15

THE MORAL PROBLEM

OF PREDATION
Jeff McMahan

Predation as a Moral Issue

Viewed from a distance, the natural world may present a vista of sublime, majes-
tic placidity. Yet beneath the foliage and concealed from the distant eye, a con-
tinuous massacre is occurring. Virtually everywhere that there is animal life,
predators are stalking, chasing, capturing, killing, and devouring their prey. The
means of killing are various: dismemberment, asphyxiation, disembowelment,
poison, and so on. This normally invisible carnage provided part of the basis for
the philosophical pessimism of Schopenhauer, who suggested that “one simple
test of the claim that the pleasure in the world outweighs the pain . . . is to com-
pare the feelings of an animal that is devouring another with those of the animal
being devoured.”1
The unceasing mass suffering of animals caused by predation is also an impor-
tant though, at least until recently, largely neglected element in the traditional
theological “problem of evil”—that is, the problem of reconciling the idea that
there is a benevolent, omnipotent deity with the existence of suffering and other
evils. Referring to “the odious scene of violence and tyranny which is exhibited
by the rest of the animal kingdom,” John Stuart Mill commented that

if there are any marks at all of special design in creation, one of the things
most evidently designed is that a large proportion of all animals should
pass their existence in tormenting and devouring other animals. They have
been lavishly fitted out with the instruments necessary for that purpose;
their strongest instincts impel them to it, and many of them seem to have
been constructed incapable of supporting themselves by any other food. If
a tenth part of the pains which have been expended in finding benevo-
lent adaptions in all nature, had been employed in collecting evidence to

6241-1145-PII-015.indd 268 8/11/2015 12:45:20 PM

 The Moral Problem of Predation 269

blacken the character of the Creator, what scope for comment would not
have been found in the entire existence of the lower animals, divided, with
scarcely an exception, into devourers and devoured, and a prey to a thou-
sand ills from which they are denied the faculties necessary for protecting
themselves! If we are not obliged to believe the animal creation to be the
work of a demon, it is because we need not suppose it to have been made
by a Being of infinite power.2

The suffering of animals is particularly challenging to the task of theodicy

because it is not amenable to the familiar palliative explanations of human suf-
fering. Animals are assumed not to have free will and are thus incapable either
of choosing evil or of deserving to suffer it. Neither are they assumed to have
immortal souls; hence there can be no expectation that they will be compen-
sated for their suffering in a celestial afterlife. Nor, finally, do they appear to be
conspicuously elevated or ennobled by the final suffering they may endure in
a predator’s jaws. Theologians have had formidable difficulties attempting to
explain to their human flocks why a loving deity permits them to suffer; but the
labors of theodicy will not be completed even if theologians are finally able, in
Milton’s words, to “justify the ways of God to men,” for their God must answer
to animals as well.
There have certainly been important religious thinkers who have found fault
with the arrangement whereby a large proportion of sensitive beings are able
to survive only by feeding upon others, and some of these thinkers have enter-
tained visions of a better order. The prophet Isaiah, for example, writing in the
8th century BCE, described some of the elements of an improved natural order,
beginning with the abandonment of war by human beings and continuing with
the conversion of predators to veganism:

The wolf also shall dwell with the lamb, and the leopard shall lie down
with the kid; and the calf and the young lion and the fatling together; and
the little child shall lead them. And the cow and the bear shall feed; their
young ones shall lie down together; and the lion shall eat straw like the ox.3

Isaiah does not mention whether the reformed, pacifist human beings would
join the other animals in their veganism, but it is doubtful that he would have
them fall below the moral standards set by wolves and lions. These are standards
that most human beings, unlike other predators, could satisfy now with no sacri-
fice of health and little if any sacrifice of happiness; yet most persist in practicing
forms of predation that are at once more refined and more dreadful than those
of other predators. Instead of having to capture their prey and kill it with their
hands and teeth, human predators tend to employ professionals to breed their
prey in captivity, slaughter them, and prepare their bodies for consumption. And
just as most human beings rarely observe acts of predation in the wild, so they
do not witness the mass torment and killing that occurs in their mechanized

6241-1145-PII-015.indd 269 8/11/2015 12:45:20 PM

 270 Jeff McMahan

farms and abattoirs, which is deliberately concealed, albeit with the collusion of
those from whom it is concealed. A veil of propriety is maintained both to avoid
putting people off their feed and to spare them the recognition that they too are
predators, red in tooth even if not in claw (though curiously some do paint their
vestigial claws the color of blood). Among our modes of sanitized predation, the
one that is most common in developed societies—factory farming—inflicts a
lifetime of misery and torment on its victims, in contrast to the relatively brief
agonies endured by the victims of predation in the wild.
There are no even remotely credible arguments for the moral permissibility
of factory farming. There is, in my view, only one argument for the permissibil-
ity of a practice of eating meat that has any plausibility, though it is restricted
in scope. It supports the permissibility of eating meat only from animals that
are caused to exist in order to be eaten, reared humanely to have lives that are
worth living, killed painlessly, and then “replaced” by new animals that are
caused to exist in a continuing cycle of production. I will not discuss this argu-
ment here, though I have done so elsewhere.4 The problem with it is that one
cannot know whether it is sound unless one can first determine whether and to
what extent creating new individuals, whether human or nonhuman, can weigh
against and compensate for killing existing individuals or allowing them to die.
This question in turn cannot be answered with confidence unless the answer
can be shown to have acceptable implications for a range of related but deeply
intractable problems in “population ethics.”5
In my view, it is only if this argument is sound, and even then only if meat is
obtained exclusively from animals that have lived contented lives and been killed
with little or no terror or pain, that there can be a permissible practice of eating
meat, at least until meat produced in vitro becomes widely available. Unless these
two conditions are met, we must fulfill our role in realizing Isaiah’s vision not
only by abandoning war (which also involves ceasing to act in ways that give
others a just cause for war against us) but also by abandoning predation, with
possible exceptions for the eating of animals that are arguably nonsentient, such
as oysters and clams.

Ending Predation?
Granting, then, at least for the sake of argument, that morality requires that we
eat straw like the ox, or at any rate the moral equivalent of straw, the question
arises whether we also have a moral reason to protect animals from predation by
nonhuman predators. This question is restricted in two important ways. First,
the question is not whether there is a moral reason to intervene against all forms
of predation, but only whether there is a reason to protect potential prey that are
capable of suffering and of having a life worth living. Second, it is not, at least in
the first instance, whether there is an obligation to prevent predation where pos-
sible, but only whether there is a moral reason to do so, and if so how strong that

6241-1145-PII-015.indd 270 8/11/2015 12:45:20 PM

 The Moral Problem of Predation 271

reason is. It may be, for example, that there is a strong moral reason but that it is
outweighed by competing considerations, or cannot be effectively acted on, in
present conditions—conditions that might, however, be susceptible to change.
There is some intuitive support for the idea that there is a moral reason to
intervene against predation. If one were to happen upon a young animal that was
about to be captured and slowly devoured alive, piece by piece, by a predator,
one’s impulse would be to frighten off the predator, if possible. One’s sympathies
are with the prey, not the predator. There is a video that can be viewed on the
Internet that shows a small group of lions about to kill a baby water buffalo but
are prevented by defensive action by a herd of adult water buffalos.6 In watching
this video, one’s sympathies are with the buffalos, and one experiences relief and
satisfaction when they succeed, even though one knows that the lions are neither
immoral nor cruel and may go hungry for having been thwarted.
Yet despite their intuitive response in this case, most people’s immediate reac-
tion to the suggestion that there is reason to reduce or eliminate predation in the
natural world is incredulity. Indeed, when the issue is raised in the philosophical
literature, the usual response to this suggestion is to argue that if a moral theory
implies that there is a moral reason to reduce or eliminate predation, that con-
stitutes a reduction of the theory to absurdity—a reductio ad absurdum. It is a
familiar objection to utilitarianism that it is excessively demanding in the sacrifices
it requires individuals to make for the sake of other people. Recently Alison Hills
has sought to strengthen this type of objection by arguing that when utilitarianism
takes account, as it must, of the sheer magnitude of the suffering experienced by
animals in the wild, it must imply that human beings have reasons, and in many
cases duties, to intervene to mitigate it, including duties to reduce the incidence of
predation. She assumes, however, that this implication counts strongly against the
plausibility of the theory. “Utilitarians,” she writes, “have severely underestimated
both how demanding their theory is, and how counter-intuitive. The demanding-
ness objection is much more damaging when we take animals seriously.”7
Lori Gruen, a sensitive critic of human practices that are harmful or degrad-
ing to animals, also considers the claim “that those who argue that other animals
deserve our ethical attention should be committed to ending predation.” She
recognizes that this claim is often cited “as a reductio ad absurdum [of ] the idea
that we have ethical obligations to animals.”8 But rather than challenging the
assumption that the implication would be absurd, Gruen argues that none of
what she considers to be the major moral theories, whether utilitarian, rights-
based, or feminist, actually implies that we ought to intervene to reduce or
eliminate predation. These theories therefore “avoid the reductio that critics have
raised.”9 She suggests at one point that “perhaps a better way to go is to figure out
how to minimize the pain prey experience when eaten by predators.”10 But this
presupposes, mistakenly in my view, that the reason to protect potential prey is
only to prevent their suffering and not to preserve their lives as well.11 In the end
she seeks to reconcile defenders of the well-being of animals to the abandonment

6241-1145-PII-015.indd 271 8/11/2015 12:45:20 PM

 272 Jeff McMahan

of prey in the wild by suggesting that any efforts that might be made to help
them would probably exacerbate rather than alleviate their predicament. “When
we consider,” she writes,

the form human intervention often takes, and the havoc it wreaks, we
may want to leave predators alone. . . . Perhaps we would do best to dis-
play more humility, to ask questions and explore options and to exercise
restraint and perhaps even try to come to terms with tragedy, if need be.12

One reason why most defenders of animals are reluctant to acknowledge a

moral reason to intervene against predation may be that they fear that embrac-
ing the conclusion of their opponents’ reductio would diminish the credibility
of their overall view in the minds of the majority. If so, their fear may well be
justified. But unlike political action, moral philosophy is not a matter of stra-
tegic calculation, manipulation, and compromise. Its aim, as I conceive it, is
to discover the truth about matters of morality. If we are ultimately to act in
conformity with the reasons given by morality, we must know whether we do
indeed have a moral reason to try to reduce the incidence of predation or perhaps
even to eliminate it, if that becomes possible. The fact that the vast majority of
people worldwide would now find it preposterous to suppose that we have such
a reason provides little reason to suppose that they are right, just as the unifor-
mity of opinion about the ethics of slavery among whites in the antebellum
South provided little reason for supposing that it can be permissible to kidnap
and enslave other people. Contemporary beliefs about the moral permissibility
of harming animals and allowing them to be harmed are just as contaminated
by self-interest, religious dogma, and prejudices masquerading as science as were
the slave-owners’ beliefs about slavery. Most commonsense intuitions about the
question whether human beings ought, if possible, to eliminate or reduce the
incidence of predation are therefore epistemically highly suspect. If the argu-
ments in favor of intervention are better than the arguments against it, we can
hope that they will eventually come to guide human action, in the same way that
moral arguments against eating meat have, in only a few decades, increased the
proportion of people who are vegetarians or vegans from negligible to substantial
in those societies in which the arguments have been published and debated. In
the meantime, there seems little reason to fear that people will be persuaded not
to become vegetarian, or not to oppose factory farming, by becoming convinced
that arguments that support vegetarianism and oppose factory farming also sup-
port the eventual elimination of predation. The moral case for vegetarianism has
many dimensions that will remain compelling even if certain arguments for veg-
etarianism have implications that people are unwilling to accept. People can, for
example, see that they bear greater responsibility for the suffering and premature
deaths of animals that are killed specifically for human consumption than they
bear for the suffering and deaths caused by predation.

6241-1145-PII-015.indd 272 8/11/2015 12:45:20 PM

 The Moral Problem of Predation 273

The case in favor of intervening against predation is quite simple. It is that

predation causes vast suffering among its innumerable victims, and to deprive
those victims of the good experiences they might have had were they not killed.
Suffering is intrinsically bad for those who experience it and there seems always
to be a reason, though not necessarily a decisive one, to prevent it—a reason that
applies to any moral agent who is capable of preventing it. (If suffering can be
deserved, deserved suffering might constitute an exception, as its intrinsic bad-
ness for the victim might be outweighed by its impersonal goodness.)13 There
seems, indeed, to be a universally applicable reason not only to prevent the pain-
ful deaths of potential prey that exist now, but also to terminate the cycle in
which new predators continuously replace the old, thereby ensuring an inex-
haustible supply of sentient beings that can avoid suffering and death themselves
only by inflicting suffering and death on others. The elimination of predation
could therefore make the difference between an indefinitely extended future in
which millions of animals die prematurely and in agony every day and an alter-
native future in which different animals would live longer and die in ways other
than in terror and agony in the jaws of a predator.
Most people who read this chapter will recognize that we have a moral reason
to avoid causing animals to suffer if we can do so without cost, and that this is
because suffering is intrinsically bad for those that experience it. But if animal
suffering is bad when we cause it, it should also be bad when it results from other
causes, including the action of other animals. As Martha Nussbaum plausibly
claims, “the death of a gazelle after painful torture is just as bad for the gazelle
when torture is inflicted by a tiger as when it is done by a human being.” This,
she continues, suggests “that we have similar reasons to prevent it, if we can do
so without doing greater harms.”14 Most of us believe, rightly in my view, that
our moral reason not to cause suffering is in general stronger than our reason to
prevent it from occurring—for example, to prevent someone or something else
from causing it. If that is correct, our moral reason not to contribute to causing
suffering by eating meat produced in factory farms is stronger than our reason
to prevent comparable amounts of animal suffering caused by others, including
predators. But that is compatible with our having a strong reason to prevent suf-
fering in animals for which we would be in no way responsible when we can do
so at little or no cost to ourselves.

The Counterproductivity Objection

There are two ways in which the incidence of predation could be significantly
reduced, perhaps eventually to none. One is to bring about the gradual extinc-
tion of some or all predatory species, preferably through sterilization, and with
the exception of the human species, which is capable of voluntarily ending its
predatory behavior. The other, which is not yet technically possible, is to intro-
duce germ-line (that is, heritable) genetic modifications into existing carnivorous

6241-1145-PII-015.indd 273 8/11/2015 12:45:20 PM

 274 Jeff McMahan

species so that their progeny would gradually evolve into herbivores, in fulfill-
ment of Isaiah’s prophecy.
Both of these methods of eliminating or reducing the incidence of predation
would obviously require substantial interventions in the natural world. Perhaps
the commonest objection to the simple moral case I sketched for intervening
against predation is that any such intervention would risk environmental catas-
trophe, for the complexity of any major ecosystem so far surpasses our under-
standing that an attempt to eliminate predation within it, however carefully
planned and well intentioned, would have unpredictable ramifications through-
out the system. The most obvious scenario is that the elimination or even sig-
nificant reduction in predation would produce a Malthusian dystopia in which
herbivore populations would expand beyond the ability of the environment to
sustain them. Instead of being killed quickly by predators, herbivores would
then die slowly, painfully, and in greater numbers from starvation and disease.
Rather than diminishing the suffering and extending the lives of herbivores, the
elimination of predation might increase their suffering overall and even dimin-
ish their average longevity. We can call this the Counterproductivity Objection.
Given the state of our knowledge at present, this is a strong objection to almost
any attempt to reduce predation now. But we should not be dismissive of Isa-
iah’s gifts as a prophet. Ecological science, like other sciences, is not stagnant.
What may now seem forever impossible may yield to the advance of science in
a surprisingly short time—as happened when Rutherford, the first scientist to
split the atom, announced in 1933 that anyone who claimed that atomic fission
could be a source of power was talking “moonshine.” Unless we use Ruther-
ford’s discovery or others like it to destroy ourselves first, we will likely be able
eventually to eliminate predation while preserving the stability and harmony of
ecosystems. It should eventually become possible to gradually convert ecosystems
that are now stabilized by predation into ones resembling island ecosystems that
have flourished for significant periods without any animals with a developed
capacity for consciousness being preyed upon by others. We should therefore
begin to think now about whether we would have moral reason to exercise the
ability to intervene against predation in an effective and discriminating way if we
were to develop it. If we conclude that we would, that gives us reason now to try
to hasten our acquisition of that ability.
One possible way to eliminate predation in an ecosystem without increasing
the suffering of herbivores through overpopulation is to limit the expansion of
herbivore populations by means other than predation. In some instances in which
predation has been diminished unintentionally, human beings have then inter-
vened to replace the original predators. In the United States, Britain, and vari-
ous other developed countries, for example, the increasing incursion of human
activities into hitherto stable ecosystems has diminished the number of animals
that once preyed on deer, resulting in an increased number of deer that have then
had difficulty finding sufficient food. Some suffer starvation or malnutrition

6241-1145-PII-015.indd 274 8/11/2015 12:45:20 PM

 The Moral Problem of Predation 275

while others survive by feeding in people’s gardens, thereby becoming a nuisance

to those whose who drove out the predators. Most human communities solve
this problem by permitting or encouraging the hunting of the deer. Hunters then
happily perform the service of culling the herds without having to be paid for it,
as they enjoy both the killing and the eating of their prey.
But when they are successful, hunters, like other predators, deprive their prey
of further life that could have been good. And they often cause great suffering
as well, particularly when they wound an animal that is able to escape being
killed. In principle there are better ways of controlling populations of herbivores
whose exposure to predators has decreased. The most obvious of these is selec-
tive sterilization.15 Scientists are already working to develop effective means of
sterilization that do not require surgery. These are mainly intended for use in
dogs and cats (in part to reduce the number of strays, which is the analogue for
domesticated animals of overpopulation among wild animals), but some com-
munities are seeking to use them to control local deer populations as well.16 If we
were ever to become serious about eliminating or reducing the incidence of pre-
dation, we could eventually develop a chemical means of sterilization that could
be administered to herbivores in the wild in a discriminating and painless way.
Presumably it will become possible at some point to regulate the size of herbivore
populations through germ-line genetic modification as well.
The question whether predation is bad is relevant to present action in ways
other than helping to guide or inform our research agendas. There are vari-
ous predatory species that are now threatened with extinction. Many people
advocate intervention to preserve those species and to restore their populations
to some prior level. An example of such a species is the Siberian tiger. Human
beings can decide now whether to allow (or cause) that species to disappear, to
enable it to continue to exist in small numbers, or to try to restore the number of
its members to a much higher level. Because the number of remaining Siberian
tigers has been low for a considerable period, any ecological disruption occa-
sioned by the great decline in their number has already occurred. If the several
hundred that remain were all to disappear, the effect on the ecology of the region
would presumably be negligible. But there might also be little ecological risk in
facilitating the gradual reintroduction of a much larger population of Siberian
tigers into the extensive region in which they once flourished, thereby greatly
diminishing the probability that they will become extinct.
If this is right, human beings can choose between two ecologically sustain-
able options. One is to complete the elimination of predation by Siberian tigers
in a large region, the other to increase the level of predation in that region by
repopulating it with tigers. If the latter option would substantially increase ter-
ror, suffering, and premature death among other animals inhabiting the region,
and maintain that increase indefinitely, then the view that there is a moral reason
to prevent suffering and premature death among animals, however they might
be caused, supports the option of allowing (or causing) the tigers to die out in

6241-1145-PII-015.indd 275 8/11/2015 12:45:20 PM

 276 Jeff McMahan

the region—unless, perhaps, their role in the food web would simply be taken
over by some other preexisting predatory species, in which case the extinction
of the tigers would be a loss without any compensating gain in the reduction of
suffering.

The Impersonal Value of Species

As the previous sentence concedes, there is generally a loss of value when a spe-
cies becomes extinct. I accept that, and I also accept that there can be value in the
continued existence of a species that is independent of or additional to the value
for each member of the species of its own continued existence. The existence and
survival of a species may, in other words, have impersonal value—that is, value in
itself, independent of whether it is good or valuable for or to anyone. This may
seem most obviously true of species, such as the Siberian tiger, whose members
are beautiful, graceful, majestic, or otherwise aesthetically impressive—though
the impersonal value of the species is entirely distinct from the aesthetic benefits
that the existence of the species provides for us. That animal species have imper-
sonal value is part of what Ronald Dworkin means when he observes that “we
tend to treat distinct animal species (though not individual animals) as sacred.
We think it very important, and worth a considerable economic expense, to pro-
tect endangered species from destruction.”17 Sacredness, as Dworkin understands
it, is impersonal in that an entity can be sacred without being good for anyone.
Yet it is unlike certain other forms of impersonal value in that it does not imply
that it would be better for there to be more of whatever it is that is sacred. Thus,
that a type of entity is sacred does not imply a moral reason to cause more enti-
ties of that type to exist. “Few people,” Dworkin comments, “would think it
important to engineer new bird species if that were possible. What we believe
important is not that there be any particular number of species but that a species
that now exists not be extinguished.”18
This understanding of the impersonal value of species seems to imply that the
loss involved in the extinction of a species cannot be wholly, or perhaps even par-
tially, compensated for by the coming-into-existence of an entirely new species,
whatever its properties might be. If that is right, it excludes one possible response
to the claim that the extinction of, for example, the Siberian tiger would involve
a significant loss in impersonal value—namely, that this loss could be made up
for by the genetic engineering of a new, equally majestic herbivorous species.
But it is doubtful that this conception of the impersonal value of species is cor-
rect. Since animals first appeared, an indefinite number of species have become
extinct and an indefinite number of new species have arisen. If the extinction of
a species involves a loss of impersonal value that cannot be made up for by the
appearance of a new species, it seems that the world must have got worse with
every extinction, even when the extinction of one species has coincided with the
appearance of more than one new species. It thus seems that, according to this

6241-1145-PII-015.indd 276 8/11/2015 12:45:20 PM

 The Moral Problem of Predation 277

understanding of the value of species, the world would have been better imper-
sonally if none of the earliest species had become extinct, even if a consequence
of that would have been that most of the newer species, perhaps including the
human species, would never have existed.
This conception of the impersonal value of species also raises the question
how the extinction of a species could involve a loss of impersonal value when the
coming-into-existence of that species did not involve a gain in impersonal value.
A parallel and perhaps more familiar question arises if we think, as Dworkin says
we do, that individual human lives are sacred—that is, that the death of a person
involves a loss of impersonal value, though the coming-into-existence of a per-
son does not involve a gain in impersonal value. These beliefs, while common,
seem to be in tension with one another.
Persons are, however, quite different from species. While individual persons
do seem morally irreplaceable, in that the loss involved in the ceasing to exist
of one cannot be counterbalanced or offset by the coming-into-existence of
another, it seems that the loss in impersonal value from the extinction of one spe-
cies could in principle be made up for by the coming-into-existence of another,
even if the new species would not make up for the loss of the instrumental value of
the previous species (that is, its value for other beings and its role in the ecosystem
of which it was a part). Suppose, hypothetically, that some primate species would
not have existed had some psychologically less developed species of reptile not
become extinct. It is plausible to suppose that the replacement of the reptilian
species by the primate species would not have been a net loss in impersonal value
but an actual improvement in impersonal terms.
As this example suggests, the impersonal values of species can vary. Indeed,
even those who are most convinced that species generally have impersonal value
might accept that there are some species that lack it, so that their extinction
would involve no loss of any kind—or even some species whose existence has
negative impersonal value. Either of these might be true, for example, of HIV or
the Ebola virus. But if one believes that even these species have impersonal value,
so that the world would be in one respect worse if they were to become extinct,
it seems clear that the impersonal loss would be vastly outweighed by the gain
to human beings. This is important because it shows that we acknowledge that
the impersonal value of a species may have to be traded off against other values.
One such value is the prevention of suffering. Although many philosophers
are skeptical of the idea that species have impersonal value (some because they
believe there is no such thing as impersonal value), almost no one denies that
suffering is intrinsically bad (which is compatible with its being instrumentally
good on occasion).19 And not only is the intrinsic badness of suffering less open
to doubt than the impersonal badness of the extinction of species, but also the
extent to which suffering is bad is more susceptible of rough measurement than
the extent to which the extinction of a species is bad. Imagine an animal species
that has been slowly declining for more than a century and is now on the verge

6241-1145-PII-015.indd 277 8/11/2015 12:45:20 PM

 278 Jeff McMahan

of extinction. Because the number of its remaining members is small, the effect
of its extinction on the equilibrium of the ecosystem of which it is a part would
be negligible. And suppose that its members are rather repellent, so that there
would be no aesthetic loss in its disappearance. How impersonally valuable is
the preservation of this species? How much ought human beings to sacrifice as
a means of preventing its extinction? Such questions are notoriously difficult to
answer. It is, moreover, quite difficult even to understand how to argue about
them—for example, to know what considerations might plausibly be advanced
in favor of one position or against another.
Imagine that the endangered species just described is a carnivorous species
that preys only on human beings who live in the remote and undeveloped area in
which the surviving members of this species are located. These isolated human
beings seek to hunt down and kill all the remaining animals that prey on them.
This is necessary for their safety and the safety of their children. No one else is
proposing to preserve the species by keeping some of its members in captivity
where they would threaten no one. The choice is therefore between eliminat-
ing the species and allowing its remaining members to continue to kill and eat
human beings. Recall that this is a species that lacks any instrumental value. By
hypothesis, the only reason to preserve it, apart from the interest its existing
members have in continuing to live, is that it has impersonal value and contrib-
utes marginally to the species. It is difficult to believe that the impersonal value
of the species could on its own outweigh the lives and well-being of the human
beings who would otherwise be its victims.
It seems, then, that even granting that most or all species have impersonal
value, this value may vary among species and may in some instances have rela-
tively little weight in relation to other values such as the prevention of human
suffering and the protection of human lives.

The Suffering of Animals

Given that the impersonal value of a species might not weigh heavily in our
deliberations about its preservation if its extinction would have no disruptive
ecological effects but its survival would cause significant harm to human beings,
it seems that the impersonal value of the survival of certain carnivorous species
could also be outweighed by the importance of preventing suffering and pre-
mature death among other animals. Yet, as I observed earlier, most people find
it implausible to suppose that the harms that predators inflict on their animal
prey constitute a significant moral reason for trying to eliminate or reduce the
incidence of predation.
It is common for people to think that, while the suffering of animals matters,
the continuation of the lives of individual animals generally does not, or does
not matter much. It is this belief that motivates the action of many who avoid
eating meat produced by factory farms but see no objection to eating the meat

6241-1145-PII-015.indd 278 8/11/2015 12:45:20 PM

 The Moral Problem of Predation 279

of animals that have been reared humanely and killed painlessly but at an early
point in their natural life span. I believe this view is mistaken and that if the
suffering of animals matters, that is because their well-being matters, in which
case it must also matter whether they live or die, if death would deprive them of
well-being they would otherwise have had. But I will not argue for this here.20
I will put aside the loss that animal prey suffer in being killed by predators and
focus entirely on the suffering they experience in being hunted and devoured.
The suffering that animals undergo while being caught and eaten may be
intense, and the process by which they are killed may last for a quarter of an hour
or more. Because the number of predators worldwide is enormous, and because,
like us, many of them must eat with considerable frequency, the aggregate amount
of suffering in the world at any time that is caused by predation is unimaginably
vast. If human beings could eliminate even one carnivorous species while ensur-
ing that its extinction would not have disruptive ecological effects, that alone
could prevent a vast amount of suffering among animals that would otherwise
have been prey for members of that species. As I argued earlier, the prevention of
that suffering could outweigh the loss in impersonal value involved in the disap-
pearance of the species. If the impersonal value of a species can be outweighed by
human suffering, it seems it can also be outweighed by animal suffering.
Some might reject this inference on the ground that the suffering of animals
does not matter, or matters much less than the comparable suffering of human
beings. But it is unlikely that the idea that the suffering of animals does not
matter at all can be reconciled with the idea that the suffering of human beings
does matter. If we concede that the suffering of all human beings matters, that
would seem to be because we recognize that it is in the nature of suffering that it
is intrinsically bad for those who experience it and that it therefore matters that
it should not occur. Thus, my recognition that the suffering of human beings
matters is not the view that my suffering matters to me, while other people’s
suffering matters to them, or that theirs might matter to me if I cared about or
were specially related to them. If that were my view, I should have the same view
about the suffering of animals—namely, that it matters to them and that it might
matter to me if the suffering animal were my pet. But my recognition is rather
that the suffering of other human beings matters because it is in the nature of
suffering that it ought not to be. As Nagel puts it, the “immediate attitude” of the
sufferer is that “this experience ought not to go on, whoever is having it.”21 There is
nothing in this thought about the species of the subject of the experience.
(The idea that the suffering of animals does not matter is also incompatible
with at least one interpretation of the counterproductivity objection. That objec-
tion is that, without predators to control their numbers, herbivores would suffer
even more than they do now, for their deaths from starvation or disease would
involve substantially more suffering than a comparatively quick death inflicted
by a predator. This would of course not be an objection if the suffering of prey
did not matter.)

6241-1145-PII-015.indd 279 8/11/2015 12:45:20 PM

 280 Jeff McMahan

One might argue that the suffering of prey is offset by the pleasure that preda-
tors derive from eating them—a claim that Schopenhauer, in the passage quoted
in my first paragraph, thought self-evidently false. But even if Schopenhauer
were wrong, most acts of predation would still produce far more loss than gain,
for each meal for a predator comes at the cost of depriving the prey of a great
many meals it would have enjoyed, perhaps as much as the predator enjoys eat-
ing its prey. I know of no reason to suppose that carnivores as a rule get more
pleasure from eating than herbivores do.
There is, perhaps, some reason for skepticism about whether animal prey
actually do suffer, or suffer much, in being killed and eaten by predators. Some
people, for example, speculate that when prey are caught, their brains release a
flood of endogenously produced analgesics. Whether or to what extent this is
true is an empirical question, for which there is some evidence on both sides. In
human beings, great physical trauma sometimes induces unconsciousness, and
there are also stories of soldiers fighting for their survival who later testify that
they became aware of some grave physical injury only after the fighting had
ceased. Yet there are also people who have survived being mauled by an animal
who report having experienced agonizing pain and terror. Given these facts, it
seems reasonable to suppose that there is some variation in the degree of suffering
experienced by prey when they are being killed. While some may immediately
become unconscious or numb, others may suffer excruciating pain. The higher
the proportion of cases that involve numbness of unconsciousness, the less mor-
ally urgent the problem of predation is. But given the evolutionary function
of pain, it seems likely that the killing of prey inflicts great suffering in a high
proportion of cases. Thus, despite the possibility of numbness or unconscious-
ness, few of us would be indifferent between the prospect of dying without
violence and dying by being torn asunder by a pack of wild dogs. Even though
it is implausible to suppose that the suffering of animals does not matter, there
are various reasons why the suffering of an animal might matter less than the
equivalent suffering of a person, or less than suffering caused in the same way in
a person. One obvious reason is that suffering in a person may be accompanied
by fear of its significance or anxiety about its continuation, whereas most ani-
mals seem to be immune to these higher-order thoughts. This does not, how-
ever, show that the same degree of suffering matters unequally in human beings
and animals. It shows only that a certain degree of suffering may be increased
or exacerbated in persons, though not in most animals, by self-conscious reflec-
tion upon it. This consideration also cuts both ways, for it is equally true that
human suffering can be mitigated somewhat by an understanding of its cause,
which can be reassuring if it indicates that the suffering is transient and without
sinister significance. (One need not suppose that these different dimensions of
suffering are crudely additive. It may be that self-conscious awareness of physical
suffering yields a qualitatively different form of suffering that is much worse than
mere physical suffering of the same degree. But, except at the high end of the

6241-1145-PII-015.indd 280 8/11/2015 12:45:20 PM

 The Moral Problem of Predation 281

spectrum of suffering, it seems that, for each degree of self-conscious awareness

of suffering, there is some degree of merely physical suffering that is just as bad.
The worst forms of physical suffering in persons tend to crowd out self-conscious
reflection altogether.)
There are, however, two important respects in which a fixed degree of suf-
fering can be worse in a person than in an animal. These are not ways in which
equivalent suffering can be intrinsically worse in a person; rather they are ways
in which suffering can be worse in a person because of its further effects on the
victim. First, human suffering can be worse because of its opportunity costs.
When suffering is distracting or debilitating, which it often or even typically
is, it may prevent the sufferer from experiencing great happiness that he or she
would otherwise have experienced, or from engaging in some valuable activ-
ity. This can be the case with animals only to a lesser extent, for their lower
psychological capacities exclude them from many of the higher dimensions of
well-being accessible to most human beings, and it is unlikely that an animal
that is incapacitated by suffering would otherwise be engaged in an activity of
substantial value.
Second, suffering can have damaging psychological effects throughout the
subsequent life of a person. These effects may be especially pronounced when
the suffering occurs early in life, but as our enhanced understanding of post-
traumatic stress disorder has revealed, certain types of suffering can be highly
damaging psychologically at any time in a person’s life. Of course, similar phe-
nomena can be observed in animals, as anyone who has known a dog that was
mistreated as a puppy is aware. But the scope for damage is much greater in most
human beings. In part this is simply because the lives of human beings continue
much longer than those of most animals, so that the damaging effects within
a human life are typically more protracted over time. But it is also because the
greater psychological depth, complexity, and unity in most human beings make
it possible for them to have lives that contain more, and arguably more impor-
tant, dimensions of the good (such as significant accomplishment, personal rela-
tions based on mutual understanding, and so on) and are therefore more worth
living than those of animals. In most cases, therefore, the psychological damage
caused by suffering is worse in human beings because the life that is damaged
matters more.
At least in most cases, however, what is primarily bad about suffering is the
suffering itself, as it is occurring, not its opportunity costs or psychological rami-
fications throughout the sufferer’s subsequent life. It is therefore important to
consider whether the sheer intrinsic badness of suffering can matter less simply
because the sufferer is an animal rather than a human being. Some think it can-
not. Thus, Peter Singer maintains that

pain and suffering are in themselves bad and should be prevented or mini-
mized irrespective of the race, sex, or species of the being that suffers. How

6241-1145-PII-015.indd 281 8/11/2015 12:45:20 PM

 282 Jeff McMahan

bad a pain is depends on how intense it is and how long it lasts, but pains
of the same intensity and duration are equally bad, whether felt by humans
or animals.22

Yet it is possible that the suffering of one individual can matter less in an objec-
tive way than the equivalent suffering of another individual. It might be, for
example, that a certain amount of suffering matters less morally when it is expe-
rienced by an individual whose moral status is lower.
We can see how this might be possible by considering the common view
about suffering that is deserved. When a person experiences deserved suffering,
the suffering is as intrinsically bad for him as it would be if he did not deserve it.
But when the suffering is deserved, it may be impersonally good rather than bad
and others may have no reason to prevent it. So how suffering matters and how
much it matters may depend on certain facts about the sufferer. If what an indi-
vidual deserves can affect the way in which that individual’s suffering matters, it
may be that facts about the moral status of the sufferer could also affect the way
in which suffering matters.
Commonsense intuition suggests that moral status can be affect the extent
to which a death matters. Suppose that one can either save the life of a human
stranger who would then live only for another day or save the life of a stray
dog that would then live for another month. Assume for the sake of argument
that, because the person would lose so little good life in dying today rather than
tomorrow, the dog’s loss in dying now would be greater. It might still matter
more to save the person—that is, one’s moral reason to save the person might be
stronger—not because the person’s continuing to live matters more to him or to
others but because the person’s higher moral status makes his continuing to live
more important morally.
If it is true that the strength of our moral reason to save an individual’s life
can vary with the individual’s moral status, it may be that the strength of our
moral reason not to cause an individual to suffer, or to prevent that individual
from suffering, also varies with the individual’s moral status. It seems that our
moral reason to prevent an individual from suffering is in general stronger than
our reason to enable that individual to enjoy further benefits. Yet to save an
individual’s life is just to enable that individual to have further benefits. It seems,
therefore, that if the strength of our reason to enable an individual to have the
benefits of further life varies with the individual’s moral status, the same should
be true of the strength of our reason to prevent that individual from suffering.
There is, however, a significant reason for doubting that the suffering of
animals matters less because their moral status is lower. It derives from the fact
that the lower moral status of animals cannot plausibly be explained by refer-
ence to their not being members of the human species. Their lower moral status
must be attributable instead to their lack of certain morally significant intrinsic
properties on which the higher moral status of persons supervenes, such as the

6241-1145-PII-015.indd 282 8/11/2015 12:45:20 PM

 The Moral Problem of Predation 283

capacities for self-consciousness, rationality, and autonomy. But whatever the

intrinsic properties are that distinguish persons morally from animals, there are
some members of the human species that lack them. Consider, for example, an
orphaned human infant with a genetic condition that both limits its potential
for psychological development, so that it can never have psychological capacities
beyond those of a day-old infant, and also makes it impossible for the infant to
live for more than a few months after birth. It is arguable that this human infant
lacks the properties that are the basis of the higher moral status of persons. Even
though almost everyone has the intuitive conviction that this infant has a higher
moral status than an animal with more highly developed psychological capaci-
ties, such as an intelligent and sociable dog, no one, to my knowledge, has given
a plausible explanation of the basis of this higher status. Suppose, then, that the
moral status of this infant is lower than that of a normal adult person. This would
help to explain one claim that I think is true—namely, that our moral reason to
save the life of a person who would live only another day would be stronger than
our reason to save the life of the infant, even if the infant might live a few more
months in complete comfort. Yet it is doubtful that the infant’s suffering would
matter less, so that, apart from the issue of opportunity costs, it would matter
more to prevent the person from experiencing a certain degree of physical suffer-
ing than to prevent the infant from experiencing the same suffering. But if this
cognitively impaired and inevitably short-lived infant would have a moral status
no higher than that of an unusually intelligent and sociable dog, the equivalent
suffering of the dog should also matter no less because of its lower moral status.
In this section I have argued that suffering is bad in itself and that there is a
moral reason to prevent it in any beings that might experience it (with the pos-
sible exception of those who allegedly deserve it). I have conceded that in many
cases our reason to prevent the suffering of a person may be stronger than our
reason to prevent equivalent suffering in an animal—for example, because of
opportunity costs, side effects, or special relations. I have considered the possibil-
ity that the suffering of animals matters less because of their lower moral status
but have suggested that this has an implication that most would be reluctant
to accept—namely, that the suffering of an orphaned human infant with psy-
chological capacities and potential no higher than those of a higher nonhuman
animal must matter less as well. The conclusion I draw is that there is a strong
moral reason to prevent the suffering of animals in the wild when this is possible.

Playing God and the Principle of Nonintervention

In addition to the counterproductivity objection, the appeal to the impersonal
value of species, and the claim that animal suffering does not matter, or matters
much less, another reason for opposing intervention against predation is that it
would be presumptuous for beings as imperfect and fallible as we are to attempt
to regulate the natural world in accordance with our own notions of what is

6241-1145-PII-015.indd 283 8/11/2015 12:45:20 PM

 284 Jeff McMahan

good and bad. Some people with religious commitments may object that any
attempt to reduce or eliminate predation would be a usurpation of a prerogative
that belongs to the deity alone.
This type of objection, which even now is sometimes also advanced against
efforts to alleviate human suffering, is quite puzzling. Is the idea that it must be
offensive to the deity if our action suggests that we can do a better job of running
things than he can? This is reminiscent of the response by the central character
in Saki’s novel, The Unbearable Bassington, to a woman who expresses the wish
that she could improve him:

You’re like a relative of mine up in Argyllshire, who spends his time pro-
ducing improved breeds of sheep and pigs and chickens. So patronizing
and irritating to the Almighty, I should think, to go about putting superior
finishing touches to Creation.23

Yet while the God of the Old Testament is certainly portrayed as abnormally
sensitive to criticism and neglect, it is nevertheless curious to suppose that the
creator of the universe could be so insecure.
Nor can those who object to our “playing God” seriously suppose that we
actually have the ability to thwart the designs of the deity. Indeed, it is one of the
burdens of theodicy to explain why the deity does not consistently exercise the
capacity to prevent our interference with his arrangements but instead stands by
while people cause some terrible harm and only later intervenes by sending them
to Hell. One might infer that a person can be guilty of playing God, usurping a
divine prerogative that we expect the deity will not exercise, by preventing a crime
rather than allowing it to be done and then punishing the offender. Yet theists
generally assume that it can be permissible to prevent people from inflicting unjus-
tified suffering on others—even by defensive force, if necessary. But if preventing
people from inflicting unjustifiable suffering on others is among our legitimate
prerogatives, it seems that merely preventing predators from coming into existence
as a means of preventing animal suffering should be among them as well.
If, as many of us believe, there is no deity guiding events, our only options are
allowing events to be determined by purposeless natural forces and guiding them
ourselves as intelligently and beneficently as we can. The latter is significantly
more likely than the former to result in better outcomes overall.
This seems an appropriate response not only to the objection to “playing
God” but also to the closely related but more secular “principle of noninterven-
tion” that has largely guided wildlife policy in the United States since the Wil-
derness Act was passed in 1964.24 Since then, conservation biologists, ecologists,
and environmentalists have generally sought to protect wilderness areas from all
forms of encroachment or intervention by human beings. This is unsurprising
given that most previous human interventions had been motivated by self-interest
and were heedless of any consequences other than benefits the interveners sought

6241-1145-PII-015.indd 284 8/11/2015 12:45:20 PM

 The Moral Problem of Predation 285

for themselves. And even when human interventions were more benignly moti-
vated, they were sometimes ill-informed or incompetent. Insofar as the opposi-
tion to intervention has been a response to this history of damaging disruption, it
will cease to be appropriate once our science enables us to intervene with a high
probability of avoiding unforeseen effects—provided, of course, that our aims
are also morally justified.
It has been argued, however, that there is a more principled basis for non-
intervention, which is that interferences with animals in the wild violate their
autonomy or self-determination. In the original edition of his landmark book,
The Case for Animal Rights, published in 1983, Tom Regan wrote:

The total amount of suffering animals cause one another in the wild is not
the concern of morally enlightened wildlife management. Being neither
accountants nor managers of felicity in nature, wildlife managers should
be principally concerned with letting animals be, keeping human predators
out of their affairs, allowing these “other nations” to carve out their own
destiny.25

A similar passage appears in the new preface to an updated edition of the book
published in 2004:

Our ruling obligation with regard to wild animals is to let them be, an
obligation grounded in a recognition of their general competence to get
on with the business of living, a competence that we find among members
of predator and prey species . . . As a general rule, they do not need help
from us in the struggle for survival, and we do not fail to discharge our
duty when we choose not to lend our assistance.26

Lori Gruen, who quotes the latter passage with approval, adds this reinforcing
summary: “Paternalism is appropriate in the case of children, but not so in the
case of individuals who are capable of exercising their freedom to live their lives
in their own ways.”27
Despite my admiration for both these authors, I find their claims about the
autonomy and competence of animals inflated. When Regan says that “members
of predator and prey species . . . do not need help from us in the struggle for sur-
vival,” he is at least obliquely invoking a familiar Darwinian notion that seems
to have no place in this discussion. It is true that predator and prey species will
continue to evolve through the competition for survival, but it is false that indi-
vidual prey can do without our help in their struggles for survival. They would
do much better were we to protect them than they do now when we leave them
to deal with predators by themselves. The fact that predators tend to be well fed
when there are prey around, together with the fact that in some species only
about 1 percent of those born survive to adulthood, indicates that the “general

6241-1145-PII-015.indd 285 8/11/2015 12:45:20 PM

 286 Jeff McMahan

competence” of prey “to get on with the business of living” is less impressive
than Regan’s comment suggests.
The claims of Regan and Gruen are reminiscent of the claims of those, from
John Stuart Mill to Michael Walzer, who have maintained a general opposi-
tion to ostensibly benevolent military interventions by arguing that they tend to
violate the rights of self-determination of people in the states in which they are
carried out.28 The core of truth in this anti-interventionist position is that, when
the internal balance of forces within a single political community has not already
been skewed by prior external intervention, it is generally better for a dispute
about the terms on which the members of this community will live together to
be settled among themselves rather being imposed by others who will not share
in the resulting common life. This is so even when the outcome is determined by
the superior forces of one side rather than by agreement. For only in this way will
the outcome be the product of a genuine process of self-determination.
The limitation of this argument, however, is that it fails to apply in situations
in which there are two or more distinct communities, or collective “selves,” liv-
ing within the same political boundaries and one seeks to rule, expel, enslave,
or exterminate another. When, as in Rwanda in 1994, one group is engaged in
genocide, nonintervention is not a matter of “allowing these ‘other nations’ to
carve out their own destiny.” (Indeed, in Rwanda the members of one nation
were allowed literally to carve up those of another.) Nor could intervening be
plausibly described as acting paternalistically toward “individuals who are capa-
ble of exercising their freedom to live their lives in their own ways.” But condi-
tions in which the members of one human group systematically exploit and kill
the members of another are the analogues in human affairs of predation among
animals, though the analogy is imperfect because the situation of prey in the
wild is generally even more hopeless in the absence of intervention. Prey are
seldom able to defend themselves and there is certainly no prospect, as there is
in conflicts involving human beings, that predators and prey will, on their own,
eventually achieve a modus vivendi that will enable them to live together peace-
fully in the manner prophesied by Isaiah. All things considered, there seems to
be no reason not to prevent the suffering or premature deaths of animal prey on
the ground that this would involve a failure of respect for the competence or self-
determination either of individual animals or of animal groups.

The “Values of Nature”

It is sometimes suggested that concern about the suffering of animals in the wild
betrays a limited, parochial, or perhaps anthropocentric perspective on issues that
are appropriately evaluated only from an ecological or environmental point of
view. Several years ago I wrote a short piece for the online Opinionator column
of the New York Times in which I argued for the moral desirability of control-
ling predation with the aim of reducing the suffering of animals.29 The present

6241-1145-PII-015.indd 286 8/11/2015 12:45:20 PM

 The Moral Problem of Predation 287

chapter is, in effect, a rewritten, significantly expanded descendant of that little

article. The online piece elicited quite a bit of commentary, the vast majority of
which was highly and often indignantly critical. One comment was from Paul
Falkowski, an eminent professor of ecology and evolution at my own university.
He wrote that

it is clear you have either never taken a course in ecology and evolution,
or forgot the message. There is this strange thing called a food web—in
which organisms are primary producers, eat primary producers, eat the
eaters of primary producers—and so on. That is called life. It has NO
ethical or moral values. Those are HUMAN values. A wolf or lion kills
another animal—the pain and suffering are not ecological issues—the life
of the wolf or lion is the issue. If the wolf or lion dies of starvation—then
the prey potentially become over populated—like the deer in Princeton.
Your values are not the values of nature.30

A prominent environmental philosopher, J. Baird Callicott, has made similar

claims:

Pain and pleasure seem to have nothing at all to do with good and evil if
our appraisal is taken from the vantage point of ecological biology. . . . The
doctrine that life is happier the freer it is from pain and that the happiest
life conceivable is one in which there is continuous pleasure uninterrupted
by pain is biologically preposterous.31

He adds that “if nature as a whole is good, then pain and death are also good.”32
Presumably, this is again “from the vantage point of ecological biology.”
Both writers object that those who are concerned about the suffering of ani-
mals are wrong to assume that suffering is an ecological issue or that it matters
from the perspective of ecological biology. But that complaint reflects an ele-
mentary confusion. While there are some philosophers who identify moral facts
and properties with natural facts and properties, even they would agree that it
would be a mistake to suppose that the question whether there is a moral reason
to prevent the suffering of animals is a scientific question that could be answered
by consulting theories of ecological biology.33 Falkowski seems to think that I
would do better if I were to put aside my values, which are perhaps merely per-
sonal or subjective, as well as what he calls human values, and instead consult
the “values of nature.” But I have no idea what he might mean by that curious
phrase. Nor are there “human values” that contrast with nonhuman values.
Falkowski is of course right that whether we have a moral reason to prevent
the suffering of animals is not an ecological issue. But that is because no moral
issue is an ecological issue. Morality is nonetheless a part of reality. And while
only science, including ecological science, can tell us what the consequences of

6241-1145-PII-015.indd 287 8/11/2015 12:45:20 PM

 288 Jeff McMahan

our interventions in the natural world are likely to be, only morality can tell us
whether those interventions are, all things considered, good or bad, justified or
unjustified.
To suggest, as Callicott does, that the claim that suffering is bad for those who
experience it is “biologically preposterous” makes no more sense than to say that
it is mathematically preposterous. The claim that suffering is bad is neither bio-
logically sensible nor biologically preposterous, for it is not a claim about biology
at all. Nor does biology have anything at all to say about it. Suffering is of course
a biological phenomenon and it occurs in the natural world. But the badness of
suffering is not a biological phenomenon. It is nevertheless neither an illusion
nor a projection; it too is a part of reality, just not a part that is accessible to the
investigative tools of natural science.
Some ecologists have begun to challenge the venerable principle of noninter-
vention in nature by appealing to what they consider to be a higher value: the
preservation of the “health” of ecosystems. They argue that we should regard
a wilderness area as “a place where concern for ecosystem health is paramount,
even if human action is required to maintain it.”34 In a recent op-ed column in
the New York Times, three ecologists cite the example of a remote island in Lake
Superior on which the number of wolves is dwindling, allowing for an ecologi-
cally disruptive surge in the moose population. The shrinkage of the wolf popu-
lation is not the result of starvation but of genetic degeneration from inbreeding.
In the past, the wolf population was replenished and diversified when wolves
crossed to the island on temporary bridges of ice; but with global warming those
bridges hardly ever form anymore. Without human intervention, wolves could
disappear from the island. Arguing in favor of such intervention, the writers cite

one of the most important findings in conservation science: that a healthy

ecosystem depends critically on the presence of top predators . . . when
large herbivores . . . are present. Without top predators, prey tend to
become overabundant and decimate plants and trees that many species
of birds, mammals, and insects depend on. Top predators maintain the
diversity of rare plants that would otherwise be eaten, and rare insects that
depend on those plants. The loss of top predators may disturb the nutrient
cycling of entire ecosystems. In addition, predators improve the health of
prey populations by weeding out the weakest individuals.35

The authors conclude by observing that if the health of the island’s ecosystem
is preserved through the reintroduction of a substantial population of predators,
the island will remain a place “where we can witness beauty while reflecting on
how to preserve it.”36
According to these writers, we face a dilemma. While in general they support
the principle of nonintervention, they claim that in this case it conflicts with
the value of “ecosystem health.” One of these guiding values must yield in this

6241-1145-PII-015.indd 288 8/11/2015 12:45:20 PM

 The Moral Problem of Predation 289

case, and they argue that it should be the principle of nonintervention. But the
sufferings and premature deaths of the moose are not mentioned. They do not
appear to be regarded as relevant considerations. Yet while it seems obvious that
suffering and premature death are bad for those to whom they happen, it is much
less clear that the “health” of an ecosystem has significant value independently
of the well-being of its sentient inhabitants. Many philosophers, for example, are
“welfarists” who believe that only the welfare of sentient beings matters.
Imagine a small island with an ideally healthy ecosystem. No one had ever
been to this island until a person arrived and somehow loaded all the sentient
beings, predators and prey alike, into his ark and transported them to another
environment into which they were integrated without any compromise of the
health of that environment’s ecosystem. Suppose that no person will ever again
go to the island. But the removal of the animals eventually fatally disrupted
the processes of nutrient cycling on the island, and the health of the ecosystem
began a decline that resulted eventually in its destruction, in that all but the most
rudimentary forms life on the island disappeared. None of this was bad for any
sentient being. The animals that were removed, along with their descendants,
flourished in their new environment at least as well they would have on the
island, and no person was or ever will be prevented by this one intervention from
witnessing the beauty of a healthy ecosystem on the island.
Let us grant that the welfarists are wrong and that, even though what has
happened to the island is not worse for any sentient being that will ever live, it is
nevertheless impersonally bad, in that the island ceased to have a flourishing eco-
system and became barren and lifeless instead. Even so, it does not seem terribly
bad. It is a comparatively minor change for the worse. Suppose that shortly after
the animals had been removed, a government learned what had been done. This
government had a fixed set of resources that it could have used either to return
the animals to the island, thereby preventing the destruction of its ecosystem, or
to produce and distribute a medicine that would prevent thousands of patients
from experiencing brief but intense suffering during minor surgical procedures.
It decided to use the resources in the second way. It does not seem a mere human
prejudice to suppose that this was the better choice.
Consider again the actual island in Lake Superior from which wolves may
vanish. Some ecologists, as I mentioned, propose to transport wolves to the
island, thereby ensuring that the cycle of predation continues, as a means of
preserving the health of the island’s ecosystem. Yet the cycle of predation they
wish to preserve and prolong indefinitely is, for the prey, a cycle of fear, suffer-
ing, and violent death. This is an essential element of what they call “health.”
In this actual case, as they point out, the alternative to reintroducing a substan-
tial wolf population may be only to allow overpopulation among the moose,
which would result in greater suffering as a result of premature but protracted
deaths from starvation and disease. That is, the counterproductivity objection
may apply to the option of not sustaining the cycle of predation. But suppose

6241-1145-PII-015.indd 289 8/11/2015 12:45:21 PM

 290 Jeff McMahan

there were another alternative: intervention to control the size of the moose
population through nonviolent means. It will eventually become possible, as
I noted earlier, to use chemical means of sterilization to regulate the size of
herbivore populations in the wild. If the counterproductivity problem could be
solved in this way, by stabilizing the moose population and thus preserving the
harmony of the ecosystem as a whole without increasing the wolf population,
it seems that the only advantage of maintaining predation on the island would
be that it might be less costly. But if the means of controlled sterilization were
available, it is unlikely that the difference in cost could justify the perpetuation
of avoidable suffering and premature death over indefinitely many generations
of moose. If we were to take the suffering and premature deaths of animals
more seriously than we do, it would probably not take long to develop effec-
tive chemical means of sterilization and techniques for administering them in
a discriminating and precisely calibrated way to regulate the size of herbivore
populations.

Conclusion
I have argued that the prevention of suffering in animals matters—arguably
almost as much as the prevention of suffering in human beings does. Some people
may concede this, and also concede the weakness of the objections to the reduc-
tion or elimination of predation I have reviewed, but nevertheless find it point-
less to press the case for intervening against predation. This is not just because
they anticipate that human beings will never become motivated to undertake so
ambitious a project merely for the sake of animals, but also because they believe
that there will always be more serious problems that will have a higher moral
priority, such as the relief of poverty, the prevention of crime, the prevention of
war, the mitigation of threats to human survival, and so on. Some argue that it
would be wrong to devote our efforts and resources to the problem of predation
when more important problems remain unresolved.
When this claim is pressed in public discussions, it is often made defensively
and hypocritically by people who are not among those who are working seri-
ously to address the moral problems cited. This is of course merely ad hominem
and does not address the concern of those who make the claim without hypoc-
risy. There are, however, at least three responses to the genuine concern about
moral priority.
Perhaps the most important of these is that the many problems that might be
cited as more important than preventing the suffering that predation causes to
animals are so vast and demanding that it is unlikely that any particular indi-
vidual is morally required to devote significant time, effort, and resources to
any one of them in particular. For any individual, making significant sacrifices
to address any of these problems is likely to be supererogatory. When that is the
case, it cannot be wrong to devote one’s efforts to preventing the suffering of

6241-1145-PII-015.indd 290 8/11/2015 12:45:21 PM

 The Moral Problem of Predation 291

animals even when it would be better if one were to devote one’s time and effort
to a more important problem instead.
Second, even if the extent to which the suffering of animals matters morally
is discounted for the lower moral status of animals, the suffering of animals in
the wild may still be one of the more important moral problems simply because
the number of animals that suffer and die from predation is so vast. The number
of animals in the world exceeds the number of human beings by many orders
of magnitude.
Of course, predation is not the only cause of suffering or premature death
in animals. Animals suffer and die from disease, parasites, malnutrition and
starvation, dehydration, freezing, and so on. But this just means that we have
moral reason to try to prevent animals from suffering and dying from these
causes as well, when we can do so at reasonable cost and without neglecting
other duties.
Finally, it may well be that any substantial efforts to mitigate the suffering
of animals in the wild through the control of predation must await advance-
ments in both our scientific and moral capacities. At present it does seem
more important to concentrate on eliminating various major sources of human
misery and premature death. We can, moreover, be more confident of our
potential effectiveness in alleviating suffering and preventing premature death
through, for example, the reduction and eventual elimination of human pov-
erty than we can be in our ability to reduce the incidence of predation without
causing unforeseen side effects. But even now there are cases, such as that of
the island in Lake Superior, in which decisions must be made that will affect
the level of predation in a certain area. In these cases, there is a strong moral
reason to do what will diminish or eliminate predation rather than what will
sustain or increase it.37

Notes
1. Arthur Schopenhauer, Essays and Aphorisms (London: Penguin, 1970), p. 42.
2. John Stuart Mill, Three Essays on Religion (London: Longmans, Green, Reader, and
Dyer, 1875), pp. 58–59.
3. Isaiah 11:6–7.
4. Jeff McMahan, “Eating Animals the Nice Way,” Daedalus (Winter 2008): 66–76, and
“The Comparative Badness of Suffering and Death for Animals,” in Tatjana Višak
and Robert Garner, eds., The Ethics of Killing Animals (Oxford: Oxford University
Press, forthcoming).
5. The seminal and still unsurpassed work in this area is Derek Parfit’s Reasons and
Persons (Oxford: Oxford University Press, 1986 reprint), pt. 4.
6. http://www.youtube.com/watch?v=LU8DDYz68kM. Last viewed December 7, 2013.
7. Alison Hills, “Utilitarianism, Contractualism, and Demandingness,” Philosophical
Quarterly 60 (2010): 225–242, p. 231.
8. Lori Gruen, Ethics and Animals: An Introduction (New York: Cambridge University
Press, 2011), p. 179.
9. Ibid., p. 183.

6241-1145-PII-015.indd 291 8/11/2015 12:45:21 PM

 292 Jeff McMahan

10. Ibid., p. 181.

11. See McMahan, “The Comparative Badness of Suffering and Death for Animals.”
12. Gruen, Ethics and Animals, pp. 184 and 187.
13. For a defense of the claim that there is a universally applicable reason to prevent suf-
fering, no matter whose it is, see Thomas Nagel, The View From Nowhere (New York:
Oxford University Press, 1986), pp. 156–162. I take no position here on whether
anyone can deserve to suffer.
14. Martha C. Nussbaum, Frontiers of Justice: Disability, Nationality, Species Membership
(Cambridge, MA: Belknap Press, 2006), p. 379. Nussbaum is among the few philoso-
phers who have suggested that there is a moral reason to intervene against predation.
The other two of whom I am aware are Eric Rakowski in Equal Justice (Oxford: Clar-
endon Press, 1991), pp. 363–367, and Tyler Cowen in “Policing Nature,” Environmen-
tal Ethics (Summer 2003): 169–182. The latter can be found in manuscript form on
his website at http://www.gmu.edu/centers/publicchoice/faculty%20pages/Tyler/
police.pdf.
15. Compare Martha Nussbaum’s claim that “any nonviolent method of population con-
trol (for example, by sterilization) is to be preferred to a violent method.” Nussbaum,
Frontiers of Justice, p. 380.
16. Douglas Quenqua, “New Strides in Spaying and Neutering,” New York Times,
December 2, 2013; Lisa W. Foderaro, “A Kinder, Gentler Way to Thin the Deer
Herd,” New York Times, July 6, 2013.
17. Ronald Dworkin, Life’s Dominion (New York: Alfred A. Knopf, 1993), p. 75.
18. Ibid. Dworkin’s claims are explicitly about what “we” believe or what “people”
believe. They are not necessarily endorsements of the beliefs that he claims people
have.
19. For a strong challenge to the concept of impersonal value, see Richard Kraut, Against
Absolute Goodness (New York: Oxford University Press, 2011).
20. For elucidation and defense, see McMahan, “The Comparative Badness of Suffering
and Death for Animals.”
21. Nagel, The View From Nowhere, p. 161.
22. Peter Singer, Animal Liberation, rev. ed. (New York: Avon Books, 1990), p. 17.
23. Saki (H. H. Munro) The Unbearable Bassington (London: Bodley Head, 1929), p. 74.
24. John A. Vucetich, Michael P. Nelson, and Rolf O. Peterson, “Predator and Prey, a
Delicate Dance,” New York Times, May 9, 2013. There have, however, been some
interventions that have been harmful to animals, such as the extermination of ani-
mals not considered to be native and the reintroduction of predators. See Jo-Ann
Shelton, Killing Animals that Don’t Fit In: Moral Dimensions of Habitat Restora-
tion,” Between the Species 13/4 (2004): 1–21; and Oscar Horta, “The Ethics of the
Ecology of Fear against the Nonspeciesist Paradigm: A Shift in the Aims of Interven-
tion in Nature, Between the Species 13/10 (2010): 163–187.
25. Tom Regan, The Case for Animal Rights (London: Routledge and Kegan Paul, 1983),
p. 357.
26. Tom Regan, The Case for Animal Rights (Berkeley: University of California Press,
2004), p. xxxvii.
27. Gruen, Ethics and Animals, p. 182.
28. John Stuart Mill, “A Few Words on Non-Intervention,” The Collected Works of John
Stuart Mill, vol. 21 (1825), accessible online at http://oll.libertyfund.org/?option=
com_staticxt&staticfile=show.php%3Ftitle=255&chapter=21666&layout=html&Ite
mid=27; and Michael Walzer, Just and Unjust Wars (New York: Basic Books, 1977),
chap. 6.
29. Jeff McMahan, “The Meat Eaters,” Opinionator (blog), New York Times, September 19,
2010, accessible at http://opinionator.blogs.nytimes.com/2010/09/19/the-meat-
eaters/#more-61873.
30. Ibid., comments section.

6241-1145-PII-015.indd 292 8/11/2015 12:45:21 PM

 The Moral Problem of Predation 293

31. J. Baird Callicott, “Animal Liberation: A Triangular Affair,” in Robert Elliott, ed.,
Environmental Ethics (New York: Oxford University Press, 1995): 29–59, pp. 52–53
and 54.
32. Ibid.
33. Prominent moral naturalists include Richard Boyd, Philippa Foot, and Peter Railton.
34. Vucetich, Nelson, and Peterson, “Predator and Prey.”
35. Ibid.
36. Ibid.
37. In preparing this substantially revised and expanded version of my original piece in
the Opinionator, cited in note 29, I have greatly benefited from insightful written
comments by Andrew Chignell, Terence Cuneo, Catia Faria, Matthew Halteman,
Oscar Horta, and Ezekiel Paez.

6241-1145-PII-015.indd 293 8/11/2015 12:45:21 PM