Parfit and Identity, does it really not matter?

The origins of the imperative, "know thyself", are lost in the sands of time, but the age-old examination of human consciousness continues here.

Moderator: MagsJ

Parfit and Identity, does it really not matter?

Postby Matt » Wed Dec 03, 2003 4:22 pm

This is one of my old uni essays on the subject of personal identity. Generally on that subject I agree with Derek Parfit's work in Reasons and Persons but I'm uncomfortable with his conclusions. Having read this essay again I already disagree with much of it myself, but it's all part of the thought process and if just makes one person think then it was worth posting it and all that, what, what! If you haven't read Parfit some of the terms might be a bit confusing but I will explain them if anyone wants me to, just ask.

The original question set was "What does the claim that identity does not matter mean? Present a critical discussion of this view", but to be honest it turned into more of a criticism of Parfit's interpretation of the term "personal identity". There were the constraints of a 2,000 word limit and that it was a uni essay so I had to demonstrate that I had looked at several angles. Section III could have been left out in my opinion but markers don't really like you saying "I'm not going to discuss this because it's wrong" without actually giving some reasons.

I should note that that course leader of this particular module was Dr. Scott Campbell and some of this was developed in conversations with him. I've included all notes at the end, any number in [] is a note further down the essay, as well as a bibliography for the sake of completeness.


I. Introduction

In Reasons and Persons, Parfit makes the claim that identity is not what matters to us when we talk about our own survival. The possibility that there could be more than one copy of us about in the future means that a unique relation like identity cannot hold if several qualitatively identical copies are created. Worse still, if one of those copies die, they should not consider it as bad as conventional dying, as there will be someone around who possess what really matters to us. This counter-intuitive view is a direct result of Parfit undermining the concept of identity.

It is not sufficient to just critique the idea that identity does not matter, we must also look at the underlying assumption that Parfit attaches to it, what he actually means when he says identity does not matter. Certain assumptions he makes about the scope of identity distort the meaning we might actually want to attach to it. I hope to show his attack on what he calls "identity"is not an attack on identity at all but something much more, an attack on the idea that consciousness is not concerned with it's own existence, just the existence of a consciousness that is connected to it by relation R. Furthermore I hope to show his attack ultimately fails because of the confusion between identity and the existence of discrete streams of consciousness.

II. The claim and its meaning

The concept that Parfit (1984) presents seems simplistic and compelling enough. Identity must be a one-one relation; there is only one thing that something can be identical with and that is itself. Imagine we have someone who is being teletransported[1] to Mars. A computer glitch means that we arrive not only at Mars but on the Moon as well. But by the definition of identity, there can only ever be one of us around for identity to hold.

How can we not have survived? As Parfit (1984) asks, "How can a double success be a failure?" (p.256). Parfit considers three cases: that you don't survive; you survive as either one or the other; or you survive as both. He dismisses, quite rightly, the idea that you survive as one or the other, and then by citing the one-one relation of identity, states that you cannot survive as both.

It is at this point where he does a philosophical sleight of hand, defining P.I. as:

[X] P.I. = R + U. (p.263)

Where U is uniqueness, expressing the identity part of the formula, with R being relation R[2]. As he has shown uniqueness cannot make our double success a failure, the only thing that can be important to us is R. So in attacking the concept of identity that does not matter, he actually means there is nothing more than relation R that matters. As long as you have a future self that bears relation R to you, you should be happy.

This shows us that Parfit moves from the initial claim that identity does not matter to the conclusion that "the best description is that I shall be neither resulting person" (p.279). This is a direct result of his statement that identity is not what matters, and it is this that many of us would feel hard to accept.

This assertion is the meaning behind the concept that identity is not what matters to us. It is more than what we could see as the softer claim that we may not be able to determine exactly one person who we are identical with in the future[3]. It is this conclusion that we shall have to attack in order to show that it is us that survives in duplication; either by saving identity, as we shall attempt to in section III, or by showing that Parfit's meaning cannot be backed up by his arguments, as we shall try to in section IV.

III. Possible defences for identity

There have been several attempts to save identity in a fission case. Parfit himself cites Lewis' attempt to show that in the case of fission we already have two people existing inside us, thus dodging Parfit's problem. We must admit that this creates all sorts of strange scenarios and if we assume the indeterminacy of the future we might have to say there were infinitely many people inside one person just to guard against any case of future fission. It is also unclear what would happen in fusion. So we shall leave Lewis' theory, conceding to Parfit that it is untenable[4].

Baillie (1996) has also argued against Parfit's view. His argument is simple enough, suggesting that his Irreplaceability Thesis[5] undermines the force of Parfit's argument. Using the example of a man losing his wedding ring, he says "as anyone who has been in this situation can attest, the ring is irreplaceable." (p.265). His argument is that we assign importance to a particular token for some objects, rather than just any token, and that this extends to persons. Other examples he gives are the importance attached to original paintings or items of sentimental value, such as a signed book. These objects are irreplaceable by T-duplicates[6]. Thus why should we accept that a T-Duplicate is acceptable in relation to own identity?

At first sight this argument can be seen as somewhat true, but there is an underlying mistake that Baillie has made that will show us why his argument doesn't work. The importance of the particular token in these examples are given to these objects because they were the ones which were actually there for the ceremony, painted by the artist, etc.

The argument doesn't work for persons because in transferring a personality we keep all that matters to us, the memories, the personality, etc. What does not matter to us is that I still have the particular hand that I used to score the winning try in a rugby game 10 years ago. I would not be concerned if an arm that I lost in an accident was replaced with an exact replica grown in a vat; this would not tarnish my fondness of that incident in any way. This argument then can extend to any part of my body and eventually all parts of our body[7].

IV. Some sort of identity does matters[8]

Identity as it is classically defined is indefensible; we saw that in the previous discussions. It just cannot be that some sort of numerical identity is what is most important in our continued survival, I would not consider myself dead in the case of an accidental double teletransport.

But it is still intuitive to me to oppose Parfit's claim that the best description we can give of that case is that I am neither person. Surely I will be both! What is it about Parfit's assumptions that bring this view about? We have already hinted at the problem, that the concept of personal identity is mixed with some sort of uniqueness.

One of the most striking things of all the discussions of personal identity is the rigid adherence that identity must be a one-one relation. It is obvious, of course, that numerical identity must be, but by allowing the very fact that our body could be duplicated, or our consciousness split into two identical consciousnesses[9], we are immediately shown that what we have always talked about, personal identity, cannot be a one-one relation. It is simplistic for Parfit to set up identity and then destroy it as he does, because he has already defined our continuance as something that cannot be a one-one relation.

Could this be arrogance on the part of the philosopher? After all, for the layman it is quite possible to be identical to both. If we accept that teletransportation allows us identity with the destroyed "original", then if two copies are created in such an incidence this identity cannot disappear. It is just that we have wrongly defined personal identity.

Parfit himself looks at a people who reproduce by division, much like amoeba[10]. When these beings initially divide they will essentially be the same, having the same thoughts, memories, etc. but as time goes on they will grow to become different. This is why when we look at the teletransporter case we cannot say they are the same person, because in the future they will not be.

But it is exactly this that shows us that there is something more to his definition of identity that he is letting on. When the alien splits, does the original not survive at all? Or does it continue on in both the "daughter" aliens.

What happens when we remove consciousness from the aliens? The new divided entities are still related to the original somehow, both still connected to it. But as we have taken the R out of [X], and there is no U there should be nothing to bind them together to the original being. But there is still some relation between them, call it direct continuity. However this relation is not shared between the resulting beings.

Thus we could easily change our P.I. formula one to utilise this notion of direct continuity (C):

[Y] P.I. = R + C

Thus the importance of P.I. is maintained without the previous one-one relation of uniqueness. R cannot supply this direct continuity as it is transitive. In the case of our malfunctioning teletransporter, the person on Mars shares relation R with the one on the Moon for some time after the split from the initial person. But they do not share C; they can only share that with the original.

C can be seen as time dependently transitive[11]. That is that it can only ever be transitive in one direction of personal time, it can move from t1 to t2 to t3, but not from t2 to t1 and then back to t2 again. Thus V at t1 can share C with Y at t2, and Y with Z at t3, and thus V shares C with Z, but if we have a fission case where an X is also is also created at t2 who shares C with V too, it does not follow that X and Y can be connected as the transitivity would have to go back and then forwards in time[12].

This saves us from the unwelcome conclusion of Parfit, no longer do we have to say that we survive as neither, we survive as both. Nor would one of the copies who was about to die just after the split be able to be told what matters to him would survive anyway[13], as C would not be present and thus what matters to him would no longer be present.

V. Prudential concern, what does it actually mean?

Prudential concern, an oft used phrase in the literature about personal identity, is a concept which has helped create the confusion that has led Parfit to concluding that no sort of continuity matters. While we talk about whether I should be scared of the future prospect of one copy dying a little while after the teletransporter has done it's job we can comfort ourselves that there will be another copy. That will be of not comfort to the copy that will die however. What sense can we actually make of concern for a future self? Being told that one copy will have a life of pleasure and the other a life of drudgery, are we supposed to be happy?

This is an absurd suggestion, as it will still be you, and only you, who will have to suffer the drudgery. The prospect should fill you with dread and no pleasure. The only sense we can make of talk of personal identity is from present to past, and not from present to future. The question "Will I survive?" can only ever be answered intelligibly at whatever point the future you are looking at, as long as at least one person replies affirmatively to the question "Are you Matt Manser?", I am still surviving.

V. Conclusion

The preceding discussions have been building up to the conclusion that Parfit's claim that "identity does not matter" is actually meaningless. He has actually created a straw doll to impale; identity could never be what matters as soon as we accept any sort of reductionist view with no physical content of what it means to be me. If I can be exactly replicated once, I can be replicated a thousand times. All this means is that what we once looked for to see our continued existence is a term that is defined differently, not identity in a numerical sense, but a weaker form of the concept of identity, direct continuity, C. There is nothing fundamentally shocking in Parfit's conclusion that identity is not what matters because by his own definition it could never be. The problem with his argument comes when he pushes the boundaries of the meaning of identity beyond that which he has actually shown to be true.


[1]Where you step into a machine, it scans your body and creates an exact replica elsewhere, destroying the original in the process. If we accept some sort of psychological criterion for identity, this would be an acceptable way of travelling; you would survive the transport. Parfit uses brain splitting to avoid some objections, but these are irrelevant for the purposes of this essay, thus I shall use the clearer example of two copies.
[2]"R is psychological connectedness and/or psychological continuity, with the right kind of cause." (Parfit (1984), p.262).
[3]More on this use of identity with more than one person later
[4]Fuller discussions of Lewis' theory can be found in Sider (1996) and Belzer (2002). I find these convincing enough, not just for the previously stated reasons, to abandon an attempt to save identity through Lewis' proposal.
[5]p. 266
[6]The term Baillie uses for teletransporter duplicates.
[7]I do concede we would have some attachment to our first body, but not to the same extent that we have attachment to a particular wedding ring, in which the specific token is what is important to us.
[8]This section is inspired by the work of Gendler (2002), though I feel Gendler has wrongly attacked relation R instead of attacking Parfit's concept of identity. There are several points I disagree with in his article and a full essay would be required to work from Gendler's ideas to those presented in this section, so with the constraint of space I shall just present my own.
[9]Apart from the position they occupy in space-time.
[10]Parfit (1984), pp.299-300
[11]It is important to stress the time factor must be from the personal perspective, to avoid any time travelling objections.
[12]Back to t1 to connect to V and then forward to t2 again to connect with Y.
[13]As in the malfunctioning scanner example Parfit (1984) uses on p.200.


Baillie, J. (1996) "Identity, Relation R, and What Matters".
Pacific Philosophical Quarterly, vol. 77, pp.263-267

Belzer, M. (1996) "Notes on Relation R"
Analysis, vol. 56, pp.56-62

Belzer, M. (2002) "Parfit and Lewis on Survival of Fission and What Matters in Survival: Why Parfit won the Debate"
Unpublished, found at

Gendler, T.S. (2002) "Personal Identity and Thought-Experiments" Philosophical Quarterly, vol. 52, pp.34-54

Parfit, D. (1984) Reasons and Persons
(Oxford : Oxford University Press)

Sider, T. (1996) "All the World's a Stage"
Australasian Journal of Philosophy, vol. 74, pp.433-453.
"The irony of the Information Age is that it has given new respectability to uninformed opinion." -John Lawton
User avatar
Posts: 954
Joined: Tue May 21, 2002 2:37 pm
Location: Nottingham, England

Return to Psychology and Mind

Who is online

Users browsing this forum: No registered users