We Are Not Human Beings

In this article, Parfit explores the questions of personal identity and numerical identity – what we are and how we might continue to exist.


We can start with some science fiction. Here on Earth, I enter the Teletransporter. When I press some button, a machine destroys my body, while recording the exact states of all my cells. This information is sent by radio to Mars, where another machine makes, out of organic materials, a perfect copy of my body. The person who wakes up on Mars seems to remember living my life up to the moment when I pressed the button, and is in every other way just like me.

Of those who have thought about such cases, some believe that it would be I who would wake up on Mars. They regard Teletransportation as merely the fastest way of travelling. Others believe that, if I chose to be Teletransported, I would be making a terrible mistake. On their view, the person who wakes up would be a mere Replica of me.

This disagreement is about personal identity. To describe such disagreements, we can first distinguish two kinds of sameness. Two black billiard balls may be qualitatively identical, or exactly similar. But they are not numerically identical, or one and the same ball. If I paint one of these balls red, it will cease to be qualitatively identical with itself as it was; but it will still be one and the same ball. Consider next a claim like, ‘Since her accident, she is no longer the same person’. This claim involves both senses of identity, since it means that she, one and the same person, is not now the same person. That is not a contradiction, since it means that this person’s character has changed. This numerically identical person is now qualitatively different.

When people discuss personal identity, they are often discussing what kind of person someone is, or wants to be. That is the question involved, for example, in an identity crisis. But I shall be discussing our numerical identity. In our concern about our own futures, that is what we have in mind. I may believe that, after my marriage, I shall be a different person. But that does not make marriage death. However much I change, I shall still be alive if there will be someone living who will be me. And in my imagined case of Teletransportation, my Replica on Mars would be qualitatively identical to me; but, on the sceptic’s view, he wouldn’t be me. I shall have ceased to exist. That, we naturally assume, is what matters.

In questions about numerical identity, we use two names or descriptions, and we ask whether these refer to the same person. In most cases, we use descriptions that refer to people at different times. Thus, when using the telephone, we might ask whether the person to whom we are speaking now is the same as the person to whom we spoke yesterday. To answer such questions, we must know the criterion of personal identity over time, by which I mean: the relation between a person at one time, and a person at another time, which makes these one and the same person. We can also ask what kind of entity we are, since entities of different kinds continue to exist in different ways.

Views about what we are, and how we might continue to exist, can be placed, roughly, in three main groups. On some views, what we are, or have as an essential part, is a soul: an immaterial persisting entity, which is indivisible, and whose continued existence must be all-or-nothing. Even if we don’t believe in immaterial souls, many of us have some beliefs about ourselves, and personal identity, that would be justified only if some such view were true. Though such views make sense, and might have been true, I shall not discuss them today, since we have strong evidence that no such view is true.

Of the other views, some can be called Lockean. Locke famously defined a person as ‘a thinking intelligent being that has reason and reflection and can consider itself as itself, the same thinking thing in different times and places’.2 Lockean criteria of identity appeal to the kind of psychological continuity that, in my imagined case, holds between me and my Replica. The Lockean view that I have earlier defended, which I called

the Narrow, Brain-Based Psychological Criterion States: If some future person would be uniquely psychologically continuous with me as I am now, and this continuity would have its normal cause, enough of the same brain, this person would be me. If some future person would neither be uniquely psychologically continuous with me as I am now, nor have enough of the same brain, this person would not be me. In all other cases, there would be no answer to the question whether some future person would be me. But there would be nothing that we did not know.

On this view, my Replica would not be me, since he would not have my brain. That, I claimed, would not matter, since being destroyed and Replicated would be as good as ordinary survival. I shall later return briefly to that claim. The other main kind of view appeals not to psychological but to biological continuity, and is now often called Animalist.

In considering this disagreement, I shall first describe some Animalist objections to the various Lockean views that were put forward, in the nineteen sixties, seventies, and eighties, by such people as Shoemaker, Quinton, Perry, Lewis, and me. As Snowdon, Olson, and other Animalists pointed out, we Lockeans said nothing about the human beings – or to use a less ambiguous phrase, the human animals – that many of us think we are.

If persons are, in the Lockean sense, entities that can think about themselves, and whose continued existence essentially involves psychological continuity, a human embryo or fetus is not a person. But this fetus is, or becomes, a human animal. This animal’s body, Lockeans claim, later becomes the body of a Lockean person. Animalists ask: What then happens to the human animal? It would be convenient for Lockeans if this animal retired from the scene, by ceasing to exist, thereby leaving its body under the sole control of the newly existing person. But that is not what happens. Most human animals continue to exist, and start to have thoughts and other experiences. So if Lockeans distinguish between persons and human animals, their view implies that whenever any person thinks some thought, a human animal also thinks this thought. Every thinking of a thought has two different thinkers. That conclusion seems absurd. As McDowell writes: ‘surely there are not two lives being led here, the life of the human being. . . and the life of the person.’3 We can call this the Too Many Thinkers Problem.

There may also be an Epistemic Problem. If there are two conscious beings thinking all my thoughts, the person and the animal, how could I know which one I am? If I think I am the person, Animalists object, I might be mistaken, since I may really be the animal.

There is a third problem. Snowdon pointed out that, on Locke’s definition, human animals qualify as persons.4 So if Lockeans distinguish persons from human animals, they must admit that, on their view, all of our thoughts and other experiences are had by two persons, one of whom is also an animal. This objection may seem decisive, by undermining the whole point of this Lockean distinction. We can call this the Too Many Persons Problem.

Several Lockeans have suggested answers to these objections. Shoemaker, for example, argues that, if we claim animals to be entities whose criterion of identity is biological, and requires the continued existence of much of their bodies, such animals could not think, or have other mental states, since the concepts that refer to mental states apply only to entities whose criterion of identity is psychological. Though these human animals might seem to have thoughts and experiences, that would not really be true.5

Baker argues that the animal and the person are both constituted by the same body, which gives them an ontological status that is in between being one and the same entity and being two, separately existing entities. For that reason, Baker claims, though there are, strictly, two different thinkers thinking each of our thoughts, we can count these thinkers as if they were one.6

We can next distinguish between concepts which are substance sortals, in the sense that they apply to some persisting entity whenever it exists, and phase sortals, which apply to some entity, in the present tense, only while this entity has certain properties. Two such phase sortals are ‘teenager’ and ‘caterpillar’. When we reach the age of 20, we cease to be teenagers, but we don’t thereby cease to exist. Nor do caterpillars cease to exist when they become butterflies.

I have earlier suggested that, in response to these Animalist objections, Lockeans should claim that the concept of a person is another phase sortal.7 On this view, we are human animals who began to exist as an embryo or fetus though we were not then persons in the Lockean sense. And if we suffered brain damage which made us irreversibly unconscious, we would continue to exist, though we would have ceased to be persons. One of Locke’s aims was to describe persons in the way that makes most sense of our practical and moral beliefs. ‘Person’, Locke writes, ‘is a forensic term’, applying only to responsible rational beings. We could keep this part of the Lockean view if we claim that we have certain reasons, and certain principles apply to us, only while we are persons. For example, I might point to an ultra-sound image of an embryo or fetus, saying ‘There I am. That was me’, but adding that, since I was not then a person, it would not have been wrong for some doctor to kill me. We might make similar claims about the concept of a human being. We might say that, just as an acorn with one green shoot sprouting is not yet an oak tree, an embryo is not yet a human being. And some other moral principles apply to us, we might claim, only after we become human beings.

Lockeans, I now believe, need not retreat to any such claim. There is another, stronger Lockean view that can answer the Animalist objections that I have described. This view also avoids some problems that face Animalist views. So I shall next describe these other problems.


Most Animalists believe that we shall continue to exist if and only if our bodies continue to exist, and to be the bodies of living animals. Williams even claimed that persons are bodies.8 But suppose that, in

Transplanted Head, my body is fatally diseased, as is Williams’s brain. Since we have, between us, only one good brain and body, surgeons bring these together. My head is successfully grafted onto the rest of Williams’s headless body.

On Williams’s view, he would wake up with my head, being psychologically just like me and mistakenly believing that he was me.

Most of us would find that claim incredible. Suppose that you knew both Williams and me, and you visit the resulting person in the post-operative recovery room. You see my head on the pillow, and have a long conversation with someone whom you assume to be me. If some nurse then lifted the blankets on the bed, and you saw the rest of what you knew to be Williams’s body, you wouldn’t conclude that you weren’t, as you assumed, talking to me. You would believe that the person with my head would be me. As many Animalists concede, this widely held belief, which some call the Transplant Intuition, provides a strong objection to their view.

Olson suggests that Animalists can explain why most of us find this objection plausible.9 In all actual cases, Olson claims, when some present person is psychologically continuous with some past person, that is strong evidence that these people have the same body, thereby being one and the same person. So it is not surprising that we mistakenly believe that, if our brain and psychology were transplanted into a different body, we would wake up in that other body. We would find this Brain-Based Psychological Criterion just as plausible, even if, as Animalists believe, this criterion is false.

These claims do not, I believe, answer this objection to Animalism. When we compare different proposed criteria of identity, we should consider cases in which these criteria would conflict. If in such imagined cases Criterion A seems much more plausible than Criterion B, we cannot defend B by saying that A seems plausible only because, in all or most actual cases, A coincides with B. Suppose that, on the Finger Print Criterion, some future person would be the same as some present person if and only if these people have qualitatively identical finger prints. In rejecting this view, we could point out that, if some plastic surgeon remoulded the tips of someone’s fingers, we would all believe that this person would continue to exist, with the same brain and psychology, though with different finger prints. Finger Printists might reply that, if this Brain-Based Psychological Criterion seems more plausible when it conflicts with the Finger Print Criterion, that is only because, in nearly all actual cases, people with the same brain and psychology also have the same finger prints. That would be a weak reply. If the Finger Print Criterion seems much less plausible when these criteria conflict, that is a strong objection to this criterion. Similar remarks apply to the Transplant Intuition. If it seems very plausible that the person with my head but the rest of Williams’s body would be me, that is a strong objection to the Animalist claim that this person would be Williams.

Some Animalists assume that all animals have the same criterion of identity over time. Since many animals, such as oysters, do not even have psychologies or brains, these Animalists could not accept a Brain-Based Psychological Criterion of animal identity. But other Animalists might claim that different kinds of animal continue to exist in different ways, and with different criteria of identity. At least in the case of human beings, they might say, the animal goes with the brain. These Animalists would then agree that, in Transplanted Head, the resulting person would be me.

This version of Animalism may now seem to coincide with this Lockean view, thereby ending this disagreement. But that is not so. We should distinguish between our cerebra, or upper brains, and our brain stems. It is our cerebrum on which all of our distinctive mental activity depends. The brain stem controls the functioning of our body in other ways. Most Animalists believe that, if our cerebrum were destroyed, but our brain stem continued to maintain the functioning of our heart, lung, and most of our other organs, we the human animal would continue to exist, though in an unconscious vegetative state, or coma.

We can therefore add some details to our imagined case. We can suppose that, in Transplanted Head, my brain stem would be left behind. It is only my head and cerebrum that would be successfully grafted onto Williams’s brain stem and the rest of Williams’s body. With its brain stem retained, my body would then continue to be the body of a living though unconscious animal. It would be implausible to claim that this would now be a different animal, because the animal that used to have this body had gone with its cerebrum into a different body.

Suppose next that in another case, which we can call

Surviving Head, my head and cerebrum are not grafted onto someone else’s brain stem and body, but are kept alive and functioning by an artificial support system.

As before, you visit the post-operative recovery room, see my head on the pillow, and talk with what you take to be me. If the nurse lifted the blankets on the bed, and you saw not a human body but an artificial support system, you wouldn’t believe that the conscious being to whom you had been talking wasn’t me. Some Animalists might claim that this conscious being would be the same animal as me. But there is now a different objection to this view. This conscious being would be a person, in the Lockean sense. But this person, whose physical basis is only an artificially supported head, would not seem to be an animal.

To strengthen this objection, suppose that, in

Surviving Cerebrum, what is removed from my body is not my head, but only my cerebrum, which is then kept functioning by an artificial support system. The resulting entity is conscious, as the neuro-physiological evidence shows. There is also some device which enables this conscious being to communicate with the outside world, since the brain activity involved in certain voluntary mental acts enables this being to spell out the words of messages to us, and some other device enables us to send replies. In this way you have conversations with this conscious being, who claims to be me, seems to have all my memories, and starts to dictate the rest of my unfinished book.

As before, this conscious rational being would be a Lockean person, whom many of us would believe to be me. But it would be harder for Animalists to defend the claim that this conscious being, whose physical basis is only a cerebrum, is an animal, and the same animal as me.

It is worth supposing, however, that some Animalists make this claim. These people might say that an early embryo is a human animal, though it lacks most of the properties of a living organism. The same would be true, they might claim, of my detached, artificially supported cerebrum.

If Animalists made this claim, their view would cease to be an alternative to Lockean views. On the Lockean Brain-Based Psychological Criterion, some future person would be me if this person would be uniquely psychologically continuous with me, because he would have enough of my brain. This criterion implies that, in Surviving Cerebrum, the conscious being would be the same person as me. When Animalists entered this debate, their main claim was that such psychological criteria of identity are seriously mistaken, because we are human animals, so that our criterion of identity must be biological. If these Animalists now claimed that, in Surviving Cerebrum, the conscious rational being would be a living animal, who would be me, these people would be claiming that the true criterion of identity for developed human animals is of this Lockean psychological kind.10 Since these Animalists would now be Lockeans, I shall here consider only those other, Non-Lockean Animalists who would believe that, in Surviving Cerebrum, the conscious being, though a Lockean person, would not be an animal.

This fact, these Animalists might say, is compatible with their view, which claims only that most persons are animals. There might be some conscious beings that are Lockean persons but aren’t animals. But we could then ask: How would this conscious being be related to the human animal, Parfit, who used to have this cerebrum?

Animalists have two alternatives. They might claim that, when my cerebrum is detached from the rest of my body, a new conscious rational being comes into existence. But, as Johnston and Olson write, that claim would be hard to believe. It is hard to see how we could create a new conscious being merely by disconnecting my cerebrum from the rest of my body.

Suppose next that, after this conscious being spends many days communicating with us, my cerebrum is detached from its artificial support system and successfully grafted onto some other human animal’s brain stem and body. The resulting being would then be a human animal. But what would happen to the conscious being which existed for a period on its own, as a non-animal? It would be convenient for Animalists if this Lockean person, who is not an animal, would cease to exist when my cerebrum was grafted into the rest of someone else’s body. But it is hard to see how, merely by connecting this conscious being to the rest of this body, we would thereby cause this being to cease to exist. Animalists object to the way in which, when Lockeans describe how a young human animal becomes a person, Lockeans implausibly assume that the animal then retires from the scene. The same problem arises here the other way round. Animalists cannot plausibly assume that, when my cerebrum is grafted into this animal’s body, this Lockean person would retire from the scene. As Olson writes:

Animalism seems to imply that the detached brain would be a person who comes into being when the brain is removed and ceases to exist when the brain goes into a new head. And that seems absurd.11

Olson calls these the Creation and Destruction Problems.

To avoid these problems, Animalists might instead claim that this conscious being already existed when my cerebrum was in my body, and that this being would continue to exist both while it is artificially supported, and after it is grafted into another human animal’s body. But if Animalists made this claim, they would face another version of the Too Many Thinkers Problem. As well as the human animal thinking my thoughts, there would be another conscious being that was not an animal, thinking all the same thoughts. This problem, moreover, isn’t raised only by this imagined case. It applies to every actual fully developed human animal. On this version of the Animalist view, all of any human animal’s thoughts are also thought by another, different conscious being.

Animalists, Olson writes, therefore face this dilemma:

if your brain thinks now, there are too many thinkers; if it doesn’t, things can gain or lose mental capacities in an utterly baffling way.12

Olson calls this the Thinking Parts Problem.

This problem, he writes, is ‘considerably more serious than animalism’s unintuitive consequences in brain-transplant cases’, adding that ‘it has no obvious solution’.13


This problem has, I believe, an obvious solution. According to some Lockeans, as I have said, the person and the animal are both constituted by the same body, in a way that makes them, though not numerically identical, not wholly separate either. On this view, though it is strictly true that each thought is thought by two thinkers, the person and the animal, we can count these thinkers as if they were one.

According to another, better view, we are not animals, or human beings. We are what McMahan calls the conscious, thinking, and controlling parts of human beings. We can call this the Embodied Part View. The Thinking Parts Problem has a thinking parts solution.

On this view, the Creation and Destruction Problems disappear. If my cerebrum were detached from the rest of my body and artificially supported, no new conscious being would mysteriously come into existence. Nor would a conscious being mysteriously disappear if my cerebrum were later successfully grafted into another human body. The same conscious being would exist throughout, first as the thinking, controlling part of one human animal, then existing for a while on its own, then becoming the thinking, controlling part of a different human animal.

This view also avoids the Too Many Thinkers Problem. Animals digest their food by having a part, their stomach, that does the digesting. Animals sneeze by having a part, their nose, that does the sneezing. These facts do not create a Too Many Digesters or Too Many Sneezer’s Problem. Human animals think, we can similarly claim, by having a part that does the thinking. There are not too many thinkers here.

Some Animalists consider this Embodied Part View. Olson writes:

If we are neither animals nor material things constituted by animals, we might be parts of animals.

But Olson then rejects this view, calling it ‘a desperate ploy’, and doubting whether anyone ‘seriously advocates’ this view.14

Olson rejects this view because he assumes that an animal’s thinking part would have to be claimed to be its brain. Though Olson calls it ‘just about conceivable that the brain view might be true’, he does not include this view among the ‘live options’ that are worth considering. Johnston similarly claims that, if we accept the view that brains can think, we shall be led to absurd conclusions.15

What Olson calls the brain view is, however, only one version of the Embodied Part View. This version is not, I believe, absurd. Some other Animalists claim that, rather than having bodies, we are bodies. On that view, it is our bodies that have our experiences, and think our thoughts. If these other Animalists came to accept the Embodied Part View, they might claim that the animal’s conscious thinking part is not its body but its cerebrum or upper brain. Rather than saying, for example, that Einstein’s body discovered the theory of general relativity, they would say that Einstein’s brain made this discovery. Many people would find that claim more plausible. In a well-known radio quiz programme, people compete for the title Brain of Britain. And Hercule Poirot says, tapping his forehead ‘These little grey cells. It is “up to them”’.

If we are Embodied Part Theorists, however, we need not make such claims. The thinking part of a human animal, we could say, is related to this animal’s cerebrum or upper brain in a way that is roughly similar to that in which this animal is related to its whole body. Most of us distinguish between ourselves and our bodies. If we deny that human animals are their bodies, we could similarly deny that the thinking part of these animals is their upper brain.

In what seems to me the best of the few published defences of the Embodied Part View, McMahan claims that we are the minds of human animals. McMahan calls this the Embodied Mind View.16

Some Animalists come close to accepting this view. Carter imagines a case in which President Nixon’s brain is transplanted into the empty skull of Senator McGovern. Nixon’s mind, Carter claims, would then become McGovern’s mind. And Carter writes:

McGovern’s mind may . . . remember being part of a person who stepped onto a certain helicopter after resigning as President.17

If we transferred Nixon’s brain into McGovern’s body, Carter adds, McGovern might bear some moral responsibility for Nixon’s decision to bomb Cambodia, since the mind that used to be Nixon’s but is now McGovern’s would be ‘the mind that once decided to do this awful thing’.

Since Carter claims that our decisions are made by our minds, we might expect him to claim that we, the decision-makers, are these minds. On that view, in Carter’s imagined case, we would not unjustly hold McGovern responsible for Nixon’s earlier decisions. But Carter rejects this view, writing

since people have arms and legs and minds don’t, people can’t be identified with minds.18

Some other people would object that, just as we shouldn’t claim that brains think, or make decisions, we shouldn’t claim that minds think, or make decisions. Johnston, for example, writes:

If we are saying that something is. . . thinking. . . the subject of predication should be an animal or person.19

Embodied Part Theorists can make similar claims. On a third version of this view, human animals think by having a conscious thinking part which is a person in the Lockean sense. We can call this the Embodied Person View. This, I believe, is the best version of the Embodied Part View.20

Though Olson claims that the Thinking Parts Problem has no obvious solution, this problem, he writes, ‘is no reason to prefer any other view to Animalism’.21 That is not so. This problem is a strong reason to prefer the Embodied Part View, since this is the only view on which the Thinking Parts Problem disappears.

Nor is this view merely a philosophical invention, since it states more clearly what many non-philosophers already believe, or would after reflection believe. Olson writes,

no one thinks that we are heads.

No one, we should agree, thinks that we are just heads. But we might be embodied heads. And most of us would believe that, for us to survive, it would be enough that our head survives, and continues to be the head of a conscious being. The body below the neck is not an essential part of us.

We can next mention the actual cases of those conjoined twins who share all or most of the same body below the neck, but have two heads, and have different thoughts and other experiences. No one doubts that these are the heads of two different people.

It may be less obvious that, for us to survive, it would be enough that our cerebrum survives. If those who love me, rather than seeing my head on a pillow, saw only an artificially supported cerebrum floating in a vat, they might doubt that I was still there. But as the dictated messages would show, the conscious being based on this cerebrum would be psychologically just like me, seem to have all my memories, etc. After reflection most of us would believe that I was still there. Whether I have continued to exist could not depend on whether my functioning cerebrum retained its outer covering of bone and skin, so that this conscious being still looked like me.

There are many actual cases of another relevant kind. One such case is that of Nancy Cruzan, whose cerebrum ceased to function, but whose brain stem maintained her body in a vegetative state for seven years until a US Supreme Court ruling granted her parents’ request to have an artificial feeding tube removed. On Cruzan’s gravestone her parents had inscribed:

Departed January 11 1983 At Peace December 26 1990

When Cruzan’s cerebrum was irreparatly damaged, her parents came to believe, Cruzan the person departed from her body, though the human animal continued to exist with its heart beating and its lungs breathing until, after the feeding tube was removed, the heart stopped and the animal was at peace.

Nor are we merely appealing to such intuitions. We have reasons to make such claims. While defending Animalism, Olson writes

if there are now two things thinking your thoughts, one doing it on its own and the other such that its thinking is done for it by something else, you are the one that thinks on its own.22

But this plausible own-thinker principle supports, not Animalism, but the Embodied Person View. The animal’s thinking is done for it by something else, the part whose physical basis is the cerebrum. This human animal could not think on its own, since without this part it could not think at all. But the conscious thinking part can think on its own, as it would do in some of the imagined cases that we have been considering. If, as Olson claims, we are ‘the one that thinks on its own’, we are not the animal but this conscious thinking being, the Lockean embodied person.

When Johnston discusses these cases, he appeals to something like the own-thinker principle. Johnston considers the suggestion that, though the person and the animal think the same thoughts, that is not puzzling, since there are not two separate thinkers here. On this view,

(1) the person ‘counts as a thinker derivatively’, since the person thinks only ‘because the animal does’.

Johnston rejects this view, claiming that it ‘gives the wrong result’. In his words:

if I had to pick which of two things I am identical with, the person or the animal, a good rule would be: Pick the thing which is non-derivatively the subject of mental acts. And on the present proposal. . . it is the animal, and not the person, that is non-derivatively the subject of mental acts. . . But the result we wanted is that I am identical with the person.23

Since this proposal gets things the wrong way round, the obvious next suggestion is that

(2) we are identical, not with the animal, but with the person, which is the non-derivative thinker, and the subject of our mental acts.

Johnston comes close to accepting (2), since he discusses the view that any human animal has ‘a mental organ dependent on its brain, whose operation constitutes the animal’s thinking’, and he also claims that, as persons, we should take ourselves to be ‘the non-derivative or primary source of the thought in us’. These claims support the view that

(3) we, who are persons, are the part of the animal that does the thinking.

Johnston, however, rejects (3). Summarizing what is shown by the Too Many Thinkers Problem, Johnston writes:

Olson has a sound argument here. . . one that must condition all further discussion of personal identity. We are animals. 24

Johnston adds only that we, who are persons, are not essentially animals, since he believes that, in cases like Surviving Head or Cerebrum, we would continue to exist as persons, though we would have ceased to be animals. Though Johnston’s other claims imply that we are the part of the animal that does the thinking, Johnston’s belief that we are animals leads him to reject that conclusion.


I turn now to possible objections to the Embodied Part View, whose best form I have claimed to be the Embodied Person View.

One objection is that, as Carter claims, we can’t be our minds, since we have arms and legs and our minds don’t. Discussing the view that we are brains rather than minds, Olson similarly writes:

Is it really a serious view. . . that we are about four inches tall and weigh about three pounds?25

We can call this the Physical Properties Objection. On this objection, we have many physical properties which cannot be had by our conscious thinking part, whether we claim this part to be a brain, or a mind, or a Lockean person. Since we have such physical properties, we must be human animals, rather than some part of these animals.

This objection can be answered. If we are embodied persons, as I believe, we can explain how and why we can intelligibly claim ourselves to have the physical properties of our bodies. We already do that now if we distinguish between ourselves and our bodies, as when I say that I am 6 foot tall and weigh 160 pounds because my body has these properties.

We sometimes use ‘I’ and ‘me’ more widely, to refer to more than our bodies. I might say, for example, that I have been splashed with mud, though it was only my trousers that were splashed. And if I were a veiled Islamic woman, I might say that someone had seen me, though this person saw only my clothes. If we are the conscious controlling part of an animal, we are very closely related to the rest of this animal’s body, in which we can feel sensations, and with which we can see, hear, smell, and touch the world around us. As Descartes wrote, while defending his soul-involving version of the Embodied Part View, this controlling part is not lodged in our body merely in the way in which a pilot is lodged in a ship. Since we can explain how and why, on the Embodied Part View, we can claim ourselves to have the properties of our bodies, the Physical Properties Objection fails.

We can next return to the Epistemic Problem, which is held to count against all views which distinguish between a person and a human animal. On such views, Olson writes:

how could you ever know which one you are? You may think you’re the person. But whatever you think, the animal thinks too. So the animal would . . believe that it is a person. . . Yet it is mistaken. If you were the animal and not the person, you’d still think you were the person. So for all you know, you’re the one making the mistake.26

Olson here assumes that pronouns like ‘I’ and ‘you’ are unambiguous, and must always refer to the same thing.

That is not, I believe, true. We use ‘I’ in different senses, or ways. It is often claimed that the word ‘I’ unambiguously refers to the speaker of the sentence in which ‘I’ is used, or to the thinker of an I-involving thought. But this claim itself illustrates the ambiguity. The speaker of any sentence may be a human animal. But when we think I-involving thoughts, we may not be intending to refer to a human animal. We can think of ourself as the direct thinker of these thoughts, whatever this thinker is. This thinker might be, not the animal, but the part of the animal that does the thinking, which I am calling the Lockean person.

If our pronouns are in this way ambiguous, the Epistemic Problem partly disappears. In describing this problem, Olson writes

Suppose you were the animal rather than the person.

But we can’t usefully suppose either that we are the animal, or that we are the person, since we would then be supposing falsely that the words ‘I’ and ‘we’ must always refer to the same thing. Some uses of these words may refer to an animal, and others to a person. The names of nations have a similar ambiguity, since they may refer to a nation-state, as in the claim ‘France declared war’, or to a part of the Earth’s surface, as in the claim ‘France is roughly hexagonal.’ We shouldn’t claim that France must be either a nation-state or a part of the Earth’s surface, though we don’t know which.

It will help to make our pronouns more precise. In our thoughts about ourselves, we can use the phrase ‘Inner-I’ to refer to the Lockean person, and ‘Outer-I’ to refer to the human animal. We can use similar senses of Inner- and Outer- me, you, he, she, we, and us. Suppose next that someone thinks both

(A) Inner-I am the person, the conscious, thinking, controlling part of this animal, which is directly thinking this thought,


(B) Outer-I am the animal that is indirectly thinking this thought, by having a part, Inner-I, that does the thinking.


On the Embodied Person View, the person and the animal both think both these thoughts. And, as thought by either thinker, both thoughts are true.

It might be objected that, when the animal believes that

(A) Inner-I am the person,

the animal would be falsely believing that it is the person, since it would be using the pronoun ‘Inner-I’ and the verb ‘am’ to have a belief about itself. But that is not so. The animal would understand these new, more precise pronouns, by having a part that understands them, and the animal believes (B) as well as (A), by having a part that does the believing. If these uses of the word ‘am’ seem misleading, we could restate these thoughts as

(C) Inner-I is the person that directly thinks these thoughts, and Outer-I is the animal that indirectly thinks them.

We can use ‘is’ rather than ‘am’ when thinking about ourselves, as General De Gaulle did whenever he thought ‘De Gaulle is the saviour of France’. As before, whether (C) is thought directly by the person, or indirectly by the animal, (C) is true.

Return now to Olson’s claim that, if there were two thinkers of all our thoughts, the person and the animal, neither thinker could know which one it was. When Olson presents this objection, he discusses the version of Teletransportation which I called the Branch Line Case.27 Suppose that the new improved replicator scans my brain and body without destroying them, and then makes a Replica of me, in a room that is just like mine. Olson claims that, because I and my Replica would be exactly similar, in exactly similar surroundings, each of us would believe that he was me, and neither could know which of us was right. Similar remarks apply, Olson claims, to the Lockean view which distinguishes between the person and the animal. But this analogy is misleading. When I and my Replica both believe ourselves to be me, and then wonder who is right, these are two different conscious mental processes, or episodes of thinking. No such claim applies to the Lockean view. On this view, just as there is only a single episode of sneezing when the animal sneezes by having a part, its nose, that does the sneezing, there is only a single episode of thinking when the animal thinks by having a part that does the thinking.

Consider next Johnston’s claim that we should take ourselves to be ‘the non-derivative or primary source of the thought in us’. We can distinguish two kinds of derivative thinking. Some of our thinking is derivative in the sense that we are merely thinking again what someone else thought first, and led us to think. Platonists, for example, might derivatively think what Plato thought. The Moon similarly shines at night, in a derivative way, by reflecting light that comes from the Sun, which is the Solar System’s non-derivative or primary source of light. But no such claim applies to the animal and its conscious, thinking part. When a human animal thinks by having a part that thinks, there is nothing that corresponds to the derivative shining of the Moon. There are not two thinkers here, one of whom thinks in a derivative way by thinking again what the other thinks. The animal’s thoughts are derivative in a second, stronger sense. When Inner-I the Lockean person thinks some thought, we can truly say that Outer-I, the animal thereby thinks this thought. But the animal does not itself do any thinking. The animal cannot think in what Olson calls the strict or non-derivative sense.

On the objection that we are discussing, if there was both a person and an animal which think all the same thoughts, neither could know whether it was the animal or the person. This objection can now be answered. When Descartes asked what he could know, despite the arguments for skepticism, he thought:

I think, therefore I am.

Descartes concluded that he could know that he was an immaterial thinking substance. As Lichtenberg objected, Descartes should have thought only:


This is the thinking of a thought, so at least some thinking is going on.

Descartes’ Cogito leaves it open in what sense, or way, any thought must have a thinker. This question we can also leave open here. We can assume that any conscious being that can think about itself, and its identity, is at least a person in the Lockean sense, whatever else this being may be. We are supposing that someone thinks


(C) Inner-I is the person that is directly thinking these thoughts, and Outer-I is the animal that is indirectly thinking them, by having a part that is doing the thinking.


To explain the meaning of the pronoun ‘Inner-I’, we can claim that, when used in some thought, this pronoun refers to the person that is the direct thinker of this very thought. When any direct thinker uses ‘Inner-I’, knowing what this phrase means, this thinker knows that it thereby refers to itself. So in thinking (C), Inner-I the person would know that it is not the animal but the person.

We can next ask what the animal could know. Just as the animal thinks some thought only by having a part, the Lockean person, that does the thinking, the animal can know something only by having a part, the person, that knows this thing. Since the person knows that Outer-I is the animal that indirectly thinks these thoughts, the animal thereby knows, in its derivative way, that Outer-I is this animal. Nor could the animal be mistaken, since the animal cannot make any mistake except by having a thinking part that makes this mistake, and this part, the Lockean person, would not be mistaken.

There is, I conclude, no Epistemic Problem. And if Lockeans appeal to the Embodied Person View, they can answer the other Animalist objections to Lockean views. Since the animal thinks only by having a part that thinks, there are not too many thinkers here. And since the animal is a person only in the derivative sense of having a Lockean person as a part, there are not too many persons here.


The title of this lecture claims that we are not human beings, in the sense that means: human animals. Some of my remarks may seem to have undermined that claim. If our pronouns are ambiguous, as I have suggested, how can I hope to show that we are not human beings, or animals, but are the conscious, thinking, controlling parts of these animals?

I have not undermined that claim. If we resolved the ambiguity of our pronouns, by distinguishing the inner and outer senses, these senses would not have equal status, or equal importance in our conceptual scheme.

Return to my imagined case in which my head and cerebrum would be successfully grafted onto someone else’s brain stem and the rest of that person’s body. My own brain stem would maintain the functioning of the rest of my body, which would remain the body of a living but unconscious human animal.

Most of us would believe that, in this case, it would be I who would later wake up, with my head and the rest of this other person’s body. If we used these more precise pronouns, we would then be believing that it would be Inner-I, the Lockean person, who would wake up, and continue my life with a new body below the neck. Outer-I the human animal would continue to exist in a vegetative state. But that would not affect Inner-Me, the person. And if we imagined ourselves about to undergo this operation, most of us would believe that we would be the person who woke up again, not the animal that lingered on in a vegetative state.

Since the inner senses of pronouns have more importance, we can now express these senses in the old familiar, briefer way. I hereby announce that, from now on, I shall use the word ‘I’ in the different, more precise sense that I have expressed with the phrase ‘Inner-I’. I invite Inner-you, the other Lockean persons in this room, to do the same. We can then truly claim that we are not human beings in the sense that refers to human animals, but are the most important parts of these animals, the parts that do all the things that are most distinctive of these human animals, as conscious, thinking, rational beings.

Olson considers the objection that, since our pronouns are ambiguous, there is no single answer to the question of which entity we are. Though he is an Animalist, Olson then writes:

If the word ‘I’ in my mouth sometimes refers to a thinking thing and sometimes to an unthinking thing. . . [such as my body] then our concern is with the thinking thing. Never mind the referential role of personal pronouns. This is an essay in metaphysics. Our question is about the nature of the beings holding the enquiry. So we can rephrase our question. . . . What sorts of beings think our thoughts. . . ?28


The answer, I have claimed, is: Lockean persons. Olson also writes:

Surely it couldn’t turn out that there is something other than me that thinks my thoughts, whereas I myself think them only in some loose, second-rate sense?29


That is true, since Olson is such a Lockean person, the part of a human animal that thinks Olson’s thoughts in the strict, first-rate sense.

If, as I have argued, we are not the animals that we call human beings, what difference does that make?

The most direct moral implications apply to the first part of every human being’s life, and to the last part of many human beings’ lives. According to the Catechism of the Catholic Church, ‘Human life must be. . . protected absolutely from the moment of conception. From the first moment of his existence, a human being must be recognized as having the rights of a person. . . The first right of the human person is his life’ (1992). Many people make similar claims.

If we, who are Lockean persons, are not human beings, these claims do not apply to us. Nor do such claims apply when our cerebrum has died, so that we have ceased to exist, though our brain stem keeps the human animal alive. As McMahan claims, neither early abortion, nor removing a feeding tube from such a human being, would kill one of us. Though such acts raise moral questions, they do not violate the rights of persons.

If we are not human animals, that in one way makes little theoretical difference. In a book whose final proofs I corrected a few months ago, my first sentence is:

We are the animals that can both understand and respond to reasons.30

It was only while preparing this talk that I came to believe that we aren’t animals, but are the conscious, thinking, controlling parts of these animals. But though my book’s first sentence is, I now believe, misleading, I would not revise that sentence, but would only add a qualifying note. My sentence is close enough to being true. We are each part of a human animal, and we make this animal able, in a derivative way, to understand and respond to reasons. Outer-We are, in that sense, rational animals, because Inner-We are rational persons.

There are some other theoretical implications. For those who believe that we don’t have souls, in the sense of persisting immaterial substances, one of the main recent philosophical disagreements has been between Lockean or psychological theories, and Animalist or biological theories. Animalism, Shoemaker writes, ‘presents a poweful challenge to neo-Lockean views’. The ‘crux of the current debate. . . [is] whether this challenge succeeds.’31 I have tried to show that it does not. Animalists rightly claimed that Lockeans should not ignore the question whether we are animals, and Animalists put forward forceful objections to most Lockean views. But if Lockeans revise their claims, by turning to the Embodied Person View, these objections can, I have claimed, be answered. And this view also avoids some strong objections to Animalism. If Animalists also turned to this view, this disagreement would be resolved, and we together would have made philosophical progress.

I shall end with a more personal remark. In my earlier writings about these questions, my main aim was not to defend a Lockean criterion of personal identity, but to argue that, in our thoughts about our identity, or what is involved in our continuing to exist, most of us have, at some level, various false beliefs. We believe, for example, that if we are about to lose consciousness, it must either be true, or be false, that we shall wake up again. Such beliefs, I argued, are mistaken. Personal identity is not as deep, and simple, as most of us take it to be. Even if we did not know whether we would ever wake up again, we might know the full truth about what was going to happen. Since we have such false beliefs about what is involved in our continued existence, we may misunderstand the rational and moral importance of personal identity. On the true view, I claimed, though we have reasons for special concern about our future, these reasons are not given, as we assume, by the fact that this will be our future. Nor will our death be as significant as most of us believe. In my somewhat misleading slogan, personal identity is not what matters.

In defending these claims, I appealed in part to the imagined case in which two future people would be psychologically continuous with me as I am now, because each person would have one half of my cerebrum. But this is only one example. And I have found it hard to convince some people that, in other cases, personal identity is not what matters. I cannot persuade these people, for example, that if they were about to be destroyed and replicated, it would not matter that their future Replica would not be them, so that they would never wake up again.

If Animalism were true, it would be easier to defend these claims. Suppose again that, because your body below the neck is fatally diseased, as is someone else’s cerebrum, doctors will successfully graft your head and cerebrum onto this other person’s brain stem and headless body. According to Animalists, it would be this other person who would later wake up with your head, being psychologically just like you, and mistakenly believing that he or she was you. If we accepted this Animalist view, it would be easier to see that personal identity is not what matters. It would be clear that, if someone would later wake up with your head, and would be psychologically just like you, it would have no practical or moral importance that this person would not be you. While defending his biological, Animalist view, Olson similarly writes:

In divorcing our identity from psychological continuity, the Biological Approach would entail that these relations of practical concern are even less reliably connected to with numerical identity than Parfit and Shoemaker have argued.32


If, as I have argued, Animalism is not true, I cannot defend my claims about what matters by appealing to this imagined case. That gives me a reason to wish that Animalism were true. But this is not a reason to believe that Animalism is true. So I regret that Animalism – a view that is highly plausible, widely accepted, and was strangely neglected until Snowdon, Olson, and others gave it the prominence that it deserves – seems not to be true.


1 This lecture is taken from material written with the support of a Foundational Research Grant from the Ammonius Foundation, for whose generosity I am very grateful. I hope to publish other parts of this material elsewhere. I have been much helped by comments from Eric Olson, Sydney Shoemaker, and Ingmar Persson.


  1. John Locke, Essay Concerning Human Understanding, Book II, Chapter XXVI, Section 9.
  2. McDowell, John, ‘Reductionism and the First Person’, in Reading Parfit, edited by Dancy, Jonathan (Blackwell, 1997), 237.
  3. ‘Persons, Animals, and Ourselves’, in The Person and the Human Mind, edited by Gill, Christopher (Oxford University Press, 1990), page 90.
  4.  ‘On What We Are’, in The Oxford Handbook of the Self, edited by Gallagher, Shaun (Oxford University Press, 2011). In what follows, I shall not be rejecting Shoemaker’s view, but proposing another, simpler way of thinking about ourselves.
  5. Baker, Lynne Rudder, Persons and Bodies (Cambridge University Press, 2000).
  6. In my ‘Persons, Bodies, and Human Beings’, written around 1992, published in Contemporary Debates in Metaphysics, edited by Hawthorne, John, Zimmerman, Dean, and Sider, Theodore (Blackwell, 2008).
  7. Williams, Bernard, ‘Are Persons Bodies?’ in Problems of the Self, (Cambridge University Press, 1973).
  8. ‘Animalism and the Remnant Person Problem’, henceforth ‘Animalism’, Sections 1 and 2, in Metaphysics of the Self, edited J. Goncalves (Peter Lang, forthcoming).
  9. These Animalists would be adding only that these Lockeans had failed to point out that their criterion also applies to human animals. That would be no objection to this Lockean view.
  10. ‘Animalism’, Abstract.
  11. ‘Animalism’, end of Section 5.
  12. Olson, EricWhat Are We? (Oxford University Press, 2007), 216.
  13. ‘Animalism’, start of Section 7.
  14. Johnston, Mark, ‘Human Beings Revisited: My Body is Not an Animal’, in Oxford Studies in Metaphysics, Volume 3, edited by Zimmerman, Dean (Oxford University Press, 2007), 54–5.
  15. McMahan, Jefferson, Killing at the Margins of Life, (Oxford University Press, 2002), Chapter 1. The Embodied Parts Proposal was independently suggested by Persson, Ingmar, in his ‘Our Identity and the Separability of Persons and Organisms’, Dialogue, 1999. But Persson rejected this view, on what seem to me overly skeptical grounds. See also his ‘Self-Doubt: Why We Are Not Identical To Things of Any Kinds’, Ratio, December 2004, and his magisterial The Retreat of Reason (Oxford University Press, 2005).
  16. Carter, William R. ‘How to Change Your Mind’ Canadian Journal of Philosophy, May 1989, 13.
  17. Ibid., 7 note 9.
  18. Johnston, Mark ‘My Body is Not an Animal’, in Zimmerman, D, ed. Oxford Studies in Metaphysics, Volume 4 (Oxford University Press, 2006) 54. There is much highly relevant material, which I hope to discuss elsewhere, in Johnston’s, Surviving Death (Princeton University Press, 2010).
  19. Though we can also apply the Embodied Part View to such animals as dogs, whose conscious, thinking and controlling part may be at best a borderline case of a Lockean person.
  20. Olson, op.cit. note 13, 216.
  21. ‘Animalism’ Section 5, quoting from Roderick Chisholm Person and Object (Open Court, 1976) 104.
  22. Johnston ‘My Body is Not an Animal’ op.cit. 50.
  23. Johnston ‘My Body is Not an Animal’, op.cit. 48.
  24. What Are We? op.cit. 76
  25. Olson, Eric ‘Thinking Animals and the Reference of ‘I’, Philosophical Topics, Spring, 2002.
  26. Olson, Eric ‘Personal Identity’, Section 6, Stanford Encyclopaedia of Philosophy, 2002.
  27. Olson, op.cit. note 13, 13.
  28. Olson, op.cit. note 13, 79.
  29. On What Matters (Oxford University Press, 2011).
  30. Shoemaker, , ‘Persons, Animals, and Identity’, Synthese 2008, 315.
  31. Olson, Eric, The Human Animal (Oxford University Press, 1997) 71–2.
  • Author

    Derek Antony Parfit FBA was a British philosopher who specialised in personal identity, rationality, and ethics. He is widely considered one of the most important and influential moral philosophers of the late 20th and early 21st centuries.

  • First published

    Philosophy, Volume 87, Issue 1, January 2012, pp. 5 - 28