Utilitarianism about animals and the moral significance of use

David Killoren1 • Robert Streiffer2

Published online: 18 December 2018

� Springer Nature B.V. 2018

Abstract The Hybrid View endorses utilitarianism about animals and rejects util-

itarianism about humans. This view has received relatively little sustained attention

in the philosophical literature. Yet, as we show, the Hybrid View underlies many

widely held beliefs about zoos, pet ownership, scientific research on animal and

human subjects, and agriculture. We develop the Hybrid View in rigorous detail and

extract several of its main commitments. Then we examine the Hybrid View in

relation to the view that human use of animals constitutes a special relationship. We

show that it is intuitively plausible that our use of animals alters our moral obli-

gations to animals. That idea is widely believed to be incompatible with the sort of

utilitarian approach in animal ethics that is prescribed by the Hybrid View. To

overturn that conventional wisdom, we develop two different principles concerning

the moral significance of human use of animals, which we call the Partiality Prin-

ciple and the Strengthening Principle. We show that the Partiality Principle is

consistent with several key commitments of the Hybrid View. And, strikingly, we

show that the Strengthening Principle is fully consistent with all of the main

commitments of the Hybrid View. Thus we establish the surprising result that

utilitarians about animals can coherently offer a robust and intuitively appealing

account of the moral significance of animal use.

Keywords Utilitarianism � Consequentialism � Animal ethics � Special relationships � Use

& David Killoren

[email protected]

Robert Streiffer

[email protected]

1 Australian Catholic University, Melbourne, Australia

2 University of Wisconsin-Madison, Madison, USA

123

Philos Stud (2020) 177:1043–1063

https://doi.org/10.1007/s11098-018-01229-1

1 Introduction

On one view, utilitarianism is mistaken about how we ought to treat human beings,

but is more or less correct about non-human animals. Robert Nozick considers

(without endorsing) a version of this view, which he calls utilitarianism for animals,

Kantianism for people (1974: 39). Nozick’s name for the view is overly narrow (not

all non-utilitarian approaches are Kantian). Here’s the name we’ll use: the Hybrid

View.

All versions of the Hybrid View (as we define it) endorse utilitarianism about

animals and reject utilitarianism about humans. We explain what it means to be a

utilitarian about animals in Sect. 3. But before we do that, we show in Sect. 2 that

the Hybrid View underlies common views about animal research, agriculture, and

zoos. The Hybrid View has been mostly overlooked in the philosophical literature,

but is at least tacitly endorsed by a great many non-philosophers, including

professionals and policy-makers in science and other areas where ethical concerns

about animals are highly practically relevant. For this reason, sustained philosoph-

ical attention to the Hybrid View is overdue.

Among recent views in philosophical ethics, Jeff McMahan’s Two-Tiered

Account (2002: 245–251) comes closest to the Hybrid View [n.b., McMahan

professes agnosticism on key features of the Two-Tiered Account (2002: 260)].

According to the Two-Tiered Account, the morality of respect governs in the realm

of persons (including most but not all humans) while the morality of interests

governs in the realm of non-persons (including most but not all animals).

McMahan’s Two-Tiered Account does not count as a version of the Hybrid View

for two reasons. First, while the morality of interests eschews certain non-utilitarian

elements [e.g., a requirement to respect dignity (2002: 260)], it is not an

unrestrictedly utilitarian principle. Second, the Two-Tiered Account distinguishes

between persons and non-persons, which is different from the Hybrid View’s

distinction between humans and animals. Nevertheless, the Two-Tiered Account

and the Hybrid View belong to the same family: both posit that different classes of

individuals belong to different moral realms and that different principles ought to

govern our treatment of individuals in those different classes.1

In developing the Two-Tiered Account, McMahan fails to satisfactorily address a

difficulty for theories of this family: what to do when our actions affect individuals

in both classes, e.g., when our actions benefit humans but harm animals. Given that

such cases are rife, there is a pressing need to clarify what it even means to posit the

sort of moral separation that is characteristic of both the Two-Tiered Account and

the Hybrid View. In Sect. 3, we’ll address that difficulty and uncover core

commitments of the Hybrid View.

Our aim is not to defend the Hybrid View. Rather, we aim to rigorously

characterize the Hybrid View and to consider the extent to which it can be

reconciled with the view that humans stand in special relationships to animals.

1 For other theories within this family, see (Cohen 1986; Cohen and Regan 2001; Singer 2011: 71–93;

Varner 2012; Regan 1983: 246).

1044 D. Killoren, R. Streiffer

123

We’re especially interested in the relationship between an animal and a human who

uses that animal for human purposes.

In Sect. 4, we explain why proponents of the Hybrid View should want to

accommodate the view that human use of animals gives rise to special relationships.

We identify two ways of understanding the moral significance of use—the Partiality

Principle (Sect. 5) and the Strengthening Principle (Sects. 6, 7)—and explore their

compatibility with the Hybrid View and its associated commitments.

As will be discussed below, both principles look like deontological principles

incompatible with a utilitarian approach in animal ethics. Yet we’ll show that the

Strengthening Principle can be incorporated into a version of the Hybrid View,

allowing the Hybrid View to be more accommodating of characteristically

deontological claims than it initially appears to be. Further, we show that the

Partiality Principle can be incorporated into a theory that shares some of the same

utilitarian attractions as the Hybrid View, although it falls short of being a version of

the Hybrid View.

Our results in this paper are significant not only for animal ethics but also for

ethical theory generally. Many philosophers believe that utilitarianism is incom-

patible with the possibility that special relationships matter morally in and of

themselves. W.D. Ross neatly expresses that idea: ‘‘If the only duty is to produce the

maximum of good, the question who is to have the good—whether it is myself, or

my benefactor, or a person to whom I have made a promise to confer that good on

him, or a mere fellow man to whom I stand in no such special relation—should

make no difference to my having a duty to produce that good’’ (Ross 1930: 22). And

this idea remains widely accepted today (Crisp m.s.). We provide a novel argument

against that conventional wisdom.

2 The Hybrid View underlies common views about zoos, research, and agriculture

To date, there has been little psychological research on whether folk intuitions

involve a version of the Hybrid View.2 But common views about and regulations

governing institutions involving the use of animals provide ample evidence that

many are inclined to some form of the Hybrid View.

Let’s start with zoos. In some zoos, animals are treated with cruelty and neglect.

But in well-run zoos, animals are treated well and are, on the whole, happy. Many

are skeptical that there are grounds for morally objecting to well-run zoos.

To be sure, some believe that even the best zoos are unethical. Lori Gruen says

that confining animals in zoos ‘‘is an exercise of domination, and it violates [the

animals’] Wild dignity, even if it doesn’t cause any obvious suffering’’ (2011: 155).

Here Gruen suggests the non-utilitarian idea that features of an animal’s captivity

can matter independently of their effects on the animal’s welfare. But Gruen’s views

2 Lucius Caviola et al. (in preparation) have obtained survey evidence that people are more willing to

harm a few animals to save many animals than to harm a few humans to save many humans, which

suggests that people hold a utilitarian view about animals but a deontological view about humans.

Utilitarianism about animals and the moral significance… 1045

123

are not the norm: vast crowds of ordinary people who consider themselves to be

morally decent routinely enjoy visiting zoos they believe to be well-run. This

suggests that many believe that as long as zoos ensure that their captives are

sufficiently happy, there is no basis for moral objection, which in turn suggests that

many people have utilitarian intuitions about zoos.

By contrast, when we consider the idea of human zoos, purely utilitarian

intuitions become scarce among decent people. It seems that there is something

seriously wrong with confining humans in zoos, even in a hypothetical case in which

the captive humans are treated exceptionally well and are very happy.

And one finds a similar combination of views in many other human-animal

relationships that involve captivity. For example, most people have no principled

moral objection to responsible ownership of non-human pets, yet ownership of

human pets seems abhorrent in principle. These views can be elegantly explained in

terms of the Hybrid View.

Comparing human research regulations with animal research regulations also

suggests a commitment to a Hybrid View, albeit one implemented through historical

accidents and political compromises. Human subjects research regulations include a

variety of deontological elements. For example, in the US and other jurisdictions,

there is a near absolute prohibition on using children in non-therapeutic research

that is more than a minor increment over minimal risk (Ross 2005; Gennet and

Altavilla 2016), even though such research could be beneficial in the aggregate. This

prohibition is largely motivated by deontological concern about harming people

without their consent.

Yet, with the exception of a widely publicized recent NIH funding decision about

chimpanzees, no remotely comparable prohibition exists in US regulations on

animal research. This reflects a broadly utilitarian approach that allows for the use

of animals in significantly harmful, non-therapeutic research despite animals’

inability to consent. As Gary Varner observes, most animal researchers agree with

Peter Singer’s utilitarian view of animals, even as they disagree with Singer about

whether that view supports or condemns wide swaths of animal research (Varner

1994: 26). The few elements of US regulations on animal research that take a non-

utilitarian form, such as the general prohibition on using paralytics without

anesthesia, would seem to be justified as wise policy from a purely utilitarian

perspective. Even the recent US decision regarding chimpanzees was largely driven

by the empirical finding that ‘‘most current use of chimpanzees for biomedical

research is unnecessary’’ (Institute of Medicine 2011: 66–67).

The Nuffield Council on Bioethics found that a utilitarian cost–benefit analysis

serves as ‘‘the cornerstone’’ of animal research regulations in the UK (Nuffield

2005: 27, 52). Admittedly, the Council shies away from classifying the regulations

as endorsing a utilitarian view of animals, noting that the regulations include a de

facto ban on research involving great apes, a prohibition on harmful experiments for

entertainment, a prohibition on using animals in cosmetics research, and a

prohibition on procedures likely to cause unalleviated severe pain or distress

(Nuffield 2005: 52, 53). Yet such prohibitions at the level of policy can also be

given a utilitarian justification. Regarding the EU more generally, Andrew Knight

1046 D. Killoren, R. Streiffer

123

goes so far as to say that a ‘‘utilitarian cost:benefit analysis underpins all

fundamental regulations governing animal experimentation’’ (2011: 3, 4).

The Hybrid View also seems implicit in widespread views about agriculture.

Consider the view that Jeff McMahan calls humane omnivorism, according to which

‘‘factory farming is objectionable because of the suffering it inflicts on animals,

[but] it is permissible to eat animals if they are reared humanely and killed with

little or no pain or terror’’ (2016: 65) [c.f. ‘‘benign carnivorism’’ in McMahan

(2008)]. The well-known agricultural industry consultant Temple Grandin is a

devotee of this view: she has devised and advocated for handling and slaughter

methods designed to reduce the suffering of farm animals, so she evidently thinks

that the happiness of animals matters. Yet she still endorses the slaughter that is

essentially involved in animal agriculture. Grandin’s humane omnivorism is thus

consistent with (though may not entail) a utilitarian approach to animal ethics.

Relatedly, Peter Sandøe observes that farm animal legislation often ‘‘has strong

affinities to a utilitarian way of thinking’’ (2003: 473). By contrast, few would

defend the use of humans for food even if it could somehow be justified on

utilitarian grounds.

Some aspects of common thinking about animal ethics are at odds with the

Hybrid View.3 Nevertheless, the points we have covered in this section establish

that the Hybrid View captures a significant fraction of common beliefs and practices

regarding humans and animals.4 For this reason the Hybrid View is deserving of

serious attention from philosophers. We turn now to the task of giving a rigorous

characterization of the Hybrid View’s core commitments.

3 The Hybrid View

What does it even mean to be a utilitarian about animals but not about humans?

A first difficulty in answering this question derives from the fact that

utilitarianism is a diverse family of theories. We’ll want to understand the Hybrid

View in a way that is neutral between the major branches of the utilitarian family

(though we’ll set aside rule-utilitarian and expected-utility variants). So we need to

identify commitments that utilitarians have in common. For starters, we suggest that

utilitarians of all major types will agree to the following consequentialist thesis:

3 Consider a scene from the early-90s movie Fierce Creatures: Willa Weston objects to Vince McCain’s

attempts to raise zoo revenue by securing sponsorships for the animals because it is ‘‘degrading to the

animals.’’ Pointing to a sign on a tiger which has a picture of a bottle of Absolut vodka next to the slogan

‘‘Absolut Fierce,’’ she says ‘‘That is unacceptable!’’ Willa’s concern about degradation seems not to

derive from concern about welfare and thus seems to be at odds with utilitarianism about animals. Insofar

as Willa’s concern would be shared by many ordinary people, it illustrates that not all judgments about

animals are in line with the Hybrid View. Thanks to an anonymous referee for pushing us on this point. 4 There is, of course, a further question as to why people who hold the Hybrid View do so. It may be an

expression of speciesism—an irrational bias toward members of one’s own species—or it may be a

consequence of views about the different kinds of values that can be realized in the lives of animals versus

the lives of humans, values that call for different kinds of moral responses. We needn’t take a stand on

this question here.

Utilitarianism about animals and the moral significance… 1047

123

Optimific Is Always Okay: It is never wrong, and hence it is always

permissible, for an agent to act optimifically (where an action is optimific iff

its consequences are at least as good as the consequences of any alternative

action open to the agent).5

To see why Optimific Is Always Okay will be ecumenically appealing among

utilitarians, consider utilitarianism’s three major varieties: maximizing, satisficing,

and scalar.

Maximizing utilitarians hold that we always and everywhere have an (all things

considered) obligation to act optimifically (Moore 1903: 76, 77, 197, 198; Brandt

1988: 342; Mogensen 2016: 215). Those who accept such a view will endorse

Optimific Is Always Okay.

Satisficing utilitarians hold that we are always and everywhere obligated to do

what is at least good enough, and that in some cases an action may be good enough

without being optimific (Slote and Pettit 1984; Jamieson and Elliot 2009). For the

satisficer, optimific actions are in many cases supererogatory, but will always be

above the ‘‘good enough’’ threshold and so will never be impermissible.

Scalar utilitarians do away with deontic properties altogether and thus do not

regard any actions as obligatory, forbidden, right, wrong, etc. (Norcross 2006). For

the scalar utilitarian, optimific actions are never wrong because no actions are

wrong.6

By contrast, deontological theories characteristically imply that it is sometimes

wrong to act optimifically (Kagan 1989; Kamm 2001: 207–289, 2007: 11–47,

130–189; Mogensen 2016; Sinclair 2017). For example, a standard deontological

claim, anathema to utilitarians, is that it can be morally wrong to kill one person in

order to harvest her organs to save five even when doing so is optimific. Thus,

Optimific Is Always Okay represents a distinction between deontologists and

utilitarians.

Given that utilitarianism is committed to Optimific Is Always Okay, the Hybrid

View should involve some restricted version of Optimific Is Always Okay that

applies narrowly within the realm of animals and does not apply in the realm of

humans. But it is not immediately obvious how to construct such a restricted

version.

As a first pass, one might think that Optimific Is Always Okay applied narrowly

to animals would look like this:

The Animal-Affecting Criterion: If u-ing affects animals, then u-ing is morally

permissible if u-ing is optimific.

But this principle is unsatisfactory for the simple reason that many actions affect

animals and humans. To see why this creates a problem, consider the following

5 For consequentialists, the goodness of an act’s consequences can include any intrinsic value the act

itself might have (Thomson 1994: 14). 6 Recently, some scalar consequentialists (Sinhababu 2018) have argued that actions have deontic

properties but that those properties are matters of degree. These theorists too will typically sign on for

Optimific Is Always Okay.

1048 D. Killoren, R. Streiffer

123

variant of the footbridge trolley case. A runaway trolley is headed toward five

innocent people. You can push a large man from a footbridge into the trolley’s path,

killing the man, but stopping the trolley and sparing the five. Here’s the twist: if the

large man is pushed into the trolley’s path, his large body will provide a hearty feast

for vultures; but the five are so small that if the trolley is allowed to kill them, the

vultures will be left with only a light snack.

A standard deontological view is that it is wrong to push the large man. So the

Hybrid View should be consistent with that view, since the Hybrid View is non-

utilitarian about humans. But the Animal-Affecting Criterion implies that it is

morally permissible to push the large man, simply because the vultures will be

affected by this decision and pushing the large man is optimific. Given this, the

Animal-Affecting Criterion cannot be built into the Hybrid View.

Here is a second principle that might be thought to represent Optimific Is Always

Okay applied within the realm of animals:

The Not-Human-Affecting Criterion: If u-ing does not affect human beings,

then it is morally permissible to u if u-ing is optimific.

This principle seems to cohere with the sorts of views that motivate the Hybrid

View, but it is largely uninformative. Nearly all actions we have an interest in

ethically evaluating will have some effects on humans, so the antecedent will rarely

be satisfied in real-world cases.

Here is a third possibility:

The Moral Significance Criterion: If u-ing affects animals in a morally

significant way but does not affect humans in any morally significant way,

then u-ing is morally permissible if u-ing is optimific.

Unlike the Not-Human-Affecting Criterion, the Moral Significance Criterion has

substantial implications for real-world cases. For there are many cases in which an

action has morally significant effects on animals but no morally significant effects

on humans.

Nevertheless, the Moral Significance Criterion does not fully capture the Hybrid

View, for it does not address actions that have morally significant effects on both

humans and animals. Thus, the Moral Significance Criterion does not capture many

of the sorts of views we discussed in Sect. 2 to motivate the Hybrid View. For

example, the Moral Significance Criterion cannot be invoked to explain the alleged

permissibility of harmful animal research performed for the sake of greater human

benefit. This means that although the Moral Significance Criterion should be

regarded as a commitment of the Hybrid View, it cannot be taken to be exhaustive

of the Hybrid View’s commitments.

Here is a criterion that captures the sorts of views we’ve just mentioned:

The Negative Criterion: If u-ing affects both humans and animals in morally

significant ways and u-ing is optimific, then u-ing is morally wrong only if u- ing is morally wrong in virtue of a relation between u-ing and those humans

who will be affected by u-ing.

Utilitarianism about animals and the moral significance… 1049

123

We believe that those who want to endorse the Hybrid View will sign on for the

Negative Criterion. And we think that a commitment to the Negative Criterion is a

distinctive (though perhaps not unique) feature of the Hybrid View.

Consider a case in which animals are harmed in research that will provide

benefits for humans and thus affects animals and humans in morally significant

ways. Suppose that the research is optimific (because the benefits exceed the harms

and there is no better alternative). Then the Negative Criterion implies that the

research is not wrong unless it is wrong in virtue of some relation between the

research and the humans who will be affected by it. This seems clearly to be a

judgment that defenders of the Hybrid View as discussed in Sect. 2 will want to

endorse.

In light of these considerations, the Hybrid View should be interpreted as

including a commitment to both the Moral Significance Criterion and the Negative

Criterion.

Thus far we’ve been considering features that utilitarianism shares with other

consequentialist theories: Optimific Is Always Okay is a consequentialist principle,

not a specifically utilitarian one. Now we need to consider some distinguishing

features of utilitarianism in particular.

We take utilitarianism (as understood in modern philosophical vernacular) to

involve two major commitments. First, utilitarians are welfarists about value:

utilitarians believe that only states that constitute welfare are intrinsically (non-

instrumentally) good or bad. Second, utilitarians are subjectivists about welfare:

utilitarians believe that mental states either constitute welfare or determine which

states constitute welfare.

There are three main theories of welfare: hedonism, preferentism, and objective-

list theories (Parfit 1984: 493–502; Kagan 1998: 29–41; c.f. Woodard 2013).

According to hedonism, only hedonic states constitute welfare. According to

preferentism, only desire satisfaction and desire frustration constit