structure. Thus, maintaining local interactions in only one direction and on only
one correspondent pair of elements, together, cases A and B form 12 4d 2 × 2 blocks as was required
in the previous section. In any case, each non-zero entry will have the same 2 terms h(i4k−1 )2d with
4
(k)
different signs depending on cijss ,k s
,jd+s involved in each factor of Hli . Clearly, blocks can be rearranged
to adequately order the GBS basis elements getting the form (11). A brief analysis shows that there are
no more diagonal-off elements in addition to last cases being generated by local terms. Additional
diagonal-off elements come from the non-local terms, such as those in Table 1.
$ %
4.3.3. Analysis of ΨI d | Hnli |ΨKd
4 4

With the correspondent parts definition and the analysis for Hli , we can identify two cases for
(k,k )
the different terms Hnl : (a) non-local interactions between correspondent parts; and (b) non-local
i
interactions between non-correspondent parts. The discussion is similar to the previous subsection.
(k,k+d)
Correspondent terms Hnl . This term in the Hamiltonian Hnli contains σ0 ⊗ ... ⊗ σi ⊗ ... ⊗ σi ⊗ ... ⊗
i $ %
σ0 with σi in positions k and k + d, and σ0 in any other. When this term is allocated in ΨI d | Hnli |ΨKd
4 4
in agreement with (18), it does not cancel if each factor in the product become different from zero,
implying is = k s ∀s = 1, 2, ..., d. Thus, this term gives non-zero entries only in the diagonal elements.
Thus, each non-zero entry of Hnli will have d different terms in each diagonal element (one for each
pair of interacting correspondent particles). Those terms will appear with different signs in each
(k,k+d) (k)
diagonal element in spite of cijss ,k s
,jd+s . At this point, note that results for Hnli and Hl were expected
i
due to the results in [11,26] and the separability of the GBS basis in their constitutive entangled pairs.
(k,k =k+d)
Non-correspondent terms Hnl . These terms have a different behavior. Each term contains
i
σ0 ⊗ ... ⊗ σi ⊗ ... ⊗ σi ⊗ ... ⊗ σ0 , with σi in positions k and k , and σ0 in any other. It defines two

157

www.pdfgrip.com
Entropy 2018, 20, 610

pairs of correspondent parts involving σi : [k, k + d, k , k + d] if k < k ≤ d or k, k + d, k − d, k if

k ≤ d < k ≤ 2d. Then, each factor in (18) related with those two pairs (s = k, k or s = k, k − d) will
now include Tr(σis σi σks ) (until unitary factors), which is non-zero only if: (a) is or k s are one of the
pairs 0 and i or i and 0; (b) i, is , k s are a permutation i, i , i of 1, 2, 3 (having two cases depending on
the parity). The last situation is similar to the local terms in the previous subsection, but in two parts
simultaneously. The remaining factors for s = k, k or s = k, k − d will require is = k s in order to
become non-zero. The latter scenario gives 16 possibilities for each term h(i(4k−1 +4k −1 ))2d , which will
4
appear
$ in diagonal-off
% positions obtained departing from the diagonal position (i1 , ..., id ; i1 , ..., id ) in
ΨI d | Hnli |ΨKd , by changing each index in the pair (ik , ik ) in the row, following the rules depicted
4 4
in cases A and B. Thus, for each column and with i, k, k fixed, only one row becomes non-zero, in
agreement with the previous rule. Each entry of this kind involves four terms, including the four
combinations of each pair of non-correspondent parts selected from the set [k, k , k + d, k + d]. Instead,
when all values i and k, k are considered, a total of 3 · 12 d(d − 1) non-zero rows appear in each column
(clearly, by considering all these terms, SU (2) decomposition is not achieved).
$ %
4.3.4. Analysis of ΨI d | Hcnli |ΨKd
4 4

(k,k+d) (k,k+d) (k )
Correspondent terms Hcnl .
For each term Hcnl , the behavior is similar as for Hl . Because
i i i
only one correspondent pair has j p = js = 0 = jd+s = k p in (18), then is , k s for s = k should be
0, i or j p , k p . For s = k , is = k s . As before, it means that each term is diagonal-off by combining
the values of index k in I and K as before: 0, i; i, 0; j p , k p ; and k p , j p . For a fixed column and i, k,
it will give four possibilities and two SU (2) blocks. Each entry will have two terms corresponding
to the different parities p. Note that only one i and k can be considered to achieve the SU (2)
decomposition. Otherwise, for each column, 3d rows different from zero could appear, breaking the
SU (2) decomposition as for the local interaction case.
(k,k =k+d) (k,k =k+d)
Non-correspondent terms Hcnl . As for Hnl , in this case the only non-zero terms have
i i
is = k s for s = k, k , k − d, k − d. Meanwhile, for the two remaining cases s ∈ {k, k , k − d, k − d} ∩
{1, 2, . . . , d}, each is , k s should be selected from the set 0, j p ; j p , 0; i, k p ; k p , i or 0, k p ; k p , 0; i, j p ; j p , i. In a
(k,k =k+d)
specific column and fixing i, it will give 16 possibilities and 8 blocks in SU (2), as for the Hnl
i
case. Note that parity p should be fixed in this case because each one gives a different decomposition.
Each entry will contain four terms for each parity p combining the four possible interaction terms.
Again, if all options for i and k, k , p are considered, then 3 · d(d − 1) non-zero rows will appear
for each column, breaking the SU (2) decomposition. These terms are not commonly introduced
in models such as Heisenberg–Ising and those related. Instead, for magnetic systems they are the
first-order approximation in the spin–orbit coupling, introducing antisymmetric exchange as in the
−
→ → − −
→
Dzyaloshinskii–Moriya model: HDM = D · (− σ1 × →
σ2 ). There, D is the Dzyaloshinskii–Moriya vector
defining the orientation of coupling. Here, as only one term can be included in order to preserve the
SU (2) reduction property, this coupling should be strictly oriented.

4.4. Explicit Analytical Formulas for Hamiltonian Components

After the last analysis, it is clear that other candidates to generate SU (2) decomposition are
possible, but they involve more than two parts at a time (as in the case of H0 ), which are non-physical
for common point-like interactions. Nevertheless, these terms could appear for the quantum
mechanical extended objects in which (1) is a mere expansion of the interactions. Therefore, we
will restrict our
$ remaining%discussion
$ to local
% or pairwise
$ interactions.
% In this section, analytical
formulas for ΨI d | Hli |ΨKd , ΨI d | Hnli |ΨKd , and ΨI d | Hcnli |ΨKd are provided to summarize the
4 4 4 4 4 4
previous findings and because of their utility for optimal computer simulation purposes for larger

158

www.pdfgrip.com
Entropy 2018, 20, 610

systems. In order to simplify the expressions, we introduce the definition of the following generalized
Kronecker delta:
d
S
δIK ≡ ∏ δis ks , (23)
s =1
s/∈S

where S is a set of scripts of the excluded parts in the product. Thus, for Hli :

$ % d
{k }
ΨI d | Hli |ΨKd
4 4
= ∑ δIK Hlki
I4d ,K4d
,
k =1
(24)
1
iδ0,t ,iδ1,t
with : Hlki I d ,Kd
4 4
= ∑ h(i4k +dt −1 )2d Fi,k
4
,
t =0

is ,is
by noting c0,0 = 1. In Hlk , [k , k + d] is the correspondent pair where each local interaction is
i I4d ,K4d
being applied. There, the exchange factor generating the diagonal-off entries in the SU (2) blocks is:

3
∑
j,k
Fi,k = δik 0 δkk i c0,i i,0
j,k + δik i δk k 0 c j,k + ii2 i δik i δkk i cij,k,i . (25)
i ,i =1

For Hnli :

$ % d d
{k } {k ,k }
ΨI d | Hnli |ΨKd
4 4
= ∑ c,k
δIK Hnl
i I4d ,K4d
+ ∑ δIK nc,k k
Hnl
i I4d ,K4d
,
k =1 k > k =1
c,k i ,i
with : Hnl = h(i(4k −1 +4k +d−1 ))2d δik kk ci,ik k , (26)
i I4d ,K4d 4

1
iδ ,iδ1,t iδ ,iδ1,t
nc,k k
Hnl
i I4d ,K4d
= ∑ h(i(4k +dt −1 +4k +dt −1 ))2d
4
Fi,k0,t Fi,k0,t .
t ,t =0

Each term belongs to correspondent and non-correspondent interactions, respectively. In

c,k nc,k k
Hnl and Hnl , [k , k ] are the parts with non-local interactions between them. Similarly,
i I4 ,K4
d d d di I4 ,K4
for Hcnli :

$ % d 1 d
{k } {k ,k }
∑ ∑ ∑
c,k nc,k k p
ΨI d | Hcnli |ΨKd = δIK Hcnl + δIK Hcnl ,
4 4 i I d ,K d i I4d ,K4d
k =1 4 4 p =0 k > k =1
1
j p ,k p
with : c,k
Hcnl
i I d ,K d
= ∑ h( jp 4k −1 +k p 4k +d−1 )2d4 Fi,k , (27)
4 4 p =0
1
jp δ ,j p δ1,t kpδ ,k p δ1,t
∑
nc,k k p
Hcnl = h( j k +dt −1 + k k +dt −1 )2d F jp ,k0,t Fk p ,k0,t .
i I4d ,K4d p4 p4 4
t ,t =0

c,k nc,k k p
Again, Hcnl and Hcnl are the correspondent and non-correspondent interactions in the
i i
Hamiltonian, [k , k ] being the parts where there are non-local interactions. This explicitly shows
c,k nc,k k nc,k k p
the existence of four (for Hlk and Hcnl ) and sixteen (for Hnl and Hcnl ) diagonal-off entries,
i i i i
respectively, in agreement with cases A and B depicted by Figures 1 and 2 (if only single specific values
of i, k , k are considered instead of the whole sum), generating 2 × 4d−1 = 12 × 4d and 8 × 4d−2 = 12 × 4d
blocks, respectively. Then, the SU (2) decomposition could be achieved only by: (a) including any
c,k
desired non-local terms Hnl (to generate the diagonal elements); and (b) including only one type of
i
nc,k k c,k nc,k k p
interaction among Hlk , Hnl , Hcnl or Hcnl for concrete values for i, k , k , and p.
i i i i

159

www.pdfgrip.com
Entropy 2018, 20, 610

j ,jd+s
An important property used later for Fi,ks is that only one term in (25) remains with the election
js ,jd+s
of ik and k k . Because each c j,k is real or imaginary, and more concretely as a brief analysis shows,
if it is not zero, then it becomes imaginary only if js or jd+s is equal to 2, this property is transferred
j ,j
to Fi,ks d+s .

5. Specific Interactions Generating SU (2) Decomposition

In this section, we summarize and organize the global findings to reach the SU (2) block structure
on the GBS basis. Finally, we conclude that there are three great types of interactions that are able to
generate the block structure depicted in Section 3.

5.1. General Depiction of Interactions Having SU (2) Decomposition for the GBS Basis
Based on the previous discussion, there are three groups of interactions that are able to generate
the SU (2) decomposition on the GBS basis. The first one (Type I) involves all kinds of non-local
and non-crossed interactions between any two correspondent parts in any direction. These terms
generate the diagonal terms depicted previously in the Hamiltonian. Together, only two local
interactions in only one specific direction and on only one pair of correspondent parts, k l , should
be included to generate the diagonal-off entries. Thus, this group of interactions generates the
SU (2) blocks. Note that local interaction terms could be intended as external driven fields as
in [11,26]. The second interaction (Type II) is obtained by substituting the previous local interactions
with non-local interactions among only those non-correspondent elements included in two pairs
of correspondent parts. This means that if k, k , k + d, k + d with k < k ≤ d are these elements in
the two correspondent parts, then only the interactions between the following non-correspondent
elements are allowed: [k, k ], [k, k + d], [k , k + d], and [k + d, k + d]. This group of four interactions
generates the diagonal-off terms to conform the SU (2) blocks. Nevertheless, the Type II interaction
should normally be understood as a non-driven process of control. Note that Type II interaction
nc,k k
could be classified into two other subclasses: (a) Type IIa for non-crossed interactions Hnl ; and
i
nc,k k p
(b) Type IIb for crossed interactions Hcnl . Finally, the third interaction (Type III) involves both the
i
non-local and non-crossed interactions, with the inclusion of crossed interactions between one specific
correspondent pair.
In order to clarify the structure of those notable interaction architectures as special cases of
Hamiltonian (1), we make some remarks as follows. Figure 3 summarizes the three types of interactions
depicted above by listing the 2d qubits involved and then relating them with arrows in agreement
with their mutual interactions. Then:

A: Curved arrows point out those qubits related through entangling operations in any case.
B: All curved arrows in the bottom refer to Heisenberg–Ising-like (non-crossed) interactions
involving the three possible spatial directions together. Those interaction relations set the
correspondent pairs.
C: For the curved arrows in the top, two kinds of entangling operations can be considered according
to the text: Heisenberg–Ising-like (non-crossed) interactions or Dzyaloshinskii–Moriya-like
(crossed) interactions. Only one characteristic spatial direction is allowed.
D: Type II interactions can be split into Type IIa and Type IIb if interactions in the top are non-crossed
or crossed (between parts of two different correspondent pairs), respectively. Type IIb interactions
in the top admits only one possible parity from the two possible.
E: Type III interactions admit only crossed interactions in the top between parts of one specific
correspondent pair, but the two possible parities together are allowed.
F: For Figure 3a, the right arrows correspond to external local interactions such as those generated
by magnetic fields on spin-based qubits. Due to their locality, they are referred to as driven
interactions, although it actually depends on the available control of the interactions.

160

www.pdfgrip.com
Entropy 2018, 20, 610

Figure 4a shows a pictorial representation of each interaction, where the pairing is graphically
represented. Therein, yellow rays with blue contour are non-crossed interactions in the three
spatial directions [B]. Yellow rays with red contour represent one interaction from non-crossed
or crossed entangling interactions in only one spatial direction [D]. Blue rays with red contour
indicate non-crossed interaction in three spatial directions together with a crossed interaction in
only one direction [E]. Yellow triangles indicate local interactions on the respective qubits in only one
correspondent pair [F].

Figure 3. Three types of physical interactions able to generate the block decomposition. Non-local and
non-crossed interactions among any correspondent parts combined with: (a) local interactions on only
two correspondent parts (k l , k l + d); (b) any two non-correspondent parts in only two specific pairs of
correspondent parts of only one subtype, non-crossed or crossed; and (c) crossed interactions between
a specific pair of correspondent parts.

In particular, note that this description is in agreement with the results in Table 1 for d = 1,
although it is a special case because diagonal-off entries for Type I, II, and III coincide in the same
diagonal-off entries, so both interactions could be combined at the same time, preserving the SU (2)
decomposition. This case has a richer structure for control in terms of the number of free parameters
involved with respect to the number of parts to be controlled. Note that while Types I and III are
only able to modify the inner entanglement of the correspondent pairs, Type II interaction (Type IIa
and IIb) allows the modification of the global entanglement between different correspondent pairs,
thus letting it spread on the entire system by switching the pairs involving interactions generating
diagonal-off entries.

161

www.pdfgrip.com
Entropy 2018, 20, 610

Figure 4. Representation of qubit interactions able to generate SU (2) decomposition: (a) Type I, II, and
III interactions among 2d qubits (Type III assumes the inclusion of crossed interactions
% in the pair k );
and (b) Distributed evolution on 22d−1 Bloch spheres, each one for the states ψj .

5.2. General Structure of SU (2) Blocks

A complementary analysis of SU (2) blocks obtained for the last interactions is given in this
subsection. Their form is particularly useful as a connection with optimal control schemes, such as
those presented in [9]. In any case (Type I, II, or III), each block S H I ,I (with I , I the rows in which is
situated) has the form:
 
h11 h12
S H I ,I = ∗
h12 h22
(28)
h11 + h22 h11 − h22
= II ,I + Re(h12 )XI ,I − Im(h12 )YI ,I + ZI ,I ,
2 2

where {II ,I , XI ,I , YI ,I , ZI ,I } is the Pauli basis for the SU (2) block. If the Hamiltonian coefficients
involved in the block are time-independent, then the corresponding SU I ,I block in the evolution
matrix becomes:
h11 + h22 h11 + h22
SU I ,I =eiSH I ,I = ei 2h̄ t eiωn·sI ,I t = ei 2h̄ t (cos ωt + i sin ωtn · sI ,I )
t
h̄
 
−h22
h +h
i 112h̄ 22 t cos ωt + i h112h̄ω sin ωt i hh̄ω
12
sin ωt
=e ∗
h12 −h22 ,
i h̄ω sin ωt cos ωt − i h112h̄ω sin ωt
1 h − h22 (29)
with :n = (Re(h12 ), −Im(h12 ), 11 ),
h̄ω 2
sI ,I = (XI ,I , YI ,I , ZI ,I ),
1
h̄ω = |h12 |2 + |h11 − h22 |2 ,
4
j ,j
clearly belonging to U (1) × SU (2) (see Appendix A.2). As stated previously, F j,ks d+s is imaginary only
if js or jd+s is 2. Thus, only one component from n1 or n2 is different from zero because non-diagonal
entries of block in (24), (26), and (27) are always real or imaginary. This reduces the optimal control to

162

www.pdfgrip.com
Entropy 2018, 20, 610

the second case reported by [9]. An additional analysis shows that h11 ± h22 = 0 in general (without
imposing restrictions on the non-local strengths hJ ). This aspect will be relevant later.

5.3. Structure of SU (2) Blocks for Each Interaction

Several classical interactions fitting in the current procedure were analyzed. All them generate
blocks (not necessarily SU (2) blocks) when they are expressed in the GBS basis, denoting a kind
of universality for this basis due to its ability to gather similar interactions through simplified
representations. For the sake of the search for SU (2) decomposition, we discuss finally closed forms
for the specific Hamiltonians able to achieve the SU (2) decomposition. These formulas are quite useful
for computer simulation purposes.

5.3.1. Blocks in Type I Interaction

This interaction includes non-crossed spin interactions between correspondent particles in all
spatial directions and external local interactions on the pair [k , k + d] of correspondent particles in
direction j. From (24)–(26), it can be written as:

( j,k )
H I = HD + HND I ,
3 d 2d
0
with :HD ≡ ∑ ∑ h(i (4k−1 +4k+d−1 ))2d4 σ(i (4k−1 +4k+d−1 ))2d ,
4,s (30)
i =1 k =1 s =1
1 2d
0
( j,k )
HND I = ∑ h( j4k +dt −1 )2d4 σ( j4k +dt −1 )2d ,
4,s
t =0 s =1

generating SU (2) blocks with the diagonal terms from non-local interactions between correspondent
parts and the non-diagonal terms from local interactions. Departing from (24)–(26), we obtain for the
Hamiltonian components:

$ % 3 d  
δi ,2 +(1−δi )(1−δ0,i )
ΨI d | H I |ΨKd
4 4
= δIK ∑ ∑ (−1) ,i
k k h ( i (4k −1 +4k +d−1 ))2d
4
+
i =1 k =1
(31)
1
{k } jδ ,jδ1,t ( j,k )
∑ h( j4k +dt −1 )2d δIK F j,k0,t
4
≡ HD IK + HND I
IK
.
t =0

i ,i δ +(1−δ )(1−δ )
The last formula is obtained noting that ci,ik k = (−1) i,2 i,i
k
0,i
k . The first term of the
last expressions denotes the diagonal terms of interaction. This formula shows that the pair of entries
in the diagonal of each SU (2) block are generally different. Because the block is formed by switching
an index ik in the row labels (or two as in the following cases) in agreement with the association rules
0 ↔ j or i ↔ k (j is the direction associated to the interaction and i, j, k a permutation of 1, 2, 3), then for
i = j the terms in HD IK have a sign change. This implies that in general h11 = h22 in (28), generating
non-diagonal S H I ,I -blocks. The second term contains the four diagonal-off elements generating two
i ,ik
blocks with two terms each. Note that Hamiltonian terms (hI ) are real together with ci,ik , so diagonal
j,0 0,j
terms are real, as expected. Diagonal-off terms will be real or imaginary depending on F j,k , F j,k .
In any case, concretely, they are imaginary only if j = 2.
Note that this interaction (when it is applied to a combination of correspondent pairs with
bipartite entangled states) generates only non-local operations on each correspondent pair, such as
those presented in [11,13]. Still switching the direction j and the correspondent pair k on which the
local interaction is applied, this kind of Hamiltonian cannot generate extended entanglement between
correspondent pairs more than that included in the initial state. This means that if the initial state is
separable by correspondent pairs, it will remain separable at this level (but should be able to entangle

163

www.pdfgrip.com
Entropy 2018, 20, 610

or untangle the parts of each pair). Conversely, it cannot disentangle each correspondent pair from the
remaining state in more complex cases. We dedicate a later section to analyzing these topics.

5.3.2. Blocks in Type II Interaction

Type IIa: In this case, the interaction is completely non-local between correspondent pairs to generate
the diagonal entries, and in only one direction between non-correspondent parts in two correspondent
pairs to generate the diagonal-off entries. The Hamiltonian becomes:

( j,k k )
H I Ia = HD + HND I Ia ,

( j,k k )
1 2d
0 (32)
with :HND I Ia = ∑ h( j(4k +dt −1 +4k +dt −1 ))2d
4
σ( j(4k +dt −1 +4k +dt −1 ))2d ,
4,s
t ,t =0 s =1

with a non-local and non-crossed interaction in the direction j for the group of non-correspondent
terms defined by k < k ≤ d. The Hamiltonian entries are similar to those in (24)–(26), but with
the last restriction for the non-correspondent terms of interaction. Due to discussion in the previous
subsection, diagonal-off entries in the Hamiltonian are now always real. The components become:
$ %
( j,k k )
ΨI d | H I Ia |ΨKd = HD IK + HND I Ia ,
4 4 IK

( j,k k )
1 (33)
{k ,k } jδ ,jδ1,t jδ ,jδ1,t
HND I Ia
IK
≡ ∑ h( j(4k +dt −1 +4k +dt −1 ))2d δIK
4
F j,k0,t F j,k0,t .
t ,t =0

Type IIb: For this interaction, the non-diagonal part generated by the non-local interaction between
non-correspondent parts is supplied by a non-local and crossed interaction among non-correspondent
parts of two correspondent pairs:

(i,k k p)
H I Ib = HD + HND ,
I Ib

(i,k k p)
1 2d
0 (34)
with :HND
I Ib
≡ ∑ h( j
p4
k +dt −1 + k
p4
k +dt −1 )2d
4
σ( j
p4
k +dt −1 + k
p4
k +dt −1 )2d
4,s
.
t ,t =0 s =1

As before, i, j p , k p is a permutation of 1, 2, 3 with parity p = 0, 1 (even and odd, respectively).

Thus, the components become:
$ %
(i,k k p)
ΨI d | H I Ib |ΨKd = HD IK + HND ,
4 4 I Ib IK

(i,k k p)
1
jp δ ,j p δ1,t kpδ ,k p δ1,t (35)
{k ,k }
HND
I Ib IK
= ∑ h( j
p4
k +dt −1 + k
p4
k +dt −1 )2d
4
δIK F jp ,k0,t Fk p ,k0,t .
t ,t =0

The non-diagonal entries are now imaginary, except for i = 2.

5.3.3. Blocks in Type III Interaction

Finally, for Type III interaction, the non-diagonal part is generated by the non-local and crossed
interaction between a pair of correspondent parts k :

(i,k )
H I I I = HD + HND I I I ,

(i,k )
1 2d
0 (36)
with :HND I I I = ∑ h( jp 4k −1 +k p 4k +d−1 )2d4 σ( j
p4
k −1 + k
p4
k +d−1 )2d
4,s
,
p =0 s =1

164

www.pdfgrip.com
Entropy 2018, 20, 610

with the Hamiltonian components:

$ %
(i,k )
ΨI d | H I I I |ΨKd = HD IK + HND I I I ,
4 4 IK

(i,k )
1
{k } j p ,k p
(37)
HND I I I
IK
= ∑ h( jp 4k −1 +k p 4k +d−1 )2d4 δIK Fi,k ,
p =0

where non-diagonal entries are imaginary only if i = 2.

Figure 4b shows a distributed evolution on 22d−1 Bloch spheres for the states
1 1 1 2d−1 1
ψj = α2j−2 Ψ2j−2 + α2j−1 Ψ2j−1 , which are part of the global state |ψ = ∑2j=1 ψj ,
1
where each |Ψk  is an element of the GBS basis. Each state ψj evolves as a different curve on each
Bloch sphere depending on parameters hJ .
Finally, we should note that each of the previous interactions involves labels to be completely
( j,k ) ( j,k ,k ) ( j,k ,k ,p) ( j,k ,k )
identified, namely: H I , H I Ia , H I Ib , and H I I I . These labels will be omitted by simplicity
unless their specification becomes needed. In any case, closed expressions (31), (33), (35), and (37) are
computationally efficient to generate matrix representations of Hamiltonians H I , H I Ia,b , H I I I , and for
their respective U, inclusively in the time-dependent case, although a numerical approach to construct
could also be necessary.

5.4. Available Parameters and Structure of Entries

The number of free parameters (coefficients hI of Hamiltonian) and their availability are important
to set control procedures. In this section, we count the entries and terms for each Hamiltonian,
summarizing the previous findings. If D ≤ 3 is the number of spatial dimensions involved in each
interaction, then the accounting of free parameters generating the SU (2) decomposition, together with
the maximum number of entries by column able to generate it (breaking the SU (2) decomposition) is
reported in Table 2. Note that the number of entries by column for all Hamiltonians (labeled with i, in
some sense the direction of the interaction) can be increased by a factor D if all directions are considered
at time. In the table, each Hamiltonian analyzed is reported, arriving at the main Hamiltonians
H I , H I Ia,b , and H I I I . Accounting shows few free parameters at time (compared with the exponential
growth of the matrix with the system size d) to set a whole control (over all blocks) in one period
of constant driven parameters, suggesting the use of time-dependent or at least constant-piecewise
parameters to increase the control.

Table 2. Rows generated and free parameters in each interaction considered in the text.

Hamiltonian Entries Type Entries by Column/Row Parameters by Entry

H0 Diagonal 1 D≤3
Hli Non-diagonal d 2
c
Hnl Diagonal 1 d
i

2 d ( d − 1)
nc 1
Hnl i
Non-diagonal 4
c
Hcnli Non-diagonal d 2
nc
Hcnl i
Non-diagonal d ( d − 1) 4
HI 2 × 2 block 2 2 + Dd ≤ 2 + 3d
H I Ia,b 2 × 2 block 2 4 + Dd ≤ 4 + 3d
HI I I 2 × 2 block 2 2 + Dd ≤ 2 + 3d

5.4.1. Structure of Diagonal Entries Belonging to a Specific Block

Other aspects should be discussed. The first is related to terms in diagonal entries generated by
c,s
non-local interactions Hnl among correspondent parts. Note that blocks are generated by interactions
j
c,k c,k nc,k k nc,k k p
other than those, which are prescribed as a difference in one (Hnl or Hcnl ) or two (Hnl or Hcnl )
i i i i
terms in the scripts labels, in agreement with the rules depicted in Figures 1 and 2. This implies

165

www.pdfgrip.com
Entropy 2018, 20, 610

that there will be two or eight blocks, each one relating rows (and columns) differing in only one
c,s
or two terms of their scripts, respectively. Note the diagonal entries for Hnl in (18) for the GBS
j

defined as in [22]: Tr(σ̃i∗s σj σ̃iTs σjT ) = 2(−1)δj,2 +(1−δj,is )(1−δ0,is ) . Then, for each three strengths for a
fixed correspondent pair, there will be only four sign combinations (none is the negative of another)
depending on: (a) the direction of the interaction involved (on the correspondent pair s) is j = 2 or
j = 2; and (b) is for the sth script is in the set {0, j} or in {i, k } (with i, j, k a permutation of 1, 2, 3).
There, the factors corresponding to other correspondent pairs will be equal to one. Then, for the 3d
terms included in all diagonal entries there will be 4d combinations for the whole terms—precisely the
number of rows. This implies that all diagonal entries are different (but not independent because there
are only 3d parameters). For two rows differing in only one or two terms in their scripts, only the three
c,s
or six terms corresponding with the strengths of Hnl for such correspondent pairs (associated with
j
those terms in the scripts) will change their signs in the diagonal terms in their block. Consequently,
for such 4d−1 or 4d−2 groups of blocks having the same scripts exchanged and generated by the whole
combinations in the other d − 1 or d − 2 terms in their scripts, they will have the same h11 − h22
parameters, respectively. Thus, it will be only two or eight different h11 − h22 parameters for the entire
H. Meanwhile, h11 + h22 parameters could be different.

5.4.2. Structure of Diagonal-Off Entries Belonging to a Specific Block

The second aspect is related to the explicit calculation of cijss ,k s
,jd+s for the basic cases of interest in
the diagonal-off entries. (a) For H I and H I Ia,b : js = j, jd+s = 0, or js = 0, jd+s = j (j being the direction
label involved in the local and non-local interactions between non-correspondent parts); and (b) for
H I I I : js = j p , jd+s = k p . Table 3 explicitly shows these values. Note the parallelism between their two
halves (vertically and horizontally).

Table 3. Values of cijss ,k

,j
s
for all exchange scripts in H I , H I Ia,b , H I I I . i, j, k is an even permutation of 1, 2, 3.
d+s

( js , js+d ) (is , ks ) cijss ,k s

,jd+s (is , ks ) cijss ,k s
,jd+s (is , ks ) cijss ,k s
,jd+s (is , ks ) cijss ,k s
,jd+s

(0, 2) (0, 2) −i (2, 0) i (1, 3) i (3, 1) −i

(2, 0) (0, 2) i (2, 0) −i (1, 3) i (3, 1) −i
(0, j = 2) (0, j) 1 ( j, 0) 1 (i, k) −(−1)δ2k (k, i ) −(−1)δ2k
( j = 2, 0) (0, j) 1 ( j, 0) 1 (i, k) (−1)δ2k (k, i ) (−1)δ2k
2 ∈ ( j, k ) ( j, k) −i (k, j) i (0, i ) −i (−1)δ2k (i, 0) i (−1)δ2k
2 ∈ (k, j) ( j, k) i (k, j) −i (0, i ) −i (−1)δ2k (i, 0) i (−1)δ2k
(1, 3) (1, 3) 1 (3, 1) 1 (0, 2) −1 (2, 0) −1
(3, 1) (1, 3) 1 (3, 1) 1 (0, 2) 1 (2, 0) 1

These cases generate the diagonal-off entries in each block in agreement with the exchange rules
depicted previously for the sth scripts of such entries’ rows: (is , k s ) ∈ {(0, j), ( j, 0); (i, k ), (k, i )}, with
i, j, k a permutation from 1, 2, 3 and j the associated direction for the corresponding interaction being
used from H I and H I I a,b ; (is , k s ) ∈ {( j p , k p ), (k p , j p ); (0, i ), (i, 0)}, with i, j p , k p a permutation of parity p
from 1, 2, 3 and j p , k p are the associated directions for the interaction H I I I .
First, we should note that the signs for each term in the diagonal-off entries do not depend on
the entries’ scripts in positions other than the parts in which the interaction is being applied, k , k in
the expressions of the previous section (30), (32), (34), and (36). This is because Tr(σ̃i∗s σjd+s σ̃kTs σjTs ) =
Tr(σ̃i∗s σ0 σ̃kTs σ0T ) = 2 . Instead, signs only depend on the type of exchange indexes shown in Table 3.
It has already been noted that cijss ,k
,j
s
is imaginary only if js = 2 or jd+s = 2. This property is then
d+s
j ,j
transferred to the corresponding F j,ss d+s , and then transformed to h12 as a function of the number of
those factors in (31), (33), (35), and (37). Thus, by exchanging is , k s (block transposing), only the cases
with h12 ∈ I will change their sign.

166

www.pdfgrip.com
Entropy 2018, 20, 610

The final fact is related with the different signs appearing in the terms of diagonal-off entries.
This will be important to analyze the number of independent blocks in the entire evolution matrix. For
H I and H I I I , the two different terms are obtained by the exchange of js , jd+s . Thus, for H I , only the
cijss ,k
,jd+s with ( js , js+d ) = (0, 2), (2, 0) and (is , k s ) ∈ {(0, 2), (2, 0)} or ( js , js+d ) = (0, j  = 2), ( j  = 2, 0) and
s

(is , k s ) ∈ {(i, k), (k, i )} will change their sign (in the first four rows of Table 3). For H I I I , if ( js , js+d ) =
(1, 3), (3, 1) and (is , k s ) ∈ {(0, 2), (2, 0)} or ( js , js+d ) = (0, 2), (2, 0) and (is , k s ) ∈ {(i, k), (k, i )}, then
cijss ,k s
,jd+s will change their sign (in the last four rows of Table 3). For H I Ia,b , two terms in the scripts are
involved, so different aspects contribute: the location of interacting parts, the type of exchange, and
their order in the scripts.
Last properties exhibits the way in which each term in h12 will change its sign. The three aspects
mentioned in the previous paragraph allow us to understand the diagonal-off structure of H I , H I Ia,b ,
and H I I I (considering that their diagonal components follow the properties discussed above). In
the following subsections, we analyze this structure for each interaction, particularly discussing the
independence of blocks in terms of the free parameters, making a distinction between the effective
parameters (those appearing in the final expression of (28)) and the physical parameters (those
appearing as coefficients hI in the Hamiltonian). They are not the same because many physical
parameters appear clustered in the same way in (28), because the entries of SU depend only on the
parameters h11 ± h22 , h12 . As a result, by grouping finally in the U (1) × SU (2) blocks, there will be
only two or eight different blocks SU in U.

5.4.3. Block Entries of H I

The diagonal-off entries have exactly the two terms h( j4s+dt−1 )2d for t = 0, 1, and there are only
4
two combinations: adding or subtracting terms. As was stated previously, they are imaginary only if
local interactions are in the direction j = 2. In this case, we separate the factor ±i for j = 2 cases in
the diagonal-off entries, and the remaining coefficients in the opposite corners in each block are equal
as expected from (28). Then, there is generally one term with the same sign through all diagonal-off
entries (when k = 2, or otherwise when js = 2 in the first four rows in the Table 3), leaving only two
possibilities for the remaining term. Thus, in each H I matrix there are blocks with only two different
diagonal-off entries, depending only on the index exchange type in the local interaction position and
not on the remaining indexes. Thus, for a fixed set of indexes for the positions unrelated to the part
on which the interaction is applied, a pair of blocks exists, one each for the exchanges (0, j), ( j, 0) and
(i, k), (k, i ), with different relative signs in their diagonal-off terms. For the corresponding diagonal
entries, in (29), only the difference h11 − h22 is relevant. As previously stated by analyzing equation (31),
it is also possible realize that in each diagonal entry there are only two terms from the 3d terms changing
their sign with respect to other rows. Block scripts differ in only one index, those corresponding with
i = j (the local interaction direction) and k = k (the correspondent pair on which the local interaction
is being applied), leaving only two terms and two different combinations for h11 − h22 . This implies that
there are only two different blocks for (29) through all U, each one operating with different exchange
rules, (0, j), ( j, 0) or (i, k ), (k, i ). Each one is the same (until unitary factors, which can be different) for
all entries with different indexes in positions other than k . This fact can be attributed, depending on the
number of disposable parameters (five, including the time and excluding the parameters in the unitary
factor of each block), to the independence between the two types of blocks in the evolution matrix (29).

5.4.4. Block Entries of H I Ia

j ,j
For the non-diagonal entries, because the exchange factor F j,ss d+s appears two times for each j, all
of them are real, so the opposite corners of each block are always equal. Each entry has four terms
with alternating signs, in agreement with the Table 3, as a function of the rows’ subscripts. Signs
only depend on both indexes exchanged: either they are the same type (0, j), ( j, 0) or (i, k ), (k, i ), or
otherwise opposite with an exchange of each type. This will give only four sign combinations (a
calculation not developed explicitly here), except for j = 2, where the appearance of two factors i will

167

www.pdfgrip.com
Entropy 2018, 20, 610

change the overall factor, giving eight combinations, one half of them with opposite overall sign to the
remaining. For the diagonal entries, based on the ideas in the previous case for H I , there will be four
terms changing their relative signs with respect to other associated diagonal entries in the same block,
but now differing in two part indexes (due to the related non-local interaction). As before, one term
has a fixed sign, so there are only eight combinations for the three remaining terms from the 16
possible. This means eight different combinations for h11 − h22 in (29), due to the values k = k , k
for the non-correspondent parts with non-local interactions in HD for this case. Thus, similar to H I ,
in this case there will be eight different blocks in U for (29): one for each one of the eight different
combinations of the exchange rules. Each type applies in the same way for all entries with different
indexes in positions other than k , k . There are nine free parameters, including the time and excluding
the parameters in the unitary factor for the block, so independence among the eight types of blocks
can be more elusive. Despite all this, located operations not involving all GBS basis states appear
as achievable.

5.4.5. Block Entries of H I Ib

j ,j
Although the exchange factors F j,ss d+s are crossed and j takes two different values in the subscripts,
the discussion regards certain similitude to that for H I Ia . For the diagonal-off entries, in agreement
with Table 3, it implies that only if j = 2 is not included in the crossed interaction (i = 2 in (35))
will they become real. Each entry will have four terms with alternating signs, in agreement with the
outcomes of products of exchange factors in Table 3 as a function of the rows’ subscript involved. Here,
there will be eight combinations (four and four with opposite overall signs), except for j = 2 with only
four combinations. For the diagonal entries, h11 − h22 in (29), the situation is identical to H I Ia . Then,
there will be eight different block types in U for each combination of exchange rules on the indexes
k , k . Again, nine free parameters for the SU (2) blocks are available.

5.4.6. Block Entries of H I I I

This is a special case exception of the previous remark where js , jd+s is not of the forms 0, j or
j p ,k p j p ,k p k p ,j p
j, 0. Nevertheless, Fi,s becomes in the same way on of c0,i i,0
j p ,k p , c j p ,k p , c j p ,k p , or c j p ,k p . However, several
aspects are identical to the H I case. A brief analysis shows that entries become real only for i = 2
(see the last four rows of Table 3). Each diagonal-off entry has two terms with alternating signs as
functions of entry labels. For the diagonal entries, again only two types of terms change their sign
in HD from (31) for the rows forming the SU (2) blocks with the exchange rules. This gives only two
types of h11 − h22 in (29), again generating only two different blocks in the whole U—each one for a
kind of exchange rule involved here, containing five free parameters.
To resume the findings, Figure 5 shows the relations exhibited in the exchange indexes for each
interaction. This figure depicts each of the exchange index relations of GBS basis states under the
interaction. Thus, Figure 5a,d, depicts the two groups of exchange states for H I and H I I I generated
by the two different blocks through the whole SU (22d ) evolution matrix, both independent up to five
parameters and with h12 in (29). Figure 5b,c depict the double exchange indexes induced by the
eight blocks generated by H I Ia and H I Ib . These eight blocks are independent up to nine parameters.
All representations in Figure 5 are for a single GBS basis state, but clearly one specific block is operating
on any of them simultaneously. Note finally that for all cases there are a complementary number of
free physical parameters in h11 + h22 : 3d + 2 − 4 = 3d − 2 for H I and H I I I and 3d + 4 − 8 = 3d − 4
for H I Ia and H I Ib (time t is not accounted because it was considered in the SU (2) fitting). Then, there
is a linearly growing space to fit the blocks into a programmed operation in terms of the physical
parameters, although there is an exponential growth of those blocks.

168

www.pdfgrip.com
Entropy 2018, 20, 610

Figure 5. Exchange index relations involved for each interaction and highlighted properties for their
correspondent SU0 : (a) H I ; (b) H I Ia ; (c) H I Ib ; and (d) H I I I . Exchange relations in (b,d) are doubled by
I ,I
considering the vertical switching in one of the indexes for each pair shown.

6. Connectedness, Superposition, Entanglement and Separability

To understand how dynamics is addressed under the interactions H I , H I Ia,b , H I I I (used
independently or combined), some complementary analysis is convenient. In order to prepare the
reader, some illustrative examples are included in Appendix A.4 for d = 1 and d = 2, depicting some
notable properties of dynamics in such cases by including several kinds of entangling operations.

6.1. Exchange Connectedness under Interactions

Under the SU (2) decomposition, pairs of states in GBS basis become related, showing a probability
exchange between them. As it was seen, each one of the H I , H I Ia,b , H I I I interactions has rules for this
exchange. In any case, it should be clear this exchange is achievable between any pair by combining
all types of interactions obtained by switching the value of: (a) interaction direction and correspondent
pair j, k in (31) for H I ; (b) interaction direction and correspondent pairs j, k , k in (33) for H I Ia ;
(c) interaction directions, correspondent pairs, and parity i, k , k , p in (35) for H I Ib ; and (d) interaction
direction and correspondent pair i, k in (37) for H I I I . Several types of interactions can be combined in
a sequence. The combination of interactions is not precise for the basis element connectedness, but it is
necessary to increase the entanglement, and thus to connect two arbitrary quantum states. In those
terms, there are only two types of states: (1) those exchanging one script (H I and H I I I ), and (2) those
exchanging two scripts (H I Ia and H I Ib ) in the GBS basis elements under the rules depicted in Figure 5

169

www.pdfgrip.com
Entropy 2018, 20, 610

(although the rules and connections are different). All basis states become connected under one or
several interactions applied consecutively, depending on the number of necessary exchanges in their
scripts. Figure 6 shows a graph with these relations for the cases d = 1, 2, 3. Green edges indicate one
script exchange and red lines indicate two script exchanges. The connection can only be achieved with
a single interaction in the first two cases, due to the low entanglement level. Figure 6a corresponds to
the figure presented in [11] for Bell states in SU (4) systems.

Figure 6. Connectedness graphs between states under SU (2) decomposition for one (green) and two
(red) exchange scripts for all generalized Bell state (GBS) basis states: (a) d = 1; (b) d = 2; and (c) d = 3.

Connectedness in a finite number of steps by applying some or all cases in each type of interaction
(piecewise with constant parameters or with time-dependent parameters in each case) warrants the
full probability exchange between the occupancy level of each state in terms of the discussion included
in Section 3. Nevertheless, not all interactions are able to reach an arbitrary evolution. As is obvious,
H I and H I I I are not able to generate extended entanglement out of the correspondent pair on which
they operate (this assumes no rearrangements are made in the correspondent pairs and their elements).
We discuss this aspect in the next subsection.

170

www.pdfgrip.com
Entropy 2018, 20, 610

6.2. Notable Quantum Processing Operations Achievable under SU (2) Decomposition

−h22 +h22
Departing from SU , then by fixing ωt = 2n2+1 π, h112h̄ω = , hh̄ω
12
= ic δ, h112h̄ω = 2(m − 12 ); n, m ∈ Z,
c ∈ {0, 1}, δ ∈ R, where  + δ = 1 in (29). Note that the parameter c depends on each kind of interaction
2 2

in the terms discussed in the previous section. Then, we get the SU I ,I block [27]:
 
 ic δ
Hcm (δ, )I ,I ≡ (−1)m , (38)
(−i )c δ −

operating on the GBS basis. Note that this form cannot always be achieved independently in all blocks
in terms of the free parameters and the possible restriction h0,0,...,0 = 0 (here det(Hcm (δ, )I ,I ) = −1,
although it is not decisive in the following development). Nevertheless, we need only achieve it
in some blocks in the immediate discussion. We are using the time-independent case, but other
more practical cases with time-dependent Hamiltonian coefficients can be implemented. The last
form is highly versatile. If s || = δ = √1 (s = sign(), referred in the notation as −, +), we get a
2
m,c,sign()
Hadamard-like gate HI ,I ≡ SU I ,I (in particular, if c = 0, but this condition can be relaxed).
When δ = 1, we get an exchange-like gate [12,26] for the pair in the SU (2) block, EIm,c ,I
≡ SU I ,I . Note
that this case is a limit case for the time-independent case (29) when h12  h11 − h22 . Otherwise, it can
be achieved in two steps of time-independent piecewise Hamiltonians (as in [12]) or as a continuous
time-dependent Hamiltonian. These gates are:
   
(−1)m s ic 0 ic
HIm,c,s
,I

= √ , EIm,c
,I
= (−1)m . (39)
2 (−i )c −s (−i )c 0

+h22
Note additionally that when h112h̄ω = ( mα − 1)π, ωt = mπ; n, m ∈ Z, we get the quasi-identity
gate SU I ,I = eiαπ II ,I ≡ IIα ,I . The combination of these blocks (allowed because the block
independence previously discussed) allows important quantum processing operations to be set.

6.3. SU (2) Decomposition in the Context of n−Qubit Controlled Gates

Transformation between quantum states can generally be achieved by means of linear and
anti-linear operators. Anti-linear operators are particularly useful to depict time-reversal operations or
the action of some Einstein-Podolsky-Rosen channels. If these kinds of operations are being considered
in the processing, an extension of the Hamiltonian (1) should be considered by the inclusion of
anti-linear operations [28]. In this work, we have restricted our development to linear operators, as
was settled in Sections 2 and 3.
Below of such context, it should be advised that SU (2) decomposition is compatible with the
most quantum information developments in the literature. Nevertheless, many of those works
do not consider that such proposed processing forms are rarely compatible with the dynamics of
physical systems if the computational basis continues to be used (the natural basis based on physical
properties of local systems such as spin and polarization). The nature of entangling operations naturally
induces both superposition and entanglement, thus generating a complex dynamics evolution in
such basis compared with the structured gates proposed in the quantum information developments
(whose authors were clearly not always concerned with the underlying physics). SU (2) decomposition
(mainly the part developed in the Sections 2 and 3) naturally proposes a better basis to set the quantum
processing grammar for certain interaction architectures (e.g., those developed in Sections 4 and 5).
The induced 2 × 2 block structure allows such processing structures to be set more easily, mainly based
on binary processing.
In the context of quantum computation, the most common trend is the settlement of universal
gates in the sense of a quantum Turing machine. A set of universal quantum gates for two-qubit
processing was established by [29] as a set of local gates together with the CNot gate. Despite

171

www.pdfgrip.com
Entropy 2018, 20, 610

universality, this trend is not optimal because for a given processing, it is not clear how to express it in
terms of those elements in the universal set. In an alternative trend, [30] has settled a more optimal
gate decomposition by factorization in terms of P−unitary matrices. In the last two trends, SU (2)
decomposition for SU (4) (d = 1, meaning two-qubit processing) has shown how to adapt those results
for the physics of Heisenberg–Ising interactions including driven magnetic fields: (a) in [31], a set
of alternative universal gates has been proposed on the grammar of Bell states; and (b) in [26], an
optimal set of six gates (P−unitary matrices) is proposed using the forms of SU (2) decomposition
on a Bell states basis to reproduce any other gate for two-quibit processing. In the current context,
those outcomes are automatically applicable to Type I and III interactions. Type II interactions are
excluded because they require at least d = 2. In any case, the contribution of the SU (2) reduction is in
the proposal of Bell basis as a grammar instead of the computational one so that the physical evolution
fulfills the forms required by the processing gates.
Although two-qubit processing is still universal, more powerful processing is possible by attaining
more than two qubits at a time. In this approach, [32,33] have stated universal processing gates in
terms of local rotations and n−qubits controlled gates. In the computational basis, rotations are
obtained by local interactions by turning off the entangling operations, but controlled gates can be
physically difficult to reproduce. In the SU (2) reduction scheme, the form of rotations in those works
(Ry (α) and Rz (α)) are achieved by the forms (29) as follows. First, Ry (α) is mainly achieved by settling
h11 = h22 and h12 ∈ I. Rz (α) is obtained by fixing h12 = 0. Notably, those rotations are not necessarily
physical neither local, they could operate among entangled states. Instead, they can be determined
as rotations on the informational states being used (elements of GBS basis). Other basic forms are
also easily obtained, for example, Ph(δ) is obtained by settling cos ωt = ±1. For the controlled gates
Λn (U ) proposed in [32], authors in [33] turn to a long factorization in terms of rotations and controlled
gates Λ1 (U ) (which can also be obtained departing from the CNot gate and rotations). In any case, if a
computational basis is used, the reproduction of the CNot gate can still bring certain difficulties in
many quantum systems [34]. In the context of SU (2) reduction, CNot gate and inclusively Λ1 (U ) are
directly obtained if the Bell basis is used as grammar:
 
SU 1 → I 0
Λ 1 (U ) = , (40)
0 SU 2 → U

where U is a general matrix in SU (2) as in (29). Because of the independence of blocks stated
in Section 5, the achievement of Λ1 (U ) is warranted. Then, the construction of Λn (U ) follows
immediately as proposed in [32,33], but considering those forms working on the grammar basis
of the Bell states or on the GBS basis in general. Clearly, in the SU (2) decomposition scheme, other
controlled gates are achievable by the alternative selection of the elements on which interaction is
being applied. If more optimal factorization methods are possible for d > 1 (where blocks are repeated
by groups), based on the set of matrices U as in (14) by including all the possible forms generated by
Type I, IIa, IIb, and III interactions, it is still an open question.

6.4. Generating Superposition and Entanglement

In the following, we will use an arrow to depict a certain group of quantum operations. On the top
of the arrow, we set the type of interaction being used. On the bottom, we set the subspace on which
they apply or the generic form of each operation, together with their prescriptions. For instance, if an
operation for d = 4 (8 qubits and 256 elements in the GBS basis) generated by the Type IIa interaction
is applied in the associated direction y and on the pairs 1 and 4 (j = 2, k = 2, k = 4 in (32)) with
prescriptions for a Hadamard gate mixing the basis states |Ψ0  = |Ψ0,0,0,0  and |Ψ130  = |Ψ2,0,0,2  (i.e.,

172

www.pdfgrip.com
Entropy 2018, 20, 610

0,0,+
H0,130 ) and an exchange gate between the basis states |Ψ1  = |Ψ1,0,0,0  and |Ψ131  = |Ψ3,0,0,2  (i.e.,
0,0
E0,131 ), we will write:

(2,2,5)
H I Ia
−−−−−−−−−
0,0,+ 0,0
→. (41)
H0,130 ⊕ E0,131

Although other operations can be defined between the remaining basis states, if they are not
specified, it is because some operations are repeated for other certain groups of scripts (e.g., for
|Ψ20  = |Ψ0,1,1,0  and |Ψ150  = |Ψ2,1,1,2 , H 0,0,+ is also being applied) or because the concrete operation
being developed does not require such specification (e.g., there is no specification for the operation
between |Ψ67  = |Ψ3,0,0,1  and |Ψ193  = |Ψ1,0,0,3 ). In some cases, complex families of subsequent
operations are required, and then one family is specified by a group of indexes defining it.

6.4.1. Generating 2−Separable Superposition

By using the general block operations SU I ,I ∈ U (1) × SU (2) (29), it is possible to arrive at a state
exhibiting complete superposition through all of the basis elements. Thus, for example, departing from
the simple state |Ψ0 2d = |Ψ0 1 |Ψ0 2 ...|Ψ0 d (easily obtained from |00...0), a couple of local operations
(i,k)
H I on each correspondent pair k are sufficient to generate a state containing representatives from
each basis element:
(1,k)
HI
k =1,2,...,d
0
d 1
|Ψ0 2d −−−−−−−−−
4
s,s SU s,s
→
∑ αi,0
k
| Ψi k ,
s − s = 4k −1 , k =1 i =0
d =0
s4,k
(3,k)
HI
k =1,2,...,d % 4d − 1 (42)
−−−−−−−−−→ 0 d 1 1
4 4
s,s SU s,s ⊕ s ,s SU s ,s ∑ ∑ k k
αi,0 β j(i),i Ψ j (i )
k
≡ ∑ γI | Ψ I  ,
−
s −s=3·4 ,s −s =4
k 1 k − 1 k =1 i = 0 j(i )∈{i,3−i } I=0
d =0,s d =1
s4,k 4,k

d
with :γI = ∏ αkj−1 (I4,kd ),0 βkI4,kd ,j−1 (I4,kd ) ,
k =1

where j−1 (i ) is the inverse of j(i ) and directions i = 1, 3 were used as examples. In addition, αi,j
k are

the components of SU s,s in the first operations, and βkj(i),i are the components of SU s,s , SU s ,s for the
second operations with i = 0, 1, respectively. Figure 7 depicts each step of the process, using the local
operations (alternatively, crossed interactions in H I I I could be considered).

Figure 7. Processes to build 2−separable states with complete superposition.

(i,{s})
The last process is a particular case of more general operations by considering O J = SU I ,I
to mix the states through the momentary associated blocks changing the indexes {s} with some
interaction H J , J ∈ { I, I Ia, I Ib, I I I } in the associated direction i. We coin the term k −local operation
when SU I ,I generates entanglement at the most in k parts. In our basic interactions scheme,

173

www.pdfgrip.com
Entropy 2018, 20, 610

there are only 2−local and 4−local operations, as was discussed previously. Thus, following the
previously-introduced notation, we set a family of procedures to develop superposition including
the previous procedure. Departing from the |Ψ0 2d , it is possible to apply several alternate 2−local
operations to generate superposition involving all GBS basis states. By defining a sequence of paired
directions for the H I evolution involving all pairs s = 1, 2, ..., d (this process can alternatively be
achieved by H I I I ): {{is , k s (is )}|{1, 2, 3} is = k s (is ) ∈ {1, 2, 3} \ {is }; s = 1, 2, ..., d}. Additionally,
js (is ) ∈ {1, 2, 3}, is = js (is ) = k s (is ). Then, following the evolution process:
(i1 ,1) (k1 (i1 ),1)
HI HI 3 1
|Ψ0 2d −−−−→
(i ,{1})
∑ α1t,0 |Ψt,0,...,0  −−−−→
(k (i ),{1})
∑ α1p
1 ( 1 ),0
β1
1 ,p1 ( 1 )
Ψ1 ,0,...,0 ,
OI 1 t∈{0,i1 } OI 1 1 1 = 0
(43)
(i ,2)
HI 2 HI
(k d (id ),d)
3 , d - 1 %
−−−−→ . . . −−−−→
(i ,{2}) (k d (id ),{d})
∑ ∏ αsps (s ),0 βss ,ps (s ) Ψ1 ,2 ,...,d ≡ Ψ fsep ,
OI 2 OI 1 ,...,d =0 s =1

where ps (s ) are the inverses of the association rules for the one index exchanges depicted in Figure 5a
(or 5d for H I I I ): ps (is ) = is = ps ( js (is )), ps (0) = 0 = ps (k s (is )). Additionally, |α0,0
s |2 + | αs |2 = 1,
is ,0
| βs0,0 |2 + | βsks (is ),0 |2 = 1, | βsis ,is |2 + | βsjs (is ),is |2 = 1. TrS (ρIJ ) represents the partial trace with respect to
%
the entire system except the s ∈ S parts. As expected, Ψ fsep is 2−separable:

%$ , 3 1 -, 3 1 -†
Tr{k } ( Ψ fsep Ψ fsep ) = ∑ αkp k
(k ),0 β k ,pk (k ) Ψ k ∑ αkp k
(k ),0 β k ,pk (k ) Ψ k (44)
k k
k =0 k =0

due to the limited nature of operations involved, which cannot be able to generate more extended
entanglement. In addition, superposition can be limited to the SU (2) blocks coverage through the
number of parameters introduced, αsp ( ),0 , βs ,p ( ) , and their physical scope. As shown in [11], a
s s s s s
richer superposition coverage on SU (22d ) can be achieved with additional 2−local operations on each
part, introducing extra parameters and probability mixing. As in [11], n in (29) is limited to take the
two forms (n x , 0, nz ) or (0, ny , nz ) (for the time-independent case), but by combining both forms we
arrive at two general forms with arbitrary n = (n x , ny , nz ) (this also fulfills the time-dependent case
with adequate hij (t)).
Although this procedure can include a general full 2−separable state together with entangled
segments between correspondent pairs, it cannot exhibit states with more extended entanglement,
requiring more extended entangling operations such as HI Ia and HI Ib . The quest is to obtain general
states departing from a simple resource, which is still an open challenge—particularly for the possible
entanglement degree there (a more ambitious challenge is the transformation between two general
states [35], but it can always be reduced in two steps of this kind). We discuss this issue in the remaining
subsection, and we develop some procedures to generate some maximal entangled states of arbitrary size.

6.4.2. Entanglement Dynamics under Interactions

Now, we analyze the entanglement generation under the interactions being considered.
We employ the partial trace criterion [19] for pure states by considering a single SU (2) combination
of two GBS basis states |φIJ  = αI |ΨI  + αJ |ΨJ . In addition, the explicit form for coefficients
will be written as αI = cos θ/2, αJ = eiφ sin θ/2. Then, we construct their associated density matrix
ρIJ = |φIJ  φIJ | to conveniently take partial traces in order to analyze the entanglement of specific
subsystems in this quantum state under concrete interactions. Because the rules in the exchange scripts
(in the GBS basis states to form the SU (2) blocks) are basically the same for the three interactions
H I , H I Ia,b , H I I I , the analysis is reduced to only two cases. The first is for a pair of GBS basis elements

174

www.pdfgrip.com
Entropy 2018, 20, 610

|ΨI  , |ΨJ  differing in only one subscript between I and J : is = js ∀s ∈ {1, . . . , d}, s = k (in H I , H I I I
interactions). Thus, in this case (omitting the base b and the size d for simplicity in the scripts):

% 1 2d − 1 , 0
d -
1
φIJ = √ ∑ σ̃is ⊗ (αI σ̃ik + αJ σ̃jk ) E ,D
|E  ⊗ |D . (45)
2d E ,D=0 k  = s =1

The second case is for a pair of elements in the GBS basis differing in two subscripts of I
and J : is = js ∀s = 1, . . . , d, s = k , k (in H I Ia,b interactions):

% 1 2d − 1 , 0
d -
2
φIJ = √ ∑ σ̃is ⊗ (αI σ̃ik ⊗ σ̃ik + αJ σ̃jk ⊗ σ̃jk ) E ,D
|E  ⊗ |D . (46)
2d E ,D=0 k ,k =s=1

Then, we analyze the entanglement of several subsystems in each case by taking the partial trace
with respect to its complement. Calculations are direct. At the end, the association rules 0 ↔ i and
j ↔ k should be applied to explicitly denote the viable relations between I and J , and to reduce some
traces on parts k , k in (45) and (46). In Table 4 we report the generalized bipartite concurrence for
pure states [36]:

C 2 (TrS (ρIJ )) = 2(1 − TrS (ρ2IJ )), (47)

where, j is assumed as the direction label of the interaction involved. If m = min(m1 , m2 ), where
m1 , m2 are the Hilbert space dimensions of each subsystem, then this measure changes smoothly from
0 for separable states to 2(m − 1)/m for maximally entangled states. Note that we take σ̃i ≡ eiφi σi ,
although it is only relevant for σ2 . With this distinction, we introduce φ = φ + φi − φj .
k k

Table 4. Bipartite concurrence C 2 (TrS (ρIJ )) for several subsystems in the SU (2) mixing of some pairs
of GBS basis states.

Case S C 2 (TrS (ρI J ))

%
(a) φIJ
1 [s ∈
/ {k , k + d}] 1
%
i
(b) φIJ
1 [s ∈ {k , k + d}] 1 − sin2 θ (cos φ δ0,ik · jk + (−1) j j
k k (1 − δ0,ik · jk ) sin φ )2
%
(c) φIJ
1 [k , k + d] 0
%
(d) φIJ
2 [k , k + d] sin2 θ
%
(e) φIJ
2 [k , k ] 3
2 − 12 sin2 θ (cos2 φ δik jk δik jk + sin2 φ (1 − δik jk δik jk ))

Table 4 includes some obvious results for “local” interactions on single parts (H I ) or on
correspondent pairs (H I I I ): (a) any part is maximally entangled with respect to the remaining system
(through its correspondent pair) if there are currently no active local or non-local crossed interactions
in H I and H I I I , respectively, so C 2 (TrS (ρIJ )) = 1; (b) nevertheless, if these local or non-local crossed
interactions act on the correspondent pair, each part of it can become separable or partially entangled
to the remaining system; and (c) any correspondent pair (as a subsystem) is separable from the
remaining system in any GBS basis state, so C 2 (TrS (ρIJ )) = 0. Note that in the cases (b) and (c) that
the subsystem comprises two parts [k , k + d] being compared with the remaining system, so the
Hilbert space dimension is four (m = 4). Similarly, the most important results here: (d) shows how
interactions between non-correspondent parts (crossed or non-crossed) affect the original separability
of each correspondent pair with respect to the remaining system, letting it become entangled with the
remaining system. Finally, (e) exhibits the change of entanglement between non-correspondent parts.
They are clearly originally entangled with their respective pair outside of the subsystem, but that
entanglement becomes reduced (C 2 (TrS (ρIJ )) ≤ 3/2) due to the non-local interactions.

175

www.pdfgrip.com
Entropy 2018, 20, 610

6.4.3. Generating Larger Maximal Entangled Systems

The generation of extended entanglement can be shown with a couple of introductory
1 1
examples [27]. If β ij = Ψ2i+i⊕ j are the GBS basis elements for d = 1 corresponding to the
Bell states [22], then considering the | GHZ  and |W  states of size 2d expressed in the GBS basis:

d , -
1 1 0 d
1 1 0
| GHZ 2d = √ ∑ |i, i  j = d+1 ∑ |Ψ0  j + (−1)i |Ψ3  j , (48)
2 i =0 j =1 2 2 i =0 j =1

d− 2
1
1 2d 0 d 1 d 0
d
|W 2d = √ ∑ δi,2j−1 , δi,2j j = d−1 ∑ (|Ψ0  j + |Ψ3  j ) ⊗ |Ψ1 i , (49)
2d i=1 j=1 2 2 i =1 j =1
j  =i

where j sums over correspondent pairs. Note that we are alternating the notation in the kets by
1
convenience: |Ψk  j is the Bell state |Ψk  on the jth correspondent pair, while Ψi1 ,i2 ,...,id = |ΨI  is the
I = 4d−1 id + . . . + 4i2 + i1 element in the GBS basis. For d = 2, they are simply:

1 1
| GHZ 4 = √ (|Ψ0,0  + |Ψ3,3 ) = √
2
∑ |ΨI  ,
2 I∈{0,15}
(50)

1, 1
|W  4 =
2
|Ψ1,0  + |Ψ0,1  + |Ψ3,1  + |Ψ1,3 ) = ∑ |ΨI  .
2 I∈{1,4,7,13
(51)
}

Then, we can depart from the basic state |0000 = 12 (|Ψ0 1 + |Ψ3 1 ) ⊗ (|Ψ0 2 + |Ψ3 2 ) for d = 2.
We arrive at the | GHZ  by applying the following operations (as before, the interaction Hamiltonian is
indicated in the upper position, while the operation is written below):
(3,1)
HI 1
|0000−−−−−−−→ √ |Ψ0 1 ⊗ (|Ψ0 2 + |Ψ3 2 ),
0,0,+ 0,0,+ 2
H0,3 ⊕ H12,15
(3,2)
HI
−−−−−−−
0,0,+
→ |Ψ0 1 ⊗ |Ψ0 2 = |Ψ0,0  , (52)
H0,12
(3,1,2)
H I Ia 1 1 4
−−−−−−−
0,0,+
→ √ (|Ψ0,0  + |Ψ3,3 ) = √ (|Ψ0  + |Ψ15 ) = | GHZ  .
H0,15 2 2

The first operation requires action on two sets of GBS basis states. They are of the same form,
so they are easily achieved in terms of prescriptions for HIm,c,s
,I

. Note that no more specifications are
needed in complementary blocks. They are free because their effect will work on states that are not
included. Similarly, for example:
(2,1,2)
H I Ia 1
| GHZ 4 −−−−−−−→ √ (|Ψ0,0  + |Ψ1,1 ),
0,0
I0,10 ⊕ E5,15
0 2
(1,2)
HI 1
−−−−−−− → √ (|Ψ0 1 ⊗ |Ψ1 2 + |Ψ1 1 ⊗ |Ψ0 2 ),
0,0 0,0
E0,4 ⊕ E1,5 2
(3,1)
HI 1 1 (53)
−−−−−−− → √ ( √ (|Ψ0 1 + |Ψ3 1 ) ⊗ |Ψ1 2 + |Ψ1 1 ⊗ |Ψ0 2 ),
0,0,+ 2p
2 2
H4,7 ⊕ I1,2
(3,2)
HI 1
−−−−−−−→ ((|Ψ0 1 + |Ψ3 1 ) ⊗ |Ψ1 2 + |Ψ1 1 ⊗ (|Ψ0 2 + |Ψ3 2 ))
2q 2r ⊕ H 0,0,+ 2
I4,8 ⊕ I7,11 1,13
1
= (|Ψ4  + |Ψ7  + |Ψ1  + |Ψ13 ) = |W 4 ,
2

176

www.pdfgrip.com
Entropy 2018, 20, 610

where p, q, r ∈ Z. In the last operations, the block independence discussed in the previous section was
applied to justify the construction of some simultaneous operations.

6.4.4. Recursive Generation of Larger Maximal Entangled Systems

In the previous subsection, we obtained the larger maximal entangled states | GHZ 4 and |W 4
departing from the more basic states such as |0000. The enlargement of entangled states can be stated
in a more impressive way as recursive processes. In each case, these processes are based on the control
of the parameters involved and the independence among block types generated in each interaction.
Thus, the process shown in Figure 8 combines some of the operations depicted previously to
develop | GHZ 2(d+1) departing from | GHZ 2d , stating a procedure to get larger versions of these
maximal entangled states. The first step begins by using the state |Ψ0 d+1 ⊗ | GHZ 2d . Then, a local
operation is applied on each pair in the original state k = 1, 2, ..., d to reduce the factors (|Ψ0 k + |Ψ3 k )
and (|Ψ0  j − |Ψ3  j ) in (48) into |Ψ0  j and |Ψ3  j , respectively. Then, we exchange the indexes 30 ↔ 21
for the non-correspondent pairs k and d + 1 with a non-local operation. This transformation is
followed by a couple of local operations changing the indexes 2 ↔ 3 for the pair k and 1 ↔ 3 for
the pair d + 1 (which adds a factor i). In this last case, we transform the index 0 by itself, but adding
the factor i . Finally, we revert for k = 1, 2, ..., d + 1 the initial transformation between |Ψ0 k , |Ψ3 k
and (|Ψ0 k ± |Ψ3 k ), respectively. All additional index transformations are settled as the identity. The
state obtained will be i | GHZ 2(d+1) . It is notable that only one 4-entangling operation between the
added pair with another arbitrary pair from the original 2d-partite system has become necessary in
this case. This reflects the low robustness of the genuine entanglement for these states. Considering
the expression for | GHZ 2d in (48), the precise prescriptions are:
(3,k)
HI
k =1,2,...,d
−−−−−−−−− → √1
|Ψ0 d+1 ⊗ | GHZ 2d 4 0,0,+ | Ψ0  d +1 + | Ψ0  d +1 ⊗ | Ψ N  d ,
s,s Hs,s 2
k −1
s − s =3·4 ,
s,s ∈{3p≤ N | p∈N}
(1,k ,d+1)
H I Ia
−−−−−−−−− → 1
0 ⊕ E0,0
I0,u
N,u
√ | Ψ0  d +1 + | Ψ u  d +1 ,
u =4 k − 1 +4 d 2
u = N − 4k −1 + 4d
(1,k )
HI
−−−−−−−−− → √1 (54)
I 0 k −1 ⊕ E0,0 | Ψ0  d +1 + | Ψ u  d +1 ,
0,4 u ,u 2
u = u + 4k −1
(2,d+1)
H
→ √i
I
−−−−−−−−−
1 | Ψ0  d +1 + | Ψ N  d +1 ,
0,1 2
I2
0,2·4 d ⊕ Eu ,N
(3,k)
HI
k =1,2,...,d+1
−−−−−−−−− →
4 0,0,+
H i | GHZ 2(d+1) ,
s,s s,s
− s = 3· 4k −1 ,
s
s,s ∈{3p≤ N | p∈N}

1 1
where |ΨI n = Ψi1 1 ⊗ Ψi2 2 ⊗ ... ⊗ |Ψin n . In addition, N = 4d − 1 and N = 4d+1 − 1. Note that
the first and last operations are actually a set of operations for k = 1, 2, ..., d and k = 1, 2, ..., d + 1
(3,k)
through several correspondent pairs. They exploit the Hadamard-like block operations for H I to
switch first the | GHZ 2d into versions where only the states |Ψ0  and |Ψ3  appear. Thus, operations
(1,k ,d+1)
generated with H I Ia between two different correspondent pairs are used as exchange operations
(1,k ) (2,d+1)
entangling the added state |Ψ0 d+1 . Then, the additional operations H I and H I generate a

177

www.pdfgrip.com
Entropy 2018, 20, 610

state expressed only in terms of |Ψ0  and |Ψ3 , to finally be transformed into | GHZ 2(d+1) with the
same kind of initial operations.

Figure 8. Processes to build recursive enlargement of | GHZ  entangled states.

To obtain the |W 2(d+1) state, we begin with |Ψ0 d+1 ⊗ |W 2d , then we use the same local
transformation to reduce the factors (|Ψ0 k + |Ψ3 k ) in (48) into |Ψ0 k for each k = 1, 2, ..., d.
Then, we apply a sequence of non-local transformations between the pairs k, d + 1 for k = 1, 2, ..., d to
transfer probability between states with indexes 01 ↔ 10 there, in such a way as to reach the coefficient
√1 in each term. Finally, we revert the initial transformation for k = 1, 2, ..., d + 1, changing |Ψ0 k
d +1
into (|Ψ0 k + |Ψ3 k ). The final result is |W 2(d+1) , as is shown in Figure 9. Note how the entangling
operations need to go through the overall pairs. It reflects the robustness of genuine entanglement
in these states. By considering the expression for |W 2d in (48), the following process gives the
prescriptions to reach |W 2(d+1) from |W 2d :
(3,k)
HI
k =1,2,...,d
4 −−−−−−−−−
4 →0,0,+
0 ⊕
s,s Is,s s ,s H
s ,s | Ψ0  d +1 d
| Ψ 0  d + 1 ⊗ |W  2d
s − s =4k −1 , √ ⊗ ∑ | Ψ 4i − 1  d ,
s−4k−1 ,s −2·4k−1 ∈{3p≤ N | p∈N} d i =1
s − s =3·4k −1 ,
s −4i−1 ,s −4i−1 ∈{3p≤ N | p∈N},
k=i ∈{1,2,...,d}
(1,k,d+1)
H I Ia
k =1,2,...,d d+1 d0 +1
−−−−−−−−−→ 1
4
u,u Iu,u ⊕H0 (δk ,k )4k−1 ,4d
0 0 √ ∑
d + 1 i =1 j =1
| Ψ0  j ⊗ | Ψ1  i , (55)
u=4i−1 ,u =u+4k−1 +4d
k =i ∈{1,2,...,d} j  =i
(3,k)
HI
k =1,2,...,d,d+1
4 −−−−−−−−−
4 →0,0,+
0 ⊕
Is,s H
s,s s ,s s ,s 2( d +1)
s − s =4k −1 , |W  .
s−4k−1 ,s −2·4k−1 ∈{3p≤ N | p∈N}
s − s = 3· 4k −1 ,
s −4i−1 ,s −4i−1 ∈{3p≤ N | p∈N},
k =i ∈{1,2,...,d+1}

(3,k)
As before, Hadamard-like block operations for H I allow the states |W 2d and |W 2(d+1) to be
switched, at the beginning and at the end, in terms of |Ψ0  and |Ψ1 . The remarkable set of operations

178

www.pdfgrip.com
Entropy 2018, 20, 610

(1,k,d+1)
are obtained with H I Ia to entangle the added state |Ψ0 d+1 through the operations (38), which
progressively transfer the probability to the state |Ψ4d , completing a state easily transformed into
|W 2(d+1) with the final set of operations. Additional exchange in the indexes is settled in the identity.
The adequate set of k values for each step of operations should fulfill the d + 1 equations:

g0 ≡ 0,
√ √
d = d + 1(  j + δ j g j −1 ), j = 1, 2, ..., d,
(56)
g j = δj −  j g j−1 , j = 1, 2, ..., d − 1,
√ √
d = d + 1(δd − d gd−1 ).
These equations can be solved numerically for any d. Figure 10 shows the − log10 i solutions for
d = 1, 2, ..., 60 by taking i , δi > 0. Note that i drops rapidly to zero when d and i grow.

Figure 9. Processes to build recursive enlargement of |W  entangled states.

Figure 10. Solutions for i in H00 (δi , i )I ,I involved in the enlargement of |W d into |W d+1 for values
of d ∈ {2, ..., 60}.

6.5. Multipartite Entanglement and General States

In a previous subsection we described how to generate extended superposition using type I
interactions. However, that process does not reach genuine entangled states. The use of type I Ia, I Ib
interactions is mandatory to extend the entanglement as a set of operations involving elements of two
pairs. Nevertheless, it is clear that many operations and combinations are necessary and possible.
For instance, by considering the permutation i, j, k from 1, 2, 3 and departing from the state |Ψ0 2d ,
the process to reach an entangled state based on a complete combination from the basis elements for
two correspondent pairs is as follows (note that the process is not unique). First, we apply a 2−local
operation on the pair s and direction i followed by another on the pair s in the direction j. A linear
combination from four basis elements is obtained. Then, we apply a 4−local operation in the direction

179

www.pdfgrip.com
Entropy 2018, 20, 610

k and for pairs s, s , obtaining a state of eight terms. Finally, we again apply a 2−local operation on
pair s in the direction j. At the end, we obtain the desired state of sixteen terms with the pairs s, s ,
genuinely entangled as it was seen in Table 4.

(i,s) ( j,s )
HI HI 1
|Ψ0 2d −−−−→
(i,{s})
∑ αst,0 |Ψ0,...,t,...,0,...,0  −−−−→
( j,{s })
∑ αst,0 αst ,0 Ψ0,...,t,...,t ,...,0 ,
OI t∈{0,i } OI t∈{0,i }
t ∈{0,j}
(k,s,s )
H I Ia 1
−−−−→
(k,{s,s })
∑ αst,0 αst ,0 βs,s δ
, ;t,t t,pt,
s,k
δ
t ,pt , Ψ0,...,,..., ,...,0 ,
O I Ia , ∈C4 s ,k
(57)
t∈{0,i }
t ∈{0,j}
( j,s)
HI 1
−−−−→
( j,{s})
∑ αst,0 αst ,0 βs,s αs δ
, ;t,t χ, t,pt,
δ
s,k t ,pt , δ,pt,χ Ψ0,...,χ,..., ,...,0 ,
s,j
OI χ,, ∈C4 s ,k
t∈{0,i }
t ∈{0,j}

where C4 = {0, 1, 2, 3} and pt,

s,j is the extension of the inverse exchange rule presented before ps (  ),
but specifying the rule j as a function of the direction of the interaction involved. The script t ∈ {0, i }
is a label specifying each possible inverse. This means that if j is the characteristic direction of the
0,j i,j
interaction, then: p0,i 0,k 0,0 i,i i,k i,0
s,j = k = ps,j , ps,j = 0 = ps,j and ps,j = i = ps,j , ps,j = j = ps,j . This single process
could be improved using additional interactions to grow the spectrum of coefficients αsβ,α , βs,s β,β ;α,α
in
order to have a wider coverage of SU (4). In addition, it is clear the last process (or another alternative)
should be repeated, varying one or two pairs in order to generate more complex entanglement.
The question about how to generate a specific state under this procedure or to generate certain kind or
level of entanglement is clearly open, mainly due to the poorly understood complexity to measure this
property for large states in general.

7. Conclusions
Quantum gate array computation is based on the transformation of quantum states under certain
universal operations. These operations are used to manipulate the information settled on quantum
systems to simulate or reproduce computer processing, and normally use separable states as primary
resources. Quantum systems involved—light or matter—are manipulated around entanglement
generation in this kind of processing. Then, commonly involved interactions are non-local, implying
that their parts become entangled when they are being manipulated. In the process, several slightly
differentiated interactions are applied, each one with a different set of eigenvalues. This does not allow
a common grammar to be set through the entire quantum information processing problem.
SU (2) decomposition provides a procedure not only to reduce control in the quantum
manipulation states. Together, it provides a common language to address the evolution through
several kinds of similar interactions in order to manage a wider processing. Upon the selection of a
compatible basis, it allows the recovery of two-state processing despite the inclusion of the necessary
entangling interactions. Although we developed the procedure for certain types of well-known
interactions (i.e., Heisenberg–Ising and Dzyaloshinskii–Moriya), the process can be extended to other
interactions and architectures (the arrangement of qubits under interaction) by the adequate selection
of the basis on which dynamics should be expressed conveniently. In addition, it is advised that other
configurations based on qudits are possible using alternative group decompositions to SU (22d ) and
SU (2). Finally, the development only proposes the change of quantum information grammar being
used as function of the physical system in the deployment, preserving their applicability for most
quantum information proposals in the literature.
Some applications of SU (2) decomposition are foreseen. It can be exploited in the quantum control
of larger systems in which control schemes are not as well-developed as those of SU (2) dynamics.

180

www.pdfgrip.com
Entropy 2018, 20, 610

The previously established decomposition allows the establishment of exact control when blocks are
reduced to the standard forms I, NOT, H, etc. The success of such strategy for exact control depends on
the number of free parameters involved, which can be reached using a sequence of pulses instead of a
single one, or otherwise time-dependent controlled parameters in the Hamiltonian, although the block
structure is conserved. Similarly, optimal control in terms of energy or time can be achieved when
procedures such as those in [9,10] are adapted to each block in the depicted structure. More ambitious
ideas about the control of quantum processing such as the use of traveling waves, ion traps, resonant
cavities, or superconducting circuits [37–40] could be adapted to the architectures presented here.
Note that the selectivity of pairing in the blocks is related to the arisen non-diagonal elements (i.e.,
with the interactions generating diagonal-off entries in all cases). This approach to quantum evolution
will allow analytical control of the flow of quantum information in different adaptive geometrical
arrangements. The use of more feasible external fields (other than stepwise fields) is compulsory,
which is completely compatible with the current SU (2) reduction scheme [41].
In a related but not necessarily equivalent direction, selective block decomposition could be
useful for unitary factorization in quantum gate design (e.g., that developed for the SU (4) case [26]),
particularly for large dedicated gates involving the processing of several qubits. A current challenge
in the mathematical arena is solving how to express certain SU (22d ) matrices as a finite product in
2d−1 2d−1
U (1)2 −1 × SU (2)2 , such as those developed here.
Finally, other applications in quantum processing could be engineered for multichannel quantum
information storage, using certain subspaces to store differentiated information which could be
processed simultaneously in other subspaces (e.g., in quantum image processing or quantum machine
learning). Additional research should be developed to adapt this procedure to specific gate operations,
and the translation of the most common algorithms into equivalent ones based on entangled resources
like those shown here.

Funding: This research received no external funding.

Acknowledgments: I gratefully acknowledge the support from Escuela de Ingeniería y Ciencias of Tecnológico
de Monterrey in the development of this research work.
Conflicts of Interest: The author declares no conflict of interest.

Appendix A
The Appendix is divided into four parts to develop a more detailed understanding of some critical
aspects in the paper. The first is the motivation of the Hamiltonian used here, which is expressed in
terms of Pauli operators (or Pauli matrices) together with the identity. Because another central aspect
is the terminology around the group theory, the second Appendix brielfy explains some terms and
developments used in the paper, always centered in the special unitary groups, SU (n). Special attention
is given to the concept of the groups product, which is central in the paper. The third Appendix explains
the GBS basis developed by Sych and Leuchs [22]—a set of quantum states with partial entanglement
setting a basis that is useful for our development. Because this paper contains sections which may
make it difficult to understand the generality of the proposal for larger values of d, the fourth Appendix
presents the two more basic examples: d = 1, which has already been indirectly presented in the
literature [11]; and d = 2, which comprises aspects not encountered with d = 1, while they are present
for the d > 1 cases.

Appendix A.1. Generic Hamiltonian Expressed in Terms of Pauli Operators

The Hamiltonian for the interaction between a magnetic object and an external magnetic field
−
→
is −−
→μ · B , where − →μ is the dipole moment of the object. For quantum particles, this dipole moment
is precisely the spin, commonly expressed in terms of the Pauli operators − →
σ = (σx , σy , σz ) as
−
→ ge −
→ geh̄ −
→ −
→ −
→
μ = 2m s = 4m σ . Thus, the interaction reads H = − σ · B by absorbing the physical constants 2h̄
−
→
in B .

181

www.pdfgrip.com
Entropy 2018, 20, 610

For quantum magnetic systems, the most common interaction between two level systems is the
Heisenberg–Ising interaction. This interaction is a low-order approximation for the far-field interaction
between two magnetic dipoles in terms of the spin particles: − s1 · J · −
→ → σ1 · J · −
s2 or −
→ →
σ2 , when the spins
−
→ h̄ −
→
si = 2 σi are expressed in terms of Pauli operators by absorbing the factors 2 in J, which is generally
h̄

a tensor. When J becomes diagonal, we get the anisotropic Heisenberg–Ising interaction. Moreover,
if those diagonal elements become equal, the interaction becomes isotropic.
In the context of this work, another kind of interaction appears—the Dzyaloshinskii–Moriya
interaction, which is a contribution to the total magnetic exchange interaction between two neighboring
−
→ → − −
→
magnetic spins [24,25]: H = D · (− σ1 × →σ2 ). D is a vector expressed in terms of the sources’ orientation.
Clearly, this Hamiltonian will contain terms such as D1 σ1y ⊗ σ2z , D2 σ1z ⊗ σ2x , D3 σ1x ⊗ σ2y , where
the script denotes the part and the subscript denotes the component (tensor product symbol ⊗ is
introduced to remark that the product is between the spins of different quantum objects).
It is clear from the previous examples that the interactions contain terms involving one or two
Pauli operators from the different physical parts. Although most terms with two spins appear there
because interactions directly occur between a pair of physical objects, the motivation to include all
classes of products of Pauli operators in the Hamiltonian (1) is to consider the most extensive types of
interaction Notably, in the development of this work, precisely the previous interactions set a special kind
of interactions, making the SU (2) decomposition possible Particularly we should note that each term
in the Heisenberg–Ising and Dzyaloshinskii–Moriya interactions are only able to generate entangling
operations between the pair of objects involved. However, extended entanglement could be generated by
including many of those interactions between other pairs, as in the case of Ising chains. A conclusion
from this paper is that non-physical terms containing more than two spin factors in the interaction could
automatically generate more extended and inclusively genuine entanglement. For instance, for 2d qubits,
one term containing 2d factors in one term of the Hamiltonian σ1z ⊗ σ2z ⊗ ...σ2dz , which can generate
genuine entanglement in fewer steps than are necessary in Section 6. Note also that powers or additional
factors for each operator are not necessary because of their algebraic properties (any product of them for
each part can be reduced to only one operator until unitary factors).
Although these examples are for magnetic systems, these kinds of Hamiltonians are not exclusive to
those systems. Thus, for instance, the dipole interaction for a two-level system (an atom or ion restricted
to excitation between two energy levels) in a radiation trap, particles in a double-well potential, etc.,
also have Hamiltonians expressed in terms of the Pauli matrices because they are the basis of SU (2)
dynamics, common for all two-level systems (see A.2). Finally, there is a mathematical reason for the form
of Hamiltonian (1). For all two-level quantum systems, dynamics are ruled by transformations given
by elements of the unitary group of order 2, U (2) (see Appendix A.2), as solutions from the Schödinger
equation for the evolution operator. In group theory, those elements can be depicted as the exponential of
the generators of the group, which are precisely the Pauli operators defining an associated Lie algebra:
exp(i∑k∈0,1,2,3 αk σk ) (see Appendix A.2). For composed systems, the set of generators (see Appendix A.2)
and the basis elements (see Appendix A.3) for their dynamics are precisely the different products between
5
the generators for each part: { dj=1 σjk |k ∈ 0, 1, 2, 3}. Thus, through the Schrödinger equation, we can
identify the exponent with the Hamiltonian in (1), thus representing the most general Hamiltonian for the
current system composed of 2d two-level quantum systems.

Appendix A.2. Group Theory Basics in the Context of the SU (2) Decomposition
In the current Appendix we deliver a minimum of the group theory context to understand some
aspects in this work. For a deeper treatment, [42,43] is a modern introductory resource. We begin by
remarking on the notion of a group. It is a set G of elements gi ∈ G together with a defined product
operation · fulfilling the properties: (a) Closure: g1 · g2 ∈ G for all g1 , g2 ∈ G (otherwise with a defined
map: G × G → G); (b) Associativity: g1 · ( g2 · g3 ) = ( g1 · g2 ) · g3 for all g1 , g2 , g3 ∈ G; (c) Identity element:
there is a unique e ∈ G such that g · e = e · g = g for all g ∈ G; (d) Inverse: for each g ∈ G there exist
g−1 ∈ G such that g · g−1 = g−1 · g = e. If G ⊂ G is itself a group, then we say G is a subgroup of G.

182

www.pdfgrip.com
Entropy 2018, 20, 610

In two-level quantum systems we are interested in states defined in terms of a superposition of two
orthonormal states |ψ0  , |ψ1 : |ψ = α0 |ψ0  + α1 |ψ1 . Equivalently, we can use the matrix notation:

 
α0
ψ = (A1)
α1

to depict such states. Those states have a time evolution |ψ(t) via the evolution operator U (t) obeying
the Schrödinger Equation (2) in terms of the Hamiltonian operator, H. U (t) is an operator acting on
the original state to evolve it: |ψ(t) = U (t) |ψ. It fulfills: (a) the outcome |ψ(t) belongs to the set
depicted by a superposition of |ψ0  , |ψ1 ; and (b) the norm of the new state is preserved: ψ(t)|ψ(t) =
2 1
ψ|U (t)† U (t)|ψ = ψ|ψ, then U (t)† = U (t)−1 . In the context of the Dirac notation, the dual U (t)†
is another operator related with U (t). The evolution operator should clearly fulfill: (i) U (0) = I, the
identity operator which leaves |ψ without change; and (ii) U (t2 )U (t1 ) = U (t1 + t2 ). Then, the reader can
easily note that the set of operators U (t) for different values of t should form a group with the property
U (t)† = U (t)−1 . This group is said the unitary group of order 2, U (2). Similarly, we can define a n−level
system, where the evolution operators define the unitary group of order n, U (n).
For U (2), because the action of U (t) on |ψ is again a linear combination of |ψ0  , |ψ1 , then we
know that:

 
1 2 u00 u01
U (t) = ∑ ui,j |ψi  ψj or : U (t) =
u10 u11
(A2)
i,j=0

(in the following, for simplicity, we will adopt both representations as equivalent). Because of the
norm definition for quantum states, we know that:
   
1 1 ∗
u00 ∗
u10 1 u11 −u01
∗
U (t) =
†
∑ ui,j ψj ψi | or : U (t) =
†
∗ ∗ = = U (t)−1 , (A3)
i,j=0
u01 u11 |U (t)| −u10 u00

which clearly shows that entries for U (t) should fulfill the restrictions:
∗ e2iφ , u
u11 = u00 = − u ∗ e2iφ , | u |2 + | u |2 = 1 (then, |U ( t )| = e2iφ , with φ ∈ R arbitrary):
10 01 00 01

   
u00 u01 e−iφ u00 e−iφ u01
U (t) = ∗ ∗ =e iφ
≡ eiφ Ũ (t). (A4)
−e2iφ u01 e2iφ u00 −(e−iφ u01 )∗ (e−iφ u00 )∗

We advise in the last structure that both eiφ and Ũ (t) form groups separately. The set of numbers
eiφ are clearly the U (1) group under the standard multiplication of complex numbers. We skip the
demonstration that Ũ (t) with the standard matrix multiplication forms a group, which is trivial for the
associativity, identity element, and inverse properties. Demonstration for the closure property is direct.
We note that elements in this group fulfill the property |Ũ (t)| = 1. This group is said to be the special
unitary group, SU (2). Normally, in quantum mechanics we select U (t) ∈ SU (2) because the phase eiφ
is non-physical. For this reason we drop the tilde indistinctly. Because (A4), we say that U (2) is the
direct product of U (1) and SU (2): U (2) = U (1) × SU (2) (the reader is advised that the term product
is not due to the scalar product underlying in (A4), but instead to a pairing in terms of the Cartesian
product of the elements of each group. For a formal definition, consult [42,43]; this concept will be
relevant later). SU (2) is clearly a subgroup of U (2).
Another important property of the SU (2) group is that any element of it can be written as a linear
combination of the Pauli matrices (this aspect is widely used in the text). This means that they are a
basis for matrices in SU (2) (then also for U (2)). In fact:

183

www.pdfgrip.com
Entropy 2018, 20, 610

Ũ (t) = !(e−iφ u00 )σ0 + i "(e−iφ u01 )σ1 + i !(e−iφ u01 )σ2 + "(e−iφ u00 )σ3 , (A5)

where ! and " are the real and imaginary part functions (in addition, note we are using here
indistinctly the notation σ0 = I, σ1 = σx = X, σ2 = σy = Y, σ3 = σz = Z, although in the text
they have several meanings as the basis of the different SU (2) elements appearing there). Moreover, in
several parts of the text, the following property is used, derived from the algebra fulfilled by Pauli
matrices (obtained from the fact: (n · −
→
σ )2s = 1, s ∈ Z):
 
−
→ cos α + in3 sin α i sin α(n1 − in2 )
eiαn· σ = cos α σ0 + i sin α n · −
→
σ = , (A6)
i sin α(n1 + in2 ) cos α − in3 sin α

where n is a unitary vector with real components. From the last expression, it is easy to demonstrate
−
→
that |eiαn· σ | = 1, so by comparing with (A4), it is advisable that (A6) is a parametrization for the
elements in SU (2). Moreover, from the Baker–Campbell–Hausdorff formula [42,43]:

−
→ −
→
ei(φσ0 +αn· σ ) = eiφ eiαn· σ . (A7)

Then, it is said that σ1 , σ2 , σ3 are the generators of SU (2), while σ0 , σ1 , σ2 , σ3 are the generators of
U (2). This fact was mentioned in Appendix A.1 to suggest the generality of Hamiltonian (1), although
some steps remain to arrive at the SU (22d ) group.
The reader can note that (13) and (29) adjust to those structures, and thus the blocks
in the decomposition belong to U (1) × SU (2). Similar arguments in group theory show that
U (n) = U (1) × SU (n), where SU (n) is the special unitary group of order n, the group of unitary
matrices U † = U −1 with determinant equal to 1, |U | = 1. Although there are generators for
such groups, the development of the article shows that the combination of 2d two-level quantum
systems requires evolution matrices in U (22d ) (or in SU (22d )). As a result of the decomposition,
2d−1
we show that the evolution matrices belonging to such groups form the product group U (2)2
(while the Hilbert space of the quantum states is decomposed into the direct sum of 22d−1 two-level
subspaces of dimension 2). Precisely, the elimination of the term h{00...0} in the Hamiltonian induces
directly in (14) that U ∈ SU (22d ). In addition, the SU (2) decomposition shows in this case that
2d−1 2d−1
U ∈ U (1)2 −1 × SU (2)2 ) (due to the dependence of one U (1)-term from the remaining). The
reader should consult the formal definition of a direct product in [42,43].

Appendix A.3. Generalized Bell States Basis in Context

GBS states (generalized Bell states) as introduced by Sych and Leuchs [22] (16) were expressed
in terms of Pauli operators. That original expression is highly convenient for the development of the
current work because it allows it to be easily connected with the form of the Hamiltonian (1) in the
same terms, allowing the important result to be easily obtained (18). Nevertheless, a more simple
expression could be given for the understanding of such states. In fact, it is easy to note that each
element in the GBS basis for 2d qubits can be written as:

% 6
0
d 1 1 1
ΨI d = ΨI d = Ψi1 ⊗ Ψi2 ⊗ ... ⊗ Ψid , (A8)
4 4,j
j =1

% 1 %
where I4,j
d = i , and Ψ
j 2γ+(γ⊕δ) = β γ,δ or | Ψi  = β ( i mod2),(i −2( i mod2))⊕( i mod2 ) , the well-known
2 2 2
single Bell states (in the last expressions, ⊕ represents the module-2 sum). Thus, each element of
the GBS basis is in reality a tensor product of d Bell states identified through their scripts in base-4.
These states are 2−separable (meaning the smallest separable subsystems still contains two entangled
parts). Thus, when we apply a Hadamard-like operation involving only one script (Type I or III

184

www.pdfgrip.com
Entropy 2018, 20, 610

interactions), we are consequently able to convert the involved Bell state into a separable state. When
we look at this version of the GBS basis, it is clearer why Type I and III interactions become in
entangling or unentagling operations on only one correspondent pair. Both types of operations actually
resemble the effect on two-qubits processing with SU (4) operations such as those developed in [11],
while the remaining system is not involved. Only the Type II operations provide more extended
entangling properties.

Appendix A.4. Illustrative Examples of SU (2) Decomposition

In the following two subsections, we develop examples of the aspect of the evolution operators
for the specific cases d = 1 and d = 2. The latter case is of special importance because it depicts how
Type II interactions extend the entanglement, as shown in Section 6.

Appendix A.4.1. Case d = 1

This case has been developed in the literature [11,13]. In the current context, only Type I and Type
III interactions are possible (because there is only one correspondent pair). The corresponding GBS
basis has four elements: |Ψ0  , |Ψ1  , |Ψ2  , |Ψ3 , the Bell states precisely. In the next expressions, we
assume a lexicographic order in the components of the basis, so any arrangement of them is supposed
(in contrast to how it was considered in (14)). The Hamiltonian H I contains at the most five terms:

3
HI = ∑ hm,m σ1m ⊗ σ2m + hk,0 σ1k + h0,k σ2k , (A9)
m =1

where k is the direction of the local interaction. Te Hamiltonian H I I I (with the crossed interaction in
the direction k) also contains utmost five terms. If i, j, k is an even permutation of 1, 2, 3:

3
HI I I = ∑ hm,m σ1m ⊗ σ2m + hi,j σ1i ⊗ σ2j + h j,i σ1j ⊗ σ2i . (A10)
m =1

Although it is an special case accepting the combination of the two Hamiltonians (Table 1), we set
them separately. Because of the space and complexity, we do not express Uk (t) in terms of the original
coefficients in (A9) and (A10). In any case, formulas (31)–(37) are sufficiently efficient to reproduce
the entries of each Hamiltonian. Instead, after expressing both Hamiltonians in the GBS basis, they
become (as in (28)):
⎛ ⎞
h111 h112 0 0
⎜ h112∗ h122 ⎟
⎜ 0 0 ⎟
H1 = ⎜ ⎟ = S H0,1 ⊕ S H2,3 =⇒ U1 (t) = SU0,1 ⊕ SU2,3 ,
⎝ 0 0 h211 h212 ⎠
0 0 h212∗ h222
⎛ ⎞
h111 0 h112 0
⎜ 0 h211 h212 ⎟
⎜ 0 ⎟
H2 = ⎜ 1∗ ⎟ = S H0,2 ⊕ S H1,3 =⇒ U2 (t) = SU0,2 ⊕ SU1,3 , (A11)
⎝ h12 0 h122 0 ⎠
0 h212∗ 0 h222
⎛ ⎞
h111 0 0 h112
⎜ h211 h212 ⎟
⎜ 0 0 ⎟
H3 = ⎜ ⎟ = S H0,3 ⊕ S H1,2 =⇒ U3 (t) = SU0,3 ⊕ SU1,2 ,
⎝ 0 h212∗ h222 0 ⎠
h112∗ 0 0 h122

185

www.pdfgrip.com
Entropy 2018, 20, 610

where the superscript i in himn points out the consecutive number of blocks and S HI ,I fulfills the
syntactic notation followed in (28) and (29). We will exploit this notation in the following section for
simplicity, where the matrix notation will become hardly extensive.

Appendix A.4.2. Case d = 2

We develop two cases for the case d = 2. The first considers the Type I interaction and the second
pertains to the Type IIa interaction. The last case involves a different situation not appearing in d = 1:
the possibility of generating extended entanglement among the four qubits involved in this case.
By considering the four qubits under the Type I interaction with the local interaction terms on the
pair k = 2 in the direction j in (30), the Hamiltonian has the form:

3 3
HI = ∑ hm,0,m,0 σ1m ⊗ σ3m + ∑ h0,m,0,m σ2m ⊗ σ4m + h0,j,0,0 σ2j + h0,0,0,j σ4j . (A12)
m =1 m =1

There are 16 GBS elements in the basis: |Ψ0,0  , |Ψ0,1  , ..., |Ψ0,3  , |Ψ1,0  , |Ψ1,1  , ..., |Ψ3,3 .
Interaction generates exchanges between the GBS basis elements as follows (forming eight blocks but
1
only two types of them). If i, j, k is a permutation of 1, 2, 3 with i < k, then: (a) |Ψm,0  ←→ Ψm,j ;
1
(b) |Ψm,i  ←→ Ψm,k , with m = 0, ..., 3. Due to the extension, we do not write matrix expressions
as in the case d = 1 (which already settled an illustrative orientation to the reader). Instead, we use
the notation of direct sum for block matrices as before. Thus, remembering the scripts are numbers
1
becoming from the base-4 scripts in the GBS basis, Ψ a,b = |ΨI  with I = a + 4b ∈ 0, 1, ..., 15,
the decomposition for the evolution operator will become for Uj (t):
   
3
3 3
3
Uj (t) = SUm,m+4j ⊕ SUm+4i,m+4k ,
m =0 m =0
then :
(A13)
U1 (t) = SU0,4 ⊕ SU1,5 ⊕ SU2,6 ⊕ SU3,7 ⊕ SU8,12 ⊕ SU9,13 ⊕ SU10,14 ⊕ SU11,15 ,
U2 (t) = SU0,8 ⊕ SU1,9 ⊕ SU2,10 ⊕ SU3,11 ⊕ SU4,12 ⊕ SU5,13 ⊕ SU6,14 ⊕ SU7,15 ,
U3 (t) = SU0,12 ⊕ SU1,13 ⊕ SU2,14 ⊕ SU3,15 ⊕ SU4,8 ⊕ SU5,9 ⊕ SU6,10 ⊕ SU7,11 .
These exchanges only involve qubits in the same correspondent pair, so they cannot extend the
entanglement beyond this pair. Additionally, we remark that the first four blocks have the same form,
as do the last four. Here, only two different types of blocks exist.
We develop the case d = 2 for a Type IIa interaction involving additional non-local and
non-crossed interactions between the pairs k = 1 and k = 2 (note that the situation will be similar for
the cases with d > 2). Assuming the interaction in the direction j and in (30), the Hamiltonian becomes:

3 3
H I Ia = ∑ hm,0,m,0 σ1m ⊗ σ3m + ∑ h0,m,0,m σ2m ⊗ σ4m +
m =1 m =1 (A14)
h j,j,0,0 σ1j ⊗ σ2j + h j,0,0,j σ1j ⊗ σ4j + h0,j,j,0 σ2j ⊗ σ3j + h0,0,j,j σ3j ⊗ σ4j .

Then, there exist eight types of % exchanges and blocks (i, j, k is a permutation of 1, 2, 3): (a)
1 1 1 1 1
|Ψ0,0  ←→ Ψ j,j ; (b) |Ψ0,i  ←→ Ψ j,k ; (c) Ψ0,j ←→ Ψ j,0 ; (d) Ψ0,k ←→ Ψ j,i ; (e) |Ψi,0  ←→
% 1 1 1 1 1
Ψk,j ; (f) |Ψi,i  ←→ Ψk,k ; (g) Ψi,j ←→ Ψk,0 ; (h) Ψi,k ←→ Ψk,i . In this case, all blocks
1
will become different, but it is not a general situation when d grows. As before Ψ a,b = |ΨI ,
with I = a + 4b ∈ 0, 1, ..., 15. Then, the evolution operator can be written as:

Uj (t) = SU0,j+4j ⊕ SU4i,j+4k ⊕ SU4j,j ⊕ SU4k,j+4i ⊕ SUi,k+4j ⊕ SUi+4i,k+4k ⊕ SUi+4j,k ⊕ SUi+4k,k+4i , (A15)

186

www.pdfgrip.com
Entropy 2018, 20, 610

noting that the exchange involving two scripts implies the generation of entanglement between the
two correspondent pairs (i.e., among the four qubits as a whole). In this case, eight blocks are different
(but not necessarily independent,—because there are only 11 parameters free, including time t).
These two examples show in detail how the SU (2) decomposition is established. For cases d > 2,
the situation becomes similar and they are easily understood using the last synthetic notation in terms
of direct sums of blocks. It should finally be remarked that formulas (30) and (37) are computationally
useful and efficient to connect the original Hamiltonian coefficients with the entries for each block.

References
1. Feynman, R.P. Simulating Physics with Computers. Int. J. Theor. Phys. 1982, 21, 467–488. [CrossRef]
2. Deutsch, D. Quantum theory, the Church-Turing principle and the universal quantum computer.
Proc. R. Soc. Lond. A 1985, 400, 97–117. [CrossRef]
3. Steane, A. Error Correcting Codes in Quantum Theory. Phys. Rev. Lett. 1996, 77, 793–797. [CrossRef]
[PubMed]
4. Bennett, C.H.; Brassard, G. Quantum cryptography: Public key distribution and coin tossing. In Proceedings
of the International Conference on Computers, Systems & Signal Processing, Bangalore, India,
9–12 December 1984; IEEE: New York, NY, USA, 1984; pp. 175–179.
5. Ekert, A. Quantum cryptography based on Bell’s theorem. Phys. Rev. Lett. 1991, 67, 661–663. [CrossRef]
[PubMed]
6. Hillery, M.; Vladimír, B.; Berthiaume, A. Quantum secret sharing. Phys. Rev. A 1999, 59, 1829–1834.
[CrossRef]
7. Long, G.; Liu, X. Theoretically efficient high-capacity quantum-key-distribution scheme. Phys. Rev. A 2002, 65, 032302.
[CrossRef]
8. Deng, F.; Long, G.; Liu, X. Two-step quantum direct communication protocol using the
Einstein-Podolsky-Rosen pair block. Phys. Rev. A 2003, 68, 042317. [CrossRef]
9. D’Alessandro, D.; Dahleh, M. Optimal control of two-level quantum systems. IEEE Trans. Autom. Control
2001, 46, 866–876. [CrossRef]
10. Boscain, U.; Mason, P. Time minimal trajectories for a spin 1/2 particle in a magnetic field. J. Math. Phys.
2006, 47, 062101. [CrossRef]
11. Delgado, F. Algebraic and group structure for bipartite anisotropic Ising model on a non-local basis.
Int. J. Quantum Inf. 2015, 13, 1550055. [CrossRef]
12. Delgado, F. Generation of non-local evolution loops and exchange operations for quantum control in three
dimensional anisotropic Ising model. ArXiv 2016, arXiv:1410.5515.
13. Delgado, F. Stability of Quantum Loops and Exchange Operations in the Construction of Quantum
Computation Gates. J. Phys. 2016, 648, 012024.
14. McConnell, R.; Bruzewicz, C.; Chiaverini, J.; Sage, J. Characterization and Mitigation of Anomalous Motional
Heating in Surface-Electrode Ion Traps. In Proceedings of the 46th Annual Meeting of the APS Division of
Atomic, Molecular and Optical Physics, Columbus, OH, USA, 8–12 June 2015; Volume 60.
15. Gambetta, J.M.; Jerry, M.; Steffen, M. Building logical qubits in a superconducting quantum computing
system. npj Quantum Inf. 2017, 3, 1–7. [CrossRef]
16. Fubini, A.; Roscilde, T.; Tognetti, V.; Tusa, M.; Verrucchi, P. Reading entanglement in terms of
spin-configuration in quantum magnet. Eur. Phys. J. D 2006, 38, 563–570. [CrossRef]
17. Pfaff, W.; Taminiau, T.; Robledo, L.; Bernien, H.; Matthew, M.; Twitchen, D.; Hanson, R. Demonstration of
entanglement-by-measurement of solid-state qubits. Nat. Phys. 2013, 9, 29–33. [CrossRef]
18. Magazzu, L.; Jamarillo, J.; Talkner, P.; Hanggi, P. Generation and stabilization of Bell states via repeated
projective measurements on a driven ancilla qubit. ArXiv 2018, arXiv:1802.04839v1.
19. Nielsen, M.; Chuang, I. Quantum Computation and Quantum Information; Cambridge University Press:
Cambridge, UK, 2011.
20. Lieb, E.; Schultz, T.; Mattis, D. Two soluble models of an antiferromagnetic chain. Ann. Phys. 1961, 16, 407–466.
[CrossRef]
21. Baxter, R.J. Exactly Solved Models in Statistical Mechanics; Academic Press: New York, NY, USA, 1982.
22. Sych, D.; Leuchs, G. A complete basis of generalized Bell states. New J. Phys. 2009, 11, 013006. [CrossRef]

187

www.pdfgrip.com
Entropy 2018, 20, 610

23. Delgado, F. Modeling the dynamics of multipartite quantum systems created departing from two-level
systems using general local and non-local interactions. J. Phys. 2017, 936, 012070. [CrossRef]
24. Dzyaloshinskii, I. A thermodynamic theory of weak ferromagnetism of antiferromagnetics. J. Phys. Chem. Solids
1958, 4, 241–255. [CrossRef]
25. Moriya, T. Anisotropic Superexchange Interaction and Weak Ferromagnetism. Phys. Rev. 1960, 120, 91–98.
[CrossRef]
26. Delgado, F. Two-qubit quantum gates construction via unitary factorization. Quantum Inf. Comput. 2017, 17,
721–746.
27. Delgado, F. Generalized Bell states map physical systems’ quantum evolution into a grammar for quantum
information processing. J. Phys. 2017, 936, 012083. [CrossRef]
28. Uhlmann, A. Anti-(conjugate) linearity. Sci. China Phys. Mech. Astron. 2016, 59, 630301. [CrossRef]
29. Boykin, P.; Mor, T.; Pulver, M.; Roychowdhury, V.; Vatan, F. On universal and fault tolerant quantum
computing. ArXiv 1999, arXiv:9906054.
30. Li, C.; Jones, R.; Yin, X. Decomposition of unitary matrices and quantum gates. Int. J. Quantum Inf. 2013, 11,
1350015. [CrossRef]
31. Delgado, F. Universal Quantum Gates for Quantum Computation on Magnetic Systems Ruled by
Heisenberg-Ising Interactions. J. Phys. 2017, 839, 012014. [CrossRef]
32. Barenco, A.; Bennett, C.; Cleve, R.; DiVincenzo, D.; Margolus, N.; Shor, P.; Sleator, T.; Smolin, J.; Weinfurter, H.
Elementary gates for quantum computation. Phys. Rev. A 1995, 52, 3457. [CrossRef] [PubMed]
33. Liu, Y.; Long, G.; Sun, Y. Analytic one-bit and CNOT gate constructions of general n-qubit controlled gates.
Int. J. Quantum Inf. 2008, 6, 447–462. [CrossRef]
34. Hou, S.; Wang, L.; Yi, X. Realization of quantum gates by Lyapunov control. Phys. Lett. A 2014, 378, 699–704.
[CrossRef]
35. Gurvits, L. Classical complexity and quantum entanglement. J. Comput. Syst. Sci. 2004, 69, 448–484.
[CrossRef]
36. Uhlmann, A. Fidelity and concurrence of conjugated states. Phys. Rev. A 2000, 62, 032307. [CrossRef]
37. Serikawa, T.; Shiozawa, Y.; Ogawa, H.; Takanashi, N.; Takeda, S.; Yoshikawa, J.; Furusawa, A.
Quantum information processing with a travelling wave of light. In Proceedings of the SPIE-OPTO,
San Francisco, CA, USA, 19–23 August 2018.
38. Britton, J.; Sawyer, B.; Keith, A.; Wang, J.; Freericks, J.; Uys, H.; Biercuk, M.; Bollinger, J. Engineered
two-dimensional Ising interactions in a trapped-ion quantum simulator with hundreds of spins. Nature
2012, 484, 489–492. [CrossRef] [PubMed]
39. Bohnet, J.; Sawyer, B.; Britton, J.; Wall, M.; Rey, A.; Foss-Feig, M.; Bollinger, J. Quantum spin dynamics and
entanglement generation with hundreds of trapped ions. Science 2016, 352, 1297–1301. [CrossRef] [PubMed]
40. De Sa Neto, O.; de Oliveira, M. Hybrid Qubit gates in circuit QED: A scheme for quantum bit encoding and
information processing. ArXiv 2011, arXiv:1110.1355.
41. Delgado, F.; Rodríguez, S. Modeling quantum information dynamics achieved with time-dependent driven
fields in the context of universal quantum processing. ArXiv 2018, arXiv:1805.05477.
42. Hall, B. Lie Groups, Lie Algebras, and Representations: An Elementary Introduction; Graduate Texts in
Mathematics 222; Springer: Cham, Switzerland, 2015.
43. Hall, B. An Elementary Introduction to Groups and Representations. ArXiv 2000, arXiv:quant-ph/0005032.

c 2018 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

188

www.pdfgrip.com
 entropy
Article
An Information-Theoretic Perspective on the
Quantum Bit Commitment Impossibility Theorem
Marius Nagy 1,2, * and Naya Nagy 3
1 College of Computer Engineering and Science, Prince Mohammad Bin Fahd University, Al Khobar 31952,
Saudi Arabia
2 School of Computing, Queen’s University, Kingston, ON K7L 2N8, Canada
3 College of Computer Science and Information Technology, Imam Abdulrahman Bin Faisal University,
Dammam 34212, Saudi Arabia; nmnagy@iau.edu.sa
* Correspondence: mnagy@pmu.edu.sa or marius@cs.queensu.ca; Tel.: +966-13-849-9272

Received: 16 January 2018; Accepted: 12 March 2018; Published: 13 March 2018

Abstract: This paper proposes a different approach to pinpoint the causes for which an unconditionally
secure quantum bit commitment protocol cannot be realized, beyond the technical details on which the
proof of Mayers’ no-go theorem is constructed. We have adopted the tools of quantum entropy analysis
to investigate the conditions under which the security properties of quantum bit commitment can
be circumvented. Our study has revealed that cheating the binding property requires the quantum
system acting as the safe to harbor the same amount of uncertainty with respect to both observers
(Alice and Bob) as well as the use of entanglement. Our analysis also suggests that the ability to cheat
one of the two fundamental properties of bit commitment by any of the two participants depends on
how much information is leaked from one side of the system to the other and how much remains
hidden from the other participant.

Keywords: quantum information theory; bit commitment; protocol; entropy; entanglement

1. Introduction
Bit commitment refers to a cryptographic protocol that can be described informally as follows. In
the first phase, Alice decides on a binary value (0 or 1) that she locks in a safe, keeping the key for
herself and then hands over the locked safe to Bob. This is referred to as the commit phase of the
protocol. Later on, during the decommit phase, Alice reveals her commitment by presenting Bob with
the safe key. Bob can now use the key to unlock the safe and learn the value of the bit that Alice has
previously committed to.
Any secure implementation of bit commitment must satisfy two crucial properties:
1. Alice should no longer be able to change the value of her commitment, once the safe is in Bob’s
hands. This requirement is known as the binding property of bit commitment.
2. Bob should not be able to learn the content of the safe before the decommit phase. This requirement
is known as the hiding property of bit commitment.

If Alice and Bob employ quantum means to work out the details of a bit commitment procedure,
then the resulting protocol falls into the category of quantum bit commitment (QBC). Unconditionally
secure QBC is known to be impossible [1], meaning that no protocol can ever be devised such that
both binding and hiding properties are guaranteed, if no restriction is placed on the capabilities of
the two participants, Alice and Bob. This result is a direct consequence of the Schmidt decomposition
theorem for composite systems [2], but the essence of the impossibility theorem for QBC is not easy to
grasp and understand intuitively, beyond the technical details outlined in the papers of Mayers [3]
and Lo and Chau [4]. That explains why even after the publication of the “no-go” theorem for QBC,

Entropy 2018, 20, 193; doi:10.3390/e20030193 189 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2018, 20, 193

quantum cryptographers were still looking for a protocol that would not fall under its scope. Therefore,
in the present investigation, we wish to gain further insight into why hiding and binding are mutually
exclusive properties for QBC by adopting the perspective and tools of quantum information theory.
The remainder of this paper is organized as follows. The next section summarizes the most
important turns and results that have shaped the history of quantum bit commitment. Section 3
defines a general framework for QBC protocols and arrives at a formulation of the hiding property in
terms of the entropy accumulated within the quantum safe. In Section 4, we show that entanglement is a
necessary condition for cheating the binding property, using the tools of quantum information theory.
A detailed procedure of how Alice is able to cheat the binding property is described in Section 5,
with a concrete exemplification for the QBC protocol proposed in the BB84 paper [5]. Section 6 extends
the analysis of the two security properties of bit commitment to protocols initiated by Bob. Finally,
Section 7 concludes the paper with a summary of our findings.

2. Brief History of Quantum Bit Commitment

The feasibility of quantum bit commitment has huge implications for the field of cryptography.
Classical cryptography is able to offer solutions (based on bit commitment) to a wide variety of situations
classified as discreet decision problems [6]. All these situations share an important characteristic, namely
that discretion is vital to achieving agreements. Examples range from negotiating arms treaties to
forming business partnerships or organizing mergers.
Classic cryptographic solutions to these applications do exist, but since they involve public-key
systems, they are inevitably based on unproven assumptions about the difficulty of factoring large
numbers and other related problems. What was expected from quantum cryptography was a totally
secure system, guaranteed by the laws of physics, similar to what was already achieved in the case of
quantum key distribution [7]. To this end, Claude Crépeau and Joe Kilian have shown how oblivious
transfer (or 1-out-of-2 oblivious transfer) can be used as a building block for solving two-party problems
requiring discretion [8]. In turn, to provide totally secure quantum oblivious transfer, one would need
a secure form of bit commitment. Consequently, much of the research effort in quantum cryptography
in the early 1990s was devoted to finding a protocol for quantum bit commitment that is absolutely
and provably secure. That result (known as BCJL after the authors’ names) was reported in 1993 [9]
and became the foundation for numerous applications in quantum cryptography, pertaining to discreet
decision making.
The surprise came in 1995 when Dominic Mayers discovered how Alice could cheat in the BCJL
bit commitment protocol by using entanglement [1]. Furthermore, Mayers [3] and, independently,
Lo and Chau [4] proved that it would be possible for Alice to cheat in any protocol for quantum bit
commitment that guarantees the hiding property. An intuitive explanation is that the description of
the quantum safe she hands over to Bob must give nothing away about the committed bit inside.
Consequently, regardless of the particular bit commitment scheme employed, the quantum states of
the safe containing either 0 or 1 must be very similar (if not identical) since otherwise Bob would be
able to discern the difference and gain knowledge about the committed bit prematurely. However,
the very fact that the two states are virtually the same gives Alice the possibility to keep her options
open and postpone her commitment for later on. Although in their 1996 review paper of quantum
cryptography, Brassard and Crépeau [10] argued that for the time being the practical implications of
the flaw discovered in the quantum bit commitment protocol are minimal, the weakness definitely
affected the entire edifice of quantum cryptography built upon quantum bit commitment.
Perhaps the importance of bit commitment for the general field of cryptography or the intuition
that the success of quantum key distribution could be replicated for quantum bit commitment still
pushed people to look for a solution. Several protocols were proposed that try to restrict the behavior of
the cheater in some way so as to obtain a secure bit commitment scheme [10–12]. It turned out that all
these protocols were falling under the scope of Mayers’ impossibility result. Building on Mayers’ work,
Spekkens and Rudolph [13] proved that the two fundamental properties of bit commitment, binding

190

www.pdfgrip.com
Entropy 2018, 20, 193

and hiding, are mutually exclusive. The more a protocol is hiding, the less it is binding and vice-versa.
This led to a general belief that the principles of quantum mechanics alone cannot be used to create an
unconditionally secure bit commitment protocol. Therefore, recent advances on the topic either exploit
realistic physical assumptions like the dishonest party being limited by “noisy storage” for quantum
information [14] or combine the power of Einstein’s relativity with quantum theory [15–17]. Yet another
direction explored by researchers in the field is the class of “cheat-sensitive” quantum bit commitment
protocols. Since the hope of designing an unconditionally secure QBC protocol had to be abandoned,
researchers focused instead on protocols in which the probability of detecting a dishonest participant is
merely required to be non-zero. Properties of such cheat-sensitive protocols are explored in [18–23].
Ultimately, secure bit commitment using quantum theory alone remains unattainable. Although
we know that entanglement and Schmidt’s decomposition theorem are key ingredients in this
impossibility result, these are technical details that fail to provide a deep and intuitive understanding on
why, for example, unconditionally secure quantum key distribution is possible, but quantum bit
commitment is not. In the following, we try to shed some light into what the conditions for successful
hiding and binding properties are and to describe the complex relationship between these properties,
with the help of quantum information theory and quantum entropy.

3. An Information-Theoretic Formulation of the Hiding Property

In a typical QBC framework, Alice encodes some classical information into a quantum system
using one out of two possible encoding bases: B0 if she decides to commit to 0 or B1 if the commit
value is 1. The quantum system represents the safe, while the classical information that Alice keeps
secret from Bob plays the role of the key in the description of bit commitment above. In order to prevent
Bob from squeezing any information from the quantum safe through some clever measurement,
B0 and B1 must be complementary bases, such as the normal computational base and the
Hadamard base. In terms of a practical implementation, these represent the rectilinear and diagonal
polarizations of a photon.
However, to characterize the hiding property formally, we also need to define more precisely
Alice’s secret key that she encodes in the quantum safe. In general, this key can be any string over
a certain alphabet, so without loss of generality, let us assume that the key is a binary string of a certain
length n. Then the hiding property is an expression of Bob’s total uncertainty over the quantum state
of the safe. From his point of view, the safe must be in a complete mixture of all possible bitstrings
that Alice might have encoded, regardless of the basis used. This is expressed formally, using density
matrices, as follows:
1 I
ρ0Bob = ∑ | B0 (k i ) B0 (k i )| = n
k i ∈(0|1)n
2n 2
(1)
1
= ∑ n
| B1 (k i ) B1 (k i )| = ρ1Bob
k ∈(0|1)n
2
i

where ρ0Bob
represents Bob’s view of the system when Alice commits to 0, and ρ1Bob represents Bob’s
view when Alice commits to 1. Since the density matrices corresponding to Alice’s two commitments
are identical, there is not even a theoretical chance for Bob to distinguish between a commitment to 0
and a commitment to 1, no matter what measurement(s) he may try to perform. This uncertainty on
Bob’s side can also be quantified using the information theoretic concept of entropy:

I I
S( B) = −tr(ρ0Bob log ρ0Bob ) = −tr(ρ1Bob log ρ1Bob ) = −tr( log n ) = n (2)
2n 2
In other words, the amount of uncertainty in the quantum state of the safe, from Bob’s perspective,
is equal to n bits and this is the maximum it can be, given that the key used by Alice is of length n.
Equation (2) therefore captures the hiding property of bit commitment in the language of quantum
information theory.

191

www.pdfgrip.com
Entropy 2018, 20, 193

So far, we have concentrated on Bob’s perspective on the quantum system representing the
safe, which is just a part of the whole ensemble Alice–Bob. The other part, which remains in Alice’s
possession, contains information about the key chosen by Alice and the bit commitment. The two
parts are not independent, as the key together with the encoding basis completely determine the
quantum state of the subsystem given to Bob. This means that the conditional entropy S( B| A) = 0.
For the reader not familiar with the various forms of entropy, we mention that S( B| A) is the amount
of information present in system B that does not come from system A, and is defined as:

S( B| A) = S( A, B) − S( A). (3)

In a hypothetical situation, where systems A and B are completely independent, S( B| A) = S( B).

An “entropy Venn diagram” of the whole system depicting the two components and the relationship
between them is given below:
The number in each region of the diagram reflects the quantum entropy or amount of uncertainty
characterizing the part of the system represented graphically by that region. The mutual information
content of A and B, depicted as the intersection of A and B in the diagram, measures how much
information systems A and B have in common and is defined as:

S( A : B) = S( A) + S( B) − S( A, B). (4)

Information which is common to both systems is counted twice in the summation S( A) + S( B),
while information which is not common is counted just once. Therefore, by subtracting the joint
information present in both systems (namely, S( A, B)) from this summation, we obtain just the
common or mutual information of A and B. We note that the whole uncertainty present in subsystem B
actually comes from subsystem A or, equivalently, there are n bits of uncertainty shared in the system
between the two components (the mutual information S( A : B) = n). On the other hand, subsystem A
has one more bit of uncertainty compared to subsystem B, which is enough to ensure the hiding
property, even if Bob is able to dispel the uncertainty characterizing the quantum state of the safe in
his view.
It is crucial here to make the observation that the above diagram is constructed solely from
Bob’s perspective on the system. Consequently, the fact that S( A| B) = 1 cannot be interpreted in the
sense that if Bob knows the exact quantum state of the safe, as prepared by Alice, there is still some
uncertainty left about the key and/or the bit commitment on Alice’s side. Dispelling the n bits of
uncertainty describing the quantum state of the safe, simply means that Bob’s knowledge of this state
advances from a complete mixture of all 2n possible terms to a single, precise, pure state characterizing
the quantum safe.
Such a sharp decrease in the entropy of the quantum safe can be triggered, for example, by Bob
learning the outcome of a projective measurement applied on all n qubits composing the quantum safe,
where each possible key k i becomes a projector Pi = |k i k i |. However, even after such a measurement
is performed, there still remains one full bit of uncertainty about Alice’s choice for the bit commitment
and the corresponding key used.
To better understand the point, let us exemplify with the trivial case of a 1-qubit safe and the two
encoding bases being the normal computational base and the Hadamard base. In this simple scenario,
Alice encodes either 0 or 1 in the basis of her choice and sends the resulting qubit to Bob. Alice keeps
a record of the basis chosen (representing her commitment) and the bit encoded (representing the key
used). From Bob’s point of view, the whole system is in a mixed state:

|0+|
√ 1 10| 0|+
√ 1|
ρ AB = 14 |00|000|0| + 14 |01|101|1| + 14 |10 + 14 |11 |0−|
√ 1 11| 0|−
√ 1| , (5)
2 2 2 2

where the first qubit denotes the encoding basis, the second one represents the encoded key and the
third qubit plays the role of the quantum safe passed over to Bob. Thus, the first two qubits come from

192

www.pdfgrip.com
Entropy 2018, 20, 193

subsystem A, while the last qubit makes up subsystem B. The amount of uncertainty present in the
system equals two bits, corresponding to the four choices Alice has with respect to the encoding basis
and key used:
S(ρ AB ) = −tr(ρ AB log ρ AB ) = 2. (6)

In order to see how the quantum safe appears to Bob, we can trace out subsystem A from the
global state ρ AB :
1 1 I
ρ B = tr A (ρ AB ) = |00| + |11| = . (7)
2 2 2
This state entails one bit of uncertainty, since it is a mixture of both possible terms |0 and |1,
each with probability 1/2. Naturally, this bit of uncertainty comes entirely from subsystem A, as it can
easily be checked that the entropy of subsystem A equals the entropy of the entire system:

1 I
ρ A = trB (ρ AB ) = (|0000| + |0101| + |1010| + |1111| = . (8)
4 4
Dispelling the bit of uncertainty that Bob sees in the quantum safe amounts to Bob being
informed or finding out through a measurement which of the two possible pure states the safe actually
finds itself in. Regardless of the answer (|0 or |1), this state can come from both commitments with
equal probability, as it can be seen from Equation (5). So the uncertainty on the bit commitment still
remains for Bob (S( A| B) = 1), thus ensuring the hiding property. On the other hand, if Alice plays by
the rules and does prepare a quantum state for the safe based on a specific key and bit commitment,
then from her perspective there is not a shred of uncertainty anywhere in the system. All entropies
depicted in Figure 1 would be zero in this case.

Figure 1. Entropy diagram of the whole system Alice–Bob viewed from Bob’s perspective. All uncertainty
present in subsystem B (quantum safe) comes from subsystem A (Alice’s own quantum register), which
has one extra bit of uncertainty.

This is reminiscent of the so-called “relational interpretation” of quantum mechanics (RQM) [24,25],
which refutes the idea of an objective or absolute reality (state of a physical system) and proclaims
that different observers can give different accounts on the properties of the same physical system.
In the case of bit commitment, different accounts arise due to different amounts of information
the participants (observers in RQM) have about a particular system, which motivates the use of
information theoretic tools in order to get further insight into the problem. It would appear that
there is a deep connection between information theory and quantum mechanics which is brought to
light in the relational interpretation of the latter.

193

www.pdfgrip.com
Entropy 2018, 20, 193

4. Cheating Requires Entanglement

In the previous section, we have clearly stated the condition that needs to be satisfied in order to
ensure the hiding property and preclude any possibility of Bob gaining premature knowledge about
Alice’s bit commitment. The condition was formulated in terms of both density matrices and using the
language of quantum information theory. In this section, we turn our attention to Alice and use the
concept of entropy to prove that, if the hiding property is guaranteed, then Alice can cheat and change
her commitment in the decommit phase if and only if she is endowed with the ability to generate and
manipulate multi-party entangled states.
In a perfectly hiding quantum bit commitment protocol, Alice can cheat in the decommit phase by
applying a transformation that will rotate the state of the whole system Alice–Bob from |ψ0  (global
quantum state corresponding to a commitment to 0) to |ψ1  (quantum state describing the status of
the entire system in the case of a commitment to 1). Since, in the decommit phase, she no longer has
access to the safe, which was given to Bob (subsystem B), Alice must be able to effect this transformation
only by acting on her side (subsystem A). This is only possible if subsystem B looks the same for both
possible commitments, not only for Bob, but from Alice’s point of view as well:

I
ρ0Alice ( B) = ρ1Alice ( B) = . (9)
2n

Then, the transformation |ψ0  −→ |ψ1  is guaranteed to exist as a direct consequence of the
Schmidt decomposition theorem. Condition (9) is not met if Alice prepares the quantum state of the
safe using a specific key k i :

ρ0Alice ( B) = | B0 (k i ) B0 (k i )| = | B1 (k i ) B1 (k i )| = ρ1Alice ( B). (10)

Consequently, in such a case, there is no transformation that can rotate |ψ0  into |ψ1  just from
Alice’s side. Therefore, maintaining a full uncertainty on the quantum state of the safe is a necessary
condition for Alice to cheat:

S( B) = −tr(ρ Alice ( B) log ρ Alice ( B)) = n. (11)

We now show that this condition implies that the entire system Alice–Bob must be in an
entangled state.
Regardless of how Alice chooses to prepare the state of the safe (subsystem B), she has full
knowledge of its state in relation to her own quantum register (subsystem A). In other words, since
she is the one preparing both subsystems, Alice has complete knowledge of the state of the ensemble
Alice–Bob. This means the whole system is in a pure state according to Alice:

S( A, B) = 0. (12)

At this point, the Schmidt decomposition theorem can be applied and it follows that both
subsystems must have the same eigenvalues. In addition, since quantum entropy is completely
determined by the eigenvalues, then it must be the case that the entropy of subsystem A is the same as
that of subsystem B:
S( A) = S( B) = n. (13)

Based on this last equality and Equation (12), the entropy diagram of the system looks like the
one depicted in Figure 2. The mutual information S( A : B) = 2n, while both conditional entropies
S( A| B) = S( B| A) = −n. A conditional entropy can be negative if and only if the two subsystems are
entangled. Equivalently put, a supercorrelation indicated by negative values of conditional entropy is
the unmistakable hallmark of entanglement.

194

www.pdfgrip.com
Entropy 2018, 20, 193

Figure 2. Entropy diagram of the whole system Alice–Bob viewed from Alice’s perspective.
The negative conditional entropies indicate that the two components of the system must be entangled.

In conclusion, we formulate the observation that the binding property can be expressed in
the language of quantum information theory as S A ( B) = 0, forcing Alice to commit to a specific
key. A value of n for the same entropy guarantees a cheating strategy that completely circumvents
the binding requirement of bit commitment. Any intermediate value for S A ( B) (between 0 and n)
denotes some degree of entanglement between the quantum safe and Alice’s register, which ultimately
leads to some probability of cheating from her part and consequently, a partial realization of the
binding requirement.
It is interesting to note that, from an information-theoretic perspective, both basic properties of
bit commitment are formulated by quantifying the entropy of subsystem B (the quantum safe), albeit
from two different points of view:
SB ( B) = n (14)

ensures the hiding requirement by maximizing the entropy of the safe from Bob’s perspective, while

S A ( B) = 0 (15)

is the condition that prevents Alice from cheating the binding property by minimizing the entropy of
the safe from her point of view.

5. Something up Her Sleeve

Having established that entanglement is an essential ingredient in Alice’s cheating strategy,
let us now detail how she can take advantage of this important quantum resource in order to avoid
commitment and keep her options open until the decommit step.
The initial state of the system prepared by Alice has to take into consideration the two requirements:
using entanglement and keeping a full uncertainty on the key used to “lock” the quantum safe.
Consequently, Alice prepares an entangled superposition in which each term corresponds to one possible
key and the encoding corresponds to a commitment to 0:

1
|ψ0  = √
2n
∑ |ki  ⊗ | B0 (ki ). (16)
i

The term to the left of the tensor product describes Alice’s own quantum register, while the
term on the right characterizes the quantum state of the safe.

195

www.pdfgrip.com
Entropy 2018, 20, 193

This departure from the original protocol in which Alice is supposed to commit to a specific key
value is sometimes labeled as the “purified” version of the protocol, due to the fact that the global
state of the system is now a pure state. However, this modification which is essential for Alice’s
cheating strategy is transparent for Bob. He cannot distinguish between the original and purified
version just based on the state of the quantum safe, in both cases the reduced density matrix being the
same and equal to ρ0Bob from Equation (1). In a way, this is part of the reason why cheating is always
possible in a QBC protocol.
In the decommit phase, if Alice wants to keep her commitment to 0, she just measures her
quantum register (in the normal computational basis) and announces to Bob the values obtained as the
encoding key. On the other hand, if she wishes to change her commitment to 1, she will have to first
apply a transformation on her register that will rotate the overall state of the system to

1
|ψ1  = √
2n
∑ |ki  ⊗ | B1 (ki ) (17)
i

and then apply the measurement. Let us illustrate this point by showing what happens in the case of
the QBC protocol proposed by Bennett and Brassard in their seminal paper which launched the field of
quantum cryptography [5].
The BB84 QBC protocol adheres to the generic framework outlined at the beginning of Section 3
with the particularizations that base B0 represents rectilinear polarization of a photon and B1 represents
diagonal polarization. For the purpose of a theoretical analysis abstracted away from implementation
details, we will use the normal computational basis and the Hadamard basis as B0 and B1 , respectively.
The cheating strategy described in the original BB84 manuscript involves Alice preparing n Bell
states √1 |00 + √1 |11, keeping the first qubit from each entangled state as her own quantum register
2 2
and sending the second ones to Bob as the quantum safe. Formally, the initial state of the ensemble
Alice–Bob is therefore:
1 1
|ψ0  = ( √ |0 A 0B  + √ |1 A 1B )⊗n , (18)
2 2
where the labels A and B are used to identify the two parts of the system. A closer look to state |ψ0 
reveals that it is actually an entangled superposition of all 2n possible keys with their encodings in the
normal computational basis:
1
|ψ0  = √ ∑ |k iA  ⊗ |k iB , (19)
2n i
so Alice is actually preparing the initial state according to the purified version of the protocol as
explained above.
Since a Bell state always yields perfectly correlated outcomes when the two qubits are measured in
the same basis, regardless of what this basis is, Alice can claim commitment to either of the two
possible bit values in the decommit step. This is how cheating is explained in the BB84 paper. Yet again,
at a closer inspection, measuring a tensor product of Bell states either in the normal computational
basis or in the Hadamard basis conforms exactly to the general cheating procedure described above.
If Alice wishes to claim a commitment to 0, she simply measures her register (subsystem A) in
the normal computational basis, collapsing the superposition of all possible keys to a specific key k i .
From Equation (19) it is obvious that when Bob measures his subsystem (the safe) also in the normal
basis, he will obtain the same outcome k i due to the entanglement present in state |ψ0 .
According to our previously described cheating procedure, if Alice wants to claim a commitment
to 1, she has to first rotate the state of the system from |ψ0  to

1
|ψ1  = √
2n
∑ | k i  ⊗ H ⊗ n | k i , (20)
i

196

www.pdfgrip.com
Entropy 2018, 20, 193

by applying a quantum transformation on her register. After that, she can measure the register in
the normal computational basis and whatever outcome k i she obtains, it will coincide with the
outcome obtained by Bob after a measurement in the Hadamard basis on his side. We claim that the
transformation that Alice has to operate on her quantum register before measurement is the Hadamard
transform. In other words, the Hadamard measurement invoked in the BB84 paper can be seen as
a Hadamard transform followed by a measurement in the normal computational basis.
We can now verify formally that the Hadamard transform does the cheating trick by proving the
following equality:
( H ⊗n ⊗ I ⊗n )|ψ0  = |ψ1 . (21)

One way to prove this is by showing that the dot product between the vectors on the left-hand
side and the right-hand side of the above equality is 1:

1 1
(|ψ1 , ( H ⊗n ⊗ I ⊗n )|ψ0 ) = (∑ √ |k i  ⊗ H ⊗n |k i , ∑ √ H ⊗n |k j  ⊗ |k j )
i 2n j 2n
1
=∑ k | H ⊗n |k j ( H ⊗n |k i )† |k j  (22)
ij
2n i
2n 2n
1
= ∑(k i | H ⊗n |k j )2 = ∑ ∑ 22n = 1.
ij i =1 j =1

Therefore, if Alice wants to change her commitment (from 0 to 1), she just has to apply H ⊗n on
her quantum register and measure it in the normal computational basis or equivalently, measure
her register in the Hadamard basis. In general, the transformation that will rotate |ψ0  into |ψ1 
depends on the particulars of the respective protocol, but it will always exist as guaranteed by the
Schmidt decomposition theorem.

6. Extensions to Protocols Initiated by Bob

We have seen that, if the entropy of the quantum safe (as it appears to both participants) is
maximal and equal to the length of the key used by Alice, then the hiding property is guaranteed,
but Alice can cheat the binding property with certainty, taking advantage of her entanglement with
the quantum safe. At the first glance, this may appear as a consequence of the fact that Alice is the one
initiating the protocol and having full control over how she prepares the quantum state of the system
acting as the safe. In other words, she has all the cards in her hand and this may be perceived as an
unfair advantage in realizing an unconditionally secure quantum bit commitment protocol.
Therefore, in this section, we extend the discussion to protocols in which the procedure is
initiated by Bob, in the hope to deny Alice any opportunity of acting dishonestly. The main idea
of the framework we take into consideration here is that Bob should have a choice in preparing the
initial state of the quantum safe, prior to Alice encoding her commitment into the safe through the
use of a key. Consequently, the commit phase of the protocol should consist of two steps, the first
performed by Bob and the second one by Alice:

1. Bob chooses one of m initial states |φj , j = 1, ..., m for the quantum system acting as the safe.
He then sends the n qubits composing the quantum system over to Alice through a quantum channel.
2. Upon receiving the quantum safe, Alice chooses a key k i , i = 1, ..., p and applies a unitary
transformation U0 (k i ) or U1 (k i ) on the qubits composing the safe, depending on whether she
wants to commit to 0 or to 1. Afterwards, she sends the qubits back to Bob through the same
communication channel and waits for the decommit phase.

In an effort to keep our model as general as possible, we allow different values for the size of the
safe (n qubits), number of possible initial states (m) and number of different quantum transformations
(p) that Alice can apply on the system prepared by Bob for a given commitment. The latter also

197

www.pdfgrip.com
Entropy 2018, 20, 193

coincides with the number of keys k i that Alice can choose from to encode in the quantum safe.
Naturally, some (or all) of these three variables can actually be related to each other, depending on the
particular characteristics of a specific protocol.
Since the choice of j is intended as a protection against cheating attempts by Alice, states |φj ,
j = 1, ..., m should be chosen such that they cannot be reliably distinguished. In one possible instance of
the protocol, Bob could prepare the initial state of the safe by encoding one of the 2n bitstrings of length
n in one of two complementary bases. This means that Bob is choosing the initial state from a set with
2 × 2n elements, thus making m equal to 2n+1 .
Similarly, the number of keys p Alice is encoding in the quantum safe may also equal 2n . Each key
may again represent a length n bitstring such that

U0 (k i ) = U0 (bn−1 ) ⊗ U0 (bn−2 ) ⊗ · · · ⊗ U0 (b0 ), (23)

where U0 (bi ) is either the identity transformation, if bi = 0, or a certain single-qubit gate G, if bi = 1.

Assuming that Alice’s strategy is to avoid commitment and encode all possible keys in quantum
parallel, the entropies of the three subsystems (from the point of view of Alice) are as follows:

S( Bob) = log m = n + 1, (24)

S( Alice) = log p = n, (25)
S(Sa f e) = n. (26)

There are n + 1 bits of uncertainty on Bob’s side, corresponding to the 2n+1 choices he has in
preparing the initial state of the safe. Similarly, Alice’s quantum register (entangled with the quantum
safe) is characterized by n bits of uncertainty, reflecting the 2n possible keys Alice can encode in the
quantum safe. Finally, the quantum safe itself appears to be in a fully mixed state following all possible
initial state preparations |φj , j = 1, ..., 2n+1 and remains in this state after Alice encodes all possible
keys k i , i = 1, ..., 2n in quantum parallel. Consequently, the entropy of the quantum safe is maximal
and equal to the number of qubits n composing the safe.
The entropy of the global system Alice–Bob-Safe is also n + 1, since the state of the entire system
(from Alice’s point of view) consists of a different entanglement for each possible initial state |φj :

2n +1 n2 n
2
1 1 1
∑ ( ∑ √ |k i U0 (k i )|φj )( ∑ √ φj |U0† (k i )k i |).
global
ρ0 = (27)
j =1 2n +1 i =1 2 n
i =1 2n

For a fixed initial state prepared by Bob, the ensemble Alice-Safe is described by an entangled
state (pure state) with zero entropy. Consequently, the Schmidt decomposition theorem dictates that
both subsystems have equal entropy:
n
S(Sa f e| Bob) = S( Alice| Bob) = . (28)
2
The full entropy diagram of the whole system and its components, as it appears from Alice’s
perspective, is depicted in Figure 3.

198

www.pdfgrip.com
Entropy 2018, 20, 193

Figure 3. Entropy diagram of the whole system Alice–Bob-Safe viewed from Alice’s perspective. The n
bits of uncertainty originally encapsulated in the safe by Bob are split between Alice’s quantum register
and the safe through entanglement.

When Bob sends to Alice the n qubits representing the quantum safe, the entropy of Bob’s
subsystem ((n + 1) bits) can be decomposed into n bits of uncertainty characterizing the quantum safe
handed over to Alice and one bit left for Bob’s own register keeping track of the actual initial state
prepared. Subsequently, through the entanglement generated by Alice, the maximal entropy of the
quantum safe is spread uniformly among the two entangled parties: Alice’s subsystem (quantum
register keeping track of the actual key encoded) and the safe. Consequently, the mutual information
between Bob and Alice, respectively, Bob and the safe is

n
S( Alice : Bob) = S(Sa f e : Bob) = , (29)
2
while the conditional entropy S( Bob| Alice, Sa f e) remains 1. We also notice in Figure 3 the negative
conditional entropies:

n
S( Alice| Bob, Sa f e) = S(Sa f e| Alice, Bob) = − , (30)
2
revealing the entanglement between Alice’s own quantum register and the quantum safe, entanglement
which ultimately allows Alice to steer the state of the system towards a commitment to 1 in the
decommit phase, if desired. However, in order for Alice to be able to cheat the binding property,
the hiding property must be enforced. In terms of the framework considered in this section, the protocol
is hiding if, no matter what its initial state was, the safe looks identical to Bob at the end of the commit
phase for both possible commit values:

2n
1
ρ0Bob = ∑ 2n U0 (ki )|φj φj |U0† (ki ) =
i =1
2n
1
= ∑ 2n U1 (ki )|φj φj |U1† (ki ) = ρ1Bob , ∀ j = 1, ..., 2n+1 . (31)
i =1

199

www.pdfgrip.com
Entropy 2018, 20, 193

In this new scenario in which Bob is the one initiating the protocol, the hiding property is not
easy to achieve, since it has to hold for every possible initial state |φj . However, one straightforward
way to ensure it, is to choose as the single-qubit gate G a Hermitian operator (like Hadamard,
for example) and then set

U1 (k i ) = G ⊗n U0 (k i ), ∀i = 1, ..., 2n . (32)

In this way, when Alice wants to change her commitment to 1 just before the decommit phase,
she just needs to flip (apply the negation operator X on) all qubits in her quantum register.
For example, consider a three-qubit safe initially in state H |0 ⊗ H |1 ⊗ H |0. Moreover, assume
that the key Alice encodes in the state of this quantum safe is 110 by applying the Hadamard gate on
the first two qubits of the safe. The state of the quantum safe will consequently change to |0 ⊗ |1 ⊗
H |0. Now, this state is the result of following the procedure corresponding to a commitment to 0.
However, we notice that the same state of the quantum safe could have resulted as a consequence of
committing to 1, if the encoded key is not 110, but its opposite 001. Furthermore, the same technique
of complementing the content of her own quantum register allows Alice to change her commitment
regardless of the initial state |φj  and/or the encoded key k i .
Once again, we can formulate the observation that the entropy of the quantum safe appears
maximized (equal to n) to both observers. This guarantees the hiding property, but leaves the door
wide open for Alice to cheat the binding property of bit commitment, despite the fact that in our new
setting Alice is forced to work with a quantum safe whose initial state is not known exactly.
In the setting considered above, Alice is still able to elude the binding requirement at the
cost of choosing an operator G that is Hermitian. This constraint is a counterpart measure to the
advantage that Bob now has through the choice of the initial state for the quantum safe. Certainly,
there are other protocol frameworks that can be designed to annihilate Bob’s advantage of trying
to prepare an initial quantum state that would allow him to distinguish between the two possible
commitments. We describe in the following an alternative protocol that adopts a different approach in
enforcing the hiding property for Bob. Subsequently, we analyze the cheating strategies available to
Alice and compare the new scenario with the framework we have just investigated, especially from
the point of view of entropies and other information-theoretic measures characterizing the various
subsystems involved.
In an effort to keep our QBC protocol as simple as possible, our next scenario considers a quantum
safe consisting of only one qubit. Bob starts off the protocol by preparing four qubits, one in each of the
states |0, |1, H |0 and H |1. He sends a random permutation of these four qubits to Alice, who then
randomly selects one as the quantum safe. For a commitment to 0, she sends the unaltered qubit
back to Bob. A commitment to 1 requires the application of the Hadamard gate before handing Bob
the single-qubit safe back.
As in the previous case, we will focus on the purification of the protocol in which Alice acts
as if committing to 0, but selects all four qubits received from Bob in quantum parallel. This is done
through an entanglement between Alice’s own quantum register (who plays the role of a pointer to
the actual qubit selected) and the quantum safe. In order to distinguish among the four possibilities,
the quantum register in Alice’s possession must span two qubits. It is easy to check that this protocol is
hiding, since, for Bob, the quantum safe appears to be in a fully mixed state, regardless of what
particular permutation he is initially sending over to Alice:

Sa f e 1 1 1 1 I
ρ0 = |00| + |11| + H |00| H + H |11| H =
4 4 4 4 2 (33)
1 1 1 1 Sa f e
= H |00| H + H |11| H + |00| + |11| = ρ1 .
4 4 4 4
However, despite the fact that the hiding requirement is realized, Alice can no longer cheat
the binding requirement, since in this setting, she needs a different transformation T to rotate |ψ0 

200

www.pdfgrip.com
Entropy 2018, 20, 193

(global quantum state corresponding to a commitment to 0) to |ψ1  (quantum state describing the
status of the entire system in the case of a commitment to 1), for each possible permutation prepared by
Bob. For instance, if the four-qubit sequence received from Bob is |0 ⊗ |1 ⊗ H |0 ⊗ H |1, then the
transformation Alice has to apply on her quantum register is:
⎡ ⎤
0 0 1 0
⎢ 0 ⎥
⎢ 0 0 1 ⎥
T1 = ⎢ ⎥. (34)
⎣ 1 0 0 0 ⎦
0 1 0 0
However, if Bob prepares and sends the sequence |0 ⊗ |1 ⊗ H |1 ⊗ H |0, then it is the following
transformation that does the job of rotating |ψ0  to |ψ1  just from Alice’s subsystem:
⎡ ⎤
0 0 0 1
⎢ 0 ⎥
⎢ 0 1 0 ⎥
T2 = ⎢ ⎥. (35)
⎣ 0 1 0 0 ⎦
1 0 0 0
Therefore, unlike in the previous instance of QBC, here, there is no unique transformation that can
help Alice elude the binding requirement, even though in both protocols she makes use of entanglement
in order to apply all her options in quantum parallel. Let us now take a look at the entropy diagram for
our current protocol, depicted in Figure 4, and see how this change is reflected in the entropies of the
three subsystems involved.

Figure 4. Entropy diagram at the end of the commit phase for an alternative Bob-initiated QBC
(quantum bit commitment) protocol, viewed from Alice’s perspective. The amount of uncertainty left
in Bob’s subsystem and not shared with the other subsystems prevents Alice from cheating.

The degree of uncertainty in Bob’s subsystem is log 24 (approximately 4.6), corresponding to the
24 degrees of freedom (number of permutations) Bob has in preparing the initial sequence given to
Alice. The same value of the entropy also characterizes the entire system, because Alice sees the
global state of the system as a mixed state having a different entanglement for each of the 24 possible
permutations of the four qubits prepared by Bob.
Assuming that Alice knows exactly the quantum state of each qubit received, the ensemble
Alice-Safe is characterized by the exact entangled state corresponding to that particular permutation.
This means that the conditional entropy S( Alice, Sa f e| Bob) = 0 and following Schmidt’s decomposition

201

www.pdfgrip.com
Entropy 2018, 20, 193

theorem, each side would entail an uncertainty of one bit. As always, the negative conditional entropies
S( Alice| Bob, Sa f e) = S(Sa f e| Alice, Bob) = −1 witness the use of entanglement by Alice.
Where the entropy diagrams for the two protocols analyzed in this section differ from one another,
is the amount of information leaked from Bob towards the other subsystems. In the first case, from the
n + 1 bits of uncertainty characterizing Bob’s subsystem, almost all of it (n bits) gets evenly distributed
among Alice and the Safe: S( Alice : Bob) = S(Sa f e : Bob) = n2 . Only one bit of information remains
unshared with Alice and the Safe: S( Bob| Alice, Sa f e) = 1. In the second protocol, on the other hand,
out of the 4.6 bits of information contained in Bob’s subsystem, 2.4 (that is, more than 50%) remains in
Bob’s hands only, and does not leak towards the other two subsystems. This is due to the fact that the
size of the quantum safe (one qubit) is much smaller than the size of the vector space spanned by Bob’s
subsystem (five qubits, in order to accommodate 24 possible initial states). To make a comparison
again, there is no mutual information between the safe and Bob’s subsystem in the second scenario,
while in the first the two subsystems share n2 bits of information.
The conditional entropy S( Bob| Alice, Sa f e) measures the amount of uncertainty left in Bob’s
subsystem once we acquire all the information in the Alice-Safe ensemble.The magnitude of this
entropy seems to be the key in preventing Alice from cheating the binding requirement. Without the
knowledge of the particular permutation prepared by Bob (most of which remains in his hands),
Alice does not have enough information to rotate the state of the global system towards a commitment to 1.
Does this mean that the particular protocol we have just investigated satisfies both crucial security
requirements of bit commitment: it is at the same time binding and hiding? The answer is no,
as this would also constitute a counterexample to Mayers’ impossibility theorem on realizing an
unconditionally secure quantum bit commitment protocol [1]. The catch is that Bob can also enlist the
help of entanglement to his own advantage. He too could not commit to a particular initial state |φj 
and instead prepare the quantum safe in a superposition of all possible initial states, each entangled
with a corresponding label (or pointer) in his own ancilla qubits. This ensemble Bob-Safe would then
be in a pure state:
24
Bob−Sa f e 1
|ψinit  = ∑ √ Pj (|0 ⊗ |1 ⊗ H |0 ⊗ H |1) ⊗ |l j , (36)
j =1 24

where each Pj represents one of the 24 possible permutation operators acting on four qubits.
After receiving these four qubits composing the safe from Bob, Alice selects all of them in superposition,
effectively creating a three-party entanglement among the three subsystems:

24 4
Alice−Sa f e− Bob 1
|ψ0 = √
4 6
∑ ∑ |ki  ⊗ U0 (ki ) Pj (|0 ⊗ |1 ⊗ H |0 ⊗ H |1) ⊗ |l j . (37)
j =1 i =1

In the equation above, U0 (k i ), for i = 0, 1, 2, 3 is to be interpreted as the selection operator acting on

the four-qubit sequence received from Bob whose effect is to select the qubit with index i. It is not
difficult to verify that by attaching a label l j to each possible permutation Pj , the hiding requirement is
Alice−Sa f e− Bob Alice−Sa f e− Bob
no longer satisfied, as the reduced density matrix obtained from |ψ0 ψ0 | by
tracing out the two qubits from Alice’s own quantum register is different from the similar reduced
Alice−Sa f e− Bob Alice−Sa f e− Bob
density matrix obtained from |ψ1 ψ1 |. Consequently, Bob would be able to
distinguish (with a certain probability p > 1/2) between a commitment to zero and a commitment to
one, which ultimately translates into a non-zero probability of detecting a dishonest Alice.

7. Conclusions
The effort that many quantum cryptographers put into developing an unconditionally secure
QBC protocol is motivated by (at least) two important factors: the intuition that the success of quantum
key distribution can be replicated for bit commitment as well, and perhaps more importantly, the key
importance of QBC as a building block in constructing more complex cryptographic applications.

202

www.pdfgrip.com
Entropy 2018, 20, 193

These two factors may, at least in part, explain why people kept trying to devise new ways of achieving
a protocol with the desired security properties, even after this was proved to be impossible. Another
factor may be the relative difficulty in grasping the essence of this impossibility result, especially since
other quantum realizations of cryptographic protocols proved superior to their classical counterparts.
In this paper, we adopted a novel approach in trying to get new insight into the problem. With the
help of the tools provided by quantum information theory, we have investigated the conditions
under which the two fundamental properties of bit commitment, the hiding property and the binding
property cannot be reconciled. Our investigation has revealed that both properties can be expressed in
terms of the entropy of the quantum system playing the role of the safe into which Alice is supposed
to lock her commitment.
The hiding property dictates that the quantum safe must appear to Bob as harboring a certain
level of uncertainty. The entropy of the quantum safe is usually maximized (especially when Alice is
the one preparing the initial state of the safe) in order to hide the commit value behind the fully mixed
state of the quantum system acting as the safe. When dealing with a dishonest Alice, the quantum
safe actually achieves maximum entropy with respect to both observers, as Alice needs to keep all
her options open. We have also proved that cheating requires enlisting the help of entanglement,
a necessary condition without which any cheating strategy is impossible, whether we are talking about
breaking the binding requirement by Alice or the hiding requirement by Bob.
Naturally, the ability to circumvent one or the other of the security requirements of bit commitment
ultimately depends on the particular details and structure of the protocol at hand. However,
our investigation, performed with the specific tools of quantum information theory, seems to point to
two important factors influencing the ability to adopt a successful cheating strategy. On one hand, it is
easier for any of the two participants to the protocol to mount an effective cheating strategy, if they
are the ones initiating the protocol and preparing the initial state of the safe. This gives them the
opportunity to avoid committing to a particular initial state and prepare a superposition of all possible
initial states, each entangled with a pointer (or label) in their own ancilla qubits. On the other hand,
devising a successful cheating scheme or successfully annihilate a cheating attempt by the adversary,
also depends on how much information they manage to keep hidden inside their own subsystem and
not share it (through the quantum safe) with the other participant.

Author Contributions: Marius Nagy and Naya Nagy analyzed the security properties of quantum bit commitment
protocols using entropic measures. Marius Nagy wrote the paper.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Mayers, D. Unconditionally secure quantum bit commitment is impossible. In Proceedings of the Fourth
Workshop on Physics and Computation—PhysComp ’96, Boston, MA, USA, 22–24 November 1996.
2. Nielsen, M.A.; Chuang, I.L. Quantum Computation and Quantum Information; Cambridge University Press:
Cambridge, UK, 2000.
3. Mayers, D. Unconditionally secure quantum bit commitment is impossible. Phys. Rev. Lett. 1997, 78, 3414–3417.
4. Lo, H.K.; Chau, H. Is quantum bit commitment really possible? Phys. Rev. Lett. 1997, 78, 3410–3413.
5. Bennett, C.H.; Brassard, G. Quantum cryptography: Public key distribution and coin tossing. In Proceedings
of the IEEE International Conference on Computers, Systems and Signal Processing, Bangalore, India,
9–12 December 1984; pp. 175–179.
6. Brown, J. The Quest for the Quantum Computer; Simon & Schuster: New York, NY, USA, 2001.
7. Biham, E.; Boyer, M.; Boykin, P.O.; Mor, T.; Roychowdhury, V. A proof of the security of quantum key
distribution. J. Cryptol. 2006, 19, 381–439.
8. Crépeau, C.; Kilian, J. Achieving oblivious transfer using weakened security assumptions. In Proceedings of the
29th Annual IEEE Symposium on Foundations of Computer Science, White Plains, NY, USA, 22–26 October 1988;
pp. 42–52.

203

www.pdfgrip.com
Entropy 2018, 20, 193

9. Brassard, G.; Crépeau, C.; Jozsa, R.; Langlois, D. A quantum bit commitment scheme provably unbreakable
by both parties. In Proceedings of the 34th Annual IEEE Symposium on Foundations of Computer Science,
Palo Alto, CA, USA, 3–5 November 1993; pp. 362–371.
10. Brassard, G.; Crépeau, C. 25 years of quantum cryptography. SIGACT News 1996, 27, 13–24.
11. Crépeau, C. What is going on with quantum bit commitment? In Proceedings of the Pragocrypt ’96:
1st International Conference on the Theory and Applications of Cryptology, Prague, Czech Republic,
30 September–3 October 1996.
12. Kent, A. Permanently secure quantum bit commitment protocol from a temporary computation bound.
arXiv 1997, arXiv:quant-ph/9712002.
13. Spekkens, R.W.; Rudolph, T. Degrees of concealment and bindingness in quantum bit commitment protocols.
Phys. Rev. A 2001, 65, 012310.
14. Ng, N.H.Y.; Joshi, S.K.; Ming, C.C.; Kurtsiefer, C.; Wehner, S. Experimental implementation of bit commitment
in the noisy-storage model. Nat. Commun. 2012, 3, 1326.
15. Kent, A. Unconditionally secure bit commitment with flying qudits. New J. Phys. 2011, 13, 113015–113029.
16. Lunghi, T.; Kaniewski, J.; Bussières, F.; Houlmann, R.; Tomamichel, M.; Kent, A.; Gisin, N.; Wehner, S.;
Zbinden, H. Experimental bit commitment based on quantum communication and special relativity.
Phys. Rev. Lett. 2013, 111, 180504.
17. Adlam, E.; Kent, A. Device-independent relativistic quantum bit commitment. Phys. Rev. A 2015, 92, 022315.
18. Hardy, L.; Kent, A. Cheat sensitive quantum bit commitment. Phys. Rev. Lett. 2004, 92, 157901.
19. Buhrman, H.; Christandl, M.; Hayden, P.; Lo, H.K.; Wehner, S. Possibility, impossibility, and cheat sensitivity
of quantum-bit string commitment. Phys. Rev. A 2008, 78, 022316.
20. Shimizu, K.; Fukasaka, H.; Tamaki, K.; Imoto, N. Cheat-sensitive commitment of a classical bit coded in
a block of m × n round-trip qubits. Phys. Rev. A 2011, 84, 022308.
21. He, G.P. Comment on “Cheat-sensitive commitment of a classical bit coded in a block of m × n round-trip
qubits”. Phys. Rev. A 2014, 89, 056301.
22. Li, Y.B.; Wen, Q.Y.; Li, Z.C.; Qin, S.J.; Yang, Y.T. Cheat sensitive quantum bit commitment via pre- and
post-selected quantum states. Quantum Inf. Process. 2014, 13, 141–149.
23. Li, Y.B.; Xu, S.W.; Huang, W.; Wan, Z.J. Quantum bit commitment with cheat sensitive binding and
approximate sealing. J. Phys. A Math. Theor. 2015, 48, 135302.
24. Rovelli, C. Relational quantum mechanics. Int. J. Theor. Phys. 1996, 35, 1637–1678.
25. Grinbaum, A. The Significance of Information in Quantum Theory. Ph.D. Thesis, Ecole Polytechnique, Paris,
France, 2004.

c 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

204

www.pdfgrip.com
 entropy
Review
Developments in Quantum Probability and the
Copenhagen Approach
Gregg Jaeger 1,2 ID

1 Quantum Communication and Measurement Laboratory, Department of Electrical and Computer

Engineering, Boston University, Boston, MA 02215, USA; jaeger@bu.edu
2 Division of Natural Science and Mathematics, Boston University, Boston, MA 02215, USA

Received: 2 May 2018; Accepted: 28 May 2018; Published: 31 May 2018

Abstract: In the Copenhagen approach to quantum mechanics as characterized by Heisenberg,

probabilities relate to the statistics of measurement outcomes on ensembles of systems and
to individual measurement events via the actualization of quantum potentiality. Here, brief
summaries are given of a series of key results of different sorts that have been obtained since
the final elements of the Copenhagen interpretation were offered and it was explicitly named so by
Heisenberg—in particular, results from the investigation of the behavior of quantum probability since
that time, the mid-1950s. This review shows that these developments have increased the value to
physics of notions characterizing the approach which were previously either less precise or mainly
symbolic in character, including complementarity, indeterminism, and unsharpness.

Keywords: quantum probability; potentiality; complementarity; uncertainty relations; Copenhagen

interpretation; indefiniteness; indeterminism; causation; randomness

1. Introduction
The orthodox approach to quantum theory emerged primarily from interactions in Copenhagen
and elsewhere from the work of Niels Bohr, Werner Heisenberg, and Wolfgang Pauli, depending also
on contributions of Max Born, and was largely set out by 1927, cf. [1–4]. After various criticisms of
the initial form, Bohr focused more strongly on complementarity in the 1930s, and Heisenberg—in
a strong response of 1955 in which the basis of the approach can be considered to have been
essentially finalized—added the new element of actualization of potentiality to its approach to quantum
probability [5]. Since then, much attention has been paid to newly emerging alternative treatments of
quantum physics, such as Bohmian mechanics and collapse-free (e.g., many-worlds) mechanics, that
use quantum probability and the quantum formalism. Indeed, with a few exceptions (e.g., [2,6–8] and
others mentioned below) relatively little attention has been paid to the implications of new research
for the more orthodox, Copenhagen approach. Here, a non-exhaustive but wide-ranging series of
theoretical results, several directly related to experiment, obtained since the work of Heisenberg related
to quantum probability, is discussed, which, by articulating more precisely and better clarifying the
application of its basic notions, including complementarity, uncertainty, and indeterminacy, which were
before either less precise or even merely symbolic, further demonstrates the value of this approach.
Beyond the basic notion that the probability of any observed future physical event in a system can
be found via the quantum state using the Born rule given the results of a complete set of measurements
on it, the character of quantum probabilities on this Copenhagen approach is discussed in the
next section; experimentally verified results demonstrating new quantum complementarities are
considered in Section 3; theoretical developments involving unsharpness and quantum measurement
are considered in Section 4; novel explications of indeterminism and randomness are considered
in Section 5.

Entropy 2018, 20, 420; doi:10.3390/e20060420 205 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2018, 20, 420

2. Probability in the Copenhagen Approach

The Copenhagen approach to quantum theory generally gives primacy to measured
phenomena—with any results regarding the measured system being given in relation to the entire
experimental situation in which each arises and with the records of measuring devices being classical
describable—without being positivist and remaining essentially realist. Although Bohr’s notion of
complementarity was the greatest influence early in the development of the approach—as noted, for
example, by Jan Faye [9] and Arkady Plotnitsky [10,11]—it was later supplanted in a development
toward more precise and mathematically advanced treatments of the effect, as exhibited in Section 3
(also cf. [12])—and even explicitly extended to situations involving entanglement as, arguably, thought
by Bohr to be the case, as Don Howard has argued [13]. The Copenhagen approach explicitly named
as such is that circumscribed by Werner Heisenberg [5] in which measurement corresponds to the
actualization of a potential physical situation where a single value appears from among a set of possible
values that were not certain (cf. [14,15] for analyses of the fundamentals of this version, specifically
considered here). The essential mathematical formalism of non-relativistic QM emerged with work
of Paul Dirac and John von Neumann, with Hilbert-space as the space of individual system states
(cf. [1–3] and Chapter. 7 in [16]), forming the context for its later extension, discussed in detail below,
cf. [2,14,17].
A succinct overview of the role of probability in the Copenhagen interpretation was given by
Heisenberg, who gave the interpretation its name: “. . . the probability function does not in itself
represent a course of events in the course of time. It represents a tendency for events and our
knowledge of events. The probability function can be connected with reality only if one essential
condition is fulfilled: if a new measurement is made to determine a certain property of the system. Only
then does the probability function allow us to calculate the probable result of the new measurement”
(pp. 46–47, [18] ; cf. [15]).

2.1. Quantum States and Probability

An (O, S, p) formulation of general physical theory serves as a basic formal framework for the
non-relativistic theory of QM [19,20]: To each physical system, one can associate with the set of all
associated sharp observables (Hermitian self-adjoint operators) O and the set of its states S, a function
p: O × S × B(R) → [0, 1], where B(R) is the set of all Borel subsets of R, cf. [21]—of values appearing
in measurements, cf. [2]. Restricting ourselves specifically to the Hilbert-space formulation of quantum
mechanics, each statistical operator ρ is decomposable into a non-trivial weighted sum of quantum
pure states represented by normalized vectors |ψi ∈ H (cf. [5,18]) as ρi = |ψi ψi | that have no further
state decomposition; each statistical operator ρ also induces an expectation functional A → tr(ρA) on
L(H), the space of linear operators on the Hilbert space H.
The probability p in the Copenhagen approach involves an explicit distinction between objective
and subjective aspects of physical states describable in this formalism and compares with that in
classical mechanics as follows. In classical mechanics, when needed at all, probability is used only in
situations where a detailed knowledge of the system is lacking, i.e., for statistical mechanics; in the
quantum context, the subjective aspect of probability also appears in such situations, which involve
state mixtures (Gemische, cf. [22] p. 9, [23] p. X) that are representable as statistical operators ρ but
not as pure ones. However, probability in QM on this approach is found also in the individual states
(Zustände, cf. [23]) as the objective aspect, representable as vectors |ψ in H [23,24]. This objective
contribution to the quantum probability of a measurement outcome is provided specifically by the state
|ψ = ∑i |ψi  via its complex amplitudes {ci }, now known as probability amplitudes, as their squared
magnitudes pi = |ci |2 .
The quantities { pi } are understood as the probabilities of the measured system to be found to
possess actual respective values of its physical properties according to the rule of Born that is elemental
to the Copenhagen approach [25,26]. The measured value of a property is considered definite (actual),
as opposed to indefinite (potential) [5], as discussed in great detail in [14,15]; a dynamical property

206

www.pdfgrip.com
Entropy 2018, 20, 420

of a quantum system S becomes actual with probability pi upon precise measurement wherein the
measuring apparatus, A, must be in contact with the greater physical environment (the “the rest of the
world") and be classically describable or macroscopic (cf. [27] for a discussion of this notion and its
use in the Copenhagen interpretation and elsewhere more recently). In general, some member from
a set of possible values must occur in measurement, but the specific actual value measured appears
randomly [28], as discussed in Sections 4 and 5 here.

2.2. Quantum Indeterminacy

Another important element in the initial success of the Copenhagen approach is that it articulates
well the behavior of joint probabilities appearing in the “uncertainty relations,” a manifestation of
complementarity. Indeed, in Born’s view, “the factor that contributed [most]... to the speedy acceptance
of [the Copenhagen] interpretation of the ψ-function was a paper by Heisenberg [29] that contained his
celebrated uncertainty relations” [30]. The result of von Neumann that (sharp) observable quantities
are simultaneously measurable if and only if they commute with one another and if and only if they are
functions of a single observable later came to play a strong role in this understanding of such relations.
It has also been shown by Pekka Lahti [31] that Heisenberg’s joint indeterminacy hypothesis—the
“uncertainty principle” [29], later generalized by H. P. Robertson, [32], providing corresponding
“uncertainty” relations discussed in Section 2.4 below—together with an axiomatic formulation of
complementarity, when considered within the (O, S, p) framework formalized by George Mackey
[19] and M. J. Maczyński [20], rigorously imply the existence of pairs of observables that cannot be
jointly sharply measured. The known set of such relations has recently expanded, as shown in the
next section.
The indeterminacy principle contrasts with the determinacy principle, that the magnitude of each
continuous quantity is determined by a real number, as is typically assumed in classical mechanics
(cf. Michael Dummett’s discussion of this principle in [33]). Note that the indeterminacy hypothesis
is a statement about the associated indeterminacy of incompatible observables not algebraically
commuting with each other (see Section 4 below), rather than the epistemic uncertainty regarding
an independent quantity per se. It is, therefore, a statement to the effect that the associated properties,
jointly considered, are objectively indefinite, as emphasized by Abner Shimony [34]; (cf. [35] Section 2.1).
It is a consequence of this indefiniteness that their measured values are also not precisely predictable,
that is, random in the sense explained in Section 5 below.

2.3. Quantum Potentia and Probability

In the Copenhagen approach, according to Heisenberg, the objective probability p, given by the
Born rule, relates to “statements about possibilities or better tendencies (‘potentia’...)” of the system
itself later to have certain actual values of measured properties [18] (p. 53). The subjective content of
these probabilities is “negligible” in the pure case, i.e., where trρ2 = 1, exactly when it is a projector,
which suffices for the maximal specification of the system’s actual properties; in this case, the elements
of the set of quantum probabilities { pi (O)} for the outcomes {oi } in a measurement of the observable
(i.e., Hermitian operator) O are

pi (O) = tr( Pi (O)ρ) = ψ| Pi (O)|ψ = ψ|φi φi |ψ (1)

which are the squared magnitudes of the corresponding complex-valued state-vector amplitudes
{ci = φi |ψ}, that is, of the components of |ψ in the eigenbasis {|φi } for O—the exclusivity of
this form being demonstrated rigorously by Andrew Gleason [21]; see Section 4.1 below. (Here, we
consider for simplicity the case of discrete properties; analogous relations hold in the continuous case.)
In particular, the complex probability amplitude ck corresponds to the potentia for actualization of
the specific property value ok upon the measurement of O, with pk (O) = |ck |2 , something discussed
further in Section 4.3 below—cf. [14]. (Very recently, the notion of quantum potentia in a sense of res

207

www.pdfgrip.com
Entropy 2018, 20, 420

potentia related to that of Heisenberg has been used by Ruth Kastner et al. to offer a novel analysis of
quantum measurement, in combination with res extensa, purely physical substance, as “implicative
constituents of every quantum measurement event” [36].)
Recall that the novel characteristic of the quantum probability first to be discovered (n.b.
a mathematically precise treatment of its novelty in general terms has been given by Luigi Accardi
in [37]), which motivates the continual reconsideration of its relationship to traditional probability,
involves ostensibly disjoint events: An empirically measurable difference between the quantum
mechanical probability and the classical mechanical probability of a disjunction of such events is that
the associated probability is not, in general, additive in the quantum case. This is exhibited in the
appearance of a particle such as an electron at a spatial location x in the basic double-slit experiment
(cf. [38]) by passing first through a slit (Slit 1) and/or the other slit (Slit 2), which occurs with a
probability p12 ( x ) that is related to the probability densities of reaching x in the alternatives of either
first passing through Slit 1, p1 ( x ), or of first passing through Slit 2, p2 ( x ):

p12 ( x ) ∝ p1 ( x ) + p2 ( x ). (2)

This quantum probability is the magnitude squared of the sum c12 ( x ) = c1 ( x ) + c2 ( x ) of the complex
amplitudes {ci ( x )} of those alternatives, rather than a simple sum of the probabilities of the two
alternative situations consistent with the future event, here detection at x.
Thus, the quantum probabilities do not arise by direct calculation from, for example, prior
probabilities of particle detection as in classical mechanics as a sum such as p12 ( x ) ∝ p1 ( x ) + p2 ( x ) in
the situation of this double-slit experiment. (Another difference of quantum probability from classical
probability is found in joint probabilities is discussed in Section 3 below.)

2.4. Quantum Interference and Dispersion

The difference between the two sorts of probability, classical and quantum, is reflected clearly in
the corresponding probability density distribution in the detection plane in the double-slit experiment:
There is an additional modulated “interference term” arising because the ci ( x ) are complex-valued,
which precludes these probabilities from being given a straightforward Kolmogorovian representation
under a single probability measure (cf. [37] and [39] p. 125, [37] for detail regarding this), so that
 , -
p12 ( x ) ∝ p1 ( x ) + p2 ( x ) + p1 ( x ) p2 ( x ) cos θ2 ( x ) − θ1 ( x ) . (3)

In this situation, there is a range of possible values for the detected position, as well as of the
momentum of the system approaching it, that is, a certain dispersion of values due to its indeterminacy.
More generally, any observable given as an Hermitian operator A will have a dispersion
Dispρ A = ( A −  AI)2  =  A2  −  A2 , for any system in state ρ; the indeterminacy relation of
Heisenberg for momentum and position relevant in this experiment was generalized by H. P. Robertson
so as to apply two any two observables A and B [29,32]:

1
(ΔA)2 (ΔB)2  ≥
|[ A, B]|2 (4)
4

where the “uncertainty” of A for state ρ is ΔA ≡ Dispρ A. These relations are connected with
single-particle interferometric complementarities, i.e., between visibility and particle path, as shown
in the 1980s and 1990s; see the following section, Jaeger et al. [40] and references therein for more on
the relation to interferometry, and a recent analysis of Paul Busch and Christopher Shilladay [41] for
a detailed discussion of the various forms of complementarity.
Significant new developments regarding quantum probability that allow for the clearer explication
of these central aspects of the Copenhagen approach are discussed in the next two sections.

208

www.pdfgrip.com
Entropy 2018, 20, 420

3. Complementarity and Entanglement

The extraordinary behavior of quantum probabilities regarding compound systems due to
entanglement was brought to the forefront relatively early in the history of quantum mechanics (QM)
by Erwin Schrödinger [42,43] and remained the subject ongoing entirely theoretical discussions until
after John Bell produced his now-famous inequality, which was subsequently rendered experimentally
testable in a reformulation by John Clauser, Michael Horne, Abner Shimony, and Richard Holt
(CHSH) [44] and shown to be violated in an interferometric setting in the early 1980s by Alain
Aspect et al. [45].
An entanglement-related manifestation of complementarity involving joint probabilities was
noted later in the 1990s: When the two-particle interference visibility is unity, the one-particle
visibility is zero, and conversely as noted in the work of Marlan Scully, Berge Englert, and Herbert
Walther [6] and of Shimony, Horne, and Anton Zeilinger [46]. In the mid-1990s, it was shown that
the interferometric phenomena involved in the violation of CHSH inequalities obey a precise trade-off
relation, later experimentally verified in the 2000s [40,47,48], over a full range of different experimental
arrangements. This is also related to the fact that entanglement can be understood as an instance of the
uncertainty of quantum properties, cf. [41]. In particular, it was found by Jaeger and Shimony that there
is a general quantum interferometric complementarity relation between single-system interference
visibility, v1 , and compound-system interference visibility, v12 , for pairs of two-level systems, further
illustrating the surprising nature of quantum correlations exhibited in two-particle interference due to
the presence of entanglement [47], as first verified at the Boston University in 2001 [48].
This novel exhibition of complementarity can be understood concretely in terms of the washing
out of photon self-interference due to indeterminacy in the initial direction of individual particles
in a doubled discrete (Mach–Zehnder, MZ) two-“slit" arrangement with a source of particle pairs at
center simultaneously feeding two Mach–Zehnder interferometers, symmetrically oriented, with one
particle moving in one interferometer involving a generalized beamsplitter (transducer) at left and
similarly for the other particle moving in a second such interferometer at right (see [47] for figures).
Consider two particles A and B in this arrangement, with A taken to be that in beams 0 and/or 1,
and similarly for particle B (but indicated by primes below). Let each particle pair of the ensemble
involved be produced by the centrally located source in a possibly entangled two-particle pure state
|Θ = γ1 |0 A |0  B + γ2 |0 A |1  B + γ3 |1 A |0  B + γ4 |1 A |1  B ,with γi ∈ C such that |γ1 |2 + |γ2 |2 +
|γ3 |2 + |γ4 |2 = 1, |0 A and |1 A being basis vectors in the Hilbert space H A of the first particle
corresponding to the propagation of A to the left in beams 0 and 1 and |0  B and |1  B being similar
vectors in space H B of B moving to the right.
Let beams 0 and 1 be brought together at a transducer, TA (inducing a general unitary
transformation in the state space, not only a phase-shift+reflection/transmission as in a simple MZ
apparatus), feeding two output beams, an upper U and lower L beam in the MZ interferometer at
left, and let similar beams in the other interferometer to the right be brought together into another
transducer, TB (inducing a similar unitary transformation in the other particle’s state space), that feeds
two corresponding output beams U and L ; the joint, local operation of this pair of transducers is
described by the general pair of local unitary operations induced by them separately: T = TA ⊗ TB .
As these transducers TA and TB are varied, the probabilities P(UU ) of coincidence detection in beams
U and U , and P(UL ), P(UL ), and P( LL ), as well as single-detection probabilities P(U ), P( L), P(U ),
and P( L )—corresponding to particle coincidence detection and single detection rates, respectively, in
the output beams of the pair of interferometer—are modulated.
The corresponding visibilities of interference were found to obey trade-off relations, quantifying
their complementaries. As T is varied over the full range of parameters for the two general local

209

www.pdfgrip.com
Entropy 2018, 20, 420

unitary transformations involved, continuously altering the apparatus, the one-particle interferometric
fringe visibility Vi (i = A, B) is found from the maximum and minimum probabilities of detection:

[ P(Y )]max − [ P(Y )]min

Vi = , (5)
[ P(Y )]max + [ P(Y )]min

where Y = U, L. V12 , the two-particle interferometric visibility in the sense of variations of detection
probability as the T is changed, is similarly calculable from the probabilities P(YY ) of occupation of
the joint-paths YY , generalizing the case of the single paths Y. For example,

[ P̄(UU )]max − [ P̄(UU )]min

V12 = (6)
[ P̄(UU )]max + [ P̄(UU )]min

where P̄(UU ) = P(UU ) − P(U ) P(U ) + 14 represents nonaccidental coincidence probabilities; likewise
for the three other possible path pairs YY [47].
When the two systems, A and B, are entangled, one has the non-factoring joint probability

P(UU ) = P(U ) P(U ), (7)

as do the other joint probabilities P(UL ), P( LU ), and P( LL ); the extraordinarily highly correlated
behavior of particles A and B arises due to entanglement, and one finds that a strong complementarity
trade-off relation, taking the form of an equality [47], holds for all |Θ, namely,

2
V12 + VA2 = 1 ; V12
2
+ VB2 = 1 ; (8)

this was subsequently experimentally confirmed by Bahaa Saleh and associates at Boston
University [48,49]. This explicitly demonstrates precise quantum complementarity involving entangled
systems of the sort violating the Bell and CHSH inequalities.

4. Quantum Measurement
In the Copenhagen approach to quantum mechanics, “the behaviour of the measuring apparatus
must be capable of being registered as something actual. . . if the measuring apparatus is to be used
as a measuring instrument at all. . . the connection with the external world is. . . necessary” [5] (p. 27);
any fully quantum A and S alone would become entangled, something Don Howard has argued Bohr
was already noting in 1927 [13], and neither S nor A+S can be considered closed during measurement.
Indeed, in a fully quantum formal treatment of such a process, the apparatus would itself have to
be measured in order to provide an outcome, and so on. In this approach, a fully formal quantum
mechanical treatment of the measurement process utilizing a closed system description or without
the use of classical descriptions for at least some elements of the measurement process is considered
impossible—quantum mechanics is literally incapable of being used to account for all details of the
process (n.b.: the adjective sometimes used in relation to this is often translated as uncontrollable in
English, but is better translated as unaccountable-for.) Moreover, the objective, probabilistic aspect
of quantum state evolution invoked upon actualization is considered irreducible to any amount of
ignorance that might be removed using the theory alone, as mentioned in Section 2.
In the Copenhagen approach, any system S, as well as the joint system of S and any other quantum
piece of apparatus A thought of as distinct from A, is an open system while being measured, because it
must be coupled to the larger world to provide an actual measurement record. A coupling of quantum
system S to classically describable devices—Heisenberg’s recording system plus the rest of the world, which
physically intervene when using an apparatus in such a way that one among a set of differing outcomes
can appear on the resulting record—is held to characterize measurement, and this coupling gives rise to
a probabilistic, indeterministic state change (corresponding to the actualization of the system property)
that relates to the actual value in a prescribed way (described by the EE link, see Section 4.2, below).

210

www.pdfgrip.com
Entropy 2018, 20, 420

Actualization (discussed in Section 4.3) begins with the coupling of the system under measurement to
the classically describable observational apparatus in the world, which is physically designed (based
on rigorous testing for reliability) to elicit an outcome, and ends upon decoupling, leaving a classically
describable physical record of the outcome.
If one does consider, contra the views of Bohr and Heisenberg, a fully quantum mechanical
closed system treatment of measurement, the behavior of the joint system A+S, or a larger chain of
interactions systems (see below), this might be expected to suffice for the description of measurement.
Such a treatment is often given as follows; cf. e.g., [17], [50] (Section 2.4). One prepares system S
via a series of physical interactions, such as filtering, in some well-defined quantum state |η , after
which it is arranged to interact with a measurement apparatus, A. This apparatus, after being similarly
arranged to be in a fiducial initial state |χ0 , would be required to enter a final state corresponding
to the value of a pointer property Z, which must be correlated with the value of the measured
property (non-degenerate observable) E of the system. We may consider, for simplicity, a discrete
measured property

E = ∑i ei |ψi ψi |

where {|ψi } is a countable orthonormal basis for the system Hilbert space H corresponding to its
eigenvalues {ei }. Another typical requirement for such a measurement is that a “calibration condition”
be satisfied, namely, that if a measured property is real, then its value must be exhibited properly,
unambiguously, and with certainty: if system S is an eigenstate |ψk  of E, then the state of apparatus
A after the interaction of the two is an eigenstate of Z (with an eigenbasis {|φi } associated with
pointer readings zi ), which serves to indicate the specific value of E present, the free-Hamiltonian
function contribution to the evolution of the system being considered negligible relative to that of
the measurement interaction. For quantum observables, the calibration condition generally takes the
form of a probability reproducibility condition, namely, that a probability measure exists for a property
be transcribed onto that of the corresponding apparatus pointer property. Finally, registration of the
measured property by the measurement apparatus is taken to include the physical reading out of the
registered value.
If one formally considers an entire chain of interacting objects connecting the system S up as far
as physically conceivable, to the brain of an experimenter, for example X, Y,. . . , in the environment in
addition to the original measurement system S and the experimenter’s apparatus A—such as focusing
elements, counter or counters, various cables, a computer, output display, etc.—a good measurement
would involve all these becoming correlated in their properties for the measurement outcome to be
physically indicated. Under the Schrödinger state evolution, which is unitary, upon completion of the
measurement interactions, one would then find

|Ψ = ∑ ci |si | ai | xi |yi  . . . , (9)

with {si }, { ai }, { xi }, {yi } etc. as the Hilbert space eigenbases for S, A, X, Y,. . ., respectively. The
result of considering all physical systems involved entirely quantum mechanically is simply a regress
backward from the prepared state, which presents and indefinite value of the quantity to be observed.
Heisenberg had already engaged this difficulty early on (in 1935 [51], cf. [52]) in a Copenhagenist
spirit by insisting on a bipartite division of a set of different systems involved, only one of which
is to be considered in any one analysis among all those possible, one for each way of making a
bipartite division of the above chain, and considering only one side of the division quantum mechanically,
as described below, cf. [51]. Again, for him, consideration of the entire measurement chain—or even
simply the system and portion of apparatus in direct contact with it—as a full accounting of the
measurement process as described within the state-vector formalism, as done in the above, without
truncation, is an inappropriate use of the quantum formalism, the proper role of which is to make
predictions of measurement outcomes; the only plausible use of the quantum formalism for the purposes
of symbolizing a measurement process requires the introduction of a cut or split (Schnitt) between

211

www.pdfgrip.com
Entropy 2018, 20, 420

what is considered the measured portion of this chain of systems on the side including the entity S and
the remainder, considered then to be a single, classically describable measuring system.
Notably, a change in the location of the cut makes no difference to the statistics obtained for
the purposes of prediction. This formal description is strictly speaking only a symbolic description
of the elements of the world involved. Heisenberg also imposed an important condition restricting
this location: “The claim . . . that it is indifferent at which location the cut between the parts of the
system to be treated quantum mechanically and the classical measuring devices should be drawn,
should thus be made more precise in the sense that this cut may indeed be shifted arbitrarily far in the
direction of the observer in the region that is otherwise described according to the laws of classical
physics; but that this cut cannot be shifted arbitrarily in the direction of the atomic system. Rather,
there are physical systems—and all atomic systems belong among these—that the classical concepts are
unsuitable to describe, and whose behaviour can therefore be expressed correctly only in the language
of wavefunctions” [51], cf.[53].
In this way, the chain of statistical correlations appearing in the formal representation of the state
|Ψ is considered to be cut in two—into a system S and the remainder A —somewhere along this
chain of interacting systems, with subsystems to one side of the cut collectively considered the system
to be measured: S subsumes S together with all other subsystems left of the cut, and A (=A + W)
is the collective of those systems right of the cut, that is, the “apparatus plus the rest of the world”
W. Thus, A , is removed from the quantum-physical description used to make predictions relating
to outcomes, with the cut always being made somewhere within S–X–Y–· · · –A–W, where a classical
description is possible for the entire portion including the recording system. The actualization of
potentia requires an interaction of S (the size of S or larger) with the classically describable measuring
apparatus, itself in interaction with the rest of the world. Formally, the change of state of the measured
system involved is sometimes said to be “projected” to the appropriate component of the initially
prepared system state, that is to be attributed by the eigenvalue-eigenstate link (discussed below in
Section 4.2). Such a projection involves a change of state differing from the unitary evolution predicted
when using the Schrödinger law of motion alone, for any non-trivial measurement.
It should be noted, however, that the Copenhagen approach to measurement can be criticized
for not offering, indeed, for denying the possibility of a complete, closed system description of the
measurement process or, for that matter, even precisely specifying the conditions under which
measurement will occur, for example, due to the unclear boundary between the classical and the
quantum realms, by its reliance on the requirements of the use of a macroscopic apparatus and
the production of classically describable records of reliable measuring instruments not precisely
characterizable by quantum mechanics. For this reason, Heisenberg’s appeal to actualization (discussed
in Section 4.3) can be considered an incomplete quantum mechanical treatment. Moreover, descriptions
of the sort given by Heisenberg in the above have been criticized for conflating measurement with
state preparation, as Henry Margenau did already in 1936 [54].
Let us turn now to Heisenberg’s indeterminacy relations. In the presentation of the indeterminacy
relations in Heisenberg’s 1929 Chicago lectures, published as The Physical Principles of the Quantum
Theory [29], in which he spoke only of the Der Kopenhagener Geist der Quantentheorie rather than a full
interpretation, the indeterminacy relations were, strictly speaking, only symbolic in nature and in the
process of being generalized (beginning with work by H. P. Robertson [32] discussed in Section 2 above).
These relations were placed on a firm mathematical grounding soon thereafter in the Hilbert-space
formalism; cf. Section 2.4. Recently, notably since the final explication of the Copenhagen approach by
Heisenberg in the 1950s [5], these relations have been analyzed, extended, and clarifed via the notion
of unsharpness in a way that captures the notion of quantum indeterminacy more efficaciously.

4.1. Unsharpness
The maximally specified state of a quantum system relative to an observable O in the Copenhagen
approach can be given as a projector ρpure = |ψψ| appearing in the spectral decomposition of

212

www.pdfgrip.com
Entropy 2018, 20, 420

an observable O, as discussed in Section 2.1; cf. [2,55]. Nonetheless, in addition to measurements

corresponding to such operators, unsharp measurements have also been formalized as the class of
quantum operations that are described by (normalized) positive-operator-valued measures (POVMs)
developed by Günther Ludwig, Karl Kraus, Busch, and Lahti [17,56,57].
Given a nonempty set S and a σ-algebra Σ of its subsets Xm , a POVM E is a collection of operators
{ E( Xm )} satisfying the conditions: (i) Positivity—E( Xm ) ≥ E(∅), for all Xm ∈ Σ; (ii) Additivity—for
all countable sequences of disjoint sets Xm in Σ, E(∪m Xm ) = ∑m E( Xm ); (iii) Completeness—E(S) = I.
If the value space (S, Σ) of a POVM E is a subspace of the real Borel space (R, B(R)), then E provides

a unique Hermitian operator on H, namely R Id dE, where Id is the identity map. The positive
operators E( Xm ) in the range of a POVM are referred to as effects E (H) = { A ∈ L(H) : O ≤ A ≤ I},
the expectation values of which provide the quantum probabilities.
Given an effect A, one can define properties in general by the following set of conditions. (i) There
exists a property A⊥ ; (ii) there exist states ρ and ρ such that both tr( Aρ) > 12 and tr( Aρ ) > 12 ; (iii) if
A is regular, for any effect B below A and A⊥ , 2B ≤ A + A⊥ = I, where a regular effect is an effect with
spectrum both above and below 12 . Thus, the set of properties is E p (H) = { A ∈ E (H)| A  12 I, A 
2 I} ∪ {O, I}; the set of unsharp properties is Eu (H) ≡ E (H) p / L(H). A POVM is an unsharp observable if
1

there exists an unsharp property in its range [2]. The POVMs are the natural correspondents in the
operator space of quantum mechanics of standard probability measures and thereby make precise the
notion of indeterminacy in the Hilbert-space setting.
The probability of a given outcome m upon a (generalized) measurement on a system in a pure
state P(|ψ) is given by
, -
p(m) = ψ| E( Xm )|ψ = tr (|ψψ|) E( Xm ) ; (10)

cf. Equation (1), which holds for the case of sharp measurement. The effects form a convex subset of
the space of linear operators on L(H) on the system Hilbert space, the extremal elements of this subset
being the projectors { Pi }. A collection of effects is said to be coexistent if the union of their ranges is
contained within the range of a POVM. Any two quantum observables E1 and E2 are representable as
sharp measures on (R, B(R)) exactly when [ E1 , E2 ] = O, following from the results of von Neumann
for Hermitian operators [23] mentioned in Section 2 above; coexistent observables are thus those that
can be measured simultaneously in a common measurement arrangement, and when two observables are
coexistent, there exists an observable, the statistics of which contain those of both observables, the joint
observable. Typically, the two observables are recoverable as marginals of a joint distribution on the
product of the corresponding two outcome spaces.
The introduction of unsharpness allows one to circumvent the requirement of commutativity of
jointly measurable properties, which captures only the extremes of complementarity, by including the
unsharp properties, and enables a continuous range of complementarity to be captured in the Hilbert
space formalism. For POVMs, commutativity remains sufficient but is not necessary for the coexistence
of effects (cf. [2]). It has been shown that “smeared versions” of two noncommuting observables can
still have a joint observable. For example, the operators F = { 12 ( I ± f σz )} and G = { 12 ( I ± gσz )}
have a joint observable precisely when f 2 + g2 ≤ 1. Therefore, as a requirement for this pair to be
jointly observable, the magnitudes | f |, | g| (their degrees of unsharpness) must be complementary,
in accordance with this trade-off relation, as demonstrated by Busch and Shilladay [41].
The introduction of POV measures and unsharpness have thus helped make indeterminacy and
mathematical complementarity more precise by exploiting the Hilbert-space setting; cf. [41] for further
detail on the intertwined connection and contrast between those two notions and more detail on
their role in understanding joint measurability of properties in experimental situations, such as the
single-particle Mach-Zehnder interferometry, not discussed here.

4.2. Linking Actual and Possible Values

In the Copenhagen approach to quantum mechanics, the state of a measured system is related to
the actual values obtained in measurement by what has come to be known as the eigenvalue-eigenstate

213

www.pdfgrip.com
Entropy 2018, 20, 420

(EE) link. The essence of the EE link was first introduced by Heisenberg and then used by others
including Dirac in 1930 and after; cf. the careful explanation of this by Marian Gilton [58]. In the 1930
version of Dirac’s Quantum mechanics, one finds the following:

“If a state ψr and an observable α are such that, when an observation is made of the
observable with the system in this state the result is certain to be the number a, we assume
this information can be expressed by the equation

αψr = aψr (11)

Conversely, when an equation of this type is given, we assume it has the physical meaning
that a measurement of the observable α with the system in state ψr will certainly give for
result the number a or that the observable α has the value a for the state ψr , to use a classical
way of speaking, which is permissible in this case” [24] (p. 30).

In 1958, in the 4th edition of his classic textbook Quantum Mechanics, which appeared after
Heisenberg’s article “The Development of the Interpretation of Quantum Mechanics,” one finds the
EE link connected explicitly with probability.

“The expression that an observable ‘has a particular value’ for a particular state is permissible
in quantum mechanics in the special case when a measurement of the observable is certain
to lead to the particular value, so that the state is in an eigenstate of the observable . . .
In the general case we cannot speak of an observable having a value for a particular
state, but we can speak of its having an average value for the state. We can go further
and speak of the probability of its having any specified value for the state, meaning the
probability of this specified value being obtained when one makes a measurement of the
observable.” [59] (p. 253)

4.3. Causation, Possibility, and Actuality

Like other sorts of probability, quantum probability can be viewed as the “graded possibility”
of the occurrence of events, as first suggested by Leibniz [60,61]. Moreover, in the Copenhagen
approach to quantum mechanics, unlike others, possibility plays a fundamental role in relating theory
to experiment. This was explicitly indicated by Heisenberg in his invocation of an aspect of Aristotle’s
theory of causation, wherein possibility appears prominently in relation to all phenomena: “. . . in
modern physics the concept of possibility, that played such a decisive role in Aristotle’s philosophy,
has moved again into a central place” [62] (p. 298).
As pointed out in Section 2 above, the isolated quantum system, described by a state-vector
|ψ ∈ H, “no longer contains features connected with the observer’s knowledge. . . it is also completely
abstract . . . and the representation becomes a part of the description of Nature only by being linked
to the question of how real or possible experiments will result” [5] (p. 26). The objective aspect of
probability is that of quantifying the likelihood |ci |2 of the appearance of each value among any set of
possible measurement outcomes as the actual result in the actualization of the potential physical state
which occurs upon measurement, which according to Heisenberg is independent of any subjectivity:
“the transition from the ‘possible’ to the ‘actual’ takes place as soon as the interaction between the
object and the measuring device, and thereby with the rest of the world, has come into play; it is not
connected with the act of registration of the result in the mind of the observer” [18] (pp. 54–55).
Heisenberg explained the objective character of this registration process as follows.“Of course,
the introduction of the observer must not be misunderstood to imply that some kind of subjective
features are to be brought into the description of Nature. The observer has rather only the function of
registering decisions, i.e., processes in space and time, and it does not matter whether the observer is
an apparatus or a human being; but the registration, i.e., the transition from the possible to the actual,
is absolutely necessary here, and cannot be omitted from the interpretation of the quantum theory.

214

www.pdfgrip.com
Entropy 2018, 20, 420

It must also be pointed out that in this respect the Copenhagen interpretation of quantum theory is
in no way positivistic. For whereas positivism is based on sensual perceptions of the observer as
elements of reality, the Copenhagen interpretation regards things and processes which are describable
in terms of classical concepts, i.e., the actual, as the foundation of any physical interpretation.” [5]
(p. 22). Thus, Heisenberg neither requires nor refers to the mind, the brain, or human knowledge for
the actualization of potential values, which appear in successful measurements as actual values of an
observable inferrable from a resulting classically describable record in accordance with the EE-link;
only the interaction of the measured system with the greater world in a way so as to produce such a
classical record is required. It is such a record from which the mind could later acquire knowledge if
the recorded, classically describable measurement outcome is later attended to.
Measurement of a quantum system precludes the state change that would otherwise occur were
it to remain isolated, that is, the time-evolution according to the Schrödinger law of motion. It is
in this way that, in the Copenhagen approach to QM, the possible becomes the actual and can be
recorded, according to Heisenberg—a way not necessarily captured by the evolution dictated by the
law of motion; it is often said that it is at this point that causation fails in QM—that it becomes acausal.
However, this non-deterministic change of state-vector evolution arises precisely with the intervention
of the measuring apparatus and the rest of the world participating in the production of a classical
record of the outcome as its cause.
Hence, the Copenhagen approach to non-relativistic quantum mechanics presented by Heisenberg
is not one in which there is a genuine lack of causation; instead, there is a form of Aristotelian causation
that is not captured by the fundamental law of motion, which only governs closed systems not being
measured—in particular, it is the form Aristotle calls chance causation, as I have argued in [14,15]:
In the actualization of potentiality, there is a single chance occurrence that lies within the set of possible
occurrences for the system in question according to the characteristics of its Hilbert-space description,
with a measured value capable of being recorded on a system that is also classically describable but
unpredictable [5]. It is in this way that the Copenhagen approach remains, in a specific sense, causative.
This approach to quantum probability, where it fundamentally involves possibility, has recently been
connected with logic in relation to the possible experience to be gained through measurement, as shown
in the next subsection.

4.4. Logic and Indeterminacy

In the Copenhagen approach, the quantum state is taken to characterize as completely as possible
the system to which it is attributed. The changes in the world occurring in any measurement are
changes of values of quantities that were theretofore potential, that is, only possibly possessed,
in accordance with the EE link connecting possible and actual observed values of observables.
Empirically, the probabilities of these various outcomes to occur accord with the likelihoods of
obtaining the set of possible outcomes for measurements in the future correspond formally to the
Hermitian operator O, in measurements on collections of identically prepared systems described by the
state |ψ with the resulting measurement record being classically describable when necessary for their
readout. These probabilities have been connected to logic; cf. [1] (Chapter 8). Indeed, the results of the
early work of von Neumann and Birkhoff associates propositions to quantum Hilbert (sub)spaces [63]
and the field of quantum logic that arose from that pioneering work; cf. [64] for a general summary of
later developments.

4.4.1. Completeness of the Quantum State Description

Whether the quantum state was indeed complete, despite the apparent incompatibility of this
with the attribution of precise values to all observable quantities, remained unclear until the work
of Simon Kochen and Ernst Specker in 1967 [65]. Their theorem now known as the Kochen-Specker
theorem precludes a consistent truth-valuation from being given to the propositions identified by von
Neumann and Birkhoff, which is what is required to satisfy the corresponding value-definiteness thesis,

215

www.pdfgrip.com
Entropy 2018, 20, 420

namely, (i) that each and every physical magnitude have a definite value at all times (see Alan Stairs’
valuable discussion [66]) and (ii) that measurements reveal those preexisting precise values; cf. [50]
(Section 2.3) for a discussion of related issues. The value-definiteness condition can be stated more
formally: Each proposition regarding the system, of the form “O ∈ Δ,” where O describes the physical
magnitude and Δ is a Borel subset of the real numbers, is given a definite value 0 or 1.
The work of Kochen and Specker followed a very general theorem of Andrew Gleason [21]: All
probability measures that can be defined on the lattice of quantum propositions from the quantum
statistical operators, that is all quantum probabilities, are of the form p( Pi ) = tr(ρPi ), for some statistical
operator ρ on Hilbert space H, for all H of dimension greater than two. This result demonstrates that
every probability measure over the set of quantum state projectors is one from a quantum state ρ on
the H attributed to the system in question; the trace measure tr(ρPi ) assigns to each projector Pi the
dimension of its range, which can then be normalized by the dimension of the (finite-d) H. It can,
therefore, be obtained by taking ρ to be the maximally mixed state on H. This shows that the only
natural generalization of Kolmogorov probability functions of the sort used in quantum mechanics
is exactly that of the Hilbert-space formulation of quantum mechanics. The values corresponding to
orthogonal projectors thus obey a Born-type rule for the assignment of probabilities.
States for which definite truth-values could be attributed to all observables are the so-called
dispersion-free states, states for which projectors take expectation values of only either 0 or 1 under
the above mapping. Following a presentation first given by Bell, we can relate this to Gleason’s
work [67]. The condition ∑i  P(|φi ) = 1 implies that both 0 and 1 occur because (1) there are no other
possible values for satisfying the condition and (2) neither alone suffices. However, then, there must be
arbitrarily close pairs |ψ, |φ having different expectation values, 0 and 1 respectively; however, such
pairs cannot be arbitrarily close, by the above lemma. Hence, there can be no dispersion-free states
providing the statistics of quantum statistics. Accordingly, no variables parameterizing dispersion-free
probability measures can exist for systems having H [67]. The set of quantum states is therefore
complete, because it provides the probability measures definable on the quantum propositional lattice.
Consider the complete set of Hermitian self-adjoint operators for the set of quantum states
describing a system attributed a Hilbert space with dimension d > 2, constraining their algebra to
reflect the values assigned, and take the assignment of the values of real numbers to the quantum
operators to reflect corresponding properties of the system. In this setting, the Kochen–Specker
theorem demonstrates the impossibility of such an assignment for a finite sublattice of quantum
propositions [65]. Take a value function, vψ connecting an observable O to a value of a physical
magnitude O when a system is in a state ψ, that is, vψ : O → R. Define F (O) to be the value
associated with F (O) for all functions F with a one-to-one, onto mapping from values of O to O. Taking
vψ ( F (O)) = F (vψ (O)) has the consequence that vψ is additive and multiplicative on commuting
operators with the consequence that vψ (I) for all states ψ as long as there is at least one magnitude O
for which vψ (O) = 0 (cf. [68], pp. 191–192). Another consequence of this is that vψ (Pi ) must be either 0
or 1 for all propositions Pi , which have corresponding projectors Pi . Thus, if one considers a resolution
of the identity into a set of projectors { Pi }, that is, this set is such that ∑i Pi = I, in an interpretation of
quantum properties where one and only one of the corresponding magnitudes Pi can take the value 1,
no such function exists except for an overly restricted class of properties.
The results of Gleason as well as Kochen and Specker thus support the basic Copenhagen
assumption that the quantum state is complete. Note also that the Kochen–Specker result can
be extended to general von Neumann algebras, as shown by Andreas Doring et al. in [69], with
implications quantum and generalized probabilistic models, as noted by Federico Holik et al. in [70].

4.4.2. Logical Quantum Indeterminacy Relations

More recent work of Itamar Pitowsky has shown that this connection can be placed in the
context of indeterminacy to provide a new class of trade-off relations exhibiting complementarity in
logical context. In his investigation, Pitowsky began by noting that George Boole, in developing his

216

www.pdfgrip.com
Entropy 2018, 20, 420

conception of probability, identified necessary and sufficient conditions for a set of rational numbers
p1 , p2 , ..., pn to represent properly the probabilities, considered (relative) frequencies, of the occurrence
of a set of n logically connected events E1 , . . . En [71] pi = prob( Ei ) i = 1, 2, ..., n to express what
Boole called “conditions of possible experience” [72]. These conditions are either linear inequalities
or equalities in p1 , p2 , . . . , pn ; if the events under consideration are entirely independent, then the
fractions corresponding to probabilities might be constrained only by the conditions pi ≥ 0, pi ≤ 1 ,
but the expression for sets within possible experience must take the simple form

N
a + ∑ ai pi ≥ 0 (12)
i

where a, ai are constants involving the logical relations constraining them [72].
This set of conditions on probabilities lie within n-dimensional polytopes in the case of
probabilities of correlation, the convex hull of a finite number of points in Rd , that is, the set of all
convex combinations of its points [73]. Any violation of these conditions is manifested geometrically
by the location of points (corresponding to probabilities) outside of the polytope dictated by them.
The conditions on possible experience can then be methodically constructed from the logical relations
among sets of possible events. Take, for example, a pair of events E1 , E2 having relative frequencies
p1 , p2 , again taking p12 to denote the frequency of the joint event E1 ∩ E2 . Being probabilities, these
numbers have the relations: p1 ≥ p12 , p2 ≥ p12 , p12 ≥ 0. The frequency of the disjoint event E1 ∪ E2
is then p1 + p2 − p12 with
p1 + p2 − p12 ≤ 1 . (13)

One then has a corresponding three-dimensional space of vectors ( p1 , p2 , p12 ) that can be viewed as
a convex polytope with vertices (0, 0, 0), (1, 0, 0), (0, 1, 0), and (1, 1, 1). Pitowksy considered a set of
measurements known to have as outcomes 0 and 1, such as the measurements on a squared value Si2
of the component of spin along orthogonal spatial directions for a spin-1 system.
In this case, the basic operators Si do not commute and so cannot be precisely measured
simultaneously, while their squares do, cf. Pitowsky’s [74]. Their sum S2 = 2I, where I is the
identity, so that in a simultaneous measurement of these spin-squared operators, one and only one of
these observables will have the value 0, while the others take value 1. This illustration of the general
situation corresponds to measurements with a triple of possible outcomes. Let the events that appear
in more than one measurement be written as Ei , and let those that appear in only one triple be Fi .
Suppose the noncontextuality of probability—the requirement that probability assignments do not
depend on the outcomes of measurements of other observables that might be measured at the same
time, cf. [75,76], and assign the same probability to each event above in all cases. Given that the
probabilities in each triple of possible outcomes must also sum to one, one then finds

p( E1 ) + p( E2 ) + p( F2 ) = 1 (14)
p( E1 ) + p( E3 ) + p( F3 ) = 1 (15)
p( E2 ) + p( E4 ) + p( E6 ) = 1 (16)
p( E3 ) + p( E5 ) + p( E7 ) = 1 (17)
p( E6 ) + p( E7 ) + p( F1 ) = 1 (18)
p( E4 ) + p( E8 ) + p( F4 ) = 1 (19)
p( E5 ) + p( E8 ) + p( F5 ) = 1. (20)

These requirements on probability then imply trade-off inequalities expressing complementarity,

for example,
p( E1 ) + p( E8 ) ≤ 3/2 . (21)

217

www.pdfgrip.com
Entropy 2018, 20, 420

One of the two outcomes E1 and E8 —which cannot arise as alternative outcomes of the same
measurement—becomes more certain, that is, there is an increased probability of occurrence. The other
outcome becomes less certain, with a decreased probability of occurrence. This is a logic-based
indeterminacy relation quantitatively expressing that the likelihoods of positive results in alternative
measurement arrangements are complementary quantities [74].
Here again, a novel sort of indeterminacy relation was obtained and given in the clear
mathematical form of a trade-off relation—this in addition to the others introduced and developed after
the advent of the Copenhagen approach. This new sort of relation arises from the consideration of a
collection of sets of alternative events in Boolean logic within single measurements, with the two events
involved resulting from a collection of such measurements as outcomes of different measurements, as
discussed in [77]. Like those discussed in the previous section, this result furthers the significance the
Copenhagen approach’s notions of complementarity and indeterminacy by revealing their appearance
beyond its original mathematical locus, further demonstrating its fundamental significance.

5. Indeterminism and Randomness

Classical physics has most often been thought to be deterministic in the following sense,
introduced by Laplace in the following statement. “We ought. . . to regard the present state of the
universe as the effect of its anterior state and as the cause of the one which is to follow. Given for one
instant an intelligence which could comprehend all the forces by which nature is animated and the
respective situation of the beings who compose it—an intelligence sufficiently vast to submit these
data to analysis—it would embrace in the same formula the movements of the greatest bodies of the
universe and those of the lightest atom; for it, nothing would be uncertain, and the future, as well
as the past, would be present to its eyes.” [78]. Under the conditions set out in this statement of the
notion, perfect predictions and retrodictions regarding the behavior of individual physical systems are
possible, in principle.
The determination of future (and past) events involved here is identified with in-principle
predictability under the assumption that unlimited resources are available to the predictor, P: It involves
in its application the existence of an intelligence with unlimited capabilities, both physical and
computational. However, this is something beyond the scope of the practice of physics from within
the universe: Finite beings—with finiteness assessed via the physical and/or cognitive resources at
their disposal—are inherently limited in their ability to predict physical events, given that prediction
requires computational resources; in any application of physical theory, such as prediction, any finite
agent can exploit only finite physical resources for this purpose.

5.1. Predictability
Although the contraints on the resources of finite beings do not always present difficulties for
precisely predicting a given future event among a given finite set of events in classical mechanics,
even unlimited resources do not suffice for quantum mechanics to be considered a deterministic theory
according to the above classic definition and that even when one among a finite set of alternatives
need be distinguished. An alternative, more physically straightforward definition of determinism
applicable to quantum mechanics would therefore be superior, in particular, one more suitable to
physics and independent of radical assumptions about the availability of computional and other
resources. One such a definition that has been suggested is: A scientific theory is deterministic if and
only if in that theory any two trajectories in the state space in models of systems overlap at one point
do overlap at every point, and it is indeterministic if and only if it is not deterministic (cf. discussion
in [79–81]).
Judging this within the (O, S, p) framework of general physics, one sees that quantum mechanics
is not a theory supporting determinism of this latter kind for individual, measured systems, but at best
provides precise predictions of the behavior of collections of identically prepared and subsequently
measured systems: On the Copenhagen approach, quantum measurement of physical properties are

218

www.pdfgrip.com
Entropy 2018, 20, 420

understood to introduce a probabilistic change of physical state that precludes the state-trajectory
overlap required by this definition. The “acausality”—as it was often called early in the history of
the approach—of this quantum state evolution is responsible for the indeterminism of the states
actually appearing from among those possible beforehand as the result of measurement. The states
connected with measurement data are always those dictated by following the EE link rule, which offers
an alternative to indeterminacy (“uncertainty”) relations for capturing the objective indefiniteness
of physical properties. As explained in Section 4.3 above, the Copenhagen approach provides
an Aristotelian form of causation for this process even in the presence of chance.

5.2. Randomness
Despite the clarity gained by the move to a trajectory-based version of determinism, predictability
remains relevant to the question of the randomness of the appearance of measurement outcomes.
Note that a crucial distinction exists between indeterminism and unpredictability: Predictability
hinges on the total context of prediction of P—be that a human or another sort of cognitive agent
such as an artificially intelligent robot—including the all the conditions of the experimental context.
Indeterminism, in the trajectory-based definition just discussed, involves only the theoretical character
of the theory involved, here QM, specifically regarding the topology of state-space trajectories,
independently of whether or how they could, if ever, come to be known. The randomness appearing
in quantum mechanics should not be identified with essentially probabilistic state evolution, that is,
indeterminism, as noted by Geoffrey Hellman [82], because indeterminism is not a necessary condition
for the appearance of randomness in a theory.
Randomness can be defined as maximal unpredictability, and the Copenhagen approach
provides a consistent understanding of the notions of indeterminism and unpredictability matching
this conception well. A notion of predictability that accords with one of the distinguishing
characteristics of the approach, namely, the involvement not only of the measurement apparatus
but also the large world (environment) of the measurement apparatus in the very definition of proper
measurement (cf. [5,10,14]), has been recently introduced by Anthony Eagle [81], namely, this physical
process definition:

“A prediction function p(S, t) takes as input the current state S of a system described by
a theory T as discerned by a predictor P and an elapsed time t, and yields a temporally
indexed probability distribution Prt over the space of possible states of the system.
A prediction is a specific use of some prediction function by some predictor on some initial
state S0 and time t0 who adopts Prt as their posterior credence function conditionally on the
evidence and the theory.” [81]

This definition of predictability when applied using the probabilities of outcomes of quantum
measurements allows their random character to be explicated.
The randomness in quantum mechanics in the Copenhagen approach can be understood
specifically as follows. Let P be any experimenter performing a measurement, let T be the quantum
mechanics, let | a0  be the the quantum state attributed to the pertinent system via the preparation
procedure used by P at time t0 , and let t be the time elapsed between its preparation and the completion
of the measurement. Predictor P will use the appropriate choice of quantum probability as his/her
p(| a0 , t), according to the circumstances of the measurement it performs, that a given outcome b
obtains, and use it in finding his posterior credence function given his entire background knowledge.
In particular, one can apply the following definition of unpredictability: “An event E (at some temporal
distance t) is unpredictable for a predictor P if and only if P’s posterior credence in E after conditioning
on current evidence and the best prediction function available to P is not 1, that is, if the prediction
function yields a posterior probability distribution that does not assign probability 1 to E” [81].

219

www.pdfgrip.com
Entropy 2018, 20, 420

Given, in particular, that measurement results are unpredictable in this sense whenever
incompatible observables are measured in succession, one then has a well defined sense of randomness
that applies to quantum mechanics:

“An event E is random for a predictor P using theory T if and only if E is maximally
unpredictable. An event E is maximally unpredictable for P and T if and only if the posterior
probability of E yielded by the prediction functions that T makes available, conditional on
current evidence, is equal to the prior probability of E. This also means that P’s posterior
credence in E, conditional on theory and current evidence (the current state of the system),
must be equal to P’s prior credence in E conditional only on theory.”

where E is the appearance of the eigenvalue b as the measurement outcome. The quantum
measurement process is then seen to be random when successive measurements of non-commuting
sharp observables are made: Knowledge of the outcome obtained for state preparation via
measurement of given quantum observable does not provide additional information about the
outcomes of measurements of a sharp observable with which it does not commute, such as when
the x-spin of a spin-1/2 particle is measured just after its z-spin is measured, for example, using
a Stern-Gerlach apparatus.
The general notion of randomness is as an extrinsic property of events that is dependent on
properties of agents such as P—and, more importantly, the scientific communities they form—and the
theories they use pertaining to all elements involved in their scientific activity. However, it only requires
that an account of how probability influences credence need not involve an interpretation of probability
that itself depends on credence, and in no way requires a subjective interpretation of the probability
attributed by quantum mechanics; it only recognizes that predictability expressed via Prt is epistemic in
character, cf. [81]. This notion is one that the reinforces the Copehagen approach: The randomness in
quantum theory is evidenced in relation to measurements when two non-commuting sharp observables
are measured in succession. The result of the second measurement is maximally unpredictable using
the law of motion to predict it on the basis of the result of the first measurement and any other
information obtained from the world external to the system.

6. Conclusions
Results from the investigation of the behavior of quantum probability in a range of novel
circumstances after Heisenberg’s clarification of the elements of the Copenhagen interpretation
in the mid-1950s were shown to clarify further the nature and identity of the forms of causation,
indeterminacy, and randomness that the interpretation attributes to quantum mechanics. A number of
novel trade-off relations quantifying complementarity in additional areas of mathematics and physics
were reviewed and described, as was the introduction of a number of mathematical constructions
making indeterminacy more precise and extending its application. Results showing that the chancy
nature of the results of measurements on quantum systems, and hence of the appearance of the
quantum state, can be explicated in a way that accords with the Aristotelian form of causation
introduced by Heisenberg in his late explication of the Copenhagen interpretation were also reviewed.
These various developments in theoretical physics, some of which also bear directly on experimental
physics, that appeared subsequent to Heisenberg’s clarification were in this way shown to demonstrate
the continually increasing value to physics of the Copenhagen approach to quantum mechanics.

Conflicts of Interest: The author declares no conflict of interest.

References and Notes

1. Jammer, M. The Philosophy of Quantum Mechanics; John Wiley and Sons: New York, NY, USA, 1974.
2. Busch, P.; Grabowski, M.; Lahti, P.J. Operational Quantum Physics; Springer: Berlin, Germany, 1995.

220

www.pdfgrip.com
Entropy 2018, 20, 420

3. Beller, M. Quantum Dialogue: The Making of a Revolution; The University of Chicago Press: Chicago, IL,
USA, 1999.
4. Schlosshauer, M.; Kofler, J.; Zeilinger, A. A Snapshot of Foundational Attitudes Toward Quantum Mechanics.
Stud. Hist. Philos. Mod. Phys. 2013, 44, 222. [CrossRef]
5. Heisenberg, W. The Development of the Interpretation of the Quantum Theory. In The Philosophy of Quantum
Mechanics; Pauli, W., Ed.; Pergammon: London, UK, 1955.
6. Scully, M.O.; Englert, B.-G.; Walther, H. Quantum Optical Tests of Complementarity. Nature 1991, 351,
111–116. [CrossRef]
7. Zeilinger, A. On the Interpretation and Philosophical Foundation of Quantum Mechanics. In Vastakohtien
Todelisuus; Ketvel, U., et al., Eds; Helsinki University Press: Helsinki, Finland, 1996, pp. 167–178;
8. Zeilinger, A. Experiment and the Foundations of Quantum Physics. Rev. Mod. Phys. 1999, 71, S288.
[CrossRef]
9. Faye, J. Niels Bohr: His Heritage and Legacy; Kluwer: Dordrecht, The Netherlands, 1991.
10. Plotnitsky, A. Epistemology and Probability; Springer: New York, NY, USA, 2010.
11. Plotnitsky, A. The Principles of Quantum Theory, from Planck’s Quanta to the Higgs Boson; Springer International
Publishing: Cham, Switzerland, 2016.
12. Jaeger, G. Entanglement, Information, and the Interpretation of Quantum Mechanics; Springer: Heidelberg,
Germany, 2009; Chapter 3.
13. Howard, D. Who Invented the ‘Copenhagen interpretation’? A study in mythology. Philos. Sci. 2004, 71,
669–682. [CrossRef]
14. Jaeger, G. Quantum Potentiality Revisited. Proc. R. Soc. A 2017, 375, 20160390. [CrossRef] [PubMed]
15. Jaeger, G. “Wave-Packet Reduction” and the Quantum Character of the Actualization of Potentiality. Entropy
2017, 19, 15. [CrossRef]
16. Bub, J. Interpreting the Quantum World; Cambridge University Press: Cambridge, UK, 1997.
17. Busch, P.; Lahti, P.; Mittelstaedt, P. The Quantum Theory of Measurement, 2nd ed.; Springer: Heidelberg,
Germany, 1996.
18. Heisenberg, W. Physics and Philosophy; Harper and Row: New York, NY, USA, 1958.
19. Mackey, G.W. Mathematical Foundations of Quantum Mechanics; W.A. Benjamin, Inc.: New York, NY, USA, 1968.
20. Maczyński, M.J. A remark on Mackey’s axiom system for quantum mechanics. Bull. Acad. Pol. Sci. Ser. Sci.
Math. Astronom. Phys. 1967, 15, 583–587.
21. Gleason, A.M. Measures on the Closed Subspaces of a Hilbert Space. J. Math. Mech. 1957, 6, 885. [CrossRef]
22. Amann, A.; Primas, H. What is the Referent of a Nonpure Quantum State. In Potentiality, Entanglement and
Passion-at-a-Distance; Cohen, R.S., Horne, M., Stachel, J., Eds; Kluwer: Dordrecht, The Netherlands, 1997;
pp. 9–30.
23. Von Neumann, J. Mathematical Foundations of Quantum Mechanics; Princeton University Press: Princeton, NJ,
USA, 1955.
24. Dirac, P.A.M. The Principles of Quantum Mechanics; Clarendon Press: Oxford, UK, 1930.
25. Born, M. Quantenmechanik der Stoßvorgänge. Z. Phys. 1926, 38, 803. [CrossRef]
26. Accardi, L. L’edificio Matematico della Meccanica Quantistica Non-relativistica: Situazione attuale.
Nuovo Cimento 1995, 110, 685–721. [CrossRef]
27. Jaeger, G. What in the (Quantum) World is Macroscopic? Am. J. Phys. 2015, 82, 896. [CrossRef]
28. Jaeger, G. Grounding the Randomness of Quantum Measurements. Proc. R. Soc. A 2016, 374, 20150238.
[CrossRef] [PubMed]
29. Heisenberg, W. The Physical Principles of Quantum Theory; University of Chicago Press: Chicago, IL, USA, 1930.
30. Born, M. Statistical Interpretation of Quantum Mechanics. Science 1955, 122, 675–679. [CrossRef] [PubMed]
31. Lahti, P. Uncertainty Principle and Complementarity in Axiomatic Quantum Mechanics. Rep. Math. Phys.
1980, 17, 287. [CrossRef]
32. Robertson, H.P. The Uncertainty Principle. Phys. Rev. 1929, 34, 163. [CrossRef]
33. Dummett, M. Is Time a Continuum of Instants? Philosophy 2000, 75, 497–515. [CrossRef]
34. Shimony, A. Physical and Philosophical Issues in the Bohr–Einstein debate. In Niels Bohr: Physics and the
World; Feshbach, H., Matsui, T.; Olesonet, A., Eds.; Routledge: London, UK, 1988; pp. 285–304.
35. Jaeger, G. Quantum Objects; Springer: Heidelberg, Germany, 2014; Chapter 2.

221

www.pdfgrip.com
Entropy 2018, 20, 420

36. Kastner, R.E.; Kauffman, S.; Epperson, M. Taking Heisenberg’s Potentia Seriously. Int. J. Quantum Found.
2018, 4, 158.
37. Accardi, L. Quantum Probability: New perspectives for the laws of chance. Milan J. Math. 2010, 2, 481–502.
[CrossRef]
38. Jaeger, G. The Double-slit Experiment. In Compendium of Quantum Physics; Greenberger, D., Hentschel, K.,
Weinert, F., Eds.; Springer: Heidelberg, Germany, 2009.
39. Khrennikov, A. Probability and Randomness: Quantum versus Classical; Imperial College Press: London, UK,
2016; Chapter 6.
40. Jaeger, G.; Shimony, A.; Vaidman, L. Two Interferometric Complementarities. Phys. Rev. Lett. 1995, 51, 54–67.
[CrossRef]
41. Busch, P.; Shilladay, C. Complementarity and Uncertainty in the Mach–Zehnder interferomenter and Beyond.
Phys. Rep. 2006, 435, 1–31. [CrossRef]
42. Schrödinger, E. Die gegenwärtige Situation in der Quantenmechanik. Die Naturwissenschaften 1935, 23, 807.
(In German) [CrossRef]
43. Schrödinger, E. Discussion of Probability Relations Between Separated Systems. Math. Proc. Camb. Philos. Soc.
1935, 31, 555–563. [CrossRef]
44. Clauser, J.F.; Horne, M.; Shimony, A.; Holt, R.A. Proposed experiments to test local hidden-variable theories.
Phys. Rev. Lett. 1973, 23, 880. [CrossRef]
45. Aspect, A.; Grangier, P.; Roger, G. Experimental Test of Realistic Theories via Bell’s Inequality. Phys. Rev. Lett.
1981 47, 460. [CrossRef]
46. Horne, M.; Shimony, A.; Zeilinger, A. Two-particle interferometry. Nature 1990, 347, 429–430. [CrossRef]
47. Jaeger, G.; Horne, M.A.; Shimony, A. Complementarity of One-particle and Two-particle Interference.
Phys. Rev. Lett. 1995, 48, 1023. [CrossRef]
48. Abouraddy, A.; Nasr, M.B.; Saleh, B.E.A.; Sergienko, A.V.; Teich, M.C. Demonstration of the Complementarity
of One- and Two-photon interference. Phys. Rev. Lett. 2001, 63, 063803. [CrossRef]
49. Saleh, B.E.A., Abouraddy, A.F., Sergienko, A.V., Teich, M.C. Duality Between Partial Coherence and Partial
Entanglement. Phys. Rev. Lett. 2000, 62, 043816.
50. Jaeger, G. Entanglement, Information, and the Interpretation of Quantum Mechanics; Springer: Heidelberg,
Germany, 2009.
51. Crull, E.; Bacciagaluppi, G. Translation of: W. Heisenberg, ‘Ist eine Deterministische Ergänzung der
Quantenmechanik möglich?’. Available online: http://philsci-archive.pitt.edu/8590/ (accessed on
4 May 2011).
52. Heisenberg, W. Ist eine deterministische Ergänzung der Quantenmechanik möglich? In Wissenschaftlicher
Briefwechsel mit Bohr, Einstein, Heisenberg u.a., Band II: 1930–1939; Meyenn, K.V., Hermann, A., Weisskopf, V.F.,
Eds.; Springer: Berlin, Germany, 1985; pp. 407–418. (In German)
53. Bacciagaluppi, G.; Crull, E. Heisenberg (and Schrödinger, and Pauli) on hidden variables. Stud. Hist. Philos.
Mod. Phys. 2009 40, 374. [CrossRef]
54. Margenau, H. Quantum-mechanical description. Phys. Rev. 1936, 49, 240. [CrossRef]
55. Srinivas, M.D. Collapse Postulate for Observables with Continuous Spectra. Commun. Math. Phys. 1980, 71,
131–158. [CrossRef]
56. Kraus, K. States, Effects and Operations; Springer: Berlin, Germany, 1983.
57. Ludwig, G. Versuch einer axiomatischen Grundlegung der Quantenmechanik und allgemeinerer
physikalischer Theorien. Z. Phys. 1964, 181, 233–260. [CrossRef]
58. Gilton, M.J.R. Whence the Eigenstate–eigenvalue Link? Stud. Hist. Philos. Mod. Phys. 2016, 55, 92–100.
[CrossRef]
59. Dirac, P.A.M. Quantum Mechanics, 4th ed.; Oxford University Press: Oxford, UK, 1958.
60. Leibniz, G.W. Lettre à la princesse Elisabeth, de fin 1678, in Leibniz, 1940, p. 58. (In France).
61. Leibniz, G.W. Lettre à Jacquelot du 20 novembre 1702 (Raisons que M. Jacquelot m’a envoyeés pour justifier
l’argument contesté de des-Cartes qui doit prouver l’existence de Dieu, avec mes réponses), GP III 444.
(In France).
62. Heisenberg, W. Sprache und Wirklichkeit in der modernen Physik. Wort Wirklichk. 1960, 1, 32.
63. Birkhoff, G.; von Neumann, J. The Logic of Quantum Mechanics. Ann. Math. 1936 37, 823. [CrossRef]

222

www.pdfgrip.com
Entropy 2018, 20, 420

64. Dalla Chiara, M.L. Quantum logic. In: Handbook of Philosophical Logic, Synthese Library; Gabbay, D.,
Guenther, F., Eds.; Springer: Dordrecht, The Netherlands, 1986; Volume 166.
65. Kochen, S.; Specker, E.P. Logical structures arising in quantum theory. In The Theory of Models; Addison, J.,
Henkin, L., Tarski, A., Eds.; North-Holland: Amsterdam, The Netherlands, 1965.
66. Stairs, A. Quantum logic, realism, and value-definiteness. Philos. Sci. 1983, 50, 578–602. [CrossRef]
67. Bell, J.S. Speakable and Unspeakable in Quantum Mechanics; Cambrige University Press: Cambridge, UK, 1987.
68. Isham, C. Lectures on Quantum Theory: Mathematical and Structural Foundations; Imperial College Press:
London, UK, 1995.
69. Doring, A. Kochen-Specker theorem for von Neumann algebras. Int. J. Theor. Phys. 2005, 44, 139–160.
[CrossRef]
70. Holik, F.; Fortin, S.; Bosyk, G.; Plastino, A. On the Interpretation of Probabilities in Generalized Probabilistic
Models. In Quantum Interaction QI 2016. Lecture Notes in Computer Science; de Barros, J., Coecke, B., Pothos, E.,
Eds.; Springer: Cham, Switzerland, 2017; Volume 10106.
71. Pitowsky, I. George Boole’s ‘Conditions of Possible Experience’ and the Quantum Puzzle. Br. J. Philos. Sci.
1994, 45, 95–125. [CrossRef]
72. Boole, G. On the Theory of probabilities. Philos. Trans. R. Soc. 1862, 152, 225. [CrossRef]
73. Pitowsky, I. From George Boole to John Bell: The Origins of Bell’s Inequality. In Bell’s Theorem, Quantum
Theory and the Conceptions of the Universe; Kafatos, M., Ed.; Kluwer: Dordrecht, The Netherlands, 1989.
74. Pitowsky, I. Betting on the Outcomes of Measurements: A Bayesian theory of quantum probability. Stud. Hist.
Philos. Mod. Phys. 2003, 34, 395–414. [CrossRef]
75. Barnum, H.; Caves, C.M.; Finkelstein, J.; Fuchs, C.A.; Schack, R. Quantum Probability from Decision Theory?
Proc. R. Soc. Lond. A 2000, 456, 1175. [CrossRef]
76. Hemmo, M.; Pitowsky, I. Quantum Probability and Many Worlds. Stud. Hist. Philos. Mod. Phys. 2007, 38,
333–350. [CrossRef]
77. Jaeger, G. Uncertainty Relations and Possible Experience. Mathematics 2016, 4, 40. [CrossRef]
78. Laplace, P.-S. A Philosophical Essay on Probabilities; Truscott, F.W., Emory, F.L., Trans.; Dover Publications:
New York, NY, USA ,1951; p. 4.
79. Earman, J. A Primer on Determinism; D. Reidel: Dordrecht, The Netherlands, 1986; pp. 137–138.
80. Montague, R. Deterministic Theories. In Formal Philosophy; Thomason, R.H., Ed.; Yale University Press:
New Haven, RI, USA, 1974.
81. Eagle, A. Randomness is unpredictability. Br. J. Philos. Sci. 2005, 56, 749. [CrossRef]
82. Hellman, G. Randomness and Reality. In PSA 1978; Asquith, P.D., Hacking, I., Eds; University of Chicago
Press: Chicago, IL, USA, 1978; Volume 2, pp. 79–97.

c 2018 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

223

www.pdfgrip.com
 entropy
Letter
Dimensional Lifting through the Generalized
Gram–Schmidt Process
Hans Havlicek 1 and Karl Svozil 2,3, * ID

1 Institute of Discrete Mathematics and Geometry, Vienna University of Technology, Wiedner Hauptstraße
8-10/104, A-1040 Vienna, Austria; havlicek@geometrie.tuwien.ac.at
2 Institute for Theoretical Physics, Vienna University of Technology, Wiedner Hauptstraße 8-10/136,
A-1040 Vienna, Austria
3 Department of Computer Science, University of Auckland, Auckland 1142, New Zealand
* Correspondence: svozil@tuwien.ac.at

Received: 26 March 2018; Accepted: 10 April 2018; Published: 14 April 2018

Abstract: A new way of orthogonalizing ensembles of vectors by “lifting” them to higher dimensions
is introduced. This method can potentially be utilized for solving quantum decision and computing
problems.

Keywords: orthogonality; quantum computation; Gram–Schmidt process

PACS: 03.65.Aa; 02.10.Ud; 02.30.Sa; 03.67.Ac

The celebrated Gram–Schmidt algorithm allows the construction of a system of orthonormal

vectors from an (ordered) system of linearly independent vectors. Let us mention that there exist a wide
variety of proposals to “generalize” the Gram–Schmidt process [1] serving many different purposes.
In contrast to these generalizations, we construct a system of orthogonal vectors from an (ordered)
system of arbitrary vectors, which may be linearly dependent. (Even repeated vectors are allowed.)
This task is accomplished by what will be called “dimensional lifting”.
Some quantum computation tasks require the orthogonalization of previously non-orthogonal
vectors. This might be best understood in terms of mutually exclusive outcomes of generalized beam
splitter experiments, where the entire array of output ports corresponds to an ensemble of mutually
orthogonal subspaces, or, equivalently, mutually orthogonal perpendicular projection operators [2].
Of course, by definition (we may define a unitary transformation in a complex Hilbert space
by the requirement that it preserves the scalar product [3] (§ 73)), any transformation or mapping of
non-orthogonal vectors into mutually orthogonal ones will be non-unitary. However, we may resort to
requiring that some sort of angles or distances (e.g., in the original Hilbert space) remain unchanged.
Suppose, for the sake of demonstration, two non-orthogonal vectors, and suppose further that
somehow one could “orthogonalize” them while at the same time retaining structural elements, such as
the angles between projections of the new, mutually orthogonal vectors onto the subspace spanned by
the original vectors. For instance, the two non-orthogonal vectors could be transformed into vectors of
some higher-dimensional Hilbert space satisfying the following properties with respect to the original
vectors: (i) the new vectors are orthogonal, and (ii) the orthogonal projection along the new, extra
dimension(s) of the two vectors render the original vectors. A straightforward three-dimensional
construction with the desired outcome can be given as follows: suppose the original vectors are
unit vectors denoted by |e1  and |e2 ; and 0 < |e1 |e2 | < 1. Suppose further a two-dimensional
coordinate frame in which |e1  and |e2  are planar; thus, we can write in terms of some orthonormal
basis |e1  = x1,1 , x1,2 as well as |e2  = x2,1 , x2,2 . Suppose we “enlarge” the vector space to include
an additional dimension, and suppose a Cartesian basis system in that greater space that includes

Entropy 2018, 20, 284; doi:10.3390/e20040284 224 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2018, 20, 284

the two vectors of the old basis (and an additional unit vector that is orthogonal with respect to the
original plane spanned by the original basis vectors).
Ad hoc, it is rather intuitive how two (not necessarily unit) vectors can be found that project onto
the original vectors, and which are orthogonal: “create” a three-dimensional vector space with one
extra dimension, assign a non-zero extra coordinate (such as 1) associated with this dimension for the
first vector, and use the extra coordinate of the second vector for compensating any nonzero value of
the scalar product of the two original vectors— in particular, whose coordinates with respect to the
new basis are
|f1  = x1,1 , x1,2 , 1 ,
(1)
|f2  = x2,1 , x2,2 , − ( x1,1 x2,1 + x1,2 x2,2 ) ,

which are orthogonal by construction.

It is not too difficult to find explicit constructions for the more general case of k vectors
|e1 , . . . , |ek  in Rn (cf. Ref. [2] for a rather inefficient method).
In the following, for the sake of construction, we shall embed Rn into Rn+k , such that we fill all
additional vector coordinates of |e1 , . . . |ek  with zeroes. For the new, mutually orthogonal, vectors
we make the following Ansatz by defining

|f1  = e1 , 1, 0, . . . , 0 ,

|f2  = e2 , x2,1 , 1, 0, . . . , 0 ,
(2)
...
|fk  = ek , xk,1 , xk,2 , . . . , xk,k−1 , 1 ,

with yet to be determined coordinates xi,j . (The symbols ei stand for all the n coordinates of |e1 .)
The unit coordinates 1 ensure that the new vectors are linearly independent. By construction,
the orthogonal projection of |fi  onto Rn renders |ei  for all 1 ≤ i ≤ k.
What remains is the recursive determination of the unknown coordinates xi,j . Note that all |f j 
must satisfy the following relations: for j > 1, orthogonality demands that f1 |f j  = 0, and therefore
e1 |e j  + 1 · x j,1 = 0, and therefore
x j,1 = −e1 |e j . (3)

In this way, all unknown coordinates x2,1 , . . . , xk,1 can be determined.

Similar constructions yield the remaining unknown coordinates in |f2 , . . . , |fk . For j > 2,
f2 |f j  = 0, and therefore e2 |e j  + x2,1 x j,1 + x j,2 = 0, yielding

x j,2 = −e2 |e j  − x2,1 x j,1 . (4)

In this way, all unknown coordinates x3,2 , . . . , xk,2 can be determined.

This procedure is repeated until one arrives at j = k − 1, and therefore at the orthogonality of
|fk−1  and |fk , encoded by the condition fk−1 |fk  = 0, and hence

xk,k−1 = − (ek−1 |ek +

(5)
+ xk−1,1 xk,1 + · · · + xk−1,k−2 xk,k−2 ) .

The approach has the advantage that, at each stage of the recursive construction, there is only
a single unknown coordinate per equation. This situation is well known from Gaussian elimination.
The Ansatz also works if one of the original vectors is the zero vector, and if some of the original vectors
are equal.

225

www.pdfgrip.com
Entropy 2018, 20, 284

The resulting system of orthogonal vectors is not the only solution of the initial problem—to find
an orthogonal vector that projects onto the original ones—which can be explicitly demonstrated by
multiplying all vectors |f1 , . . . , |fk  with the matrix

diag In , cT , (6)

whereby In stands for the n-dimensional unit matrix, c can be a real nonzero constant, and T is
a k-dimensional orthogonal matrix. (For complex Hilbert space, the orthogonal matrix needs to be
substituted by a unitary matrix, and by a complex constant c = 0.)
On the other hand, we may reinterpret our procedure as follows: Let |e1 , . . . , |ek  be a system
of vectors in Rn , not necessarily spanning Rn , and not necessarily being linearly independent.
(The ordering of the vectors in this system will be essential throughout.) We embed Rn in Rn+k
as we did above and denote the orthogonal complement of Rn by C ∼ = Rk . Therefore, the first n
coordinates of all vectors in C vanish, and R n + k can be represented by a direct sum Rn+k = Rn ⊕ Rk .
Additionally, we choose some (ordered) orthonormal basis of C, say, |g1 , . . . , |gk .
Then, there is a unique system of orthogonal vectors |f1 , . . . , |fk  in Rn+k such that the following
conditions are satisfied:
1. For all 1 ≤ i ≤ k, the orthogonal projection of Rn+k onto Rn sends |fi  to |ei .
2. The orthogonal projection of Rn+k onto C sends |f1 , . . . , |fk  to some (ordered) basis of the
subspace C. Applying the Gram–Schmidt process to this (ordered) basis gives the orthonormal
basis |g1 , . . . , |gk .
Indeed, in our previous Ansatz, we tacitly assumed the orthonormal basis |g1 , . . . , |gk  of C to
comprise the orthogonal projections of the last k vectors of the standard basis |b1 , . . . , |bn+k  of Rn+k
onto C. Condition 2 enforces the presence of all the 1s and 0s in Formula (2), since the Gram–Schmidt
process, applied to the vectors
| f1  − | e1 , . . . , | f k  − | e k ,
has to result in |bn+1 , . . . , |bn+k . Notice that the usual Gram–Schmidt process gives merely an
orthogonal basis, whose vectors can be normalized in a second step in order to obtain an orthonormal
basis. In our setting, however, such a second step is not allowed. As we saw above, now Condition 1
guarantees that |f1 , . . . , |fk  are uniquely determined.
Besides uniqueness, this construction has the additional advantage that the dot product in Rn+k
“decays” into the sum of dot products in Rn and in Rk : any basis vector fi ∈ Rn+k can be uniquely
written as fi = ei + hi , where ei and hi represent the projection of fi along hi onto the original subspace
Rn , and the projection of fi along ei onto C, respectively. Since ei is orthogonal to hi , for i = j,
fi · f j = ei · e j + hi · h j = 0, and thus
ei · e j = − hi · h j . (7)

Let us, for the sake of a physical example, study configurations associated with decision problems
that can be efficiently (that is, with some speedup with respect to purely classical means [2]) encoded
quantum mechanically. The inverse problem is the projection of orthogonal systems of vectors onto
lower dimensions. This method renders a system of non-orthogonal rays, also called eutactic stars [4–8],
which can be effectively levied to mutually exclusive outcomes in a generalized beam splitter
configurations [9,10] reflecting the higher dimensional Hilbert space.
One instance of such a quantum computation involving the reduction to ensembles of orthogonal
vectors (and their associated span or projection operators) is the Deutsch–Jozsa algorithm, as reviewed
in Ref. [2]. Another, somewhat contrived, problem can be constructed in three dimensions from
a eutactic star
1  &√ ' & √ '
√ 1, 1 , 12 3i − 1 , 21 − 3i − 1 ,
3 (8)
& √ ' &√ ' 
1
2 − 3i − 1 , 1
2 3i − 1 ,

226

www.pdfgrip.com
Entropy 2018, 20, 284

which is the projection onto the plane formed by the first two coordinates of a three-dimensional
orthormal basis
1 
B3 = √ 1, 1, 1 ,
3
&√ ' & √ '
1
3i − 1 , 12 − 3i − 1 , 1 , (9)
2
& √ ' &√ ' 
2 − 3i − 1 , 2 3i − 1 , 1 ,
1 1

which, together with the Cartesian standard basis, forms a pair of unbiased bases [11].
Still another decision configuration is the eutactic star

1
1, 1, 1 , 1, 1, −1 ,
2  (10)
1, −1, 1 , 1, −1, −1 ,

which is the projection onto the subspace formed by the first three coordinates of a four-dimensional
orthormal basis
1
B4 = 1, 1, 1, 1 , 1, 1, −1, −1 ,
2  (11)
1, −1, 1, −1 , 1, −1, −1, 1 .

More concretely, suppose some, admittedly construed, function f , and some quantum encoding
| x f (y), where x and y stand for (sequences of) auxiliary and input bits, respectively, would yield one
of the basis systems B3 or B4 . By reducing the auxiliary bits x, one might end up with the eutactic
stars introduced above. Alas, so far, no candidate of this kind has been proposed.
In summary, a new method of orthogonalizing ensembles of vectors has been introduced. Thereby,
the original vectors are “lifted” to or “completed” in higher dimensions. This method could be utilized
for solving quantum decision and computing problems if the original problem does not allow an
orthogonal encoding, and if extra bits can be introduced that render the equivalent of the extra
dimensions in which the original state vectors can be lifted and orthogonalized.
Compared with methods that were introduced [12–14] previously to optimally differentiate
between two non-orthogonal states, the scheme suggested here is similar in the sense that, in order
to obtain a better resolution, the effective dimensionality of the problem is increased. However,
our scheme is not limited to the differentiation between two states, as it uses arbitrary dimensionality.
More importantly, whereas our scheme is capable of separating different states precisely, but in general
is non-unitary—indeed, the original vectors are not mutually orthogonal, but the lifted vectors are,
thereby changing the angles among vectors, resulting in transformations that cannot be unitary—the
former methods are unitary and probabilistic.

Acknowledgments: This work was supported in part by the European Union, Research Executive Agency (REA),
Marie Curie FP7-PEOPLE-2010-IRSES-269151-RANPHYS grant.
Author Contributions: Both authors have read and approved the final manuscript.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Leon, S.J.; Björck, Å.; Gander, W. Gram–Schmidt orthogonalization: 100 years and more. Numer. Linear
Algebra Appl. 2013, 20, 492–532.
2. Svozil, K. Orthogonal Vector Computations. Entropy 2016, 18, 156.
3. Halmos, P. Finite-Dimensional Vector Spaces; Springer: Berlin/Heidelberg, Germany, 1974.
4. Schläfli, L. Theorie der vielfachen Kontinuität. In Gesammelte Mathematische Abhandlungen. Springer: Basel,
Switzerland, 1950; pp. 167–387. (In German)

227

www.pdfgrip.com
Entropy 2018, 20, 284

5. Hadwiger, H. Über ausgezeichnete Vektorsterne und reguläre Polytope. Comm. Math. Helv. 1940, 13, 90–107.
(In German)
6. Coxeter, H.S.M. Regular Polytopes, 3rd ed.; Dover Publications: New York, NY, USA, 1973
7. Seidel, J.J. Eutactic stars. In Colloquia Mathematica Societatis János Bolyai; North-Holland Publishing Co.:
Amsterdam, the Netherlands, 1976; Volume 18, pp. 983–999.
8. Haase, D.; Stachel, D. Almost-orthonormal vector systems. Beitr. Algebra Geom. 1996, 37, 367–382.
9. Reck, M.; Zeilinger, A.; Bernstein, H.J.; Bertani, P. Experimental realization of any discrete unitary operator.
Phys. Rev. Lett. 1994, 73, 58–61.
10. Żukowski, M.; Zeilinger, A.; Horne, M.A. Realizable higher-dimensional two-particle entanglements via
multiport beam splitters. Phys. Rev. A 1997, 55, 2564–2579.
11. Schwinger, J. Unitary operator bases. Proc. Natl. Acad. Sci. 1960, 46, 570–579.
12. Ivanovic, I.D. How to differentiate between non-orthogonal states. Phys. Rev. A 1987, 123, 257259.
13. Peres, A. How to differentiate between non-orthogonal states. Phys. Rev. A 1988, 128, 19.
14. Jaeger, G.; Shimony, A. Optimal distinction between two non-orthogonal quantum states. Phys. Rev. A 1995,
197, 83–87.

c 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

228

www.pdfgrip.com
 entropy
Concept Paper
On Interpretational Questions for Quantum-Like
Modeling of Social Lasing
Andrei Khrennikov 1,2, *, Alexander Alodjants 1 , Anastasiia Trofimova 1 and Dmitry Tsarev 1
1 Mechanics and Optics (ITMO) Department, National Research University for Information Technology,
197101 St. Petersburg, Russia; alexander_ap@list.ru (A.A.); nastasiatm@gmail.com (A.T.);
dmitriy_93@mail.ru (D.T.)
2 International Center for Mathematical Modeling in Physics and Cognitive Sciences, Linnaeus University,
SE-351 95 Växjö, Sweden
* Correspondence: Andrei.Khrennikov@lnu.se

Received: 27 October 2018 ; Accepted: 1 December 2018; Published: 2 December 2018

Abstract: The recent years were characterized by increasing interest to applications of the quantum
formalism outside physics, e.g., in psychology, decision-making, socio-political studies. To distinguish
such approach from quantum physics, it is called quantum-like. It is applied to modeling socio-political
processes on the basis of the social laser model describing stimulated amplification of social actions.
The main aim of this paper is establishing the socio-psychological interpretations of the quantum
notions playing the basic role in lasing modeling. By using the Copenhagen interpretation and the
operational approach to the quantum formalism, we analyze the notion of the social energy. Quantum
formalizations of such notions as a social atom, s-atom, and an information field are presented.
The operational approach based on the creation and annihilation operators is used. We also introduce
the notion of the social color of information excitations representing characteristics linked to lasing
coherence of the type of collimation. The Bose–Einstein statistics of excitations is coupled with the
bandwagon effect, one of the basic effects of social psychology. By using the operational interpretation
of the social energy, we present the thermodynamical derivation of this quantum statistics. The crucial
role of information overload generated by the modern mass-media is emphasized. In physics laser’s
resonator, the optical cavity, plays the crucial role in amplification. We model the functioning of social
laser’s resonator by “distilling” the physical scheme from connection with optics. As the mathematical
basis, we use the master equation for the density operator for the quantum information field.

Keywords: quantum-like models; operational approach; information interpretation of quantum theory;

social laser; social energy; quantum information field; social atom; Bose–Einstein statistics; bandwagon
effect; social thermodynamics; resonator of social laser; master equation for socio-information excitations

1. Introduction
From the very beginning, it has to be pointed out that we tried to make this paper readable for
people working in psychology, decision-making, cognitive, social, and political science, and having
minimal knowledge about the mathematical apparatus of quantum physics. Therefore, we try to
minimize the number of mathematical expressions, except Sections 4.4 and 9. The introduction is
very detailed and its aim is to describe the general state of the art in applications of quantum theory
to humanities. One one hand, we want to convince experts in humanities that quantum theory can
resolve the well known problems, in particular, in decision theory. On the other hand, we want to
convince physicists and especially those who work in quantum foundations and quantum information
and probability that applications of quantum theory to humanities are not an exoticities: many top
level experts, e.g., psychologists, work actively on quantum-like modeling.

Entropy 2018, 20, 921; doi:10.3390/e20120921 229 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2018, 20, 921

1.1. Quantum versus Quantum-Like Models

Since the first days of quantum mechanics, the analogy between quantum physical and psychical
processes sporadically attracted attention of leading physicists, psychologists, and philosophers.
There can be mentioned the Pauli–Jung correspondence [1], see also [2], and Whitehead’s attempt
to unify physical and mental processes within the quantum picture [3,4]. There should be also
emphasized the numerous contributions to creation of quantum models of brain’s functioning [5–15].
However, we point out that the topic of this paper has nothing to do with their attempts to model
cognition and consciousness from the genuine quantum physical processes in the brain. In the present
paper, we proceed with the quantum-like approach, the operational application of the mathematical
formalism of quantum theory, especially quantum probability and information, outside physics.

1.2. Quantum-Like Modeling of the Process of Decision-Making

Nowadays, quantum-like modeling is widely used in mathematical modeling in cognitive science,
psychology, decision-making, game theory, economics and finances, social and political sciences
(see, for example, monographs [16–23] and a few representative papers [24–38]). These applications
are typically based on the use of quantum probabilistic calculus, instead of the classical one. It was
demonstrated that experimental statistical data collected in, cognitive psychology, decision-making,
social science, game theory, molecular biology, and epigenetic demonstrates quantum probabilistic
features [18,19,22,23]. For example, such data can violate the classical formula of total probability and
this violation can be mathematically expressed in the form of quantum interference. Some cognitive
experiments demonstrating violation of the Bell type inequalities have been performed [35–38].
The violation can be interpreted as contextuality of mental observables represented as questions or tasks.
In theoretical studies in cognitive psychology and decision-making, quantum probability provides
the adequate mathematical models for the basic psychological effects such as conjunction, disjunction,
and order effects. Its use also resolves the fundamental paradoxes of the decision theory such as
Allais, Ellsberg, and Machina paradoxes [39–41]. We remark that this resolution of the paradoxes
of the classical decision theory (including expected utility, subjective utility, and prospect theories)
and creation of the paradox-free (at least up to now) decision theory played a very important role in
justification of applications of quantum probability outside physics.

1.3. Operational Formalism: Creation and Annihilation Operators

We emphasize that the quantum formalism provides only a formal operational description of
physical processes; in particular, the spontaneous emission and stimulated absorption and emission
which play the fundamental role in lasing theory. One of the important mathematical representations
of the operational formalism is given in terms of the creation and annihilation operators â , â. In quantum
optics, â represents creation of a photon through its emission; â represents disappearance of a photon
from the field resulting from absorption of this photon by an atom. These are linear operators acting
in complex Hilbert space representing the states of the quantum field. The operator â is adjoint to
the operator â. The operator n̂ = â â plays the important role in the quantum field theory. This is
the operator of the photons number. Its eigenstates are states |n of the quantum field with the fixed
number n of photons. Similar operational description can be given for the processes of state transitions
in atoms.
The operational formalism is so useful in quantum theory, since here one cannot construct a
more detailed description in terms of classical-like variables, known as hidden variables. Therefore,
one proceeds with the operational formalism, representing preparations of system and observations.
In this paper, as well in papers cited in Section 1.2, the operational formalism, including creation and
annihilation operators (see, especially, [20,34]), is used to model the process of social lasing, stimulated
amplification of social actions.

230

www.pdfgrip.com
Entropy 2018, 20, 921

One of the reasons for application of the quantum operational formalism to modeling social
processes is that, similarly to quantum physics, it is (practically) impossible to present the detailed
account of all socio-psychological factors involved in stimulated amplification of social actions. In this
situation, the operational description in terms of formal absorption and emission of information
excitations can be fruitful.
We can refer to stormy discussions in socio-political literature (see [42–46] and further references
in [47–49]) on the origin of color revolutions. These discussions are characterized by the diversity of
opinions and the impossibility to present the detailed account of interrelation of all social, psychological,
political, and economical and financial factors leading to such protest waves. The operational quantum
model of “social laser” is based on ignoring these factors (playing the role of hidden variables). We are
not interested in difference in, e.g., the emotional states of people in different countries or difference in
the political situations in these countries. The operational formalism describes formally the processes
of absorption and emission of portions of social energy; see Sections 1.5 and 4.4 for further justification
(from the information-theoretical and thermodynamical principles) of applying the quantum formalism
to modeling of social lasing.
Nowadays, the big-data approach is widely applied for analysis of social processes. The deep
learning is one of the cornerstones of the big-data project. We remark that the deep learning can be
considered as a form of the operational description. The intermediate layers of networks performing
deep learning are just the operational components having no real social or cognitive meaning.
In contrast to the big-data approach, which is based on extensive consumption of computational
resources, the quantum operational approach is endowed with the powerful mathematical formalism
of analytic analysis. This formalism can be used for determination of the key-parameters playing the
crucial role in stimulated amplification of social actions (see Section 9).

1.4. Social Laser as a Fruit of the Quantum Information Revolution

In this paper, we are interested in the information component of quantum theory. We recall that
the recent quantum information revolution generated the information interpretation of the quantum theory.
In fact, there were proposed a few often competing information interpretations [50–63]. However,
all of them are characterized by the paradigm shift: quantum systems are treated as merely carriers of
information. Thus, from this viewpoint, quantum theory is about information flows. Hence, it may
be possible to apply the quantum information approach even outside physics by identifying the
quantum-like features of processes under consideration.
In [47,48], the author proposed the quantum-like model of the social (or more generally
information) laser, the social analog of the physical laser. The social laser theory was used for the
mathematical description of stimulated amplification of social actions [49].
The model of the social laser is based on treatment of humans as carriers of social energy. In [47,48],
the notion of the social energy was (rather schematically) discussed in the relation to the social lasing
model. In this paper, we shall present the detailed analysis of this notion based on the operational
approach to quantum theory and its information interpretation (Section 3).
In the simplest setting, we consider a two level cognitive system with the relaxation and
excitation states. We call such a system a (two level) social atom, s-atom. In this paper, we restrict our
considerations to two energy level systems. In the quantum-like modeling, the discrete structure of the
social energy levels of s-atoms, energy quantization, has no straightforward relation to neurophysiology.
In particular, the two levels structure is based on so to say “to be or not to be” scaling of the social
energy. This is a kind of emotional quantity. Of course, emotions are coupled to physical and
neurophysiological processes in human’s body, but this coupling is very complicated. In addition, it is
not important for us in the present paper.
A social gain medium is composed of s-atoms. They interact with the information field generated
by the mass-media and the Internet. This field is modeled as a quantum information field carrying
quantized portions of social energy, social excitations. This field was formally introduced in [47,48].

231

www.pdfgrip.com
Entropy 2018, 20, 921

Now, we put essential efforts to propose the proper interpretation of excitations of this quantum field
as carriers of the social energy (Section 3.2).
Additionally to the social energy, we introduce the notion of a social color of an information
excitation (Section 4.2). This quantity represents additional characteristics of information excitations
linked to lasing coherence of the type of polarization and collimation (Section 4.5). We couple the
bandwagon effect, one of the fundamental effects of social psychology, with the Bose–Einstein statistics
of indistinguishable social excitations (Section 5).

1.5. Powerful Information Flows as the Basic Condition of Social Laser Functioning
The basic feature of quantum information fields is indistinguishability of excitations carried by
them. This is indistinguishability with respect to the social energy carried by excitations (see Section 4.3
for further consideration).
The Bose–Einstein statistics of information excitations playing the crucial role in social
lasing can be derived (by appealing to thermodynamical considerations) from such energy based
indistinguishability; see Section 4.4.
People overloaded by information do not put essential efforts to analyze contents of
communications delivered the by mass-media. They function as absorbers of social energy delivered
by communications emitted by TV, newspapers, and the Internet. In particular, to approach population
inversion, the mass-media can pump into a human gain medium a flow of shock news about various
catastrophes, violence, and political scandals. Agents absorb the social energy. Later, this energy can
be liberated and directed through the laser-like process of stimulated emission. Therefore, the power
of information flows plays the crucial role in creation of quantum-like processes of absorption and emission
of the social energy. Nowadays, this flow is extremely powerful. Hence, it is easier to create social
(information) lasers than say 150 years ago.

1.6. Resonators of Physical and Social Lasers

In physics, the laser’s resonator in the form of the optical cavity plays the crucial role in the process
of amplification of the output beam and making it coherent. In this paper, we model the functioning of
social laser’s resonator by “distilling” the physical scheme to exclude the straightforward connection
with optcis. We proceed with the quantum master equation for the density operator describing excitations
of the quantum information field inside the resonator of the social laser; see Section 9.3. The main
aim is establishing the proper social interpretations of the basic quantities and parameters of this
dynamical system.
The Internet based Echo Chamber is considered as the important example of social resonators.
Its functioning is mathematically represented by the field of social excitations in the form of posts and
comments interacting with the human gain medium.
One of the basic consequence of the quantum-like dynamics is the existence of the threshold
value for the pump parameter (see Section 9.5): if the power of pumping is essentially higher than this
threshold, then practically all social energy pumped into the human gain medium is transferred into
the output beam of social actions.
This is the good place to emphasize once again the role of huge power of the information flows
generated by modern mass-media.

2. Physical Laser: Schematic Presentation

Since we really hope that this paper would be interesting for researchers working in cognitive,
social, and political sciences and having no educational background in quantum physics, we present
schematically the basic scheme of the physical laser functioning. For the moment, we do not consider
the optical cavity component of the physical laser.

232

www.pdfgrip.com
Entropy 2018, 20, 921

2.1. Spontaneous and Stimulated Emission

We start with description of the processes of spontaneous and stimulated emission for excited two
level atoms with energy levels E = E1 , E2 . Denote the resonance energy of an atom by E A = E2 − E1 .
Typically, in the physical literature, one proceeds with frequencies and speaks about the resonance
EA
frequency ω A = 2πh , where h is the Planck constant. To proceed to the social modeling, we try to
eliminate the space-time picture from the model. In particular, we want to operate solely with energies.
We consider the quantum electromagnetic field interacting with atoms and describe the physical
processes generated by the interaction:

1. An atom in the excited state can spontaneously emit a photon. This process is irreducibly random,
i.e., even for a single atom, it is impossible to predict neither the instance of time nor the direction
of emitted photon.
2. Atoms in the ground state can absorb from this field only photons having the resonance energy
E A = E2 − E1 .
3. An excited atom interacting with photons with energy E A emits a photon of the same energy
4. This output flow of photons is coherent. All photons produced from the “seed-photon” have the
same features: direction of flow, polarization, and energy.

Thus, one photon produces two, these two interact with two atoms and produce four photons,
after n-step, this process generates N = 2n photons. We note that this description and its illustration
by Figure 1, although typical for physics textbooks, is too straightforward. In fact, an atom interacts
not with a single photon, but with all photons in the field (see Section 5 and Equation (10)).

Figure 1. Emission and absorption of photons.

2.2. Population Inversion

The cascade process described above plays the crucial role in generating laser beams carrying
huge energy. One of the problems is spontaneous emission. (Of course, the primary problem is that
the thermodynamic heat-bath Boltzmann distribution leads to higher population in the lower levels
than in the upper levels.)
Pumping photons into a gain medium (an ensemble of atoms in the ground state) transfers
ground state atoms into excited atoms. However, spontaneously, they can fall back to the ground
level. The basic step in generating lasing is approaching population inversion. This is transition from
an ensemble of atoms in the ground state to an ensemble in which more than 50% of atoms are in the
excited state.
We remark that, for a gain medium composed of two level atoms, population inversion
cannot be approached (at least straightforwardly) because the transition probabilities are equal:
p( E1 → E2 ) = p( E2 → E1 ). These probabilities are known as the Einstein coefficients. Their equality
for the electromagnetic field can be proven by using thermodynamics for indistinguishable systems
following Bose–Einstein statistics. For the electromagnetic field, one needs a gain medium composed
of atoms with at least three energy levels, E1 < E2 < E3 .

233

www.pdfgrip.com
Entropy 2018, 20, 921

The above considerations are about the physical laser based on the quantum electromagnetic field.
However, in general, information thermodynamics [47] does not imply the coincidence of the Einstein
coefficients for a two-level system. In principle, the social laser can be based on the simplest s-atoms
having two levels of social energy.

3. Social Energy
The notion of social energy plays the crucial role in our framework. We start considerations with a
general remark that in quantum theory value a of observable A obtained in its measurement cannot be
interpreted as the property of system S on which the measurement is performed. By the Copenhagen
interpretation, this value is generated in the complex process of interaction of S and a device used for
A-measurement. In particular, this interpretation has to be applied to energy observable E. Although
one often says, e.g., “the energy of the electron”, the correct meaning of this statement is about the
output of the E-measurement. The Copenhagen interpretation is well accommodated to the notion
of states’ superposition. Quantum system S can be in state ψ of superposition of two different energy
levels E1 , E2
ψ = c1 | E1  + c2 | E2 , (1)

where ci are complex numbers, probability amplitudes, such that |c1 |2 + |c2 |2 = 1. If system S is in this
state, then the probability to get the value Ei of the energy observable is equal to pi = |ci |2 . The states
| E1 , | E2  correspond to the definite energy levels.
Supported by the quantum interpretation of the notion of energy, we are ready to consider
the very complex notion of a social (mental) energy. This notion has been actively used in cognition,
psychology, social and political science (since the works of James [64] and Freud [65,66], later by
Jung [67], see also [68]) and recently in economics and finances, multi-agent modeling, evolution theory
and industrial dynamics [69–71], see also [47–49] for details. In addition, of course, these previous
studies are supporting for our model. However, we emphasize once again that the application of the
Copenhagen methodology simplifies and clarifies essentially the issue of the social energy.

3.1. Energy of s-Atoms

In this framework, the simplest quantification of the social (mental) energy can be done by the
question (observable) E = “Are you in the state of relaxation or excitement?” This observable takes two
values, say E1 = 0, E2 = 1. The Copenhagen interpretation is strongly involved. Before being asked the
E-question, s-atoms can be in superposition of these two states. Only by confronting the E-question
s-atom determines his/her state. The social energy observable can be represented in different forms;
for example, in the form of the question E = “Shall you go to demonstration against Trump or Brexit?”
Of course, we need not be restricted to the simplest “yes”–“no”, to be or not to be, quantification.
Finer quantifications of the social energy can considered as well. Different types of s-atoms can
emit and absorb social energy portions of different magnitudes. Here, a type of an s-atom is determined
by her/his psyche and social environment. In the operational formalism, we can proceed with some
grading of the possible social energy levels for s-atoms. Since we restrict consideration to two-level
s-atoms, they are characterized by the social energy levels E = E1 , E2 , the ground and excited states,
and the resonance energy E A = E2 − E1 . We remark that in principle the ground level energy for one
type of s-atoms can be higher than the excitation level energy for another type. Social lasing is possible
only for a gain medium with the homogeneous structure of energy levels.
This methodology demystifies the notion of social (mental) energy. Of course, the same
Copenhagen methodology can be applied to any social (mental) observable. One should not be
surprised that the methodology of quantum physics is applicable outside it. The Copenhagen
interpretation presents the very general methodology which is applicable to any kind of measurement.
We remark that the use of this measurement methodology does not imply that the whole apparatus
of quantum theory can be applicable. One should be careful by checking constraints on the class

234

www.pdfgrip.com
Entropy 2018, 20, 921

of systems and observables leading to applicability of one or another part of quantum theory.
The quantum-like approach does not mean to copy straightforwardly the complete quantum
theory to say social science. For example, to derive the Bose–Einstein statistics, we have to
assume indistinguishability of information quanta, excitations of the quantum information field
(see Section 4.1).

3.2. Energy of the Quantum Information Field

In physics, energy can be assigned not only to atoms, material systems, but also to carriers
of interactions, e.g., photons or neutrino, which are excitations of corresponding quantum fields.
(Here “assigned” has the operational meaning: “can be measured”). In social lasing, the interaction
processes are formally modeled with the aid of a quantum information field generated by a variety of
information sources (see Section 4.1). Communications “emitted” by them carry portions of the social
energy. These quanta of social energy are interpreted as excitations of the quantum information field.
Again, as in the case of s-atoms, we can proceed with “to be or not to be” quantification of
the social energy carried by communications. If an s-atom in the ground state absorbs energy from
communication C (and becomes excited), then C carries social energy E2 = 1. If C cannot excite a social
atom, then C ’s energy E1 = 0.
This social energy quantification depends on the concrete ensemble of s-atoms, a social group. It is
easy to give examples of social and political communications which would excite average Englishman
or American, but not Russian or Chinese, and vice versa. Thus, the definition of information field’s
energy is purely operational. In some sense, it is even “more operational” than the definition of
the energy for a quantum physical field. It is meaningless to speak about the social energy of the
information field without to describe the class of “detectors”; in our case, these are s-atoms. As in the
case of s-atoms, it is possible to proceed with models based on finer scales of the social energy assigned
to excitations of the information field. Each communication C is characterized by social energy EC .
It can be absorbed by an s-atom with the resonance energy E A = EC .
A variety of communications can carry the same portion of the social energy. All communications
with EC = E, where E is the fixed portion of the social energy are considered as equivalent from the
viewpoint of energy delivering. They can be represented by the same field’s state | E, the ket-vector in
field’s state space. We say that E determines a mode of the quantum information field: E is the analog of
characteristic energy Eω = hω of the electromagnetic mode with frequency ω.
How many elementary excitations of energy E are carried by the E-mode of the quantum
information field? It depends on the power of information sources emitting communications belonging
to the E-mode.

4. Social Laser

4.1. Quantum Field Representation of Information Flow Generated by Mass-Media

For the physical laser, the electromagnetic field is the basic energy source. The Bose–Einstein
statistics of excitations (quanta) of this field plays the crucial role in laser’s functioning.
For a social analog of the physical laser, the basic (social) energy source is the information field
generated by the mass-media and the Internet. As was discussed in Section 3.2, this field can be
considered as composed of information quanta field’s excitations. These excitations are emitted by
a variety of information channels and absorbed by humans, s-atoms. The information field is not
a classical physical field defined on the physical space-time, as, e.g., the classical electromagnetic
field given by the vector of the electric and magnetic fields ( E( x ), B( x )), x ∈ R3 . We model the
information field operationally with the aid of creation and annihilation operators, as a quantum field,
see Section 1.3.

235

www.pdfgrip.com
Entropy 2018, 20, 921

We remark that even in physics only the electromagnetic field has its classical counterpart with
the space-time representation. If we consider, for example, the neutrino field, this field has only the
quantum representation.
In the quantum field theory, a state of a field is mathematically described by a normalized vector
belonging to the Fock space—the complex Hilbert space representing superpositions of possible
excitations of the field. Consider the E-mode of the quantum information field. It represents excitations
in the form of communications carrying the portion of social energy E: communications which
are graded by the social group under consideration as having social energy E. In the Fock space
representation, the quantum state of this field is represented as superposition
cn
Φ= ∑ cn |n, E = ∑ √n! âEn |0, (2)
n

where ∑ |cn |2 = 1. Here, âE is the creation operator for E-excitations; |0 denotes the vacuum state of
the field—no information excitations. For the real information sources, the sum is always finite. The use
of infinite series is the price for mathematical consistency—the possibility to use a Hilbert space.
Generally, a social group is exposed to radiation of the quantum information field containing
different modes of the social energy. It is natural to consider the discrete grading of the social energy.
Here, we follow von Neumann who pointed out [72] that quantum observables with continuous
spectrum are just mathematical idealizations of real measurement procedures. For such discrete
grading, the Fock state can be written in the standard form
c
∑ √kk!
k
Φ= â E11 · · · âEknn · · · |0, (3)

where k = (k j ) and, for each multi-index k, only a finite number of k j differs from zero; the squared
sum of absolute values of the coefficients equals to one.

4.2. Coloring Information Excitations

For the quantum electromagnetic field, excitations are photons and photon’s type is determined by
index λ = ( E, α). Here, E is photon’s energy and α encodes additional characteristics linked to lasing
coherence of the type of polarization and collimation. In the same way, for the quantum information
field, the type of an excitation is determined by index λ = ( E, α). Here, E is the social energy carried by
an excitation of the information field and α is a social color of this excitation encoding its basic social
characteristics. The color of an information excitation is linked to coherence of social actions generated
by a social laser. Thus, creation and annihilation operators have to be labeled not by just the energy
index E, but by index λ, including the state representation, cf. Equation (3): |n, λ = âλn |0.
We can mention a few examples of social colors of excitations of the quantum information field:
war in Iraq, elections in USA, Brexit, tsunami in Japan, sex scandal, anti-globalism, climate change,
rasism, sexism, Trump, and so on. Determination of social colors is a socio-psychological phenomenon.
It depends on a social group. Thus social colors are not internal characteristics of the information field.
Their depend on a social group exposed to “information radiation” generated by mass media.
The social analog of lasing can be initiated only in a social group, a social gain medium,
with sufficiently rough coloring structure—to approach high concentration of excitations of the same
social color mode in the output beam of excitations. For example, social color α = “sex scandal in
Conservative and Unionist Party (UK)” is appropriate for social lasing. This is the proper social
color for the communication: “A BOMBSHELL dossier naming and shaming 36 Tories suspected of
inappropriate sexual behaviour has emerged as Westminster remains engulfed by a sex abuse scandal”.
(Express, 30 October 2017). However, if somebody would spit this color and started to use 36 colors
corresponding to concrete Tory-executives involved in the sex scandals, then such s-atom is not a
proper subject for social lasing.

236

www.pdfgrip.com
Entropy 2018, 20, 921

4.3. Indistinguishability from Information Overload and Complexity

For s-atoms, excitations of the information field carrying the same social energy and color,
λ = ( E, α), are indistinguishable. Indistinguishability is the basic condition leading to quantum statistics
through thermodynamical analysis, both for physical and information systems (see Section 4.4).
Indistinguishability of excitations generated by the mass-media is relative. It corresponds to coarse
graining of coloring and depends on a social group, a collection of s-atoms. If s-atoms do not operate in
the indistinguashability regime, i.e., they perform the detailed analysis of contents of communications,
then the quantum statistical description is inapplicable for them. In this case, social lasing is impossible.
We remark that the detailed analysis of communication’s content preassumes the use of Boolean logic.
Thus, the indistinguishability regime implies deviations from Boolean logic, cf. [73,74].
The information overload is one of the basic reasons for operation at the indistinguishability
regime. People simply do not have time and information processing resources for the deep analysis of
communications’ contents. They proceed with coarse graining leading to relative indistinduishability.
Here, “relative” has the meaning of relative with respect to a social group. In the modern human
society, the information overload is combined with complexity of information delivered by a variety of
information sources. The majority of population simply does not have mental capacity to perform the
detailed analysis of complex socio-political, financial, and ecological problems.
Coarse graining, “rough coloring” and the indistinguishability regime are the basic features of
human cognition. However, the information overload and complexity led to tremendous increase of
their role.

4.4. From Statistical Mechanics to Thermodynamics of Indistinguishable Systems

Here, we follow [47], but with emphasis of the operational meaning of the notion of the
social energy. In turn, the presentation in paper [47] on transition from statistical mechanics of
indistinguishable systems to thermodynamics is based on Schrödinger’s book [75].
Consider a system which is composed of m indistinguishable subsystems. Compound system
will be denoted by S and its subsystems by S with indexes. It is assumed that an observer can assign
to all systems some quantity called energy and satisfying the additivity requirement with respect to its
distribution over subsystems of a system. We shall denote the energy of S by E and the energy of S by
E. These quantities should not be interpreted as objective properties of systems. Energy E = E(S) can
be assigned to S as the output of observation performed by an observer on S The same interpretation
is used for energy E = E (S) assigned to S .
As always derivation of thermodynamical quantities from ensemble statistics is started with
construction of partition function Z. The possible energy levels of a subsystem S are denoted by
E1 , ..., Ej , .... An energy level Ek of m-particle compound system S is characterized by a sequence of
natural numbers m1 , ..., m j , ... of subsystems on the corresponding levels. The latter means “subsystems
with measurement outputs” E1 , ..., Ej , .... Here, it is crucial that subsystems with the same value of
energy Ej are not distinguished one from another. This indistinguishability determines the form of Z
and, hence, all thermodynamical quantities. We have

Ek = ∑ Es ms . (4)
s

Thus, the partition function is given by the sum

Z= ∑ e−μ ∑s Es ms , (5)
(ms )

where symbol (ms ) denotes an admissible set of numbers ms . For the moment, μ > 0 is just a parameter
of the model.

237

www.pdfgrip.com
Entropy 2018, 20, 921

From ln Z, it is possible to deduce the basic thermodynamical quantities; in particular, the average
value of ms
1 ∂ ln Z
m̄s = − . (6)
μ ∂Es
Now, different statistics for systems composed of indistinguishable subsystems can be obtained
by consideration of different possible ranges of values of natural numbers ms :

1. ms = 0, 1, 2, .... (Bose–Einstein statistics),

2. ms = 0, ..., q, where q ≥ 1 is a natural number (parastatistics).

In physics, q = 1, i.e., ms = 0, 1; this is the case of the Fermi–Dirac statistics. We remark that in
quantum physics selection of the Fermi–Dirac statistics from a bunch of parastistics is just postulated.
Of course, it is confirmed by the experimental situation. One cannot exclude that in social science
parastistics with q > 1 can find applications.
By restricting considerations to the Bose–Einstein and Fermi–Dirac statistics and following
Schrödinger [75], we find that the corresponding partition functions can be expressed as

1
Z = ZBE = ∏ 1 − e−μEs , Z = ZFD = ∏(1 + e−μEs ). (7)
s s

This leads to the following basic expression for the average value of ms

1
m̄s = 1 μEs
, (8)
ξe ∓1

where 0 < ξ ≤ 1 is so called parameter of degeneration, ξ = ξ (m). We remark that, for photons, ξ = 1,
with the Bose–Einstein statistics.
Then, one gets the average energy as
αs
U= ∑ 1 μEs
e ∓1
. (9)
s ξ

In physics, the quantity T inverse to parameter μ is interpreted as temperature. In social modeling,

we can speak about a kind of the social temperature. This is a complicated notion and we are not ready
to discuss it in detail in this paper. On one hand, we can try to proceed as in classical thermodynamics.
However, even by mimicking classical thermodynamics, it is important to remember that such
a quantity is of the socio-emotional type. Thus, it cannot be considered as the objective feature
of a social system. One can try to define social temperature through consideration of classes of
equivalent thermometers, measurement procedures for the social temperature. However, the above
thermodynamical considerations for indistinguishable systems represent the quantum situation. Hence,
the classical definition of temperature does not match them. As in quantum physics, we can try to
introduce a kind of social temperature as a parameter characterizing phase transitions. This is a
complicated mathematical theory and we postpone such considerations to one of further publications.

4.5. Coloring Role: Pumping versus Emission

We also point to the following striking similarity in behavior of atoms and s-atoms. A portion of
social energy absorbed by s-atom generally “lost its color”.
Somebody, say Elena, living in Moscow absorbed a social excitation emitted by the Russian
radio-station, Echo of Moscow. Typically, such excitations have the anti-corruption colors. However,
this does not mean that her spontaneous relaxation would be directed against corruption. She can
emit the portion of social energy absorbed from Echo of Moscow in a family scandal or another kind of
private or social action.

238

www.pdfgrip.com
Entropy 2018, 20, 921

Such behavior is similar to behavior of atoms interacting with photons. If an atom absorbs a
photon carrying some concrete physical characteristics, “color”, such as direction or polarization,
then it immediately forgets about its pre-absorption value. Later, it can emit a photon with a different
color, i.e., a photon flying in the direction different from the direction of the pre-absorption photon.
However, we remind that the quantum theory is about observational quantities. We do not know
what happens inside the atom between absorption and emission of photons. We neither know what
happens inside the head between hearing the news from Echo of Moscow and going to the kitchen to
start scandaling or to the center of Moscow to protest.
This kind of memory lost can be very useful in approaching population inversion (see [47]).
There is no need to pump in a gain medium the social energy of the same color as in laser’s output beam.
In principle, the social energy pumping need not be based solely on information about corruption
and other dysfunctions of the government. Shocking news about catastrophes, tornado, killers are the
important part of energy-pumping in human gain medias.
Now, we turn to physics. In contrast to the process of absorption-emission, the process of
stimulated emission of photons is characterized by “color” conservation: the “colors” of the emitted
photon and photons stimulating emission coincide. We also point out that in the process of stimulated
emission, the crucial role is played by intensity of the stimulating electromagnetic field, so-called
Bosonic effect: increase of probability of stimulated emission with increase of intensity of the flow
bosonic excitations (see Section 5 and Equation (10)).
In Section 5, we compare this effect with the bandwagon effect in psychology. The operational
identity of these two effects, physical and psychological, supports application of the quantum field
formalism and methodology to social processes. Stimulated emission of social excitations by excited
s-atoms can be considered as exhibition of the bandwagon effect: s-atom in the excited state exposed by
radiation compounded of α-colored social excitations would emit a social excitations of the same color.

5. Comparing Stimulated Emission in Quantum Physics and Bandwagon Effect in Psychology

and Social Science
We stress that the quantum field description of the stimulated emission is a collective effect, i.e.,
an atom interacts with a bunch of photons and not just with an individual photon. It interacts
with all excitations of the quantum electromagnetic field having the resonance energy of this atom,
E A = E2 − E1 .
The crucial role is played by the Bose–Einstein statistics of the photons. We consider the fixed energy
(frequency) mode of the quantum electromagnetic field. For fixed color mode α, n-photon state |n, α,
can be represented in the form of the action of the photon creation operator aα corresponding to this
mode on the vacuum state |0: √
|n, α = [( aα )n / n!]|0. (10)

This representation gives the possibility ( to find that the transition probability amplitude from
the state |n, α to the state |n + 1, α equals to (n + 1). On the other hand, it is well known that the
reverse process of absorption characterized by the transition probability amplitude from the state |n, α
√
to the state |(n − 1), α equals to n. .
Thus, in a quantum Bosonic field increasing photons’ number leads to increasing the probability of
generation of one more photon in the same state. This constitutes one of the basic quantum advantages
of laser stimulated emission showing that the emission of a coherent photon is more probable than the
absorption. This is the strong argument in favor of using the quantum modeling of lasing.
This behavior of photons or more generally excitations of any quantum Bosonic field matches
the cognitive bias known as the bandwagon effect [76]. This effect is characterized by the probability of
individual adoption increasing with respect to the proportion who have already done so. People are
not interested in underlying rational justification based on Boolean logic. They “hop on the bandwagon”
by taking into account only the number n of those who are already seating on it. It is important to stress

239

www.pdfgrip.com
Entropy 2018, 20, 921

that, for an agent interacting with bandwagon’s population, personalities of people on bandwagon
play no role: these people are indistinguishable.
This indistinguishability is only with respect to this concrete interaction: social, financial, racial,
gender, or political action characterizing this “bandwagon.” Of course, people seating on a bandwagon
are individual agents who can differ crucially: biologically, mentally, culturally.
As we have already emphasized, indistinguishability is the crucial assumption leading to quantum
statistics. In the case of bandwagon effect, this is the Bose–Einstein statistics. Of course, it need not be
exactly the photon statistics.
Thus, the bandwagon effect can be considered as social exhibition of the Bose–Einstein statistics caused by
indistinguishability.
To model social lasing, we consider the information version of the bandwagon effect: s-atom
interacts mainly not with other s-atoms, but with excitations of the information field generated
by the mass-media and the Internet as well as emitted by other agents. For the concrete social
color, these excitations are indistinguishable. In addition, s-atom in the excited state surrounded by
information excitations of social color α emits excitation of the same color with probability proportional
to the number of excitations. The crucial role is played by the field coherence, with respect to the
concrete color.
We can conclude that the formal mathematical model is the same for physical and social structures.
This is the model of stimulated emission of a system, physical or human, interacting with some
Bosonic field.

6. Social Lasing Schematically

Each class of information communications is characterized by the social energy. Coherence
corresponds to social colour sharpness (ideally one single mode α) generating a coherent beam of
social actions. People in the excited state interacting with α-colored excitations of the information field
would also emit α-colored excitations.
For example, a gain medium consisting of agents in the excited state and stimulated by the
anti-corruption coloured information field would “radiate” a wave of anti-corruption protests.
The same gain medium stimulated by some information field carrying another social colour would
generate the wave of actions corresponding this last colour.
The amount of the social energy carried by communications stimulating lasing should match the
resonance energy of s-atoms in the human gain medium.
To approach the population inversion, the social energy is pumped into the gain medium.
This energy pumping is generated by the mass-media and the Internet sources. The gain medium
should be homogeneous with respect to the social energy spectrum. In the ideal case, all s-atoms in the
gain medium should have the same spectrum, E = E1 , E2 . In reality, it is impossible to create such a
human gain medium. As in physics, the spectral line broadening has to be taken into account.
Social colors of excitations in the energy pumping beam have no straightforward connection with
the social color of excitations in the output beam.

7. Resonators of Physical Lasers

In laser physics one of the main problems in lasing initiation is approaching the population
inversion. However, population inversion is not enough to generate lasing. Stimulated and
spontaneous emissions are competing with each other. Thus, before becoming an amplifying device,
a gain medium pumped by an external energy source is first radiated as a usual electric “lamp”.
Here, spontaneous emission is dominating. The light power is distributed over a variety of frequencies
and directions of propagation, generally uniformly distributed. It is the optical cavity, laser’s
resonator, that creates the conditions necessary for stimulated emission to become predominant
over spontaneous emission.
In further considerations, it is assumed that the gain medium has approached population inversion.

240

www.pdfgrip.com
Entropy 2018, 20, 921

The cavity or resonator is composed of two mirrors that bounce the beam back and forth through the
gain medium. The cascade process of increasing photons’ number inside the cavity can be initiated
either by spontaneous emission from an atom in the gain medium or by a bunch of photons injected
in the optical cavity. (The photons carry the energy-quanta matching the energy levels of atoms in
the gain medium.) In the latter case, these photons interact with atoms and generate stimulated
emission. It is crucial that these photons have the same phase. One can imagine them as a cloud of
exponentially increasing size moving between mirrors. The concentration of the field inside the cavity
increases the probability of stimulated emission rather than spontaneous emission occurring. This is the basic
feature of bosons (see Section 5). We repeat that bosons’ behavior is similar to human’s behavior known as the
bandwagon effect (see again Section 5).

8. Resonators of Social Lasers

The same competition between spontaneous and stimulated emission plays the crucial role in
social processes. People in the excited state may “radiate” social energy spontaneously, say in debates
with relatives and friends about the political and social problems. Social colors of excitations in such
spontaneous radiation are typically randomly distributed, often uniformly distributed. Such emission
of social energy cannot lead to coherent social actions.

8.1. Structure and Functioning of Social Resonator

A social resonator consists of a gain medium composed of s-atoms which has already approached
population inversion and say an Internet based communication system, e.g., some social network.
We call such a system Echo Chamber. We restrict modeling to the Internet based Echo Chambers.
Consider the following idealized model.
Each s-atom in the excited state can emit a quantum of social energy in the form of a post or a
comment on some post. We call them, posts and comments, excitations of the social resonator. By posting
or commenting, i.e., emitting an excitation, s-atom falls to the ground state. Resonator’s excitations
play the role of photons in the optical cavity, the resonator of the physical laser. Moreover, to simplify
the model, we assume that the social resonator under consideration accepts only excitations of the
concrete color α. This is the strong constraint that is in visible contradiction with functioning of typical
Internet based social networks. We shall relax it in later modeling. The social color of an excitation
plays the role of the direction of propagation of output beam of photons emitted by laser’s resonator,
the x-axis of the optical cavity.
Suppose that, at the fixed instance of time in the social resonator, there are n excitations.
Each member of the gain medium interacts with all these excitations—with the information field.
The boson behavior of excitations implies that the probability that the concrete agent would fall to the
ground state and emit an excitation increases with n, Section 5. It is crucial that, if all excitations of the
social resonator have fixed color α, the color of excitation emitted by this agent is also α. These dynamics
lead to the exponential increasing the number n of excitations having the α-color inside this social
resonator (see Section 9 for modeling of temporal dynamics). Excitations of colors different from α
also can be spontaneously emitted by the gain medium. However, they cannot generate the cascade
process, since in the present model they are simply blocked.

Output Beam from Echo Chamber

When n becomes sufficiently large, see Section 9, it is possible to open the output channel of the
Echo Chamber and generate the stable flow of high intensity of excitations of the fixed α-color. In “outer
space”, this flow is realized in the form of meetings, demonstrations, and brutal protest actions.

8.2. Stimulated Initiation of Social Lasing

As was stressed, the straightforward blocking of excitations with colors different from one
fixed color α is the strong assumption. Of course, moderators of social networks block some posts

241

www.pdfgrip.com
Entropy 2018, 20, 921

and comments, e.g., having extremist, racist, or sexist content. However, the proportion of filtered
excitations seems not to be so high, in any event it is far from 100%. Therefore, we have to improve the
above model. As was presented in Section 7, there are two possible scenarios for initiating lasing:

1. spontaneous emission and filtering photons with momentum vectors deviating from the x-axis
by using the optical cavity;
2. stimulated emission generated by a coherently injected ensemble of photons with the
x-momentum vector.

In social lasing, the second scenario is preferable because the social mechanism of filtering of
“wrongly directed and spontaneously emitted social excitations” is not so straightforward as in optics.
Actually, the components which are not coherent with the beam are not eliminated, but become
insignificant and can be considered as “noise”, as if they were in some way “ignored” by the
mainstream social movement.
Thus, generation of the beam of social excitations having the same color α is started by injection
a block of α-colored posts into the Internet Echo Chamber. They are injected in the same moment of
time. (Of course, one has to take into account the temporal scale of Echo Chamber’s functioning).
This initializing block of excitations generates the cascade of stimulated emissions. After a few
interactions, the propagating wave of excitations is so big that the probability of stimulated emission
becomes very close to one. This is a good time to open the output channel of the Echo Chamber and to
transform information excitations into physical social actions.
Of course, spontaneously posted excitations of colors different from α can also be generated in this
Echo Chamber. However, they are generated in different moments and have a variety of colors. Even
if such a post starts to generate its own cascade, its power is negligible comparing with the dominating
cascade started with injection of α-posts.

9. Dynamics the Quantum Information Field in Social Laser’s Resonator

We proceed with the standard formalism of theory of open quantum systems by using the
quantum master equation for the state of a subsystem of a compound system. In our considerations,
the latter consists of the quantum information field interacting with the s-atom gain medium. The basic
dynamical equation is given by the quantum Markov approximation of the Schrödinger dynamics
for the state of the compound system. This is the Gorini–Kossakowski–Sudarshan–Lindblad (GKSL)
equation [77]. We point out that the using the Markov approximation is an important assumption.
Its meaning and validity for social systems is a complex question (see [78] for the corresponding
analysis of the general socio-political situation). Applicability of the quantum Markov approximation
to modeling social lasing should be studied in more detail. We have no possibility to do this in the
present paper but plan to turn to this problem in one of the future works.
As always in theory of open quantum systems, we should extend the notion of the state of a social
system by considering mixed states represented by density operators.
We follow presentations in physical works, e.g., [79–83]. One of the essential differences in
transition from quantum modeling of real physical processes to quantum-like modeling outside
physics is that the Planck constant h cannot be considered as social action quantum. In quantum
physics, the constant h couples photon’s energy E and angular frequency ω as

E = h̄ω, (11)

where h̄ = h/2π is the reduced Planck constant. As was already mentioned, we were not able to find
a natural social analog of angular frequency ω. We tried to proceed without it. Equation (11) can be
written as
τ = 1/ω = h̄/E. (12)

242

www.pdfgrip.com
Entropy 2018, 20, 921

Here, τ has the physical dimension of time and it can be interpreted as the time scale of the
quantum dynamics for the E-mode of the quantum electromagnetic field.
By transition from genuine quantum to quantum-like modeling, we have to introduce an analog
of the Planck constant, say γ. It is interpreted similarly to Equation (12), as a constant coupling the time
scale and energy. There is no reason to assume that it is equal to the physical constant h̄. Moreover,
we cannot assume that γ is the same for all social processes. If it were the case, it would be really
surprising! This is just the time scale of a social process modeled with the aid of the quantum formalism.

9.1. Creation-Annihilation Algebras for s-Atoms and Quantum Information Field

The quantum information field carrying the fixed amount of social energy EF can be represented
in the following operator form
EF EF
−it it
Ê(t) = ue γ â + ūe γ â , (13)

â
where u is the complex field amplitude and â and are annihilation and creation operators for social
excitations of Ê(t), i.e., √
√
â|n = n|n − 1, â |n = n + 1|n + 1. (14)
√ √
Here, scaling constants n, n + 1 are selected in such a way that the operator

n̂ = â â (15)

can be interpreted as the operator of the excitations’ number: n̂|n = n|n.

The creation–annihilation operators satisfy the canonical commutation relation:

[ â, â ] = â â − â â = I, (16)

where I is the unit operator.

The information field Hamiltonian is given by

ĤF = EF n̂. (17)

The n-excitation state |n is the eigenstate of the field Hamiltonian. In this state, the field energy
equals nEF . In the absence of interactions, it is preserved in the process of field’s evolution.
As everywhere in this paper, we consider s-atoms with the two-level structure of the social
energy, E = E1 , E2 and transition energy E A = E2 − E1 . Energy lowering and rising can be formally
described by creation and annihilation operators. In contrast to the information field, these are
fermionic operators:

b̂| E1  = 0, b̂ | E1  = | E2 , b̂| E2  = | E1 , b̂ | E2  = 0. (18)

Hence, they satisfy the fermionic canonical commutation relation:

{b̂, b̂ } = b̂b̂ + b̂ b̂ = I. (19)

These operators are also known as level lowering and rising operators. Here, the vectors | E1  and
| E2  represent the ground and excited states, respectively. Similarly to the field’s excitations number
operator n̂, we set n̂ A = b̂ b. We have n̂ A | E1  = 0 and n̂ A | E2  = | E2 .
We also note that the field and s-atom’s operators commute:

[ â, b̂] = [ â, b̂ ] = [ â , b̂] = [ â , b̂ ] = 0. (20)

The reader should not be disappointed that to model transitions between s-atom’s states as well
as information filed’s states, we use the same operator algebra as in physics. The formalism of creation

243

www.pdfgrip.com
Entropy 2018, 20, 921

and annihilation operators can be used in all models describing transitions between states. For example,
besides quantum physics, this formalism plays the important role in analysis of reaction–diffusion
equations. This formalism is widely used to model human cognition, decision-making, in finances,
and political studies [20,29,34].
Hamiltonian of an s-atom is given by

Ĥ A = E1 + E A n̂ A . (21)

Since Ĥ A | Eα  = Eα | Eα , α = 0, 1, in this model, an s-atom who is isolated from the information

field and being in the ground state cannot become excited by herself and being in the excited state
cannot emit spontaneously a social excitation and relax.
In fact, the forms of the field and s-atom Hamiltonians are selected to preserve the states of
the concrete social energy (in the absence of interactions). Thus, the forms of these Hamiltonians
express the law of energy conservation, but only for the states of the concrete energy. The same energy
conservation constraint is supposed in quantum physics. Therefore, the reader should not surprised
that we proceed with Hamiltonians of the same form as in physics.

9.2. Dynamics of the Compound System s-Atom-Field

Interaction between the quantum information field and s-atoms is described by Hamiltonian

Ĥ I = −γg(b̂(t) + b̂ (t)) Ê(t), (22)

where the parameter g > 0 expresses the strength of coupling between s-atoms and the
information field.
The dynamics of s-atom’s creation and annihilation operators is given by
EA EA
−it it
b̂(t) = e γ b̂, b̂ (t) = e γ b . (23)

This dynamics is just the Heisenberg picture of the state dynamics given by the
Schrödinger equation.
Consider action of the s-atom component of the interaction Hamiltonian at t = 0,

(b̂ + b̂ )| E1  = | E2 , (b̂ + b̂ )| E2  = | E1 . (24)

This is the flipping-operator representing transitions between states of an s-atom. Such transitions
should be compensated by modification of the information field state. Mathematically, this interaction
process is represented as composition of two operators.
Formally, we should also use the time-dependent creation and annihilation operators in free
Hamiltonians Ĥ A , ĤF . However, these Hamiltonians contain only compositions of operators of the
forms b̂ (t)b̂(t), â (t) â(t) and complex exponents containing time dependence cancel each other.
Finally, the interaction Hamiltonian can be represented in the form:
& E +E E +E
−it F γ A it F γ A  
Ĥ I = −γg ue âb̂ + ūe â b̂ + (25)

EF − E A EF − E A '
−it it
ue γ âb̂ + ūe γ â b̂ .

The Hamiltonian of the compound system, the quantum information field carrying the social
energy EF and interacting with an s-atom having the social energy spectrum E = E1 , E2 , is given by

Ĥ = Ĥ A + ĤF + Ĥ I .

244

www.pdfgrip.com
Entropy 2018, 20, 921

Denote by ρ the state of this compound system. Its dynamics is described by the
von Neumann equation:
∂ρ
iγ (t) = [ Ĥ, ρ(t)], (26)
∂t
with the initial condition
ρ ( t0 ) = ρ A ( t0 ) ⊗ ρ F ( t0 ). (27)

Here, it is supposed that at t = t0 s-atom’s and field’s information states are uncorrelated.
Mathematically, this absence of correlations is represented by factorization of the compound system
state into the states of the s-atom and the field.
As typical in physical considerations, it is convenient to move to the interaction representation.
In this representation, the two components of dynamics, one given by Hamiltonian Hfree = Ĥ A + ĤF
and another by interaction Hamiltonian H I , are spit. The first part is used for transformation of the
density operator ρ(t) of the form ρ̃(t) = U  (t)ρ(t)U (t), where U (t) is the unitary one parametric
group describing the dynamics generated by Hfree , i.e., U (t) = e−iHfree /γ . This is the transformation to
the interaction representation. In addition, in the latter state’s dynamics is generated by H I .
To simplify notations, in the interaction representation, we will use the symbol ρ(t) to denote the
state (instead of symbol ρ̃(t)). In this representation, the von Neumann equation has the form:

∂ρ
iγ (t) = [ Ĥ I , ρ(t)]. (28)
∂t
This equation can be solved approximately by iterated integration starting with
 t
1
ρ (1) ( t ) = ρ ( t 0 ) + [ Ĥ I , ρ(s)]ds. (29)
iγ t0

9.3. Gorini–Kossakowski–Sudarshan–Lindblad Equation for the State of the Quantum Information Field
By using the above integral iterations and under some assumptions (see Section 9.4), we can
derive the approximate quantum master equation for the reduced density operator of the field,

ρ F (t) ≡ TrA æ(t),

where the partial trace is with respect to basis | E1 , | E2  in the state space of s-atom’s social energy.
This is the special case of the quantum Markov approximation for the dynamics of a subsystem—of a
compound quantum system described by the GKSL-equation.
Our main task is to present the proper social interpretations for the parameters of the
GKSL-equation. For s-atoms composing the gain medium, denote by Ti , i = 1, 2, the average times of
being in the ground and excited states, respectively. We can call them the lifetimes of relaxation and
excitation. Denote by r2 the rate of excitation of s-atoms generated by social energy pumping into this
gain medium and by r1 the absorption rate.
The quantum master equation has the form:

∂ρ F
(t) = − A[ â â ρ F (t) + â ρ F (t) â + h.c] + B[ââ æF (t) + 3ââ æF (t)ââ − 4â ââ æF (t)â + h.c] (30)
∂t

−C [ â âρ F (t) + âρ F (t) â + h.c],

where h.c is the abbreviation for “Hermitian conjugate.” Here,

A = r2 ( gT2 )2 |u|2 (31)

and
C = r1 ( gT1 )2 |u|2 (32)

245

www.pdfgrip.com
Entropy 2018, 20, 921

are the gain and loss coefficients, respectively. These coefficients depend quadratically on the amplitude
of the information field u, the interaction coefficient g, and excitation and relaxation lifetimes T2
and T1 . They are linearly proportional to excitation and absorption rates, r2 , r1 . The term with the
saturation coefficient
B = r2 ( gT2 )4 |u|4 (33)

plays the crucial role in generation of exponential increase of the number of excitations in the field.
We shall turn to this question later by considering the dynamics of probabilities.
We remark that, since in Equation (30) the operator coefficients are time-independent,
the dynamical state update is not based on the long term memory. If we consider a discrete time
approximation of this dynamics, ρ F (tk ), tk = t0 + kδt, then the state at the moment tk+1 is completely
determined by the state at t = tk . Such property is known as the Markov property.

9.4. Social Interpretation of Assumptions for Derivation of Quantum Master Equation

There are two basic assumptions for derivation of master Equation (30) from von Neumann
Equation (26) (see [83]). To formulate these two assumptions, it is convenient to introduce the time scale
γ
τF = (34)
EF

of the evolution of field’s mode with the energy EF and the time scale
γ
τA = (35)
EA

of transition between the states | Ei , i = 1, 2.

Assumption 1. Analysis of dynamics (30) can be essentially simplified under the following condition:
γ
2π << T2 . (36)
EA

In this situation, the first two terms in the expression (25) of the interaction Hamiltonian H I can be neglected in
the process of integration with respect to time (see (29)).
Inequality (39) can be written as
T2
>> 2π. (37)
τA
Thus, the lifetime of the excited state of an s-atom should be long enough comparing with the transition
time scale τA . We note that the latter is inversely proportional to the resonance energy E A = E2 − E1 .
For a gain medium with a large gap of the social energy, it is easier to satisfy the condition (37). For such a
social group, the master Equation (30) gives a better approximation. Of course, the condition (37) expresses the
complex interplay between the magnitudes of the lifetime for the excited state | E2  and the size of the energy
gap E A .
A good gain medium is characterized by the long lifetime of the excited state and the big gap between the
states of relaxation and excitation.
For such a gain medium, the master equation approximation gives the adequate picture of social processes
in the gain medium (under additional Assumption 2).

Assumption 2. The difference between the energies E A and EF is very small, so the social energy of quanta
associated with communications compounding the field differs not so much from the resonance energy of
s-atoms—the social energy for transition between s-atoms’ states:

T2
| EF − E A | << 1. (38)
γ

246

www.pdfgrip.com
Entropy 2018, 20, 921

This condition of matching social energies is natural for s-atoms interacting with the information field.
The social energy EF carried by excitations of the information field need not be exactly equal to the resonance
energy of s-atoms in the gain medium. It can deviate from the latter, but not so much. The right formulation
of this statement can be done in the probabilistic terms. The emission spectrum of the gain medium has the
Gaussian distribution with the mean value EF and sufficiently small standard deviation σA , where σA Tγ2 << 1.

Technically, the condition (38) justifies the following approximation:

EF − E A
±it
e γ ≈ 1, t0 ≤ t ≤ T2 , (39)

in the last two terms of interaction Hamiltonian (25). By taking into account that the first two
terms in expression (25) can be neglected due to Assumption 1 interaction Hamiltonian H I , can be
approximately treated as a time-independent operator.
The condition (38) can be rewritten as

EF τ
|1 − | << A . (40)
EA T2

Therefore, for the gain medium which is “good” from the viewpoint of Assumption 1, i.e.,
τA
T2 << 1, the lifetime is long and the social energy gap is high, this condition is satisfied only for very
sharp Gaussian distribution, with the mean value EF , of the emission spectrum of the gain medium.

9.5. Probabilistic Consequences of the Quantum Markov Dynamics

Now, by using quantum master Equation (30), we can describe the dynamics of probability
p(t, n) to find n excitations in social laser’s resonator, e.g., in the form of an Internet Echo Chamber.
We have p(t, n) = n|ρ F (t)|n. By averaging Equation (30) with respect to the state |n, we obtain the
probabilistic dynamics (see [84]):

∂p
(t, n) = − A[(n + 1) p(t, n) − np(t, n − 1)] − C [np(t, n) − (n + 1) p(t, n + 1)] (41)
∂t

+ B[(n + 1)2 p(t, n) − n2 p(t, n − 1)].

As in physics, we are interested in the steady state of this dynamics, the state of equilibrium in
the Echo Chamber. After approaching this state, the social resonator can be used to emit the powerful
∂p
wave of social actions. To find a steady state, we set ∂t (t, n) = 0 and obtain the recurrence equation:

p(n) = ( A/C )(1 − nB/A) p(n − 1). (42)

In spite of simplicity of this recurrence equation, its solution cannot be represented in a compact
analytic form. In physics, one considers approximations corresponding different ranges of values of
the parameter A/C, the pump parameter, representing interrelation of the gain and loss. The equality
A/C = 1 is the threshold condition for the laser.

1. If A/C < 1, then the solution of Equation (42) can be approximately represented in the form
p(n) ≈ (1 − A/C )( A/C )n . Thus, for this region of variation of the parameter A/C, the field is
characterized by a small number of excitations: the probability that, in the resonator of the social
laser, there can be found n excitations decreases exponentially, p(n) ∼ e−(ln C−ln A) .
2. If A/C ≈ 1, then the solution has no simple analytical representation. This range of variation of
parameter A/C is characterized by large fluctuations of number n of field’s excitations.
3. If A/C >> 1, then p(n) can be approximated by the Poission distribution:

n̄n
p(n) ≈ e−n̄ , (43)
n!

247

www.pdfgrip.com
Entropy 2018, 20, 921

where n̄ = A2 /CB is the average number of excitations. It is crucial that the standard deviation
√
of the Poission distribution σ = n̄. This implies that the Gaussian distribution approximating
the Poisson distribution with the mean value n̄ = A2 /CB >> 1 is concentrated around n̄. Thus
(with the high probability), the number of excitations n present in the resonator of the social laser
is very large.
In some papers, one sets A/C = eb , i.e., b = ln( A/C ). Then, the threshold value equals to zero.

10. Conclusions
This paper is a new step towards clarification of the basic notions of the quantum-like social
lasing model. We heavily refer to the information interpretations of quantum mechanics and the
operational approach. The latter is based on the Copenhagen interpretation of quantum mechanics
and Bohr’s emphasis that quantum theory is about observations and not genuine physical processes in
the micro-world. This approach is applied to formalization of the notion of the social (mental) energy.
Although this notion has been discussed by psychologists and philosophers [64–66] for a few hundred
years, its proper formalization was missed. It seems that this notion can be handled properly only in
the quantum framework, cf., however, with other recent attempts [69–71].
The Bose–Einstein statistics of excitations of the quantum electromagnetic field, photons, plays
the crucial role in generation of the cascade process of stimulated emission. Following Schrödinger [75]
who used Gibbs ideal ensembles to derive thermodynamical quantities from statistical mechanics,
in [47], one of the authors of this paper considered thermodynamics of excitations of the quantum
information field. The Bose–Einstein statistics can be derived for them under the assumption of
their indistinguishability with respect to the social energy. In this paper, we analyze the meaning
of such indistinguishability and emphasize the crucial role of information overload in its creation.
The Bose–Einstein statistics of information excitations matches well with one of the basic effect of
social psychology, the bandwagon effect. We also present briefly thermodynamical derivation of the
Bose–Einstein statistics for information excitations [47] by emphasizing the operational meaning of the
notion of the social energy.
We discuss in a lot of detail the notion of the social color of an excitation of the quantum
information field. This quantity represents characteristics of information excitations linked to social
lasing coherence. The role of the social color in the process of energy pumping versus amplified
coherent emission is clarified.
As in physical lasing, a resonator is the basic component of social lasing. The standard resonator
of a physical laser is the optical cavity. Therefore, modeling of functioning of a physical laser resonator
is typically presented in the framework of quantum optics. Our aim was to “distill” the physical
scheme from connection with optics. We proceeded with a theory of open quantum systems and the
quantum Markov approximation (given by the GKSL-equation) of the dynamics of the compound
system, s-atoms interacting with the quantum information field. As is typical in quantum-like
modeling, we borrowed the mathematical formalism of quantum physics, but assigned new (social)
interpretations to the basic quantities and parameters. This interpretational analysis of the (standard)
mathematical expressions highlights the following social lasing constraints on the human gain medium
and the quantum information field:
• A gain medium is characterized by the long lifetime of the excited state and the big gap between
the states of relaxation and excitation.
• The social energy carried by excitations of the information field has to match the resonance energy
of s-atoms in the gain medium.
• The interrelation of the magnitudes of the excitation and absorption rates r2 , r1 and the lifetimes
of the corresponding levels T2 , T1 has to imply inequality A > C, where A and C are gain and
loss coefficients, respectively.
• The nonlinear character of interactions between excitations in a social laser resonator (encoded in
the B-coefficient) plays the crucial role in initiation of stable social lasing.

248

www.pdfgrip.com
Entropy 2018, 20, 921

• The quantum-like regime of lasing is characterized by the threshold value of the pump parameter.

The presented study on the foundational side of the quantum-like modeling of social lasing is
important for its further development, since, as was mentioned in the Introduction, application of the
quantum theory outside physics requests reanalyzing of quantum methodology.
This paper may even have some impact for quantum foundations, especially the information
interpretations of quantum mechanics [51–63]. The application of the mathematical formalism of
quantum mechanics outside physics supports (at least indirectly) the claim that this formalism basically
reflects the special laws of information processing. (The latter is confirmed by its derivation from the
natural informational principles [57–59].) In particular, the thermodynamical derivation of quantum
statistics (Section 4.4) highlights the role of indistinguishability of information quanta.

Author Contributions: Conceptualization, A.K.; Methodology, A.K. and A.A.; Validation, A.T. and D.T.; Formal
Analysis, A.K.; Investigation, A.K. and A.A.; Writing, Original Draft Preparation, A.K.; Writing, Review & Editing,
A.A., A.T., and D.T.; Supervision, A.K.; Project Administration, A.A.; Funding Acquisition, A.A.
Funding: This research was funded by the Government of the Russian Federation, Grant 08-08 and by the Ministry
of Education and Science of the Russian Federation within the Federal Program Research and development in
priority areas for the development of the scientific and technological complex of Russia for 2014-2020, Activity 1.1,
Agreement on Grant No. 14.572.21.0008 of 23 October, 2017, unique identifier: RFMEFI57217X0008.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Jung, C.G.; Pauli, W. Atom and Archetype. The Pauli/Jung Letters 1932–1958; Princeton University Press:
Princeton, NJ, USA, 2014.
2. Atmanspacher, H.; Fuchs, C.A. (Eds.) The Pauli–Jung Conjecture: And Its Impact Today; Imprint Academic:
Exeter, UK, 2014.
3. Whitehead, A.N. Process and Reality; Macmillan: New York, NY, USA, 1929.
4. Whitehead, A.N. Adventures of Ideas; Macmillan: New York, NY, USA, 1933.
5. Ricciardi, L.M.; Umezawa, H. Brain Physics and Many-Body Problems. Kibernetik 1967, 4, 44–48. [CrossRef]
6. Penrose, R. The Emperor’s New Mind; Oxford University Press: Oxford, UK, 1989.
7. Umezawa, H. Advanced Field Theory: Micro, Macro and Thermal Concepts; AIP: New York, NY, USA, 1993.
8. Hameroff, S. Quantum coherence in microtubules. A neural basis for emergent consciousness? J. Cons. Stud.
1994, 1, 91–118.
9. Vitiello, G. Dissipation and memory capacity in the quantum brain model. Int. J. Mod. Phys. 1995, B9, 973.
[CrossRef]
10. Vitiello, G. My Double Unveiled: The Dissipative Quantum Model of Brain; Advances in Consciousness Research;
John Benjamins Publishing Company: Amsterdam, The Netherlands, 2001.
11. Ezhov, A.A.; Berman, G.P. Introduction to Quantum Neural Technologies; Rinton Press: Paramus, NJ, USA, 2003.
12. Bernroider, G.; Summhammer, J. Can quantum entanglement between ion transition states effect action
potential initiation? Cogn. Comput. 2012, 4, 29–37. [CrossRef]
13. Igamberdiev, A.U.; Shklovskiy-Kordi, N.E. The quantum basis of spatiotemporality in perception and
consciousness. Prog. Biophys. Mol. Biol. 2017, 130A, 15–25. [CrossRef]
14. Bernroider, G. Neuroecology: Modeling neural systems and environments, from the quantum to the classical
level and the question of consciousness. J. Adv. Neurosc. Res. 2017, 4, 1–9. [CrossRef]
15. Basti, G.; Capolupo, A.; Vitiello, G. Quantum field theory and coalgebraic logic in theoretical computer
science. Prog. Biophys. Mol. Biol. 2017, 130A, 39–52. [CrossRef]
16. Khrennikov, A. Information Dynamics in Cognitive, Psychological, Social, and Anomalous Phenomena; Series
Fundamental Theories of Physics; Kluwer: Dordreht, The Netherlands, 2004.
17. Pylkkänen, P. Mind, Matter and the Implicate Order; Springer Frontiers Collection: Berlin/Heidelberg,
Germany, 2007.
18. Khrennikov, A. Ubiquitous Quantum Structure: From Psychology to Finances; Springer: Berlin/Heidelberg,
Germany, 2010.

249

www.pdfgrip.com
Entropy 2018, 20, 921

19. Busemeyer, J.R.; Bruza, P.D. Quantum Models of Cognition and Decision; Cambridge Press: Cambridge,
UK, 2012.
20. Bagarello, F. Quantum Dynamics for Classical Systems: With Applications of the Number Operator; Wiley Ed.:
New York, NY, USA, 2012.
21. Haven, E.; Khrennikov, A. Quantum Social Science; Cambridge University Press: Cambridge, UK 2013.
22. Asano, M.; Khrennikov, A.; Ohya, M.; Tanaka, Y.; Yamato, I. Quantum Adaptivity in Biology: From Genetics to
Cognition; Springer: Dordrecht, The Netherlands, 2015.
23. Haven, E.; Khrennikov, A.; Robinson, T.R. Quantum Methods in Social Science: A First Course; WSP:
Singapore, 2017.
24. Khrennikov, A. Classical and quantum mechanics on information spaces with applications to cognitive,
psychological, social and anomalous phenomena. Found. Phys. 1999, 29, 1065–1098. [CrossRef]
25. Busemeyer, J.B.; Wang, Z.; Townsend, J.T. Quantum dynamics of human decision-making. J. Math. Psych.
2006, 50, 220–241. [CrossRef]
26. Haven, E. Private information and the ‘information function’: A survey of possible uses. Theor. Decis. 2008,
64, 193–228. [CrossRef]
27. Pothos, E.M.; Busemeyer, J.R. A quantum probability explanation for violation of rational decision theory.
Proc. Royal. Soc. B 2009, 276, 2171–2178. [CrossRef] [PubMed]
28. Brandenburger, A. The Relationship between quantum and classical correlation in games. Game Econ. Behav.
2010, 69, 175–183. [CrossRef]
29. Bagarello, F.; Oliveri, F. A phenomenological operator description of interactions between populations with
applications to migration. Math. Models Methods Appl. Sci. 2013, 23, 471–492. [CrossRef]
30. Plotnitsky, A. Are quantum-mechanical-like models possible, or necessary, outside quantum physics?
Phys. Scripta 2014, T163, 014011. [CrossRef]
31. Boyer-Kassem, T.; Duchene, S.; Guerci, E. Quantum-like models cannot account for the conjunction fallacy.
Theor. Decis. 2015, 10, 1–32. [CrossRef]
32. Broekaert, J.; Basieva, I.; Blasiak, P.; Pothos, E.M. Quantum dynamics applied to cognition: A consideration
of available options. Phil. Trans. R. Soc. A 2017, 375, 20160387. [CrossRef]
33. Khrennikov, A. Quantum Bayesianism as the basis of general theory of decision-making. Phil. Trans. R. Soc.
A 2016, 374, 20150245. [CrossRef]
34. Khrennikova, P. Modeling behavior of decision makers with the aid of algebra of qubit creation-annihilation
operators. J. Math. Psych. 2017, 78, 76–85. [CrossRef]
35. Conte, E.; Khrennikov, A.; Todarello, O.; Federici, A. A preliminary experimental verification on the
possibility of Bell inequality violation in mental states. Neuroquantology 2008, 6, 214–221. [CrossRef]
36. Asano, M.; Khrennikov, A.; Ohya, M.; Tanaka, Y.; Yamato, I. Violation of contextual generalization of the
Leggett-Garg inequality for recognition of ambiguous figures. Phys. Scr. 2014, T163, 014006. [CrossRef]
37. Dzhafarov, E.N.; Kujala, J.V. On selective influences, marginal selectivity, and Bell/CHSH inequalities.
Top. Cogn. Sc. 2014, 6, 121–128. [CrossRef] [PubMed]
38. Barros, J.; Toffano, Z.; Meguebli, Y.; Doan, B.-L. Contextual Query Using Bell Tests. In Proceedings of the
Quantum Interaction, Leicester, UK, 25–27 July 2013; Atmanspacher, H., Haven, E., Kitto, K., Raine, D., Eds.;
Springer: Berlin/Heidelberg, Germany; ISBN 978-3-642-54942-7.
39. Allais, M. Le comportement de l’homme rationnel devant le risque: Critique des postulats et axiomes de l’
cole amricaine. Econometrica 1953, 21, 503–536. [CrossRef]
40. Ellsberg, D. Risk, ambiguity and the Savage axioms. Q. J. Econ. 1961, 75, 643–669. [CrossRef]
41. Machina, M. Risk, Ambiguity and the Dark-dependence Axiom. Am. Econ. Rev. 2009, 99, 385–392. [CrossRef]
42. Krastev, I. In Mistrust We Trust: Can Democracy Survive When We Don’t Trust Our Leaders? TED Conferences
Press: Monterey, CA, USA, 2013.
43. Mason, P. Why It’s Kicking off Everywhere: The New Global Revolutions; Verso: London, UK, 2012.
44. Fukuyama, F. The middle-class revolution. The Wall Street Journal, 28 June 2013.
45. Schmidt, E.; Cohen, J. The New Digital Age: Reshaping the Future of People, Nations and Business; Knopf:
New York, NY, USA, 2013.
46. Krastev, I. Democracy Disrupted. the Global Politics of Protest; Penn Press: Philadelphia, PA, USA, 2014.
47. Khrennikov, A. Towards information lasers. Entropy 2015, 17, 6969–6994. [CrossRef]

250

www.pdfgrip.com
Entropy 2018, 20, 921

48. Khrennikov, A. ‘Social laser’: Action amplification by stimulated emission of social energy. Phil. Trans. R. Soc.
2016, 374, 20150094. [CrossRef] [PubMed]
49. Khrennikov, A. Social laser model: From color revolutions to Brexit and election of Donald Trump. Kybernetes
2018, 47, 273–278. [CrossRef]
50. Khrennikov, A. (Ed.) Quantum Theory: Reconsideration of Foundations; Serise Math. Modelling in Physics,
Engineering, and Cognitive Science; Vaxjo University Press: Vaxjo, Sweden, 2002; Volume 2.
51. Zeilinger, A. A foundational principle for quantum mechanics. Found. Phys. 1999, 29, 631–643. [CrossRef]
52. Brukner, C.; Zeilinger, A. Information invariance and quantum probabilities. Found. Phys. 2009, 39, 677–689.
[CrossRef]
53. Caves, C.M.; Fuchs, C.A.; Schack, R. Quantum probabilities as Bayesian probabilities. Phys. Rev. A 2002,
65, 022305. [CrossRef]
54. Fuchs, C.A. Quantum mechanics as quantum information (and only a little more). In Proceedings of Quantum
Theory: Reconsideration of Foundations; Vaxjo University Press: Vaxjo, Sweden, 2002; pp. 463–543.
55. Fuchs, C.A.; Schack, R. QBism and the Greeks: Why a quantum state does not represent an element of
physical reality. Phys. Scr. 2014, 90, 015104. [CrossRef]
56. Fuchs, C.A.; Mermin, N.D.; Schack, R. An introduction to QBism with and application to the locality of
quantum mechanics. Am. J. Phys. 2014, 82, 749–754. [CrossRef]
57. Chiribella, G.; D’Ariano, G.M.; Perinotti, P. Probabilistic theories with purification. Phys. Rev. A 2010,
81, 062348. [CrossRef]
58. D’Ariano, G.M. Operational axioms for quantum mechanics. In Proceedings of the Foundations of Probability
and Physics-4, Vaxjo, Sweden, 4–9 June 2006; Volume 889, pp. 79–105.
59. Chiribella, G.; D’Ariano, G.M.; Perinotti, P. Informational axioms for quantum theory. In Proceedings of
the Foundations of Probability and Physics-6, Vaxjo, Sweden, 14–16 June 2011; Volume 1424, pp. 270–279,
ISBN 978-0-7354-1004-6.
60. Plotnitsky, A. Reading Bohr: Physics and Philosophy; Springer: Dordrecht, The Netherlands, 2006.
61. Plotnitsky, A. Epistemology and Probability: Bohr, Heisenberg, Schrödinger, and the Nature of Quantum-Theoretical
Thinking; Springer: New York, NY, USA, 2009.
62. Plotnitsky, A. Niels Bohr and Complementarity: An Introduction; Springer: New York, NY, USA, 2012.
63. Khrennikov, A. External Observer Reflections on QBism, Its Possible Modifications, and Novel Applications.
In Quantum Foundations, STEAM-H: Science, Technology, Engineering, Agriculture, Mathematics & Health;
Khrennikov, A., Toni, B., Eds.; Springer: Cham, Switzerland, 2018; pp. 93–118.
64. James, W. The Principles of Psychology; Henry Holt and Co.: New York, NY, USA, 1890; Reprinted Harvard
University Press: Boston, MA, USA, 1983.
65. Freud, S. The Ego and the Id; W.W. Norton and Company: New York, NY, USA, 1923.
66. Freud, S. The Interpretation of Dreams; In Standard edition 1954–1974; Hogarth Press: London, UK, 1900;
Volumes 4, 5, pp. 1–627.
67. Jung, C.G. On the Nature of the Psyche; Routledge Classics: London, UK, 2001.
68. Khrennikov, A. Quantum-like modeling of cognition. Front. Phys. Interdisciplin. Phys. 2015, 3, 77. [CrossRef]
69. Grauwin, S.; Bertin, E.; Lemoy, R.; Jensen, P. Competition between collective and individual dynamics.
Proc. Natl. Acad. Sci. USA 2009, 106, 20622–20626. [CrossRef] [PubMed]
70. Jaeger, C.; Horn, G.; Lux, T.; Fricke, D.; Frurst, S.; Lass, W.; Lin, L.; Mandel, A.; Meissner, F.; Schreiber, S.; et al.
From the Financial Crisis to Sustainability? Potsdam, European Climate Forum. Available online: https:
//globalclimateforum.org/?id=ecfreports (accessed on 1 December 2018).
71. Collins, R. Interaction Ritual Chains; Princeton University Press: Princeton, NJ, USA, 2004.
72. Von Neuman, J. Mathematical Foundations of Quantum Mechanics; Princeton University Press: Princeton, NJ,
USA, 1955.
73. Khrennikov, A.; Basieva, I. Possibility to agree on disagree from quantum information and decision-making.
J. Math. Psychology 2014, 62, 1–5. [CrossRef]
74. Dubois, F.; Toffano, Z. Eigenlogic: A Quantum View for Multiple-Valued and Fuzzy Systems. In Lecture Notes
in Computer Science, Proceedings of the Quantum Interaction, San Francisco, CA, USA, 20–22 July 2016; de Barros J.,
Coecke, B., Pothos, E., Eds.; Springer: Cham, Switzerland, 2016; Volume 10106, ISBN 978-3-319-52288-3.
75. Schrödinger, E. Statistical Thermodynamics; Dover Publications: Mineola, NY, USA, 1989.
76. Colman, A. Oxford Dictionary of Psychology; Oxford University Press: Oxford, UK, 2003; p. 77.

251

www.pdfgrip.com
Entropy 2018, 20, 921

77. Ingarden, R.S.; Kossakowski, A.; Ohya, M. Information Dynamics and Open Systems: Classical and Quantum
Approach; Springer: Berlin/Heidelberg, Germany, 1997.
78. Khrennikova, P.; Haven, E.; Khrennikov, A. An application of the theory of open quantum systems to
model the dynamics of party governance in the US political system. Int. J. Theor. Phys. 2014, 53, 1346–1360.
[CrossRef]
79. Fleck, J.A., Jr. Quantum theory of laser radiation. Many-atom effects. Phys. Rev. 1966, 149, 309–319. [CrossRef]
80. Haken, H. Light, Vol. II: Laser Light Dynamics; North-Holland Phyics Publishing: Amsterdamn,
The Netherlands, 1985.
81. Erneux, T.; Glorieux, P. Laser Dynamics; Cambridge University Press: Cambridge, UK, 2010.
82. Barlow, T.M.; Bennett, R.; Beige, A. A master equation for a two-sided optical cavity. J. Mod. Opt. 2015, 62,
S11–S20. [CrossRef] [PubMed]
83. Priyashanka, K.M.; Wijewardena Gamalath, K.A.I.L. Coherent states in a laser cavity. Int. Lett. Chem.
Phys. Astron. 2015, 56, 47–62. [CrossRef]
84. Greenberger, D.M.; Erez, N.; Scully, M.O.; Svidzinsky, A.A.; Zubairy, M.S. Planck, photon statistics and
Bose–Einstein condensation. Prog. Opt. 2007, 50, 273–330.

c 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

252

www.pdfgrip.com
 entropy
Article
Paths of Cultural Systems
ID
Paul Ballonoff
Owner and Operator, Ballonoff Consulting, 9307 Kings Charter Drive, Richmond, VA 23116, USA;
Paul@Ballonoff.net; Tel.: +1-703-780-1761

Received: 6 November 2017; Accepted: 17 December 2017; Published: 25 December 2017

Abstract: A theory of cultural structures predicts the objects observed by anthropologists. We here
define those which use kinship relationships to define systems. A finite structure we call a partially
defined quasigroup (or pdq, as stated by Definition 1 below) on a dictionary (called a natural
language) allows prediction of certain anthropological descriptions, using homomorphisms of pdqs
onto finite groups. A viable history (defined using pdqs) states how an individual in a population
following such history may perform culturally allowed associations, which allows a viable history to
continue to survive. The vector states on sets of viable histories identify demographic observables on
descent sequences. Paths of vector states on sets of viable histories may determine which histories
can exist empirically.

Keywords: quantum logic; groups; partially defined algebras; quasigroups; viable cultures

1. Ethnographic Foundation
The structures described here and their consequences imply much of what may be predicted
about empirical cultures. Anthropologists very often draw illustrations of structures using methods
discussed here, but based on intuition, thus have little notion of what their commonly used diagrams
might predict. While [1,2] defined mathematical means to describe the current and future demographic
organization of lineage organizations, with empirical examples, we here specify the demography of
kinship-based systems [3–8] with some related definitions in our Appendix A; and empirical examples
in [9–13]. We follow the inspiration of [14]. In an empirical culture many other relations may also
occur; we note some of those in our Part 7, Discussion at the end.
Our notion of studying viable minimal structures—which are the smallest minimal cultural
structures that can “reproduce” the ascribed social relations in one generation—follows from [15].
Our history describes how sustaining those relations allow the culture to reproduce the rules. Cultural
rules describing histories may be stated in natural languages, which label the individuals in a descent
sequence with a subset called a kinship terminology. Our term viable embodies what anthropologist
Radcliff-Brown called “persistent cultural systems” ([10], p. 124). Radcliff-Brown and others often
described histories using discrete generations, as do we. Empirical cultures are almost ubiquitously
described by many anthropologists using viable histories, typically represented in an ethnography by
its minimal structure. The nearly ubiquitous presence in ethnographies of viable histories implies they
may be the only observed histories.
An example is Figure 1 (whose source is [9]) which is an actual viable minimal structure of a history
(that is, a persistent cultural system). The triangles in the illustration are males, the circles are females,
descent moves in the downward direction, the labels are names used in the kinship terminology.
The sign “=” means “marriage” between the two individuals attached to it; the “=” on the far right
in the illustration shows a marriage to the partner on the far left in each generation. The horizontal
line which connects two individuals shows that those two are assigned descendants of the marriage
above them. The fifth generation in this illustration is equal to the first (by its labels in vertical descent
in the diagram), showing that minimal structure shown here reproduces the labelled culture here

Entropy 2018, 20, 8; doi:10.3390/e20010008 253 www.mdpi.com/journal/entropy

www.pdfgrip.com
Entropy 2018, 20, 8

in four generations, but reproduces the minimal structure (the graphs without the kinship labels)
in each generation. This minimal structure has 4 marriages in each generation hence has structural
number s = 4. The minimal structure is not intended to illustrate the actual empirical relations of each
empirical generation of individuals, instead it shows “how the rule operates”—it describes the minimal
representation of the kinship and marriage rules (see Appendix A Definitions A3). The minimal
structure describes the “principles” used in the rules of the culture.

Figure 1. Example of a typical anthropological illustration of a persistent cultural system per

Radcliff-Brown, which is also a group per Levi-Strauss and Weil.

Figure 1 is also obviously an example of a group. Claude Levi-Strauss [9] initially used groups to
demonstrate histories in an appendix by A. Weil. Ref. [9] mainly discusses groups of orders 2, 4 and 8,
though its illustration 1.17 shows a helical structure of order 4. Other studies include orders 3, 6 and
others. Ref. [16] also shows a helical structure of order 7; helical minimal structures are also groups
and have surjective descent sequences, so our modal demography discussed in [8] and Appendix A
Definition A7 also apply to helices.

2. Basic Definitions
While lineage organizations [1,2] predicted examples of population measures including the “local”
village size given the lineage structure, the kinship examples described here use values found in [8]
to predict values associated to the structural number of the history, which apply no matter what the
empirical size of the total population (so long as it is at or above the minimal size). Our definitions are
stated in our Appendix A from previous articles (see also Appendix A) and those below.

Definition 1. Let D be a finite non-empty set (called a dictionary) and let * be a partially defined binary
operation on D, such that when x, y ∈ D:

(1) If there exists an a ∈ D such that a*x and a*y are defined and a*x = b and a*y = b then x = y, we call such
object (D, *) a partially defined quasigroup, or pdq.
(2) If (D, *) is a pdq and * is fully defined on D, then (D, *) is a (complete) quasigroup.
(3) The pair L = (D, *) is a natural language with dictionary D whenever (D, *) is a pdq.
(4) If L = (D, *) is a natural language, a kinship terminology is a quasigroup subset k ⊆ L.

254

www.pdfgrip.com
Entropy 2018, 20, 8

Definition 2. Let X, Y and Z be non-empty finite sets and let (X, *), (Y, ◦ ) and (Z, ▪) be quasigroups with
binary relations *, ◦ and ▪ respectively. Then:

(1) A function f: X → Y is a homomorphism if f for all b, c ∈ X, f(b * c) = f(b) ◦ f(c).

(2) If f: X → Z and g: Y → Z are homomorphisms then f and g are isotopic.

All empirical languages are natural languages [7]. Under Definition 2(2) if Y and Z are isotopic
they are also istotopic dictionaries: two possibly different descriptions, thus typically of distinct natural
languages, of the “same” objects or illustrations. Our definition of kinship terminologies based on
quasigroups follows from [17], which refined the discussion of Weil in [9]. While empirical kinship
systems are non-associative [18], a large class is associative and complete, form finite permutations,
indeed groups [19–21] hence form kinship terminologies as defined here. Groups thus arise in
anthropology because a set of all 1-1 mappings of a finite subset set k of a quasigroup (a kinship
terminology k) onto itself forms a group (the symmetric group on k). If (X, *), (Y, ◦ ) and (Z, ▪) are
complete pdqs then (isotopic) homomorphisms classify kinship terminologies by the form of the pdq
onto which they are mapped. For example, the isotopic terminologies classified as Dravidian [12,22,23]
and others are often discussed, in part because they have interesting group theoretical structures.

Definition 3. Let H be a non-empty finite set of viable histories. Let G be a non-empty descent sequence using
H, let Gt ∈ G be a generation of G at t, and let Ht ⊆ H be a subset of t. Then then for each α ∈ Ht , the real
numbers 0 ≤ vα (t) ≤ 1 such that Σα vα (t) = 1 is the vector state of Gt .

Adopting a standard order for listing the histories, we write the vector state at t as v(t) := (vα (t),
. . . , vχ (t)), or when |Ht | = h, as v(t) = (v1 (t), . . . , vh (t)). Let H be a finite non-empty set of viable
histories, let α ∈ H, let Gt ∈ G be a generation of G, and let v(t) be the vector state of Gt (see also
Appendix A Definitions A2–A5). Then:

a. From [2,5,6,8] and Appendix A Definition A4 each structural number s has a set of values ns and
ps where ns ps = 2, where ns is the average family size of a pure system of structural number s and
ps is the proportion of reproducing adults of a pure system of structural number s. If history α
has structural number s, then each α has modal demography (nα , pα ) = (ns , ps ) (see Appendix A
Definition A7) where ps = 2/ns ; for s ≥ 3 and sα = sχ then (nα , pα ) = (nχ , pχ ).
b. Determination of the (ns , ps ) values are based on the Stirling Number of the Second Kind (SNSK)
see [8,24]. We assume here the (ns , ps ) pairs determined by [8].
c. Since H is finite, each non-empty set of viable histories H thus has a largest structural number
smax with modal demography (nmax , pmax ), and a smallest structural number smin with modal
demography (nmin , pmin ). Note that if nmax increases then pmax decreases (and as nmin decreases
then pmin increases, since given s, ns ps = 2, with 0 < p ≤ 1. Structural numbers s = 2 or 3 have
identical modal demography (ns , ps ) = (2, 1); all others structural numbers have distinct modal
demographies see [5,8].
d. The modal demography of history α with structural number sα is (nα , pα ) = (ns , ps ) is a set of
values that represent the history α maintaining its modal demography with neither increase
nor decrease in total empirical population size; it is prediction of nα and pα based on the
determination that the structural number is s, and maintains the structural number s.
e. n(t) = Σα vα (t)nα , α ∈ H, is the predicted average family size of Gt at t, given the vector state at t
see [8]. Note that this is the average family size of the population at time t, given the vector state
of each α ∈ Ht . This while the “size” of the minimal structure might be small, the size predicted
by n(t) is the predicted actual size of the total population at t, not of the minimal structure;
the minimal structure illustration “size” is dependent on the rules, not on the empirical size of
the population.

255

www.pdfgrip.com
Entropy 2018, 20, 8

f. p(t) = Σα vα (t)pα , α ∈ H, is the predicted proportion of reproducing adults of Gt at t ascribed as

married and reproducing, given the vector state of sα at t [8].
g. Thus, all of the “demographics” of cultural theory discussed here are predictions on the result
of maintaining or changing the vector states of t, given the SNSK determined values for each
modal demography (nα , pα ) at time t. Thus, [8] defines

er(t) = 1/2n(t)p(t) (1)

where r(t) ∈ R predicts an average rate of change of total population size between two generations
of G, based on the vector state of structural numbers of the histories Ht ⊆ H. [8] showed that r(t)
predicts changes in the probabilities v(t) imply cultural change is adiabatic.
h. Let H be a finite non-empty set of viable histories, and α, χ ∈ H. Using vα (t) = 1 − vχ (t), nα = 2/pα
and nχ = 2/pχ , then Equation (1) becomes

er(t) = 1 + (nαχ − 2)vα (t) + (2 − nαχ )vα (t)2 (2)

where:
nαχ := (nα 2 + nχ 2 )/(nα nχ ) (3)

is a constant determined by the values of nα and nχ ; note nαχ = nχα .

3. Paths of Descent Sequences

Definition 4. Let H be a finite set of non-empty viable histories. Let α, χ, etc ∈ Ht ⊆ H and let the structural
number of α = χ, etc. If for any such set, |Ht | > 1, vt (α) = 1 or 0, then H is not full; otherwise H is full.

Definition 5. If H is a finite non-empty set of viable histories, then F ⊆ H is a face of H see [2,4,8]. Let I = [1, 0].
A path from point a to point b in a set X is a function f: I → X with f(0) = a and f(1) = b, in which case a is
called the initial point of the path and b is called the terminal point of the path. Given a path, in case a = b then
such path is a closed path. If [x, y] ∈ I and f(x) = a and f(y) = b then f[x, y] is called an interval and a sub-path of
I. A reverse path from point a to point b in X is a function f: I → X with f(1) = a and f(0) = b. Given a path
(or reverse path) from t0 to t1 , if t1 ≥ tk ≥ t0 we say that tk is in the path.

Lemma 1. Let Ht be a finite non-empty set of full viable histories and α, χ ∈ Ht . Then:

1. r(t) ≥ 0, and r(t) = 0 if all structural numbers have the same modal demography or if all have structural
numbers 2 or 3.
2. If α and χ are distinct structural numbers and at least one has structural number >3, then r(t) > 0 and r(t)
has a maximum at v(t) = (0.5, 0.5).
3. Given any finite non-empty set H of two or more viable histories, there is a unique maximum r(t),
given by (ii).

Proof 1. Assume first the modal demographies of histories α and χ are equal (which occurs if sα = sχ or
if sα , sχ < 4). Then nα = nχ . Then nαχ = 2, so the sum in Equation (2) is er(t) = 1, or r(t) = 0. Assume now
sα = sχ and at least one structural number is >3, so the modal demographies of α and χ are distinct,
and we do not have vα (t) = 0 or 1. Then in Equation (2) (nαχ − 2) = −(2 − nαχ ), but vα (t) > 0, so then
vα (t) > vα (t)2 . Thus, (nαχ − 2)vα (t) > (2 − nαχ )vα (t)2 , so Equation (2) states r(t) > 0. 

Proof 2. To show r(t) has a maximum when v(t) = (0.5, 0.5), we use the first two terms of a Taylor
expansion from (1) to find er(t) = 1 + r(t). Differentiating twice gives a2 exp[r(t)]/δy2 = 2 − 2nαχ where
y = v(t). When sα = sχ and at least one of sα , sχ is > 3, nαχ > 2, then exp[r(t)] is concave down; so, r(t) is
also concave down. 

256

www.pdfgrip.com
Entropy 2018, 20, 8

Proof 3. There is a finite number of two-history pairs in H. Since ns increases as s increases, then
Lemma 1 part 1 shows that the largest value of r(t) will be set by that two-history subset α, χ of H
having the largest difference between their structural numbers, hence the largest nαχ in Equation (3).
Combining the ns of the largest s with any other combination of ns values will result in a smaller nαχ
hence smaller er(t) , from Equation (3). 
Observation 1. Assume a finite non-empty set of viable histories H acting on a finite non-empty
descent sequence G. Let α, β, χ ∈ Ht ⊆ H act on generation Gt ∈ G. See Figure 2:

Figure 2. Illustration of the curve np = 2 showing three histories and connections among those three
histories at or above that curve.

The curved bottom-line in Figure 2 is the locus wherever np = 2, so includes the modal
demography of each history in H; that is, the modal demography for each of α, β and χ each appear
on the line np = 2, since ns ps = 2. If for any α, vα (t) = 1, then n(t) = nα , p(t) = pα , so er(t) = 12 n(t)p(t) =
1 1
2 nα pα = 2 2 = 1 so r(t) = 0; this occurs if all histories in Ht have the same structural number (or all have
structural numbers 2 and 3). When that does not occur we have a set of two or more histories in Ht
each with 0 < vα (t) < 1 and thus r(t) > 0; see also Lemma 1. When H is full, assume α, β and χ have
distinct structural numbers, at least two of α, β, χ have structural numbers >3, and α is has the lowest
structural number of those three histories. Then the modal demography of α, β, χ have (nα , bα ) =
(nβ , bβ ) = (nχ , bχ ), and the computation of r(t) appears in the values in the triangle area of Figure 2.
However, if Ht is not full then events in the triangle area might not occur. Even if paths allow α with β,
α with χ and β with χ (thus the boundaries of the triangle area), values of r(t) within the triangle only
occur if all three histories are allowed by H, which may be prohibited by not-full H. We call such area
an un-accessed region. Thus, we study change using both full and non-full sets of histories.

4. Pictures of States on Descent Sequences

Definition 6. Let H be a finite non-empty set of histories and let (nα , pα ) be the modal demography for history
α ∈ H. Let G be a finite non-empty descent sequence using H, and let Gt ∈ G be the generation at time t. Let Ht
be a face of H specified at t. Let St := { sα |α ∈ Hi } be the set of structural numbers St ⊆ S of histories Ht available
at t. Let At := {(nα , pα )|sα ∈ Si , α ∈ Hi , (nα , pα ) = (ns , ps )} be the set of modal demographies At of the histories
in Ht ; and let v(t) be the vector state of Gt . List the histories in H in a defined order from α to χ. Then for all
α ∈ Ht and all (nα , pα ) ∈ At , let:

1. (nt |:= (nα , . . . , nχ ) be a row vector;

2. |pt ) := (pα , . . . , . . . , pχ ) be a column vector;
3. for all α, χ∈ H, arranging the sum of the inner product (nt |pt ) as a square matrix then for all α, χ∈ H,
H(t): = [nα pχ ] is a demographic picture (analogous to a Heisenberg picture in physics) at t;

257

www.pdfgrip.com
Entropy 2018, 20, 8

4. the square matrix we get by arranging the products of v(t)v(t)T as V(t) := [vαχ (t)] is a probability picture
(analogous to a Schroedinger picture in physics) of the vector state of a descent sequence at t;
5. for ε ≥ 0 let V(Δ(t)) := V(t + ε) − V(t) = [vαχ (t + ε) − vαχ (t)] := [Δαχ (t)]. (Notice that −1 ≤ Δαχ (t) ≤ 1).
We note [25] for our analogy of terminology.

Then for all α, χ ∈ H we can rewrite Equation (1) as:

1
er(t) = (n|V(t)|p) (4)
2
using a probability picture which focuses on the vector states; and

1
er(t) = v(t)H(t)v(t)T (5)
2
using a demographic picture which focuses on demographic properties of the histories.

5. Comments on Demographic Pictures

Given a finite non-empty set of full viable histories H, observing a face Ht ⊆ H at t produces a
list of the available Ht ⊆ H and thus creates a list of possible modal demographies (nα , pα ) ∈ At for all
α ∈ Ht . Let |Ht | = h. List the h histories in a fixed order from α to χ, with row vector (nt | = (nα , . . . , nχ )
and column vector |pt ) = (pα , . . . , pχ )T . So for histories α, . . . , χ ∈ Ht , we can write the demographic
picture for |Ht | = h at t as (nt |pt ) = H(t) where:
⎛ ⎞ ⎛ ⎞
n1 p1 ... n1 p h 2 ... n1 p h
⎜ .. ⎟ = ⎜ .. .. ⎟
H (t) = ⎝ ... ..
. . ⎠ ⎝ .
..
. . ⎠ (6)
n h p1 ··· nh ph n h p1 ··· 2

Each diagonal entry = 2 because a diagonal entry nα pα is determined by the modal demography
(ns , ps ) for each history, and ns ps = 2. Thus, using 12 H(t) we can restate Equation (5):
⎛ ⎞
2 ... n1 p h
⎜ .. .. .. ⎟v t T
er (t ) T
= 12 v(t) H (t)v(t) = 12 v(t)⎝ . . . ⎠ ()
n h p1 ··· 2
⎛ 1 ⎞ (7)
1 ... 2 n1 p h
⎜ .. .. .. ⎟ T
= v(t)⎝ . . . ⎠v(t)
1
2 h p1
n ··· 1

Because the two-history case has some useful properties, we present much of our discussion on
the two history version, which becomes:
⎛ ⎞ ⎛ ⎞
1 1
1 2 n1 p2 1 2 n1 p2
r (t) ⎜ ⎟ T ⎜ ⎟
e = v(t)⎝ ⎠v(t) , where H (t) = ⎝ ⎠. (8)
1 1
2 n2 p1 1 2 n2 p1 1

Lemma 2. Let H be a finite non-empty set of full viable histories. Let G be a non-trivial descent sequence using
H, let Ht ⊆ H be the face of H observed at t, and let G(t) ∈ G be the generation at t with vector state v(t). Then
r(t) = 0 only if 12 H(t) = [1] at all entries.

Proof of Lemma 2. Assume the premises. Equations (4)–(7) simply rearrange terms in
1
2 Σi Σj vi (t)vj (t)ni pj . From the definition of modal demography, pi = 2/ni and pj = 2/nj . The values on the

258

www.pdfgrip.com
Entropy 2018, 20, 8

diagonal of 12 H(t) are for each history α, 12 nα pα = 1. We thus examine the off-diagonal products nα pχ
and nχ pα . Then ni pj = 2ni /nj and nj pi = 2nj /ni . Thus, 2ni /nj = 2nj /ni occurs only if ni = nj , in which
case ni pj = nj pi = 2. This occurs only if all histories i and j have the same structural number or both have
s = 2 or 3, and thus 12 H(t) = [1] in all entries. Otherwise stated, in this case the value from Equation (3)
is nαχ = 2. 

Implications of Lemma 2: knowing the modal demography of histories in H we can compute a

proposed population growth rate r(t).

1. The result nα = nχ occurs if structural numbers sα , sχ are <4 or whenever sα = sχ ; so r(t) = 0.

Otherwise, then Lemma 2 implies Lemma 1, which says that er(t) = 1, and thus r(t) > 0. This occurs
since nα pχ does not equal nχ pα ; thus from Lemma 1 and [8] the off-diagonal elements of 12 H(t)
implies adiabatic change in r(t).
2. In discussions in physics, when nα pχ = nχ pα some claim that the resulting r(t) is
“not commutative”. In physics, the “non-commutative” result actually means switching which
experiment is taken, then comparing their results; in physics when changing the order of the
products it also means changing the experiment; but this comparison of the two results also
creates an equation that looks like our Equations (1), (2), (7) or (8). However, in physics reversing
the experiment causes different measurements, which causes the physical uncertainty between
the two results. In contrast, the seemingly “non-commuting” values in culture theory exist
because the equation for computing r(t) requires computing both “directions” of the modal
demography of histories in H (similar to comparing both directions of the physics model), and if
any two (or more) of those have histories of distinct structural numbers (at least one >3), so that
one or more nαχ > 2 (see Equation (3)), then nχ pα = nα pχ . Culture theory thus predicts adiabatic
demographic change, not uncertainty, from a mechanism similar to that which causes uncertainty
in physics.

6. Comments on Probability Pictures

Lemma 3. A probability picture V(t) is symmetric, Σi Σj vi (t)vj (t) = 1 and Σi Σj (Δij (t)) = 0.

Proof of Lemma 3. In Equation (4) V(t) is symmetric since each pair vi (t)vj (t) = vj (t)vi (t).
Since Σα vα (t) = 1 then v(t)v(t)T = Σi Σj vi (t)vj (t) = 1. At t + ε ≥ t (ε < t − (t − 1)) then Σα vα (t + ε) = 1: so
Σi Σj vi (t + ε)vj (t + ε) = 1; so Σi Σj (vij (t + ε) − vij (t)) = Σi Σj (Δij (t)) = 1 − 1 = 0. 

Since we discuss paths of histories, a frequency-domain representation of vector states is useful.

Definition 7. Let r1 , r2 , r3 be real numbers such that r1 2 + r2 2 + r3 2 = 1. Let R be a set of 2 by 2 matrices

with complex entries that forms a ring with respect to matrix addition and multiplication. Let R ⊆ R be a set
of hermitian idempotent matrices of R; and let R ∈ R be such that R = 12 [rij ] where r11 = 1 + r3 , r22 = 1 − r3 ,
r21 = r1 + ir2 , r12 = r1 − ir2 . That is:
 
1= > 1 1 + r3 r1 − ir2
R = rij =
2 2 r1 + ir2 1 − r3

Following ([26], p. 30) we define matrices

       
1 0 0 1 0 −i 1 0
1= , Σ1 = , Σ2 = , Σ3 =
0 1 1 0 +i 0 0 −1

259

www.pdfgrip.com
Entropy 2018, 20, 8

and let z0 , z1 , z2 , z3 , be complex numbers such that R = z0 1 + z1 Σ1 + z2 Σ2 + z3 Σ3 where:

1 1 1 1
z0 = (r11 + r22 ), z1 = (r21 + r12 ), z2 = i (r21 − r12 ), z3 = (r11 − r22 ).
2 2 2 2
The four matrices 1, Σ1 , Σ2 , and Σ3 are the standard Pauli spin matrices, where for R then z0 = 1,
z1 = r1 , z2 = r2 and z3 = r3 . Note that −1 ≤ r1 , r2 , r3 ≤ 1. From ([27], p. 104) R is a set of non-trivial 2
by 2 version of R; a ring of such forms an orthomodular poset and indeed an atomic orthomodular
lattice with the covering property, that is in 1-1 correspondence with the set of closed subspaces of a
two-dimensional complex Hilbert space.

Definition 8. Let H be a finite non-empty set of viable histories, let G be a non-trivial viable descent sequence
using histories Ht ∈ H, let Gt ∈ G be a generation of G using a face Ht ∈ H at t, and let v(t) be the vector state
of Gt .

1. Let 2 Ht = {α, χ} ⊆ Ht be a two-history subset of Ht . Let R(t) = 12 [rij (t)] be a projection, let r1 (t), r2 (t),
and r3 (t) be real numbers such that r1 (t)2 + r2 (t)2 + r3 (t)2 = 1, such that 0 ≤ r1 (t) < 1, 0 ≤ r2 (t) < 1,
and such that vα (t) = 12 r1 (t) = 12 (1 + r3 (t)). Then R(t) is the status of Gt .
2. A unit circle C is meant a set of points (x, y) in the plane R2 which satisfy the equation x2 + y2 = 1.

Theorem 1. Assume the premises of Definition 8. Let H be a finite non-empty set of viable histories having
structural numbers s < 152 (see [8] for use of this limit). Let G be a descent sequence using H. Let R(t) be
the status of Ht , and let v(t) be the vector state of Ht . Let 2 Ht = {α, χ} ⊆ Ht be a non-empty subset of Ht .
Let t2 > t1 > t0 define a path of vα (t) from t = t0 to t = t2 such that vα (t0 ) = 1 changes monotonically to vα (t1 ) = 0
and then monotonically back to vα (t2 ) = 1. That is, let r3 move from r3 (t0 ) = 1 to r3 (t1 ) = − 1 and then back to
r3 (t2 ) = 1. Let O(t) = (n(t), p(t), r(t)). Then:

(1) trR(t) = 1;
(2) vχ (t) = 12 (1 − r3 (t));
(3) the vector state v(t) of 2 Ht is given by the main diagonal of R(t);
(4) r(t) is a maximum when r3 = 0.

Theorem 2. Let r1 (t) = 0. Then: (i) R(t) has ΣΣij rij (t) = 1; and (ii) the sum Σ r(t)dv(t) = 0 when summed over
all paths (all variants of paths) of for pairs 2 Ht .

Theorem 3. Let r2 (t) = 0. Then Σ r(t)dv(t) = 0 when summed over all paths (all variants of paths) of all
pairs 2 Ht .

Proof of Theorem 1. Assume the premises of Theorem 1. In a two history system, 2 Ht = {α, χ} ⊆ Ht
is the vector state v(t) = (vα (t), vχ (t)) where vχ (t) = 1 − vα (t). R(t) is a status and since in a status
vα (t) = 12 r11 = 12 (1 + r3 ), and since vα (t) + vχ (t) = 1 in a two-history state, then vχ (t) = 12 r22 = 12 (1 − r3 ).
In addition, also then vα (t) + vχ (t) = 12 (1 + r3 ) + 12 (1 − r3 ) = 1 = trR(t), which establishes Theorems 1, 2,
and 3. Establishing 4: We find r(t) is a maximum when r3 = 0, given Theorem 1(1) and 1(3), and Lemma
1(2), so when r3 = 0 then v(t) = (0.5, 0.5).

Let r1 (t) = 0 so  
1 1 + r3 −ir2
R(t) = [r ] =
2 ii ir2 1 − r3

and thus 21 Σi Σj rij (t) = 12 2 = 1 which establishes 1(1).

Let Ht = {α, χ}. At time t, Ht picks a set of modal demographies At = {(nα , pα ), (nχ , pχ )} and v(t)
acts as a linear operator on At ; so we get

v(t)At = Σα v(t)(nα , pα ) = (n(t), p(t)) for all α ∈ Ht .

260

www.pdfgrip.com
Entropy 2018, 20, 8

From Lemma 2, O(t) = (n(t), p(t), e(t)) are the predicted results at t; when sα = sχ then
(nα , pα ) = (nχ , pχ ). R(t) is an idempotent Hermitian matrix per Definition 7, and under the premises
has r1 = 0. Then r2 2 + r3 2 = 1. We have a two history system with vector state v(t) = (vα (t), vχ (t))
where vχ (t) = 1 − vα (t), and where vα (t) = 12 r11 = 12 (1 + r3 (t)). We let t2 > t1 > t0 define a path from
t = t0 to t = t2 such that vα (t0 ) = 1 changes monotonically to vα (t1 ) = 0 and then monotonically again
to vα (t2 ) = 1, which occurs as r3 (t) moves monotonically from r3 (t) = 1 to r3 (t) = −1 and then back to
r3 (t) = 1. At each t, given r3 (t), we compute r2 (t)2 = 1 − r3 (t)2 . Then r2 (t)2 + r3 (t)2 = 1 traces a unit circle
C. Theorems 2 and 3 then follow from Green’s theorem. 
Observation 2. Let G be a population, Gt ∈ G with the sub-populations Gt using the set of histories using face
Ht ∈ H, and let v(t) be the vector state of Gt . Assume history α ∈ Ht , α ∈ Ht+1 , and vα (t) = vα (t + 1) = 1. Then
from t to t + 1, v(t) forms a loop. That is, the minimal descent sequence of any viable pure system α also forms a
loop, indicated also since the minimal structure of α is a group. So any pure system is a loop. Diagrams like
Figure 1 could occur when v(t) is not simply a pure system. Describing probability pictures by complex Hilbert
spaces (Definition 7) can assist predicting demographic pictures, using pure systems as the basis of computing
n(t), p(t) hence r(t).

7. Discussion
Following the examples of [1,2] we here study systems in which the cultural organization
is based on kinship descriptions using natural languages. In both cases, our theory makes
predictions on population measures on the observed outcome of the kinship systems at stated times.
Our Observation 1 and Lemma 1 predict what is found empirically: either single history systems
and specific (ns , ps ) pairs by the structural number of the identified history; or systems undergoing
change in their culture. In that case the (ns , ps ) pairs are changing and we can predict that rate of
change yielding both the n(t) and p(t) for the given t, and the value of the adiabatic growth rate r(t) at t.
An example of this prediction of rate of change in western Europe for about 1000 years from about AD
1000 to 1950 is given in [1]. The time period of that study was about 1000 years of human history in a
defined area.
Thus, study of the homotopy groups resulting from Definitions 5, 7 and 8 may thus tell us a lot
about the possible paths of the empirical demography of cultures. Definitions 7 and 8 assume no
physical model, but we can use their math to study the changes in vector states on histories on the
predicted n(t), p(t) and r(t) of the society per generation. The methods of [28,29] and many other current
works such as [30] in social sciences use complex Hilbert spaces to describe models of how “cognition”
works, using much shorter time periods, and to otherwise interpret how societies of individuals can
describe and change the world around them. Hilbert space probability models per [31], which is a
foundation paper for [13], are quite close to the Pauli model used here to describe changes in cultural
systems; they differ from ours in their application. In particular, [31], Postulate 4 does not apply here
since the applications are distinct. However, the probabilities of [30,31] are averages of probabilities on
a population, not predictions of individual probabilities. There may be thus be many ways to discuss
evolution of cultural systems using complex Hilbert spaces that have simply not yet been tried.
In this paper, in [17], and in both [30,31] the mathematical foundation starts with representation
of the basic objects as languages; ours are natural languages. Kinship systems are derived from
non-associative algebras [20,32] which in natural languages may allow groups to occur. Cultural
systems with different dictionaries but similar groups are studied as isotopic kinship terminologies
for example [8,11], which is a separate topic mathematically and empirically from study of
languages [9–13]. Ref. [33] says “. . . kinship organizations are based on terminologies, which have
their own distinctive logical structure centered on a “self” or I position. Language does not have a
structure of this kind . . . ”. So while kinship terminologies occur as part of natural languages, kinship
analysis is not the same as the study of the language.

261

www.pdfgrip.com
Entropy 2018, 20, 8

Our study also helps identify what cannot be predicted by this method. For example, sociologists
and anthropologists use relationship studies to describe how individuals are “related”; the minimal
structure defined here based on assignments made based on the “principles” used to arrange or avoid
marriages, given the natural language and the history; they do not define which specific individuals
are in fact assigned to each relation. In contrast, in genetic inbreeding experiments Sewall Wright [34]
at diagrams 7.1(a), 7.12, 7.16 and others used the minimal structure of kinship relations for illustrating
inbreeding arrangements; but in those situations, the individuals are not “assigned”—they are the
actual kin of the identified sources.
The ability to derive population measures from the language-based statement of rules is something
new to science, and should be explored. Many other things also affect population change, and are not
explored here.

Conflicts of Interest: The author declares no conflict of interest.

Appendix A. Mathematical Background from Previous Papers

For convenience of use here, this appendix adopts background previously presented mainly in [8].

Definition A1. General mathematical usages. Let L be a finite non-empty set. A partial order ≤ is a binary
relation on L such that for a, b, d ∈ L, a ≤ a; a ≤ b and b ≤ a implies a = b; and a ≤ b and b ≤ d implies
a ≤ d. Then (L, ≤) is a partially ordered set or poset. A lattice is a poset on which is defined two binary relations
join ∪ and meet ∩ such that for a, b ∈ L then a ∪ b ∈ L and a ∩ b ∈ L. A lattice (L, ≤) is bounded if L contains
special elements 0 and 1 such that for b ∈ L, 0 ≤ b ≤ 1, which we denote as (L, ≤, 0, 1). An involution is a
unary relation ’ on L such that for b ∈ L, then b’ ∈ L, b = b” and for b, d ⊆ L if b ≤ d then d’ ≤ b’. An object
(L, ≤, ’, 0, 1) is a bounded involution poset. If L is a bounded involution poset an orthogonality relation ⊥ on L
is a binary relation such that for b, d ∈ L, b ⊥ d if b ≤ d’.

Definition A2. Properties of populations and descent sequences:

Definition A2.1. Let G be a finite non-empty set called a population whose members d ∈ G are called individuals.
Let D (descent), B (sibling of) and M (marriage) be binary relations on G, satisfying these four axioms: (1) D is
anti-symmetric and transitive; (2) M is symmetric; (3) if bDc and there exists no d ∈ G, d = b, c for which bDd
and dDc, then we write cPb, and require bBc iff for b, c, d ∈ G, dPb and dPc; (4)|bM | ≤ 2.

Definition A2.2. Let G(t) = {Gt |Gt ⊆ G, t ∈ T, Gi ∩ Gj = ∅ for i = j} is a family of subsets of G, indeed a
partition of G, where t ∈ T is a set of consecutive non-negative integers starting with 0; such G(t) is a descent
sequence of G.

Definition A2.3. Gt ⊆ G is called the tth generation of G, in case, for all Gt ∈ G, each cell bB occurs in only
one generation, each subset bM occurs in only one generation, and for t > 0 when Gt ∈ G, b ∈ Gt , and cPb, then
c ∈ Gt−1 (that is, Gt contains all of and only the immediate descendants of individuals in Gt−1 ). Let |Gt | = δt .

Definition A2.4. Let Gt be a descent sequence of G. Let Bt := {bB|b ∈ Gt , Gt ∈ G, t > 0} be a partition of Gt .

in which each bB is a sibship; and let Mt−1 := {bM|b ∈ Gt , Gt ∈ G, t ≥ 0} be a set of disjoint subsets of Gt−1 in
which each bM is a marriage. Let |Bt | := βt , and |Mt | := μt .

We allow that only at t = 0 may there be individuals in a generation that did not arise by descent
from a previous generation of G. For b ∈ G, any set bM is assumed to be reproducing. Other than t = 0,
members of Gt arise from (assignment of offspring to) marriages in Gt− 1 , thus βt = μt −1 .

Definition A3. Properties of configurations:

262

www.pdfgrip.com
Entropy 2018, 20, 8

Definition A3.1. Let G(t) be a descent sequence, Gt ∈ G be a generation of G, and let a, b, c, ..., k ∈ Gt , a = b =
c = . . . = k be individuals in Gt . Then a regular structure is a closed cycle aBb, bMc, cBd, . . . , kMa of a finite
number of alternating B and M relations, in which each a ∈ Gt occurs exactly twice in such a list, being exactly
once on the left of a B followed immediately by once on the right of an M, or once on the right of a B preceded
immediately by once on the left of an M, and in such cycle each |bB| = |bM| = 2. If there are j instances of M
in such a cycle, then the regular structure is of type Mj.

Definition A3.2. Given a finite positive integer k, we assume a set of unit basis vectors ei , 1 ≤ i ≤ n, such that
in ei the ith position = 1 and all others = 0; and write c = (m1 , m2 , . . . , mk ). If Gt ∈ G is a generation and mi is
the number of regular structures of type Mi in Gt , then such a c is called a configuration.

Definition A3.3. Let c = (m1 , m2 , . . . , mk ) be a configuration. Then μc := ∑i (mi i) is the number of marriages of c.

Definition A3.4. Let C := (c|c is a configuration} be the set of configurations. Let M ∈ R+ . Let CM := {c|μc
≤ M} be a set of configurations of order M.

For example, a configuration with a single M2 structure would be written (0, 1, 0, 0, . . . ). Note that
the null configuration 0 := (0, 0, . . . , 0) ∈ CM . If c ∈ CM , c = 0, then such a c is non-null, and if B ⊆ CM
contains at least one non-null configuration, then such a B is non-null.
Configurations are often used in ethnography when describing kinship systems. A set bM
identifies a reproducing marriage in a configuration which has offspring in the succeeding generation.
In a configuration we ignore all non-reproducing individuals who may exist “empirically” in Gt .
So while |bB| ≥ 1 in general, |bB| = 2 is required in a configuration. If we let |Gt | = δt , the number of
individuals in a configuration on Gt is γt = 2μt ; so δt ≥ γt and δt ≥ 2μt ; so nt = δt /βt = δt /μt −1 .
Since individuals in Gt arise only from reproducing marriages among individuals of Gt-1 , then
βt = μt −1 . Thus:

Definition A4. If Gt is a generation of G, then nt := δt /βt is the average family size of Gt .

Definition A5. Properties of histories:

Definition A5.1. A history α is a binary relation on P(CM ), that is, (c,d) ∈ α ⊆ P(CM ) × P(CM ); such an
α induces a function of the power set of P(CM ) which we also call α, defined, for B ∈ P(CM ) , by (c,d) ∈ α(B)
= {d ∈ CM |c α d for some c ∈ CM }. We let HM := {α|α is a history defined on CM } be the set of all histories
on CM .

Definition A5.2. A configuration c ∈ CM is viable under α if there exists an integer k > 0 such that c ∈ αk (c).

Definition A5.3. A history α ∈ HM is viable if there is at least one c ∈ CM , c = 0, such that c is viable under
α. Let V(α) := {c|c ∈ CM and c is viable under α} be the set of all viable configurations under α. If 0 is the only
configuration viable under a history α, then such an α is called trivial; otherwise α is called non-trivial.

Definition A5.4. If α is a viable history then sα := min({μc |c ∈ V(α)\{0}) is the structural number of α},
where S := {sα |α ∈ HM } is a set of structural numbers of viable histories in HM . If cα ∈ V(α) such that
μc = sα then cα is a minimal structure of α. If α is a viable history, let Minα := {c ∈ V(α)|μc = sα } be the set of
minimal structures.

Definition A5.5. Let c, d ∈ CM , and let η c ∈ HM be a viable history such that η c (c) = {c} and η c (d) for c = d is
not defined; then η c is a pure system. Let Hp := {η c |c ∈ CM } be the set of pure systems on CM . If α is a history,
G(t) is a descent sequence of α, then c ∈ Min α and c = ct for all Gt ∈ G then G is a pure system of α.

263

www.pdfgrip.com
Entropy 2018, 20, 8

With the usual set union ∪ and intersection ∩ then the power set (P(CM ), ∪, ∩) is a Boolean
algebra, and using ≤ as a partial order by set inclusion, (P(CM ), ≤, 0, CM ) is a poset, indeed a bounded
involution poset. A history is thus a natural language describing “how α works to create a history”.
Here, a history is a rule; specifically a marriage rule. The structural number sα of a viable history α
is simply the value of μ of a smallest configuration which is viable under α; so also sα > 0, indeed
|V(α)| ≥ 2. Notice that Minα ⊆ V(α) and |Minα | ≥ 1.

Definition A6. Let Gt ⊆ G be a non-empty generation of G at time t, let |Gt | = n, and let Gt be partitioned
into 1 ≤ k ≤ n subsets. Then:

(i) We call a pair (n, k) an assignment.

(ii) A set of assignments is a selection denoted by A with subsets A ⊆ A. To specify more detail of the
membership of a set A we may also use subscripts or a square bracket notation [n, k] with subscripts
as required.
(iii) [n, k] := {(n, k)|given a positive integer n, (n, k) where 1 ≤ k ≤ n}.
(iv) [n, k]j := {(n, k)|given a finite positive integer j, (n, k) where 1 ≤ n ≤ j and for each n, 1 ≤ k ≤ n}.
(v) [n, k]j,i := {(n, k)|given finite integers i, j where i ≥ j, (n, k) for 1 ≤ n ≤ j and 1 ≤ k ≤ i}.
(vi) Pj := P([n, k]j ) denotes the set of subsets of [n, k]j .
(vii) If (n1 , k1 ), (n2 , k2 ) are assignments such that n1 = n2 or k1 = k2 , then (n1 , k1 ) and (n2 , k2 ) are
distinct assignments.
(viii) If A is a set of assignments, is a unary relation on A ⊆ A such that A’ := A\A.
If (n, k) is an assignment, then

(ix) n := n/k is the average family size of (n, k).

(x) p := 2/n is the reproductive ratio of (n, k).

Since np = 2, then 0 < p ≤ 1.

Definition A7. Let (n, k) be an assignment.

(i) S(n, k) is a Stirling Number of the Second Kind, where

 
k
1 k− j k
S(n, k) =
k! ∑ (−1) j
jn (9)
j =0

is the number of ways to partition a set of n distinct elements into k nonempty subsets. S(n, k) computes
the number of ways to achieve an assignment (n, k) for n individuals in a generation partitioned into
k = βt = μt−1 ≥ sα non-empty subsets [J] which we call families. Then:
(ii) S[n, k] := {S(n, k)|for given n, S(n, k), k = 1, . . . , n} is called a distribution.
(iii) Given a distribution S[n, k], then [n, k] := {(n, k)| for given n, k = 1, . . . , n} is the underlying selection
of S[n, k].

Since S(n, k) = 1 when n = k then (uniquely) for n = 2 the distribution S[2, k] is bimodal, with modes at
k = 1, 2. Therefore for n > 2:

(iv) n↑ := {j | given n}.

(v) n↑ s := {n↑ | j = s, for a given s > 0}.
(vi) S[n, n↑ s ] := {S(n, n↑ )|given s, n↑ ∈ n↑ s }.
(vii) Ns := n|S(n, n↑ ) = max(S[n, n↑ s ]).
(viii) A[s] := ∪ [n, k] for n such that n↑ ∈ n↑ s and for each such n, 1 ≤ k ≤ n, called the minimal collection of s.
(ix) AM := ∪ A[s] , given a positive integer M, for structural numbers 1 < s ≤ M.

264

www.pdfgrip.com
Entropy 2018, 20, 8

(x) ms := (Ns , s) is the modal assignment for s.

(xi) As := {ms | s ∈ S} is the set of modal assignments for s ∈ S.
(xii) ns := Ns /s is the modal average family size for s.
(xiii) ps := 2/ns is the modal reproductive ratio for s.
(xiv) (ns , ps ) is the modal demography of s.

References
1. Ballonoff, P. Structural statistics: Models relating demography and social structure with applications to
Apache and Hopi. Soc. Biol. 1973, 10, 421–426. [CrossRef]
2. Ballonoff, P. Theory of Lineage Organizations. Am. Anthropol. 1983, 85, 79–91. [CrossRef]
3. Ballonoff, P. Mathematical Demography of Social Systems. In Progress in Cybernetics and Systems Research;
Trappl, R., Pichler, F.R., Eds.; Hemisphere Publishing: Washington, DC, USA, 1982; Volume 10, pp. 101–112.
4. Ballonoff, P. Mathematical Demography of Social Systems, II. In Cybernetics and Systems Research;
Trappl, R., Ed.; North Holland Publishing Co: Amsterdam, The Netherlands, 1982; pp. 555–560.
5. Ballonoff, P. MV-Algebra for Cultural Rules. Int. J. Theor. Phys. 2008, 47, 223–235. [CrossRef]
6. Ballonoff, P. Some Properties of Transforms in Culture Theory. Int. J. Theor. Phys. 2010, 49, 2998–3004.
[CrossRef]
7. Ballonoff, P. Cross-Comment on Terminologies and Natural Languages. In Mathematical Anthropology and
Cultural Theory; 1and1.com: Philadelphia, PA, USA, 2012; Volume 3, pp. 1–15.
8. Ballonoff, P. Manuals of Cultural Systems. Int. J. Theor. Phys. 2014, 53, 3613–3627. [CrossRef]
9. Levi-Strauss, C. Elementary Structures of Kinship; Beacon Press: Boston, MA, USA, 1969.
10. Radcliffe-Brown, A.R. A Natural Science of Society; Free Press: Glencoe, UK, 1948.
11. Special Issue on Australian Systems. In Mathematical Anthropology and Cultural Theory; 1and1.com:
Philadelphia, PA, USA, 2013; Volume 5. Available online: http://mathematicalanthropology.org/toc.html
(accessed on 18 December 2017).
12. Dravidian Kinship Analysis. In Mathematical Anthropology and Cultural Theory; 1and1.com: Philadelphia, PA,
USA, 2010; Volume 3. Available online: http://mathematicalanthropology.org/toc.html (accessed on 18
December 2017).
13. White, H.C.; Coleman, J. An Anatomy of Kinship; Prentice Hall: Englewood Cliffs, NJ, USA, 1963.
14. Chiara, D.M.; Guintini, R.; Greechie, R. Reasoning in Quantum Theory, Sharp and Unsharp Quantum Logics;
Kluwer Academic Publishers: Dordrecht, The Netherlands, 2004.
15. Hirshleifer, J. Economics from a biological viewpoint. J. Law Econ. 1977, 20, 1–52. [CrossRef]
16. Denham, W. Kinship, Marriage and Age in Aboriginal Australia. In Mathematical Anthropology and Cultural
Theory; 1and1.com: Philadelphia, PA, USA, 2012; Volume 4, pp. 1–79.
17. Cargal, J.M. An Analysis of the Marriage Structure of the Murngin Tribe of Australia. Behav. Sci. 1983, 23,
157–168. [CrossRef]
18. Greechie, R.; Ottenheimer, M. An Introduction to a Mathematical Approach to the Study of Kinship.
In Genealogical Mathematics; Ballonoff, P., Ed.; Mouton: Paris, France, 1974; pp. 63–84.
19. Gottscheiner, A. On some classes of kinship systems, I: Abelian systems. In Mathematical Anthropology and
Cultural Theory; 1and1.com: Philadelphia, PA, USA, 2008.
20. Gottscheiner, A. On some classes of kinship systems, II: Non-Abelian systems. In Mathematical Anthropology
and Cultural Theory; 1and1.com: Philadelphia, PA, USA, 2008.
21. Kronenfeld, D. (Ed.) A New System for the Formal Analysis of Kinship; University Press of America: Lanham,
MD, USA, 2000.
22. Barbosa de Almeida, M.W. On the Structure of Dravidian Relationship Systems. In Mathematical Anthropology
and Cultural Theory; 1and1.com: Philadelphia, PA, USA, 2010; Volume 3, pp. 1–43.
23. Barbosa de Almeida, M.W. Comment on Vaz’ “Relatives, Molecules and Particles”. In Mathematical
Anthropology and Cultural Theory; 1and1.com: Philadelphia, PA, USA, 2014; Volume 6, pp. 1–9.
24. Hildon, P.; Peterson, J.; Stiger, J. On Partitions, Surjections and Stirling Numbers. In Bulletin of the Belgian
Mathematical Society; Belgian Mathematical Society: Brussels, Belgium, 1994; Volume 1, pp. 713–735.
25. Foulis, D.J. Effects, Observables, States and Symmetries in Physics. Found. Phys. 2007, 37, 1421–1446.
[CrossRef]

265

www.pdfgrip.com
Entropy 2018, 20, 8

26. Jordan, T.F.; Jagannathan, K. Quantum Mechanics in Simple Matrix Form; Dover Publications: Mineola, NY,
USA, 1986.
27. Beltrametti, E.; Cassinelli, G. The Logic of Quantum Mechanics; Addisson-Wessley: London, UK, 1981.
28. Beim Graben, P. Comment on Gil’s “What are the best hierarchical organisations for the success of a common
endeavor”. In Mathematical Anthropology and Cultural Theory; 1and1.com: Philadelphia, PA, USA, 2016;
Volume 9, pp. 1–4.
29. Gil, L. What are the best hierarchical organizations for the success of a common endeavor? In Mathematical
Anthropology and Cultural Theory; 1and1.com: Philadelphia, 2016; Volume 9, pp. 1–24.
30. Blutner, R.; Beim Graben, P. Dynamic Semantics and the Geometry of Meaning. Available online: https:
//pdfs.semanticscholar.org/a4ce/22edc62ff93305dde18d70edaad96fc220d3.pdf (accessed on 21 December
2017).
31. Wang, Z.; Busemeyer, J.R. A Quantum Question Order Model Supported by Empirical Texts of an A Priori
and Precise Prediction. Top. Cogn. Sci. 2013, 5, 689–710. [PubMed]
32. Cargal, J.M. Reflections on the Algebraic Representations of Kinship Structure. In Mathematical Anthropology
and Cultural Theory; 1and1.com: Philadelphia, PA, USA, 2017; Volume 11, pp. 1–5.
33. Leaf, M. Personal communication, 2015.
34. Wright, S. Evolution and the Genetics of Populations. In The Theory of Gene Frequencies; University of Chicago
Press: Chicago, IL, USA, 1969.

© 2017 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

266

www.pdfgrip.com
 MDPI
St. Alban-Anlage 66
4052 Basel
Switzerland
Tel. +41 61 683 77 34
Fax +41 61 302 89 18
www.mdpi.com

Entropy Editorial Office

E-mail: entropy@mdpi.com
www.mdpi.com/journal/entropy

www.pdfgrip.com
www.pdfgrip.com
MDPI
St. Alban-Anlage 66
4052 Basel
Switzerland
Tel: +41 61 683 77 34
Fax: +41 61 302 89 18
www.mdpi.com ISBN 978-3-03897-715-5

www.pdfgrip.com