SB logo
en's Homepage

Is Personal Identity A Useful Concept?

In this essay I shall be addressing the question of what defines a particular person, whether we can identify such a trait, and whether to do so would prove to be a useful exercise. There are a variety of different views as to what a person is, which usually follow from someone's opinion on a variety of other metaphysical questions. I hope to show that it is not simply coincidence that views on personal identity follow from other ideas, rather than preceding them, and hope to show that this implies personal identity to be less important as a characteristic.

I shall suggest that the whole concept of personal identity is a bit of a red herring, and that it is too confused to be useful. I shall do this two ways: firstly by asserting that there isn't such a thing as personal identity, that we never are the same person from one moment to the next; and secondly that even if you are unwilling to accept this, any weakened form of personal identity that may survive will fail to be of any use in answering any question - specifically I hope to show that we are able to answer any question that affects how we think or act - practical, useful questions - without a concept of personal identity.

As seems customary in the debate about personal identity, this essay will involve the discussion of a number of thought-experiments (or 'intuition pumps' as Dennett likes to call them), which are discussed in the literature in an attempt to bring out intuitions about the field, and by discovering what these intuitions are formulating a conception of what personal identity consists in. I shall argue that the complex nature of responses to these experiments and the irresolvable nature of the conflicts amongst intuitions about cases indicate that no concrete conception of personal identity is possible, and also (the lesser claim) that without a concrete conception we will never find personal identity useful in answering questions - especially since any practical question can be answered without.

Why Do We Care?

What does it matter what makes up a person? Why should we care what personal identity consists in? I would suggest that answers to these two questions will shed a little light on the answer to the question of what personal identity is - after all, we could always invent a concept, call it arbitrarily 'Personal Identity', and thus be able to say exactly what it is, but we do not want to do this, since we are trying to understand what our current conception of a person is.

I would like to offer three possible reasons for wanting to know:


What I would call 'Special Relationships'.

For Punishment. To know whom society should punish for a crime that has been committed.

We might want to know what personal identity consists in to be able to distinguish events that will allow us to continue to exist from those which result in our existence ending - personal death. For example, I might want to know whether I can survive a major operation (and there is likely to be particular debate in the case of brain-surgery), whether I could survive an accident, or even if I might be able to survive death.

In a way, interest in whether special relationships still hold is a question resulting from that about survival. What I mean by 'special relationships' are those people about whom you seem to have a particular interest because of who they are - so you might want to know who are your parents, or who is your son, or something more specific, such as who your GP is (since he's the one that knows about your health and medical history, especially if we doubt that doctors are able to note down everything they discover about you in medical notes). These are people whose loss we believe would be an additional tragedy beyond simply someone dying, because you care particularly about their survival, and hold a particular bond with them.

Finally, as Locke believed1, punishment, and ethics in general, seems to me raise very important questions involving personal identity, and I would argue more important that the two above when looked at from an altruistic rather than selfish perspective. The reason that personal identity seem to be important in ethics is that it doesn't seem fair to punish someone for a crime that someone else committed (presuming, of course, we do not deny that punishment is ever considered to be fair). Many argue that 'justice' demands that a crime should be punished - but they do not simply mean that when a crime occurs someone should be punished (as if the overall karma of the human race would be re-balanced by inconveniencing some of its members, to make up for its members inconveniencing the world), but that a crime requires the punishment of the person that committed the crime. So to decide who should be punished for a crime, it seems to be the case that we need to have some idea of what it is for someone to be the person that committed the crime at an earlier time than when we are able to capture them.

If any of these questions are serious ones - that is if we actually need to know about survival and what it is to be the same person as before - then it seems likely that we will need a concept of personal identity. My suggestion is that the questions are deceptively similar to ones that we really do care about, but that as they are formulated they simply prey upon laziness in our consideration of what we're really asking about, since the questions we are really interested in are slightly different.

The Possible Answers

Before I discuss the tests - the thought-experiments I shall use to investigate the question of what personal identity consists in - I should briefly introduce the possible answers to the question:

The Simple View

The simple view holds that what a person consists of in an indivisible thing, unrelated to their body and its functions - though frequently connected in some contingent way to a body - such as a soul, mind or ego.

The Biological View

The biological view holds that for someone to be the same person they must have the same body, or at least some form of bodily 'continuity'. Some philosophers like to say that this view holds that being the same person is simply being the same animal, yet this seems a bad move to me, since clearly some cultures hold animals to have just the same dualistic nature - having minds or souls - as humans, such as those of the Native Americans.

The Brain-based View

The Brain-based view holds that rather than identity of the whole body, only the same brain is necessary for personal identity, or at least some form of brain 'continuity'. This is usually favoured because it allows someone to lose a leg, for example, without ceasing to be the same person, as the biological view holds, whereas it seems to be the case that the brain remains almost constant (and there is far more debate about someone being the same person with a damaged brain than with a damaged leg).

The Psychological View

The psychological view holds that personal identity consists in having the same psychology or personality or at least some form of psychological 'continuity'. Perhaps the most popular view currently held, this view suggests that it doesn't matter what vessel you are held in, so long as you act and react in similar ways.

I should explain now the view I hold is, since I shall be looking at it in just the same places where I shall be considering the views above.

The Denial View

Whilst this shouldn't really be classed as a view about what personal identity consists in, I would simply deny that there is such a thing. Whilst we may loosely use the term 'person' as a simple way of talking, I cannot see any consistent way of formulating personal identity as a real concept, and would not consider which 'person' something is to be an actual property of any thing.

Strict Identity

I should make it clear straight away that I feel that the suggestion that personal identity is a matter of strict identity seems clearly incorrect to me, as Reid2 emphasises. By Leibniz's Law:

(ForAll F)((Fx & x = y) -> Fy)

That is if x is strictly identical to y then any property x has, y has. But if we are talking about people being identical to another, then clearly we have problems if we are using a notion of biological or psychological identity. I cannot be biologically identical with the person I was ten years ago, because the property that I have now of being 5'7" I did not possess back then, being a lot shorter, and likewise I have (hopefully) developed enormously psychologically, and thus could not be psychologically identical. Only if we follow the simple view do we have any chance of maintaining a strict-identity relation, since it could be argued that my essence - my soul or mind or ego - is exactly as it always was, and perhaps always will be. Any of the other theories as to what my person consists in will require a retreat to some other sort of identity.

This may not, however, be a problem, since, it is argued, we do have other notions of identity other than the strict one. As Olson points out, 'Clinton is now the same elected official as Reagan was [once]' - there are certainly relations that need not be as demanding as that of strict identity. I am not convinced that this is an adequate response. The problem is that the very point of strict identity is the maxim that any property that something holds is held by anything that is strictly identical to it, and if we move to some other form of identity it seems unlikely that we can discuss any property other than that which makes them the same. Consider the case of 'same elected official' - if we wanted to know who was president of the USA surely we would ask who was president, and not who held the same position as Reagan once held? The reason we would do this is that any similarity we are concerned about can only be a result of their shared position, since, if we deny strict identity (as obviously we must) they are obviously quite independent people. Whilst we can know if each was president that each was commander-in-chief of the US army - and presuming we are aware of the premise that all American presidents before at least the year 2000 AD were white and male, we can know that they are each white males - but these are facts we can know only from them holding a particular office, not from any relation of 'sameness' between them. More importantly, any facts that we can know about them due to them holding the same office (if we equate, as might be controversial, the sameness-relation and the property which they each hold) can only be about the aspect of them that holds that office. Namely whilst we can know that if the president is the commander-in-chief of the US army and Reagan was president and Clinton was president they were both commander-in-chief of the US army, if the presidency makes no requirements about hair colour, then if Reagan was president and Clinton was president and Reagan has brown hair we still know nothing about Clinton's hair colour. So:

(ForAll F)(ForAll x)(ForAll y)(((Ax -> Fx) & (a @ b)) -> (Fa & Fb)) where 'x @ y' is the property of x being similar to y due to A-sameness.

But surely this is not what we're looking for with personal identity. We don't just want to know that if a is the same person as b in respect of both being Ben that they both have Ben-ness. Surely we also want to know that they both have the same parents, or are both called 'Ben', or both went to Fleet Primary School, or any number of other properties that things with strict identity share, but that things with another sort of same-ness need not. Likewise, taking one view as an example, we don't want to just be able to say that if 'a is the same person as b' in respect of both having the same body that they have the same colour hair, or skin colour, or scar - we want to be able to say other things about them, using the same-ness of body just as an indication of all these other things hold, as it would do if same-ness of body indicated strict identity.

Perhaps there is a relation that would work for this purpose - or perhaps we could artificially create one simply by artificially conjoining 'Same-ness of body means same-ness of person' and 'Same-ness of person means same-ness of parents'. Regardless, however, of whether we could formulate such a relation, I still don't believe it would follow our confused intuitions about what constitutes personal identity, as shown by the problem of increase.

The Problem Of Increase

If I own a house, and someone comes during the night, and noticing a loose brick steals it, would it still be the same house? If we are talking about strict identity, then presumably not, since the later house would not have the property of having that particular brick, which the earlier house had. This is made even clearer if the brick that was stolen was the only one made of gold, since obviously then the later house would not have a golden brick, whereas the earlier one did. Arguably then this house wouldn't even be the same house - people might cease to come from miles around to see the 'House With The Golden Brick' - but I'd suggest there is always going to be some reason that each brick is intrinsic to the house, and there will always be someone who considers the brick to have been a tangible property of the house. Presumably, however, the answer to the question is actually 'yes', it is the same house, and we wouldn't consider it to be a new house if some repair work was done and the loose brick replaced. So presumably we do not consider 'Same-ness of house' to be a relation of strict identity. Suppose, however, the house continued to degrade, and we continued to replace bricks as they became loose, until we have in fact replaced every brick the house was originally made up of - is this still the same house? Instinctively I want to answer 'yes', that if it is the same house from one brick-change to the next; you can never make it a different house just by changing single bricks - after all, what's to say that the brick I changed when I first bought the house wasn't the final remaining original brick itself?

Consider, however, another possibility. Rather than changing the house brick-by-brick, I knock the house down and build another just the same in its place. Is this the same house? Well surely not, I would argue, in exactly the same way as rebuilding bridges bombed during the war which look the same does not mean you have brought back the bridge that was destroyed. It seems to be the case that gradual, slow change can preserve same-ness of house, whereas a complete destruction and recreation will not.

So perhaps we should view personal identity as being preserved when the change is slow and gradual, since we are unable to demand that no change can occur at all (unless we take the simple view, but I shall come back to this in the conclusion), as bodies will change, personalities will develop, and even some brain-cells will die. As plausible as this sounds, I don't think it actually does tie in with our conceptions of what constitutes personal identity. The problem is that we don't seem to view an accident which removes an arm to change the person, nor we consider amnesia to (and even if we were to, we would end up with some pretty strange conclusions given that people can recover completely from amnesia, and we would have to say the person had ceased to exist and later reappeared), and we don't view a non-fatal bullet through the brain to necessarily change the person - nor even, as we will discover in the tests below, the removal of an entire hemisphere of the brain - so even a fast change can preserve personal identity, and it appears to be the case that whatever criterion for personal identity we take, the definition of what constitutes 'continuity' will be very strange. It was Hume who suggested that, since personal identity appears to depend on the speed of change and (he believed) nothing can depend on time in this way, personal identity couldn't be a real property3.

The Tests

I hope that I have already shown that a discussion of personal identity is better abandoned than pursued, since it appears to be an entirely incoherent science, but if this is not yet convincing, I shall outline a number of experiments that seem to show just how contradictory a claim to be able to formulate a real criterion for personal identity would be. Many more thought-experiments are used about personal identity than I could possibly hope to discuss here, so I have taken three sets of tests which I hope are representative and can serve to produce enough contrasting intuitions at least to show that there is a great deal of debate to be had, and hopefully enough to render a coherent formulation of personal identity impossible.

Brain Transplant / Hemispherectomy / Split-brains

Usually argued as a series of thought-experiments, the brain transplant / hemispherectomy / split-brain tests involve three different brain-surgery operations. The first operation, which has in fact been performed, is a hemispherectomy - where someone has an entire hemisphere of the brain removed, due to a tumour or irreparable physical damage (for example by a bullet), or simply by a mad scientist wanting to answer a philosophy question. It turns out that a very large amount of brain mass can be removed without killing the whole brain, so long, of course, as you avoid the major arteries and don't thus starve the remainder of the brain of oxygen. In fact, people are able not just to survive with a single hemisphere of the brain, but to thrive, able to learn the skills usually associated with one hemisphere when only the other one remains, though perhaps less efficiently, with the personality intact. This is claimed to show that the psychological view is correct, since we claim that so long as the personality is intact, so is the person.

The second operation is the brain transplant, which, whilst not possible with today's level of medical knowledge, is believed by many biologists and doctors to be possible in theory, and some believe it is just a matter of time before one is performed. If, however, it were possible to transfer a brain into a different body, it would have very interesting consequences for personal identity - where would the person be, with the body or the brain? If we hold either the brain-based or psychological views, we must say the person has moved to a new body, otherwise, according to the biological view, they are a vegetable, and presumable a depleted person (in the new body) has been reactivated - I suggest that the former view is much more in tune with our intuitions than the latter.

If this were possible, however, then (the argument continues) we could imagine another experiment (presumably in this case performed by our hypothetical mad scientist) where a healthy hemisphere is extracted and implanted into another body, thus producing two bodies both exhibiting the same personality. You could even extract both hemispheres and implant them each into new bodies.

It is argued that this produces a paradox, since we now appear to have two people produced for a single one - yet to deny that the person has survived in two bodies would require us to deny one of the earlier two claims, since it would show that you cannot survive both a hemispherectomy and a brain-transplant. I'm not convinced by this conclusion, since it seems to me that if this really were possible then all it would show is that a brain is in fact two people (usually) inextricably connected, and that during this strange operation, like Siamese twins the two people are separated. An even more plausible response would simply be to deny that such an operation would work, since in a hemispherectomy the brain-stem is preserved, and it could be there that the person is 'located'. Overall, therefore, I would suggest the split-brain test comes out in favour of the psychological view.

The Teletransporter

This test involves the question of whether it is wise to use a 'Star Trek' style teleportation device, which we shall call the 'teletransporter'. Originally invented purely to avoid filming complicated and costly sequences with landing-pods leaving the ship and landing on a planet below, the devices in 'Star Trek' would examine and read a body standing on a pad into a computer, storing every detail of the object in a databank, and then destroy the object on the pad, reassembling it on the planet below, and reconstructing the object using atoms local to the planet. The question is whether this is actually a method of transportation or simply a way to die. If we hold the psychological view then the person survives such a transfer; the biological or brain-based view probably suggest that they die.

What if the person were tricked, however, and whilst an identical-looking body was reconstructed on the planet, they did not destroy the original body, and put that to hard labour until the person returns from the planet, at which point that body is destroyed. The experience would be exactly the same for the person going to the planet, but if we look at it from the point of view of the original body a very different picture is painted - it would seem to them that they walked into a room, nothing happened (presumably they would assume that the device was broken) and that then for no reason they were forced into punishment. In this case it would seem to me that the original body would be considered to be the same person - after all, if the device really were broken, they would surely be given that title. Either, therefore, this seems a case equally in favour of the brain-based or biological views, or we should be beginning to see cracks forming in the whole idea.

Swapped Babies

A much simpler case that I would like to offer, and one that can (and has) happened, is that of babies swapped at birth. If it were discovered, several years later that such a switch had occurred, would the parents demand that the babies be switched back? Namely, how important would the pure biology be, in contrast to the personalities of the children they had respectively been calling their own. This in fact happened quite recently in a case in Italy, and despite the fact that the children had grown past two years old, the parents decided to swap them back. They ignored the relative personalities of the children, deciding that their's was the one biologically similar to themselves - ignoring the fact that in all likelihood whichever child they had taken might have grown up very similarly personality-wise.

If this doesn't seem enough like a question about personal identity, or you are on the nature side of the nature / nurture debate, consider instead the same thing happening but this time with identical twins - if they were physically similar (yet, clearly, different entities) and the mother took one and went to live in London, and the father took the other and went to live in New York, if someone kidnapped one baby from the hospital, and swapped it in the airport waiting lounge for the other, which he returned to the hospital, would there be any sense in claiming that the father had taken the wrong 'person' with him? I would suggest, again, there would be. Even if the Prectoria had grown up to have exactly the same personality as baby Helen would have had she been brought up by her father instead of her mother, our instinct is still that Prectoria is a different person from the person Helen would have been. This can only suggest that in this case we believe the body or the brain to be more important that the personality - surely a case in favour of the brain-based or biological views.


It seems to me that these tests show us that the whole concept of a criterion for personal identity is incoherent, and given more space I could give many more examples each of which would show clearly that we ought to take a particular view of personal identity. Unfortunately each of these examples could come out in favour of a different view. The only coherent view seems to be that held by the dualist - the simple view - which holds that there is an essence of the person, separate from the body or psychology. It is probably clear that I have not said much against the simple view, and that is, quite simply, because I hold it to be the only coherent and consistent view of personal identity. This does not mean, however, that I should go against my opinion that personal identity is not a useful concept, because as coherent as the simple view might be, I do not consider it to be useful at all. For any theory to be useful, it must have a practical effect (people tend not to defend the sort of metaphysics that deals in question which make no empirical difference by saying that really are useful, but by saying we shouldn't only be concerned about usefulness) - and the simple view fails in that respect. If there are souls / minds / egos, then for them to be truly independent of body, brain and personality they would have to be undetectable (since otherwise they would, presumably, affect these things in their host) so even if the simple view is correct, it will never enable us to answer the question of who a person is.

So to return to my initial ideas about what it would mean for personal identity to be useful - what questions would need to be answerable - I feel that none of the current conceptions of personal identity answer them as well as the denial view. Taking them in turn, rather than being unable to answer the question of what constitutes survival, the denial view holds that nothing does, that no 'person' survives from one moment to the next, but that they are constantly dying and being reborn. This is something like Parfit's view that personal identity is not what matters in survival, since he believe that what we talk about when we discuss survival is the continued existence of aspects of what we call a 'person', but not an actual thing surviving - and of course, it is these aspects that I would agree we use to answer any questions someone's personal identity might answer. If we are asked, for example, who will feel the pain if we swap two people's brains, and then torture one, we deny that there are persons who will experience it, but argue that the thing with psychological continuity will experience the pain.

This likewise answers our second question, since the useful answer need not be in terms of personal identity. We can tell someone who will do their washing, without telling them which person their maid will be tomorrow, and we can tell a parent who will act in the way their baby does, without claiming that that is the same person as their baby was before.

As to punishment, we have to deny an abstract idea of justice, and consider things simply in terms of the ethical system we choose. If our ethical system demands we punish people to reform them, clearly we should punish those with psychological continuity with criminals, if we are doing it as a deterrent, most likely people will connect a body visually with a crime, and it seems unlikely that people will use the logic necessary to link up the body which has had a brain transplanted into it with the crime, and if we are doing it for revenge, presumably we choose whoever the victims associate with the crime, since clearly that's the only person they will feel better for punishing, regardless of whether there is any truth is the claim that they are the person who committed the crime!

1c.f. John Locke 'Essay Concerning Human Understanding' Chapter 27, 18-20.

2c.f. Thomas Reid 'Essay on the Intellectual Powers of Man' Essay 3 'Of Memory' Chapters 4&6.

3c.f. David Hume 'Treatise of Human Nature' Book 1 Part IV Sections 2&6.