In an applied ethics course I taught many times at Arizona State University, I had my students read the following article:
Singer, Peter. 1974. All animals are equal. Philosophical Exchange 1. Reprinted in LaFollette, Hugh (ed.). 2007. Ethics in Practice: Third Edition. Malden, MA: Blackwell. pp. 171-180.
This is a version of the notes I took on that article for my students. I am continuing to update the notes occasionally, even though right now I no longer work for ASU or teach this course.
Peter Singer probably is one of the most well-known philosophers alive today. He has been in the public spotlight a number of times for taking controversial moral positions based on an uncompromising utilitarianism. He was one of the pioneers of the animal welfare movement, starting principally with his 1975 book, Animal Liberation (New York: New York Review). I mention all of this because you are much more likely to encounter his name outside this course than most of the other authors we will be reading. The article we are reading is early work, published a year before Animal Liberation.
Singer’s thesis is that we ought to extend to (non-human) animals the same equality of consideration that we extend to human beings.
Here are what I think are the main contours of his argument (I am not putting his argument in premise-conclusion form here; I am just describing how it develops through his article):
- The only criterion of moral importance that succeeds in including all humans, and excluding all non-humans, is simple membership in the species Homo sapiens.
- However, using simple membership in the species Homo sapiens as a criterion of moral importance is completely arbitrary.
- Of the remaining criteria we might consider, only sentience―the capacity of a being to experience things like pleasure and pain―is a plausible criterion of moral importance.
- Using sentience as a criterion of moral importance entails that we extend to other sentient creatures the same basic moral consideration (i.e. “basic principle of equality”) that we extend to (typical, sentient) human beings.
- Therefore, we ought to extend to animals the same equality of consideration that we extend to human beings.
1. The only criterion of moral importance that succeeds in including all humans, and excluding all non-humans, is simple membership in the species Homo sapiens.
Singer argues for this simply by pointing to variation among humans. If we examine the usual characteristics that people say all humans, and only humans, share, we always find that there are human beings who lack those characteristics:
Like it or not, we must face the fact that humans come in different shapes and sizes; they come with differing moral capacities, differing intellectual abilities, differing amounts of benevolent feeling and sensitivity to the needs of others, differing abilities to communicate effectively, and differing capacities to experience pleasure and pain. In short, if the demand for equality were based on the actual equality of all human beings, we would have to stop demanding equality. (173)
The only characteristic that every single human being has in common, that no other animal has, is membership in the human species. If you want to say that every human is morally important, and that humans are the only creatures that are morally important, then your criterion for moral importance must be based simply on species membership.
2. However, using simple membership in the species Homo sapiens as a criterion of moral importance is completely arbitrary.
Singer seems to think this is fairly obvious once it is stated. If there is, in fact, no relevant difference between your group and some other group, then there is no rational ground for thinking that those who belong to your group deserve greater consideration than those who belong to the other group. Although it is fairly natural for people to use species membership as a criterion of moral importance, Singer thinks the obvious parallels with racism are so striking as to clearly invalidate that natural impulse. (This is a typically consequentialist way of thinking; you should consider how deontologists might look at this given the role of duties of special relationships in deontological ethics. Should common species membership be considered an appropriate special relationship?)
3. Of the remaining criteria we might consider, only sentience―the capacity of a being to experience things like pleasure and pain―is a plausible criterion of moral importance.
Singer argues for this in two ways. First, he argues by example that the other criteria are bad, because (again) they will exclude people who we think ought not be excluded. For instance, many people claim that the well-being of animals is unimportant because animals are not as intelligent as humans. However, clearly some human beings (say, those with very late-stage dementia) are less intelligent than some animals, but no one thinks that this makes the well-being of those human beings unimportant. Hence, intelligence is not a plausible criterion of moral importance.
Second, Singer argues that it is only by virtue of being sentient that anything can be said to have interests in the first place, so this puts sentience in a different category than the other criteria: “The capacity for suffering and enjoying things is a prerequisite for having interests at all, a condition that must be satisfied before we can speak of interests in any meaningful way” (175). Singer is trying to establish that if a being is not sentient, then the idea of extending moral consideration to it makes no sense. This negative argument is important, because one common criticism of Singer is that his criterion ends up excluding humans who are no longer sentient (like those in an irreversible coma). Singer is content to accept that consequence, but it is important that he show why the exclusion of some humans by his criterion is not problematic, given that he has criticized other criteria for their exclusions: this is what his negative argument tries to do.
However, even if he establishes that non-sentient creatures do not merit moral consideration, this is not enough to demonstrate that sentient creatures do merit moral consideration. Singer doesn’t supply much argument for the latter claim, but probably has the same kind of rationale in mind as other utilitarians; namely, that our recognition that it is bad for us to suffer should automatically lead us to conclude that suffering is bad, period. If suffering is bad, then, prima facie, it must be bad for any creature to suffer. So, if we want to make a radical distinction between our suffering and the suffering of other beings, the burden of proof is on us. Looking again at the arguments above, it is clear that Singer thinks this burden of proof cannot be met―our criterion of moral importance must be sentience.
4. Using sentience as a criterion of moral importance entails that we extend to other sentient creatures the same basic moral consideration (i.e. “basic principle of equality”) that we extend to (typical, sentient) human beings.
I need to unpack all of this carefully for you, because what Singer means by it often is misunderstood, and he doesn’t explain it in enough detail in this article to prevent misunderstandings. Let’s be clear: Singer is not saying that we are required, in practice, to treat humans and (nonhuman) animals the same. Extending to animals the same moral consideration we extend to humans means that we give the interests of animals the same weight as comparable interests of humans. However, not all interests necessarily are comparable. In other words, we cannot give the interests of animals less weight just because the beings that have them are animals, but it may be that the interests animals happen to have are the kinds of interests that do have less weight.
To clarify, let me introduce a distinction Singer makes outside of this article: that between persons and sentient non-persons. Singer defines a person (or a creature with personhood) as a creature that is aware of its own persistence over time. Personhood is not, incidentally, coextensive with humanity: Singer contends that adult chimpanzees are persons, but that human infants are not. Because they are aware of their own existence over time, persons are vulnerable to special forms of suffering that sentient non-persons cannot experience. A sentient non-person, for instance, can feel fear while dying, but only a person can experience the extra dread that comes from comprehension of its own mortality. A sentient non-person can experience pain, but only a person can feel the hopelessness brought on by awareness that his or her pain will last for weeks, months, or years into the future. In other words, persons have some important interests that sentient non-persons do not. So, taking into equal consideration the comparable interests of all sentient beings does not necessarily force us to treat all sentient beings equally. If, for instance, one can save either a dog or a human adult from a burning building, Singer would say you must save the human adult, because the balance of interests satisfied will be greater if you save the adult, than if you save the dog.1
However, Singer argues, situations where one is faced with that kind of choice are unusual. There are indeed difficult cases, such as animal experimentation, where we need to sit down and weigh carefully how much suffering each of our possible courses of action will cause; all Singer asks is that we do the calculation in those cases, instead of dismissing the suffering of non-humans as unimportant from the very start. Most of our practices toward animals, however, are very easy cases to Singer: in the vast majority of our practices toward animals, we sacrifice important animal interests (such as life and freedom) for the sake of absolutely trivial human interests (such as satisfying a taste for meat). In these cases, the interests in question still are noncomparable, but in the other direction: the suffering inflicted, for instance, on factory farm animals vastly outweighs our enjoyment of cheap and convenient meat. All such practices, Singer concludes, quite clearly lack any moral justification, and must be eliminated as quickly as possible.
The scope of livestock slaughter
I will close these notes with some statistics. The following are the United States livestock and poultry slaughter statistics for the year of 2014 (with a bit of rounding), from the USDA National Agricultural Statistics Service:2
- 30.2 million cattle
- 565,800 calves
- 106.9 million hogs
- 2.3 million sheep and lambs
- 8.66 billion chickens (with a “b”—not a typo)
- 235 million turkeys
- 26 million ducks
If one agrees with Singer’s argument, these statistics offer a glimpse of a moral problem so immense that it must dwarf virtually any other moral problem imaginable. If not, then perhaps these statistics represent the pinnacle of human achievement in efficiency and mastery over the world. I’ll leave it at that.
1 In some earlier iterations of these notes, I referred to the balance of pleasure over pain rather than the balance of interests satisfied. However, these are not quite the same thing: even though every sentient creature has an interest in experiencing pleasure and avoiding pain, and the interests of at least some creatures (perhaps all nonpersons) are exhausted by this, it is possible for persons to have interests that go beyond this. It is unclear whether Singer would have drawn this distinction when he originally wrote his article, but he would now: he is on record as a preference utilitarian, not a classical utilitarian, so he defines the good in terms of satisfaction of interests regardless of whether or not pleasure and pain are involved.
2 Livestock stats: http://usda.mannlib.cornell.edu/usda/nass/LiveSlauSu//2010s/2015/LiveSlauSu-04-27-2015.pdf (there is a summary on p. 6). Poultry stats: http://usda.mannlib.cornell.edu/usda/nass/PoulSlauSu//2010s/2015/PoulSlauSu-02-25-2015.pdf.