TICS 1501 No.

of Pages 15

Feature Review
Structures, Not Strings:
Linguistics as Part of the
Cognitive Sciences
Martin B.H. Everaert,1 Marinus A.C. Huybregts,1
Noam Chomsky,2 Robert C. Berwick,3 and
Johan J. Bolhuis4,5,*
There are many questions one can ask about human language: its distinctive Trends
properties, neural representation, characteristic uses including use in commu- The computations of the mind rely on
nicative contexts, variation, growth in the individual, and origin. Every such the structural organization of phrases
but are blind to the linear organization
inquiry is guided by some concept of what ‘language’ is. Sharpening the core of words that are articulated and per-
question – what is language? – and paying close attention to the basic property ceived by input and output systems at
of the language faculty and its biological foundations makes it clear how the sensorimotor interface (speech/
sign). The computational procedure
linguistics is firmly positioned within the cognitive sciences. Here we will show that is universally adopted is computa-
how recent developments in generative grammar, taking language as a compu- tionally much more complex than an
tational cognitive mechanism seriously, allow us to address issues left unex- alternative that relies on linear order.

plained in the increasingly popular surface-oriented approaches to language. Linear order is not available to the sys-
tems of syntax and semantics. It is an
Grammar from a Cognitive Science Perspective: Generative Grammar ancillary feature of language, probably
a reflex of properties of the sensorimo-
Language is a structured and accessible product of the human mind. We choose to study tor system that requires it for externa-
language for this reason, as one possible way to gain understanding about the human mind. This lization, and constrained by conditions
particular choice – language as part of the mind, so cognitive science – arose as the result of the imposed by sensorimotor modalities.
seminal discoveries by the mid-20th century regarding the mathematics of computation, which
It follows that language is primarily an
permitted a shift from the more conventional perspective of language as a cultural/social object instrument for the expression of
of study. This new perspective regarding computation [1–4] enabled for the first time a clear thought. Language is neither speech/
formulation of what we should recognize as the most basic property of language: providing a sign (externalized expression) nor com-
munication (one of its many possible
discretely infinite array of hierarchically structured expressions that receive systematic inter-
uses).
pretations at two interfaces, roughly, thought and sound [5–8]. We take externalization (see
Glossary) at the sensory–motor level (for instance, speech) as an ancillary process, reflecting
properties of the sensory modality, sign or speech. Therefore, communication, a particular use of 1
Utrecht Institute of Linguistics,
externalized language, cannot be the primary function of language, a defining property of the Utrecht University, 3512 JK Utrecht,
language faculty, suggesting that a traditional conception of language as an instrument of The Netherlands
2
thought might be more appropriate. At a minimum then, each language incorporates via its Department of Linguistics and
Philosophy, Massachusetts Institute of
syntax computational procedures (Box 1) satisfying this basic property. As a result, every theory Technology, Cambridge, MA 02139,
of a particular language constitutes by definition what is called a generative grammar: a USA
3
description of the tacit knowledge of the speaker–hearer that underlies their actual production Department of Electrical Engineering
and Computer Science and
and perception (understanding) of speech. We take the property of structure dependence of Department of Brain and Cognitive
grammatical rules to be central. We will illustrate the puzzling feature that the computational rules Sciences, Massachusetts Institute of
of language rely on the much more complex property of hierarchical structure rather than the Technology, Cambridge, MA 02139,
USA
much simpler surface property of linear order. 4
Cognitive Neurobiology and
Helmholtz Institute, Departments of
Viewing sentences as just linear word strings has long held a prominent place in areas of natural Psychology and Biology, Utrecht
University, 3584 CH Utrecht, The
language processing such as speech recognition and machine translation. Warren Weaver Netherlands

Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy http://dx.doi.org/10.1016/j.tics.2015.09.008 1

© 2015 Elsevier Ltd. All rights reserved.
 TICS 1501 No. of Pages 15

5
Box 1. Merge: The Basic Property of Language Department of Zoology and Sidney
Sussex College, University of
Merge is a (dyadic) operation that takes two syntactic objects, call them X and Y, and constructs from them a single
Cambridge, Cambridge, UK
new syntactic object, call it Z. X,Y can be building blocks that are drawn from the lexicon or previously constructed
objects. Put simply, Merge (X,Y) just forms the set containing X and Y. Neither X nor Y is modified in the course of the
operation Merge.
*Correspondence: j.j.bolhuis@uu.nl
(J.J. Bolhuis).
If X and Y are merged there are only two logical possibilities. Either X and Y are distinct, and neither one is a term of the
other, or else one of the two elements X or Y is a term of the other, where Z is a term of W if it is a subset of the other or the
subset of a term of the other. We can call the former operation ‘External Merge’: two distinct objects are combined[27_TD$IF].

(i)[28_TD$IF] Merge (read, that book) ) {read, that book}

If alternatively X is a term of Y or vice versa and X and Y are merged, we call this ‘Internal Merge’. So for example, we can
(Internal) Merge which book and John read which book, yielding the following:

(ii)[29_TD$IF] Merge (which book, John read which book) ) {which book, John read which book}

In this case, the result of merging X and Y contains two copies of Y. Following further operations, this structure will surface
as in (iii), under a constraint to externalize (‘pronounce’) only the structurally most prominent copy of which book:

(iii)[30_TD$IF] (Guess) which book John read

This sentence may be understood as (iv):

(iv)[31_TD$IF] (Guess) for which book x, John read the book x

Internal merge is a ubiquitous property of language, sometimes called displacement. Phrases are heard in one place but
they are interpreted both there and somewhere else.

Human language generates a digitally infinite array of hierarchically structured expressions with systematic interpretations
at the interfaces with a sensory–motor (sound/sign) and a conceptual–intentional (meaning) system. Thus, language
comprises a system to generate hierarchical syntax along with asymmetric mappings to the interfaces, a basic mapping
to the conceptual–intentional interface and an ancillary mapping to the sensory–motor interface. Merge is the basic
operation underpinning the human capacity for language, UG, connecting these interface systems. Characterizing UG in
[32_TD$IF]terms of recursive merge is just a way of saying that whatever is going on in the brain neurologically can be properly
understood in these terms.

famously made the case for a string-based approach to machine translation as a type of code
breaking, using statistical methods [9]. This position seems intuitively plausible because it
parallels the familiar way foreign language travel guides are organized, with phrases in one
language matched to corresponding phrases in another. The intuition is that simply pairing
matching sentence strings that are selected on the basis of statistical likelihood suffices, and that
accuracy does not require linguistic analysis, simply the compilation of a database of larger and
longer sentence pairs along with more powerful computers for data storage and selection.
Boosted by exactly this increased computing power along with innovative statistical work at IBM
Research (the late Fred Jelinek and John Lafferty among many others) [10], this approach rapidly
gained ascendancy in the late 1980s, gradually pushing out rule-based machine translation
approaches. But this surface-oriented ‘big data’ approach is now all encompassing, not only in
computational linguistics.

The focus on the non-hierarchical aspects of language is evident in the work of some
typologists [11], and is at the basis of usage-based, constructionist linguistic theories
[12,13]. These approaches focus on inductive mechanisms that explain the acquisition and
use of ‘low-level patterns’, ‘not predictable from general rules or principles’, allowing us to
‘create novel utterances based on [constructional] schemas’. [14]. Such approaches focus on
words or word-like constructions, usage patterns, do not acknowledge the relevance of
structure, and view acquisition essentially as statistical [15]. Introductions to psycholinguistics
generally do not mention notions such as hierarchy, structure, or constituent.

2 Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy

 TICS 1501 No. of Pages 15

A variety of evidence can be brought to bear on the study of language. This can include language Glossary
use, acquisition, cognitive dissociations, other detailed neuroscience investigations, cross- Clitic: a syntactic element that
language comparisons, and much else besides. All this follows[6_TD$IF] from the well-confirmed cannot occur freely in syntax but is in
assumption that[4_TD$IF] the human capacity for language rests on shared biological properties. need of a ‘host’. A typical clitic will
attach itself to a host, that is, a (fully
However, for the development of generative grammar one particular type of evidence has inflected) word or phrase, for
proved most useful: the ability of children to rapidly and effortlessly acquire the intricate principles example, French te ‘you’ in Je t’aime.
and properties of the language of their environment. All normally developing children acquire Compositionality: a principle that
most of the crucial elements of their language long before school age. By contrast, adults exhibit constrains the relation between form
and meaning by requiring that the
a very different developmental path when they attempt to acquire a second language [16]. Often meaning of a complex expression is
they do not come close to the level of native speakers, even after a much longer time frame for built up from the meanings of its
learning. Most researchers would agree that the distinctive ontogenesis of child language arises constituent expressions and the way
they are combined. This principle
from the interplay of several factors, including innate mechanisms, language-independent
plays an important role in formal
properties, and external experience. On our view, the ability of children to rapidly and effortlessly semantic theories.
acquire the intricate principles and properties of their native language can best be explained by C(onstituent)-command: c-
looking for innate, language-dedicated cognitive structures (collectively known as Universal command is a binary relation
between nodes in a tree structure
Grammar) that guide learning.
that is defined as follows: node / c-
commands node b iff (i) / 6¼ b, (ii) /
Defined in this way, the study of language focuses on three questions: does not dominate b and b does not
(i) What constitutes knowledge of language? This amounts to understanding the nature of the dominate /, and (iii) every g that
dominates / also dominates b.
computational system behind human language.
Context-free language: a language
(ii) How is knowledge of language acquired? This amounts to unraveling the cognitive pro- (set of sentences) generated by a
cesses underlying primary language acquisition, so as to understand how primary language context-free grammar, namely, a
acquisition differs from secondary, subsequent language acquisition. grammar whose rules are all
restricted to be in the form X ! w,
(iii) How is knowledge of language put to use? This amounts to studying the linguistic processes where X is a single phrase name
underlying language production, perception, and interpretation – under varying conditions, (such as VP or NP), and w is some
such as modality, social environment, and speech context – and the way in which language string of phrase names or words.
helps fulfill our communicative needs. Externalization: the mapping from
internal linguistic representations to
their ordered output form, either
A commitment to some answer to (i) is a logical precondition for addressing (ii) and (iii). Inquiry spoken or manually gestured.
into language acquisition and language use can proceed most effectively insofar as it is based on Gap: any node in the phrase
structure that has semantic content
careful description and understanding of the system that has evolved. The study of language –
but is without phonological content,
we believe – has made sufficient progress answering question (i) to attempt to pursue answers to for example, ‘children should be seen
questions (ii) and (iii). and – not heard’.
Generative grammar: generative
grammar is a research program that
The Infinite Use of Finite Means
includes different competing
One feature of language that distinguishes it from all non-human communication systems we frameworks, and takes linguistics as
know of is its ability to yield an unbounded array of hierarchically structured expressions, a science whose goal it is to try to
permitting ‘infinite use of finite means’ [17]. To see how and why, we need to introduce the provide a precise (explicit and formal)
model of a cognitively embedded
notion of recursion, which underlies this finite–infinite distinction. Much has been written about
computational system of human
recursion from different perspectives. There is no need to repeat this here [18–20]. What is more language, and to explain how it is
important to understand is that recursion in its original context – based on the recursive function acquired.
theory developed by Gödel, Church, and Turing [21–24] – served as the formal grounding for Merge: in human language, the
computational operation that
generative grammar and the solution to the finite–infinite puzzle. The picture of Turing machine constructs new syntactic objects Z
computation provides a useful explanation for why this is so. In a Turing machine, the output of a (e.g., ‘ate the apples’) from already
function f on some input x is determined via stepwise computation from some previously defined constructed syntactic objects X
value, by carrying forward or ‘recursing’ on the Turing machine's tape previously defined (‘ate’), Y (‘the apples’), without
changing X or Y, or adding to Z, that
information. This enabled for the first time a precise, computational account of the notion of is, set formation.
definition by induction (definition by recursion), with f(x) defined by prior computations on some Negative concord items: negative
earlier input y, f(y), y < x – crucially so as to strongly generate arbitrarily complex structures [19]. polarity items with a more restricted
distribution. They can only be
licensed by clausemate sentential
Why is recursion important? As it is formulated above, recursion is important because it supplies negation and can sometimes express
part of an answer to the seemingly unbounded creativity of language, so central to linguistic

Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy 3

 TICS 1501 No. of Pages 15

theorizing since the mid-20th century. This essential property of language provides a means for negation on their own as in fragment
expressing indefinitely many thoughts and for reacting appropriately in an indefinite range of new answers.
Negative polarity items: a word or
situations [25]. word group that is restricted to
negative contexts – needing the
This approach to the unbounded character of language may be contrasted with the conventional scope of a negation (or more
empiricist position that assumes inductive generalizations from observable distributional regu- precisely a monotone decreasing
word/phrase).
larities to be sufficient for learning and use of language. For American structuralism, a standard Parasitic gap (PG): is a gap (a null
concept was that of Leonard Bloomfield, the leading theoretician, for whom language is ‘an array variable) that depends on the
of habits to respond to situations with conventional speech sounds and to respond to these existence of another gap RG, sharing
with it the same operator that locally
sounds with actions’ [26]. Another leading figure, Charles Hockett, attributed language use to
binds both variables. PG must
‘analogy’, and this meant that we construct and understand novel sentences on the basis of conform to a binding condition
those we have constructed and understood before. For Hockett, ‘similarity’ played the central asserting that PG cannot be c-
role in language learning, production, and use [25]. This line of thought is still at the forefront of commanded by RG.
Parsers: a natural language parser is
many modern-day stochastic learning algorithms, generalized learning procedures, and natural
a program for analyzing a string of
language parsers. The crucial question, however, is whether a notion of analogy can be words (sentence) and assigning it
properly defined so as to adequately explain how children acquire language (Box 2). syntactic structure in accordance
with the rules of grammar. Ideally, the
relation between basic parsing
Syntax: What You See is Not What You Get operations and basic operations of
Given the view set out above, Aristotle's dictum that ‘language is sound with meaning’ could grammar approximates the identity
arguably be reformulated as ‘language is meaning with sound’, since the mappings of expres- function. Probabilistic parsers use
sions to the two interfaces are asymmetric, as noted above. The mapping to the systems of statistical information to provide the
most likely grammatical analyses of
inference, interpretation, and the like we assume to be simple, principled, and close to invariant, new sentences.
following structural principles unexceptionally and possibly in harmony with the methodological Phonology: the study of the abstract
principle of compositionality [27]. The mapping to the sensory modalities (speech, sign) is sound patterns of a particular
language, usually according to some
more complex, clearly subject to parameterization and is more likely to have exceptions [28].
system of rules.
Linking a cognitive system to one or other of the sensory modalities amounts to the difficult Phrase structure rules: rewrite
problem of relating two different categories of systems with different properties and different rules that generate phrase structure.
evolutionary histories. But the syntactic operations that map linguistic objects to the semantic These have the general form of (i),
where X is the name of the phrase
interface do not use the simple properties of sequential string order, that is, linear precedence.
and Y Z W defines its structure. Y, Z,
Instead, they rely exclusively on the hierarchical structural position of phrases, that is, hierarchical and W are either phrases, and
structural distance and hierarchical structural relations (Box 3). In the following we illustrate the therefore must themselves occur to
reliance of language on hierarchical structure rather than linear precedence in all areas of the left of the arrow in other rules of
this type, or non-phrasal (terminal)
language – by providing examples from semantics, syntax, morphology, and phonology.
categories (such as noun, verb, or
determiner). (i) X ! YZW
The Syntax of Semantics Prosody: the description of rhythm,
A simple textbook illustration of the reliance of language on hierarchical structure is provided by loudness, pitch, and tempo. It is
often used as a synonym for
syntactic properties of negative polarity items (NPIs) such as the English word [45_TD$IF]anybody or suprasegmentals, although its
negative concord items such as the Japanese word nani-mo (‘anything’). These items require meaning is narrower: it only refers to
an overt negative element such as not or nakat. If we omit the negative items, the sentences the features mentioned above.
become ill-formed[3_TD$IF] (‘*’), cf. (1a,b) and (2a,b): Recursion: a property of a finitely
specified generative procedure that
allows an operation to reapply to the
(1) [4_TD$IF]a. [4_TD$IF]The book I bought did not appeal to [45_TD$IF]anybody.
result of an earlier application of the
b. *The book I bought appealed to [45_TD$IF]anybody. same operation. Since natural
(2) [4_TD$IF]a. [4_TD$IF]Taroo-wa [4_TD$IF]nani-mo [4_TD$IF]tabe-nakat-ta. language is unbounded, at least one
combinatorial operation must be
Taroo-TOP what-MO [4_TD$IF]eat-NEG-PST applicable to its own output (via
[5_TD$IF]‘Taro didn’t eat anything’ recursion or some logical equivalent).
And given such an operation, any
[5_TD$IF]b. *Taroo-wa [4_TD$IF]nani-mo [4_TD$IF]tabe-ta.
derivational sequence for a generable
Taroo-TOP what-MO [4_TD$IF]eat-PST string will determine a hierarchical
structure, thus providing one notion
of structure generation (‘strong
[5_TD$IF]From (1a,b) one might also conclude, wrongly, that the English NPI anybody must appear in the generation’) distinct from the weakly
sentence sequentially after not. This conclusion is immediately refuted by the Japanese example generated string.

4 Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy

 TICS 1501 No. of Pages 15

Box 2. Simple Rules Selectional properties: the

Consider the following noun phrases (i), and their description in terms of context-free phrase structure rules (ii), and semantic restrictions that a word
accompanying figures (Figure I): imposes on the syntactic context in
which it occurs: a verb such as eat
(i) a man (ii) [3_TD$IF]a man on the moon requires that its subject refers to an
animate entity and its object to
something edible.
Syntax: the rules for arranging items
NP NP
(sounds, words, word parts, phrases)
into their possible permissible
Det N Det N Prep Det N combinations in a language.
a man a man on the moon Universal Grammar (UG): is the
theory of the genetic component of
the faculty of language, the human
capacity for language that makes it
Figure I. Structures for (i) and (ii) on the basis of Grammar G. possible for human infants to acquire
and use any internalized language
(G) [4_TD$IF]a. [4_TD$IF]N(oun) P(hrase) ! Det(erminer) N(oun) without instruction and on the basis
b. [4_TD$IF]NP ! Det N Prep Det N of limited, fragmentary, and often
poor linguistic input. UG is the general
Our ‘grammar’ in (Ga,b) (in which ‘!’ means ‘consists of’) would allow one to create an enormous variety of noun theory of internalized languages and
phrases given a vocabulary of determiners, nouns, and prepositions. However, observing that (iii) is also possible, we determines the class of generative
would have to add a rule (Gc) to our grammar: procedures that satisfy the basic
property, besides the atomic
(iii) [4_TD$IF]a girlfriend of the man from the team elements that enter into these
computations.
(G) [4_TD$IF]c. [4_TD$IF]NP ! Det N Prep Det N Prep Det N

But now we are missing a linguistically significant generalization: every noun phrase can have a prepositional phrase
tacked on the end, which is accounted for by replacing grammar G by the following simpler set of rules:

(G0 ) [4_TD$IF]a. [4_TD$IF]NP ! Det N (PP)[34_TD$IF] (noun phrases consist of a determiner and a noun and may be followed by a prepositional
phrase)
b. [4_TD$IF]PP ! Prep NP[35_TD$IF] (prepositional phrases consist of a preposition followed by a noun phrase)

(G0 ) is a simpler grammar. But note that (G0 ) represents (part of) a grammar yielding a ‘discrete infinity’ of possible phrases,
allowing us to generate ever longer noun phrases taking prepositional phrases. We could only circumvent this
unboundedness by returning to a grammar that explicitly lists the configurations we actually observe, such as (G).
But such a list would be arbitrarily limited and would fail to characterize the linguistic knowledge we know native speakers
have. This recursive generation of potential structures (‘linguistic competence’) should not be incorrectly equated with
real-time production or parsing of actual utterances (‘linguistic performance’). Note that this distinction is no different from
the rules for addition or multiplication. The rules are finite, but the number of addition or multiplication problems we can
solve is unbounded (given enough internal or external resources of time and memory).

Grammar (G0 ) also reflects the fact that phrases are not simple concatenations of words, but constitute structured
objects. (G0 ), contrary to (G), therefore correctly reflects properties of constituency as illustrated in (v):

(v) [4_TD$IF]He gave me [a book [about [the pope]]]

It is [the pope]x, he gave me [a book [about[36_TD$IF] X]]
It is [about the pope]x, he gave me [a book [37_TD$IF]X].

in (2a), where nakat follows the negative concord item nani-mo. Example (3) also shows that the
operative constraint cannot be linear order since (3) is ill-formed despite the fact that not appears
sequentially before anybody, just as it does in the well-formed example (1a).

(3) [4_TD$IF]*The book I did not buy appealed to [45_TD$IF]anybody.

What is the correct constraint governing this pattern? It depends on hierarchical structure and
not on sequential or linear structure [29,30].

Consider Figure 1A, which shows the hierarchical structure corresponding to example (1a): the
hierarchical structure dominating not also immediately dominates the hierarchical structure

Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy 5

 TICS 1501 No. of Pages 15

(A) (B)

X X

X X X X

The book X X X The book X appealed to anybody

did not
that I bought appeal to anybody that I did not buy

Figure 1. Negative Polarity. (A) Negative polarity licensed: negative element c-commands negative polarity item.
(B) Negative polarity not licensed. Negative element does not c-command negative polarity item.

containing anybody. (This structural configuration is called c(onstituent)-command in the

linguistics literature [31].) When the relationship between not and anybody adheres to this
structural configuration, the sentence is well-formed.

In sentence (3), by contrast, not sequentially precedes anybody, but the triangle dominating not
in Figure 1B fails to also dominate the structure containing anybody. Consequently, the sentence
is not well-formed.

The reader may confirm that the same hierarchical constraint dictates whether the examples in
([47_TD$IF]4–5) are well-formed or not, where we have depicted the hierarchical sentence structure in
terms of conventional labeled brackets:

([48_TD$IF]4) [7_TD$IF][S1 [NP The book [S2 I bought]S2]NP did not [VP appeal to anyone]VP]S1
[49_TD$IF](5) *[S1 [NP The book [S2 I did not buy]S2]NP [VP appealed to anyone]VP]S1

Only in example ([50_TD$IF]4) does the hierarchical structure containing not (corresponding to the sentence
The book [51_TD$IF]I bought did not appeal to anyone) also immediately dominate the NPI anybody. In (5[8_TD$IF])
not is embedded in at least one phrase that does not also include the NPI. So ([50_TD$IF]4) is well-formed
and (5[8_TD$IF]) is not, exactly the predicted result if the hierarchical constraint is correct.

Even more strikingly, the same constraint appears to hold across languages and in many other
syntactic contexts. Note that Japanese-type languages follow this same pattern if we assume
that these languages have hierarchically structured expressions similar to English, but linearize
these structures somewhat differently – verbs come at the end of sentences, and so forth [32].
Linear order, then, should not enter into the syntactic–semantic computation [33,34]. This is
rather independent of possible effects of linearly intervening negation that modulate acceptability
in NPI contexts [35].

The Syntax of Syntax

Observe an example as in (6):

(6) [4_TD$IF]Guess which politician your interest in clearly appeals to[52_TD$IF].

The construction in (6) is remarkable because a single wh-phrase is associated both
with the prepositional object gap of to and with the prepositional object gap of in, as in
(7a). We talk about ‘gaps’ because a possible response to (6) might be as in (7b):

(7) [ .
4_TD$IF]a [ uess which politician your interest in GAP clearly appeals to GAP[52_TD$IF].
4_TD$IF]G
b. response to (7a): Your interest in [53_TD$IF]Donald Trump clearly appeals to [53_TD$IF]Donald Trump

6 Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy

 TICS 1501 No. of Pages 15

(A) (B)
X X

which polician X which polician X

X X RG X

your interest X clearly appeals X clearly loves X

in PG to RG your interest X

in PG

Figure 2. Parasitic Gap. (A) Well-formed parasitic gap construction: which politician c-commands both real gap (RG) and
parasitic gap (PG). RG does not c-command PG (and PG does not c-command RG either). (B) Ill-formed parasitic gap
construction: which politician c-commands both real gap (RG) and parasitic gap (PG). RG c-commands PG.

The construction is called ‘parasitic gap’ (PG) because the ‘first’ gap in the nominal expression,
the subject, is parasitic on the ‘real gap’ (RG) in the verbal expression: (8b) is well-formed and
occurs independently of (6), while (8a) is ill-formed and does not occur independently of (6).

(8) [4_TD$IF]a. 4_TD$IF]*[ Guess which politician [S [NP your interest in PG]NP clearly appeals to Jane]S
b. Guess which politician [S [NP your interest in Jane]NP clearly appeals to RG]S

In other words, the gap in (8a) cannot exist unless it co-occurs with the independently licensed
gap of (8b), resulting in (6/7a). Parasitic gap constructions are rarely attested, virtually absent
from the empirical record. Nevertheless, language learners attain robust knowledge of parasitic
gap constructions. Although such constructions had been observed to exist long ago (J.R.
Ross, PhD thesis, Massachusetts Institute of Technology, 1967; [36]), the properties of parasitic
gaps were predicted to exist on theoretical grounds [37], and were (re)discovered as a result of
precise generative analysis [38–42]. Applying analytical or statistical tools to huge corpora of
data in an effort to elucidate the intriguing properties of parasitic gaps will not work.

However, not every co-occurrence of RG and PG yields a grammatical result:

(9) [4_TD$IF]a. [4_TD$IF]*Guess which politician clearly loves your interest in[52_TD$IF].
b. Guess which politician [S RG clearly loves [NP your interest in PG]NP]S

Hierarchical structure and structure dependence of rules are basic factors in explaining parasitic
gaps and the asymmetry between (6) and (9), a subject–object asymmetry. The PG is parasitic on
an independently occurring RG but may not be linked to a RG that is in a structurally higher
position. This is illustrated in Figure 2A and 2B for (6) and (9), respectively.

In Figure 2A who is structurally higher than both the RG and the PG, but the PG, being embedded in the
noun phrase subject, is not structurally higher than the RG. In Figure 2B, by contrast, the RG in the
subject position is in a hierarchically higher position than the PG in lower prepositional object position.
The contrasting filler-gap cases of (6) and (9) cannot be characterized by their linear properties. It
would be incorrect to state that PGs must precede their licensing RGs, as shown by (10):

(10) [4_TD$IF]Who did you [[talk to RG] without recognizing PG][54_TD$IF]?

Crucially, the RG licensing the PG is not in a structurally higher position in (10): the verb phrase
dominating the RG does not dominate the adverbial phrase containing the PG. Why this restriction
precisely holds we leave undiscussed here, but is[5_TD$IF] discussed at length[9_TD$IF] in the literature on
parasitic gaps.

Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy 7

 TICS 1501 No. of Pages 15

The same concepts apply across empirical domains in language. For example, adopting these
concepts enables us to explain certain unexpected and surprising phenomena in Dutch.
Compare (11a) to its counterpart (11b) with a phonologically weak pronoun (clitic).

(11) [4_TD$IF]a. [4_TD$IF]Ik ben speciaal voor het klimaat [4_TD$IF]naar de Provence toe [56_TD$IF]gereden.
I am especially for the climate [4_TD$IF]to the Provence [57_TD$IF]driven.
‘I [58_TD$IF]drove [59_TD$IF]to Provence especially for the climate’.
b. Ik ben er speciaal voor [4_TD$IF]naar toe [4_TD$IF]vertrokken.
I am it especially for [4_TD$IF]to [57_TD$IF]driven.
‘I [60_TD$IF]drove there especially for it’.

The clitic er ‘it/there’ is linked to two gaps, the NP [61_TD$IF]complements of the preposition voor and the
complex preposition/postposition naar. . .toe. A single clitic position er simultaneously binds two
structural positions that have different selectional properties but meet the structural con-
ditions of standard parasitic gaps. This old puzzle of structuralist and generative grammar,
sometimes referred to as ‘Bech's Problem’ [43–45], may now turn out to be explainable as a
special case of a parasitic gap construction (if a language-specific property of Dutch morphology
is added to the equation). The simple lesson to take home is that a few assumptions about the
structure of language suffice to give a unified account of superficially unrelated and disparate
phenomena that are left unexplained in models that are restricted to concepts such as linear
precedence. In fact, proposals that restrict themselves to just linear order are both too weak
(incorrectly permitting ill-formed PGs) and too strong (incorrectly ruling out well-formed PGs). They
are therefore neither sufficient nor necessary to deal with natural language and should be
dismissed.

The Syntax of Morphology

Sound and meaning in morphology can also be shown to be dependent on hierarchical
structure. But there is an asymmetry. As discussed above, computational rules of language
invariably keep to the complex property of hierarchical structure and never use the far simpler
option of linear order. But, of course, linear order must be available for externalization since the
sensory–motor system requires that whatever structure is generated must pass through some
type of filter that makes it [62_TD$IF]come out in linear order.

For further evidence of the relevance of hierarchical structure, consider the compounds in (12)
and their respective structures in Figure 3A,B.

(A) (B)

N N

N1 N2 N1 N2

thEatre N N N N prEsident

tIcket Office lAbor Union

s w s w

w s s w

Figure 3. Prosodic Prominence. Right-branching (A) and left-branching (B) nominal compound structures. Bold capital
letters in initial syllables of each word denote position of primary word stress. Compound stress rule is applied successively,
first to the lower, embedded compound, then to the next higher compound containing it. The syllable consistently assigned
strong prosodic prominence (‘s’) on each application of the rule carries compound stress.

8 Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy

 TICS 1501 No. of Pages 15

(12) [4_TD$IF]a. [4_TD$IF]lábor union president, kítchen towel rack (rack for kitchen [63_TD$IF]towels)
b. theatre tícket office[5_TD$IF], kitchen tówel rack (towel rack in the kitchen)

The correct interpretations of these compounds, both at the sensory–motor interface (namely,
different[10_TD$IF] [65_TD$IF]prosodies) and at the semantic interface (namely, different meanings) follow directly
from applying the relevant rules to their radically different hierarchical structures. Here we will limit
our illustration to prosodic prominence. The rule describing prosodic prominence is given in (13):

(13) [4_TD$IF]Assign prosodic prominence to the first noun N1 of a compound [N N1 N2] if and only
if the second noun N2 does not branch.
(More precisely: In a compound N, [N N1 N2], assign prosodic prominence (‘s’) to the
primary stressed syllable of N1 if N2 does not branch.)

The recursive application of this structure-dependent rule, based on [46–48], to the different
hierarchically structured expressions in Figure 3A and 3B yields the correct prosodic prominence
patterns in each case. If none of the parts of a compound branches, as in ticket office or labor
union, prosodic prominence (‘s’) is assigned by (13) to the left-hand noun N1 (tícket, lábor)
because its right-hand noun N2 (óffice, únion) does not branch. As a corollary effect, the N2
becomes prosodically weak ([6_TD$IF]‘w[67_TD$IF]’). The noun theatre tícket office (Figure 3A) is a compound N
consisting of a simple noun N1 (théatre) and a noun N2 (tícket office), which is itself a compound
noun with prosodic prominence already assigned by previous application of (13), as just
discussed. It is a right-branching hierarchical structure. Therefore, the N1 cannot be prosodically
prominent because N2 branches. Consequently, prominence must be assigned to N2, the inner
compound noun. The repeated application of (13) yields the correct result. Analogously, the
compound noun lábor union president has a left-branching hierarchical structure (Figure 3B).
Prosodic prominence, again, falls on the left-hand noun of the inner compound, which, in this
case, is the left-hand member of the full compound structure. The reason is that the right-hand
[68_TD$IF]member [69_TD$IF]is non-branching and must therefore be prosodically weak. A derivation working from
the bottom up guarantees a correct result.

If prosodic prominence would have been constrained by conditions on linear structure we would
have expected stress to fall uniformly and rigidly on a fixed linear position in the string. But language
does not work that way. Patterns of prosodical prominence are neither random nor rigid but
determinate and they universally depend on a more complex hierarchical structure of compounds
such as lábor union president election, evening compúter class teacher, community centre
búilding council, which have each a different stress pattern that is sensitive to structure and is
assigned in accordance with (13). Depending on specific hierarchical structure, stress falls on a
word-stressed vowel of the first, second, or penultimate noun but never on the final noun. These
results would be totally unexpected if we just assume conditions on linear properties of language.

The Syntax of Phonology

In spoken English certain sequences of words can be contracted, for example, don’t vs do not.
Similarly, want to can be contracted to wanna:

(14) a. I want to persuade the [6_TD$IF]biologist vs c. I wanna persuade the biologist.

b. Who do you want to persuade? [7_TD$IF]vs d. Who do you wanna persuade?

But this contraction is not always possible. There are some cases where one cannot substitute
wanna for want to, as in (15):

(15) [4_TD$IF]a. 4_TD$IF]I[ want my colleague to persuade the biologist.

b. *I wanna my colleague persuade the biologist.

Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy 9

 TICS 1501 No. of Pages 15

Here the constraint seems clear: one can only contract to wanna if no words intervene between
them. Apparently, the phonological process of contraction is sensitive to an adjacency condition.
However, some examples such as in (16a) and (17a) below seem to meet this adjacency
constraint, yet the contraction is still blocked, as in (16b) and (17b):

(16) [4_TD$IF]a. [4_TD$IF]Who do you want to persuade the biologist?

b. *Who do you wanna persuade the biologist?
(17) [4_TD$IF]a. [4_TD$IF]We expect parents who want to long for luxury
[72_TD$IF](that is, want[13_TD$IF] meaning ‘to be needy’[73_TD$IF])
b. *We expect parents who wanna long for luxury

Why is this so? (16a) asks ‘Who should persuade the biologist?’ – in other words, who is the
subject of persuade. In (14b) who is the object of persuade. The hierarchical syntactic structure
for these two sentences is therefore different, and it is this difference that allows contraction in
(14d) while blocking it in (16b). The syntactic structure of the two examples is representable as
(14b0 ) and (16b0 ), where we have struck through the original position of who, its place of
interpretation, before the basic operation of generative grammar has applied that put who at the
front of the sentence. The crossed-out who is not pronounced, which is why the externalized
output appears only as who do you want to persuade.

(14b0 ) [Who [do you want [to persuade who]]]?

(16b0 ) [Who [do you want [who to persuade the biologist]]]

Note that in (16b0 ) the crossed-out who[74_TD$IF] (i.e. not pronounced) intervenes between want and to,
just as my colleague does in (15a). But as we have seen, the contraction rule that yields wanna
does not tolerate any elements intervening between want and to. The complex case of (16b) thus
reduces to the simple case of (15b), and contraction is blocked [49,50].

The examples in (17), from [51], show that for contraction c-command between the verb want
and to is also a necessary condition. Contraction is not allowed in (17) because want (in the
meaning ‘to be needy’) is part of the subject and, therefore, structurally not higher than to[75_TD$IF] (cf.
17b0 ). Absence of c-command is the relevant factor blocking contraction despite the availability
of linear adjacency.

(17b0 ) [4_TD$IF]We expect [[NP parents who want] to long for luxury]

Once again then, it is ultimately the structural properties of a sentence that run the show. For
speakers, the ‘hidden’ properties, non-pronounced words (like in 16b0 ) are just as substantial as
pronounced words. The linguistic computations of the mind ‘hear’ what the ear does not. Just as
color and edges do not exist out ‘in the world’ but rather are internal constructions of the mind,
language is not a property of external sound sequences and does not exist apart from mind-
internal computations (Box 1). In this sense, language behaves just like every other cognitive
ability that scientists have so far uncovered.

Summarizing the discussion above, we have shown that for

(i) the mapping to the conceptual–intentional interface, our discussion on negative polarity
items and parasitic gaps:
! hierarchical structure is necessary and sufficient
! linear structure is irrelevant, that is, order is inaccessible

(ii) the mapping to the sensory–motor interface, our discussion of stress assignment and
contraction:

10 Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy

 TICS 1501 No. of Pages 15

Box 3. Constituents: Weak versus Strong Generative Capacity

We experience language, written or spoken, linearly, and therefore it seems straightforward to take order as a central
feature of language. But take the example a blue striped suit. We are instantaneously capable of assessing that this
phrase is ambiguous between a reading in which the suit is both blue and striped (Figure I) and a reading where the suit is
blue-striped (Figure [38_TD$IF]I).

(A) NP (B) NP

Det N Det N
a a
Adj N Adj N
blue suit
Adj N blue striped
striped suit

Figure I. [23_TD$IF]Constituency Natural Language. Two structures for the ambiguous a blue striped suit, reflecting its syntax
and semantics: ([24_TD$IF]A) a reading in which the suit is both blue and striped, and ([25_TD$IF]B) a reading where the suit is blue-striped.

[5_TD$IF]In the trees above this meaning difference is reflected in a different structuring of the same words with the same linear
order. In generative grammar these aspects (structure and order) are distinguished by the notions of weak and strong
generative capacity. In weak generative capacity, what counts is whether a grammar will generate correct strings of
words; strong generative capacity adds the requirement that the right hierarchical structure is accounted for. And this
latter point is of the essence for the study of natural language as we just illustrated.

Let us explain the difference more precisely. For example, the context-free language characterized as anbn can be
correctly generated by the grammars GA and GB in (i).

(i) a. [4_TD$IF]GA [4_TD$IF]S ) a B

B)Sb
S)ab
b. [4_TD$IF]GB [4_TD$IF]S ) A b
A)aS
S)ab

These two grammars are weakly equivalent in that they both generate exactly the same string set, accepting the string
aabb, but not aabbb. However, these two grammars differ in their strong generative capacity. For example, the substring
aab is a constituent in G[39_TD$IF]B, but it is not in G[40_TD$IF]A ([41_TD$IF]Figure II).

GA GB

S S

a B A b

S b a S

ab ab

Figure [26_TD$IF]II. Constituency Formal Language. The string aabb on the basis of grammar GA[2_TD$IF] and grammar GB[3_TD$IF].

Weak generative capacity may play a significant role in formal language theory, where it is stipulated, as in formal
arithmetic. But for natural language the concept of weak generative capacity is unnatural, unformulable, and inapplicable.
It is important to realize that many possible phrase structure grammars that weakly generate some set of words or linear
pattern fail as soon as strong generative capacity is taken into account. The main text illustrates serious challenges for any
system based solely on weak generative capacity, as was forcibly argued from the very beginning of the modern
generative enterprise [[42_TD$IF]1,73]. In this respect, natural languages behave very differently from formal languages.

Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy 11

 TICS 1501 No. of Pages 15

! hierarchical structure is necessary, but not sufficient

! linear structure is relevant, that is, order is needed for externalization.

What reaches the mind is unordered, what reaches the ear is ordered.

Language and Communication

The generative research tradition has never assumed that the communicative function of
language underpins the essential properties of language. Note that generative grammar does
not claim that language cannot be used for communicative purposes, rather that its design
features are not to be understood in communicative terms [52]. For many, both linguists and
non-linguists, it is difficult to imagine that some of the core properties of human language are not
derived from its communicative functions. This seems to follow from the observation that
language is so deeply embedded in human social interaction, facilitating the communicative
and social needs of a community of speakers to share information. Communication provides a
vehicle for sharing information with others. Viewed this way, language is closely intertwined with
non-verbal modes of communication, such as gestures, eye contact, pointing, facial expres-
sions, music, and the like, any of which may have communicative significance. For this approach
to be well-founded, one must be precise about what ‘communication’ means. One can, for
instance, somewhat naturally talk about flowers communicating with bees. The (often tacit)
assumption is that one can pursue non-human comparisons by comparing human communi-
cation to animal communication, and more precisely the natural communication systems that
use auditory, visual, or audiovisual signals [53]. And it is this notion of communication that one
has in mind when one defines language as ‘The systematic, conventional use of sounds, signs,
or written symbols in a human society for communication and self-expression.’ [54].

What then makes such verbal behavior, ‘language’, different from non-verbal systems of
communication? Communicating how to assemble an Ikea bookcase proceeds without (much)
language, via a manual consisting of just pictures, or by a video manual combining picture and
accompanying speech. But explaining what compositionality or impeachment mean is not done
via music, or facial expressions. So could it be that language as we know it might be particularly
useful in ‘hard’ communicative situations, and is, therefore, ‘far more complex than any animal
communication system’? [55]. On such a view, animal communication systems would not be so
far removed from what humans do: less complex, but not qualitatively different. By contrast, we
believe that animal communication systems differ qualitatively from human language [56–58]:
animal communication systems lack the rich expressive and open-ended power of human
language, the creative aspect of normal language use in the Cartesian sense. Moreover, even the
‘atoms’ of natural language and animal communication systems are crucially different. For animal
systems, ‘symbols’ (e.g., vervet calls) are linked directly to detectable physical events, associ-
ated with some mind-independent entity. For natural language it is radically different [59]. The
evolutionary puzzle, therefore, lies in working out how this apparent discontinuity arose [60,61],
demonstrating how the basic property fits this discontinuity both to the known evolutionary facts
and evolutionary theory [62].

As illustrated above, structure dependency is a paramount feature of natural language, which

only makes sense if solutions that rely on linear order are not available to the system that
computes the mapping to the conceptual–intentional system. But if this is the case, using
language for communicative purposes can only be a secondary property, making externalization
(e.g., as speech or sign) an ancillary process, a reflection of properties of the sensory–motor
system that might have nothing special to do with language in the restricted sense we take it to
be: uniquely human (species-specific) and uniquely linguistic (domain-specific). The fact that we
share a wide variety of cognitive and perceptual mechanisms with other species, for instance,
vocal learning in songbirds, would then come as no surprise [63]. It would also follow that what is

12 Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy

 TICS 1501 No. of Pages 15

externally produced might yield difficulties for perception, hence communication. For example,
consider the sentence They asked if the mechanics fixed the cars. In response to this statement,
one can ask how many cars? yielding How many cars did they ask if the mechanics fixed?
However, one cannot ask how many mechanics, yielding How many mechanics did they ask if
fixed the cars, even though it is a perfectly fine thought. To ask about the number of mechanics,
one has to use some circumlocution, one that impedes communication. In this case, commu-
nicative efficiency is sacrificed for the sake of internal computational efficiency, and there are
many instances of this sort. Examples running in the other direction, where communicative
function is favored over internal computational function (Box 1), seem impossible to find. Thus,
the functional relationship between efficient language-as-internal-computation versus language-
as-communication is asymmetric – in every case that can be carefully posed. The asymmetry is:
the mapping to meaning is primary and is blind to order (language as a system for thought), the
mapping to sound/sign is secondary and needs order (imposed by externalization of language).
The empirical claim is, therefore, that linear order is available for the mapping to sound/sign, but
not for the mapping to meaning.

Structures, Not Strings

The examples we have just given illustrate what is perhaps the most significant aspect of
language: utterances are not simple linear concatenations of simpler building blocks (words,
morphemes, phonemes). Rather, utterances are hierarchically structured objects built out of
these simpler elements. We have to take this property into account if we want to correctly
describe linguistic phenomena, whether semantic, syntactic, morphological, or phonological in
nature. Structure dependence of rules is a general property of language that has been exten-
sively discussed from the 1950s onwards and is not just restricted to the examples we have

Box 4. String Linguistics

To illustrate the type of problems an approach to human language that adopts a purely sequential structure is confronted
with, we use Google Translate, a powerful string-based machine translation service that supports the non-hierarchical,
linear view on language. Google Translate [used through Firefox on June 8, 2015] maps the French La pomme mange le
garçon, lit. the apple eats the boy, into the boy eats the apple, precisely because the ‘most likely’ output sentence is the
product of the probabilities of linear word strings or pairs, and the probability of the latter string vastly dominates the
probability of the former. This problem pervades the entire approach. For example, observe Dutch (i) and its Google
translation:

(i) De man van mijn tante kust de vrouw.

(ii) The husband of my aunt kissing the woman.

While not perfect – it should be The husband of my aunt is kissing the woman – this certainly approximates what one
would like. But the system fails dismally when translating the question equivalent: Dutch (iii) becomes (iv), rather than (v)[27_TD$IF].

(iii) Kust de man van mijn tante de vrouw?

(iv) Shore man of my aunt's wife?
(v) Is the husband of my aunt kissing the woman?

Here, kust (‘kisses’), derived from kussen (‘to kiss’), is translated as shore, having been misinterpreted as the Dutch noun
kust for shore/coast. Moreover, the subject de man van mijn tante is analyzed as the possessive of the object de vrouw.
What has gone wrong? Omitting much detail along with trade secrets, what such systems do is roughly this: given a
particular Dutch sentence, notated (Diii), iterate over all English strings of words to find that ‘best’ English string, E0 , which
maximizes the probability of E0  the probability (Diii j E0 ), that is, the probability of the Dutch (iii) given E0 . Note that this
statistical decomposition is linear. It will tend to select commonly occurring word pairs, for instance, kust/coast, if no
longer pairing is readily available or inferred. For example, no English pairing for the Dutch kust de man because the
‘phrase book’ is still not dense enough in the space of pairings.

Adopting the view that hierarchy is only relevant ‘when the language user is particularly attentive, when it is important for
the task at hand’ [[43_TD$IF]71] comes at a price. For a practical business solution the price is right, for a scientific approach to the
study of language the price is wrong.

Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy 13

 TICS 1501 No. of Pages 15

presented so far. These are phenomena that, in our view, must be explained in terms of intrinsic Outstanding Questions
and domain-specific properties of a biolinguistic system. What operating principles are there
besides SIMPLEST MERGE (yielding hierar-
chical, structure-preserving structure
Native speakers have robust knowledge of the constraints that we discussed above, and often without linear order) and MINIMAL SEARCH
that knowledge is tacit – again analogous to the reconstruction of ‘color’ and ‘edges’. Some- [79_TD$IF](a domain-general condition of minimal
times relevant examples are rarely attested in adult language, but children acquire them computation that restricts application
of rules of agreement and displace-
nonetheless. Furthermore, it has been shown repeatedly that infants acquiring language do
ment to strictly local domains and min-
not solely engage in statistical learning by approximating the target language [64–70]. For these imal structural[16_TD$IF] distance[80_TD$IF])?
and other reasons, usage-based approaches that reject generative procedures, and apply
statistical methods of analysis to unanalyzed data (Box 4), probing into huge but finite lists of data What can we find out about the neural
organization underlying higher-order
that are not extendable, fail to distinguish these cases properly. By contrast, generative
computation of merge-based hierar-
procedures succeed in amalgamating a large, diverse set of individual examples into just a chical structure of language and what
few constraints such as the hierarchical dominance example. are its evolutionary roots? Concentrat-
ing on the basic property, how does
the discontinuity fit the known evolu-
Linear statistical analysis fails to account for how semantic readings are specifically linked to
tionary facts and evolutionary theory?
syntactic structures or to explain why ambiguity is constrained in some cases but not in others. A
major problem is not just the failure to succeed, but more importantly the apparent unwillingness What is the precise division of labor
to come to terms with simple core puzzles of language structure such as those we have noted between domain-general and
domain-specific learning systems that
[71]. There have been a handful of other efforts to provide alternative accounts for structure enter into the explanation of learnability
dependence [[76_TD$IF]72,74], but these have been shown to fail [69]. However, if we are really interested and evolvability of natural language?
in the actual mechanisms of the internal system we should ask about the properties that How does the Strong Minimalist Thesis
determine how and why the syntax–semantics mappings are established in the way they – the conjecture that, optimally, UG
reduces to the simplest computational
are and not otherwise (see Outstanding Questions). principles that operate in accordance
with conditions of computational effi-
[7_TD$IF]Concluding Remarks ciency – enhance the prospects of
explaining the emergence and learning
Approximating observational phenomena is very different from formulating an explanatory
of human language, permitting acqui-
account of a significant body of empirical data. Equating likelihood probabilities of language sition of rich languages from poor
use with grammaticality properties of internal systems does not succeed because structural inputs (poverty of stimulus)?
properties of phrases and the generative capacity of internal systems to build structure cannot
How can we attain a better under-
be reduced to linear properties of strings. These somewhat elementary but important insights
standing of the mind-dependent
have been recognized since the very origins of generative grammar[78_TD$IF] [1,18], but seem to have nature, development, and evolutionary
been forgotten, ignored, or even denied[14_TD$IF] without serious argument in recent times. origins of the word-like elements
(‘atoms’) of human language that enter
into core computational operations of
Acknowledgements
language, yielding its basic property[54_TD$IF]?
J.J.B. is part of the Consortium on Individual Development (CID), which is funded through the Gravitation program of the
Dutch Ministry of Education, Culture, and Science and the Netherlands Organization for Scientific Research (NWO; grant What is the role of morphosyntactic
number 024.001.003). features in identifying phrases of exo-
centric constructions, that is, phrases
References not containing a head capable of
1. Chomsky, N. (1956) Three models for the description of language. 9. Weaver, W. (1947) Translation. In Machine Translation of Lan- uniquely identifying them, and demar-
IRE Trans. Inform. Theor. IT–2, 113–124 guages (Locke, W.N. and Booth, D.A., eds), pp. 15–23, MIT Press cating minimal domains of computa-
2. Miller, G.A. (1956) The magical number seven, plus or minus two: 10. Brown, P. et al. (1988) A statistical approach to language transla- tion? How do these features function
some limits on our capacity for processing information. Psychol. tion. In COLING ‘88 Proceedings of the 12th Conference on in the language architecture?
Rev. 63, 81–97 Computational Linguistics (Vol. 1), pp. 71–76, Association for
Computational Linguistics
3. Newell, A. and Simon, H.A. (1956) Logic Theory Machine: a [81_TD$IF]If an improved understanding of the
complex information processing system. IRE Trans. Inform. Theor. 11. Evans, N. and Levinson, S. (2009) The myth of language univer-
sources of complexity, diversity, and
IT–2, 61–79 sals. Behav. Brain Sci. 32, 429–492
malleability of languages helps us
4. Shannon, C.E. (1956) The zero error capacity of a noisy channel. 12. Tomasello, M. (2003) Constructing A language: A Usage-Based
explain their significance for the exter-
IRE Trans. Inform. Theor. IT–2, 8–19 Theory of Language Acquisition, Harvard University Press
nalization process[17_TD$IF], which linearization
5. Chomsky, N. (1995) The Minimalist Program, MIT Press 13. Langacker, W. (2008) Cognitive Grammar: A Basic Introduction,
Oxford University Press principles and strategies govern the
6. Reinhart, T. (2006) Interface Strategies: Optimal and Costly Com-
putations, MIT Press 14. Da˛browska, E. (2015) What exactly is Universal Grammar, and has
externalization of the syntactic prod-
anyone seen it? Front. Psychol. 6, 852 ucts generated by the basic combina-
7. Rizzi, L. (2012) Core linguistic computations: how are they
expressed in the mind/brain? J. Neuroling. 25, 489–499 15. Elman, J.L. et al. (1996) Rethinking Innateness: A Connectionist torial operation of language[82_TD$IF]?
8. Selkirk, E. (2011) The syntax–phonology interface. In The Hand- Perspective on Development, MIT Press
book of Phonological Theory (2nd edn) (Goldsmith, J. et al., eds), 16. Meisel, J. (2011) First and Second Language Acquisition, Cam-
pp. 435–484, Blackwell bridge University Press

14 Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy

 TICS 1501 No. of Pages 15

17. Moro, A. (2014) On the similarity between syntax and actions. 47. Chomsky, N. and Halle, M. (1968) The Sound Pattern of English,
Trends Cogn. Sci. 18, 109–110 Harper and Row
18. Chomsky, N. (1959) On certain formal properties of grammars. 48. Liberman, M. and Prince, A. (1977) On stress and linguistic
Inform. Control 2, 137–167 rhythm. Ling. Inq. 8, 249–336
19. Watumull, J. et al. (2014) On recursion. Front. Psychol. 4, 1–7 49. Lakoff, G. (1970) Global rules. Language 46, 627–639
20. Lobina, D.J. (2011) ‘A running back’; and forth: a review of 50. Chomsky, N. and Lasnik, H. (1978) A remark on contraction. Ling.
Recursion and Human Language. Biolinguistics 5, 151–169 Inq. 9, 268–274
21. Church, A. (1936) An unsolvable problem of elementary number 51. Aoun, J. and Lightfoot, D. (1984) Government and contraction.
theory. Am. J. Math. 58, 345–363 Ling. Inq. 15, 465–473
22. Gödel, K. (1986) On undecidable propositions of formal mathe- 52. Chomsky, N. (2013) What kind of creatures are we? The Dewey
matical systems. In Kurt Gödel: Collected Works Vol. I: Publica- Lectures. Lecture I: What is language? Lecture II: What can we
tions 1929–1936 (Feferman, S. et al., eds), pp. 346–371, Oxford understand? J. Philos. 12, 645–700
University Press 53. Hauser, M.D. (1997) The Evolution of Communication, MIT Press
23. Turing, A.M. (1936) On computable numbers, with an application 54. Crystal, D. (1992) An Encyclopedic Dictionary of Language and
to the Entscheidungsproblem. Proc. Lond. Math. Soc. 42, Languages, Blackwell
230–265
55. Hurford, J. (2008) The evolution of human communication and
24. Kleene, S.C. (1936) General recursive functions of natural num- language. In Sociobiology of Communication: An Interdisciplinary
bers. Math. Ann. 112, 727–742 Perspective (D’Ettorre, P. and Hughes, D., eds), pp. 249–264,
25. Chomsky, N. (1966) Cartesian Linguistics, Harper & Row Oxford University Press
26. Bloomfield, L. (1933) Language, Holt 56. Hauser, M. et al. (2002) The faculty of language: What is it, who has
27. Hauser, M.D. et al. (2014) The mystery of language evolution. it, and how did it evolve? Science 298, 1569–1579
Front. Psychol. 5, 401 57. Berwick, R.C. et al. (2013) Evolution, brain, and the nature of
28. Arregui, K. and Nevins, A. (2012) Morphotactics: Basque Auxil- language. Trends Cogn. Sci. 17, 89–98
iaries and the Structure of Spellout, Springer 58. Bolhuis, J.J. and Everaert, M.B.H. (2013) Birdsong, Speech and
29. Giannakidou, A. (2011) Negative polarity and positive polarity: Language. Exploring the Evolution of Mind and Brain, MIT Press
licensing, variation, and compositionality. In The Handbook of 59. Chomsky, N. (2013) Notes on denotation and denoting. In From
Natural Language Meaning (2nd edn) (von Heisinger, K. et al., Grammar to Meaning: The Spontaneous Logicality of Language
eds), pp. 1660–1712, Mouton de Gruyter (Caponigro, I. and Cecchetto, C., eds), pp. 38–46, Cambridge
30. Kuno, M. (2008) Negation, focus, and negative concord in Japa- University Press
nese. Toronto Work. Pap. Ling. 28, 195–211 60. Berwick, R.C. (2010) All you need is merge: a biolinguistic opera in
31. Reinhart, T. (1981) Definite NP-anaphora and c-command two acts. In Biolinguistic Approaches to Language Evolution (Di
domains. Ling. Inq. 12, 605–635 Sciullo, A.M. and Boeckx, C., eds), pp. 461–491, Oxford Univer-
sity Press
32. Baker, M. (2003) Language differences and language design.
Trends Cogn. Sci. 7, 349–353 61. Bolhuis, J.J. et al. (2014) How could language have evolved? PLoS
Biol. 12, e1001934
33. Musso, M. et al. (2003) Broca's area and the language instinct.
Nat. Neurosci. 6, 774–781 62. Berwick, R.C. and Chomsky, N. (2016) Why Only Us: Language
and Evolution, MIT Press
34. Smith, N. and Tsimpli, I. (1995) The Mind of a Savant: Language
Learning and Modularity, Oxford University Press 63. Chomsky, N. (2005) Three factors in language design. Ling. Inq.
36, 1–22
35. Vasishth, S. et al. (2008) Processing polarity: how the ungram-
matical intrudes on the grammatical. Cogn. Sci. 32, 685–712 64. Crain, S. (2012) The Emergence of Meaning, Cambridge University
Press
36. Ross, J.R. (1986) Infinite Syntax! Ablex
65. Lidz, J. and Gagliardi, A. (2015) How nature meets nurture: Universal
37. Chomsky, N. (1981) Lectures on Government and Binding, Foris
Grammar and statistical learning. Annu. Rev. Ling. 1, 333–353
Publications
66. Medina, T.N. et al. (2011) How words can and cannot be learned
38. Taraldsen, K.T. (1980) The theoretical interpretation of a class of
by observation. Proc. Natl. Acad. Sci. U.S.A. 108, 9014–9019
marked extractions. In The Theory of Markedness in Generative
Grammar (Belletti, A. et al., eds), pp. 475–516, Scuola Normale 67. Gleitman, L. and Landau, B. (2012) Every child an isolate: Nature's
Superiore di Pisa experiments in language learning. In Rich Languages from Poor
Inputs (Piattelli-Palmarini, M. and Berwick, R.C., eds), pp. 91–104,
39. Engdahl, E. (1983) Parasitic gaps. Ling. Philos. 6, 5–34
Oxford University Press
40. Chomsky, N. (1982) Some Concepts and Consequences of the
68. Yang, C. (2016) Negative knowledge from positive evidence.
Theory of Government and Binding (LI Monograph 6), MIT Press
Language 92, in press
41. Huybregts, M.A.C. and van Riemsdijk, H.C. (1985) Parasitic gaps
69. Berwick, R.C. et al. (2011) Poverty of the stimulus revisited. Cogn.
and ATB. In Proceedings of the NELS XV Conference, pp. 168–
Sci. 35, 1207–1242
187, GSLA, University of Massachusetts
70. Chomsky, N. (2011) Language and other cognitive systems. What
42. Hoekstra, T. and Bennis, H. (1984) Gaps and parasitic gaps. Ling.
is special about language? Lang. Learn. Dev. 7, 263–278
Rev. 4, 29–87
71. Frank, S. et al. (2012) How hierarchical is language use? Proc. R.
43. Bech, G. (1952) Über das Niederländische Adverbialpronomen er.
Soc. B 297, 4522–4531
Travaux du Cercle Linguistique de Copenhague 8, 5–32
72. Reali, F. and Christiansen, M.H. (2005) Uncovering the richness of
44. Bennis, H. (1986) Gaps and Dummies, Foris Publications
the stimulus: structure dependence and indirect statistical evi-
45. Huybregts, M.A.C. (1991) Clitics. In Grammatische Analyse dence. Cogn. Sci. 29, 1007–1028
(Model, J., ed.), pp. 279–330, Foris Publications
73. Chomsky, N. (1965) Aspects of the Theory of Syntax, MIT Press
46. Chomsky, N. et al. (1956) On accent and juncture in English. In For
74. Perfors, A. et al. (2011) Poverty of the stimulus: a rational
Roman Jakobson: Essays on the Occasion of his Sixtieth Birthday
approach. Cognition 118, 306–338
(Halle, M. et al., eds), pp. 65–80, Mouton

Trends in Cognitive Sciences, Month Year, Vol. xx, No. yy 15