91003

What matters in survival?

Derek Parfit argues in his essay “Personal Identity” from Reasons and Persons that when

I say I want to survive, i.e. I care about my personal identity persisting through time, I really

mean that I care about the existence of someone in the future who will be directly

psychologically connected with me. (The difference will become clear later.) Any inclination I

have to believe that I actually care about the persistence of my personal identity can be explained

by the fact that in normal cases, persistence of personal identity goes hand in hand with

psychological connectedness. In this paper, I outline Parfit’s argument for what matters in

survival, then connect this issue to the problem of whether it is better endure pain or to die.

Introducing Parfit’s claim about what matters in survival

It is necessary to start by explaining the definition of identity that will henceforth be

assumed. Parfit gives this set of criteria that define personal identity:

The Psychological Criterion: (1) There is psychological continuity if and only if there are

overlapping chains of strong connectedness. [Psychological connectedness is “the

holding of particular direct psychological connections” such as memory and intention

(562).] X today is one and the same person as Y at some past time if and only if (2) X is

psychologically continuous with Y, (3) this continuity has the right kind of cause, and (4)

there does not exist a different person who is also psychologically continuous with Y. (5)

Personal identity over time just consists in the holding of facts like (2) to (4) (563).

For convenience, let us denote conditions (2) and (3) as relation R, as Parfit does. That is to say,

person X is R-related to person Y if and only if X is psychologically continuous with Y for the

right reason. It is worth pointing out here that person X may be R-related to multiple people. Let

us denote condition (4) as the uniqueness condition, U; that is to say, person X is the same
person as (i.e. has identity with) person Y if and only if X and Y are R-related and the

uniqueness condition holds, or as Parfit puts it, personal identity = R + U (567). Let us define

survival of person A as the existence of a person B at a later time if the personal identity of A is

the same as that of person B.

Suppose now that we have a magic people copier: if you step inside, the machine will

scan your body for its chemical composition and physical structure, destroy your body, then

create an arbitrary number (determined beforehand) of identical living and thinking people. (By

“identical” here, I mean that on the whole, each of these people initially looks, thinks, and

behaves exactly like you did before the destruction of your body.) This is an example of what

Parfit calls fission or division. In the case that multiple people are produced (label them A, B,

…), Parfit would say that it makes no sense to say that you are the same person as any of the

copies; although A and B are psychologically continuous with you, the relation between you and

each of them fails the uniqueness condition.1 In other words, you do not survive fission.

Parfit uses the example of fission to claim that what matters in survival is not personal

identity; instead, it is the relation R (567). What does it mean for something to matter in

survival? To illustrate what this means, suppose I lose my wallet one day (a cheap wallet, easily

replaceable). Of course I care about losing my wallet; however, the reason that I care about it is

that my credit card, drivers license, etc. are in the wallet. That is to say, what matters when I lose

(or have) my wallet is the stuff inside the wallet; I don’t care about losing my wallet just by

virtue of it being my wallet. Similarly, what matters in survival is whatever I actually care about

when I survive under normal circumstances (i.e. when there is exactly one person R-related to

1
Actually, Parfit argues this solely based on the transitive property of identity, which says that if
A = B and A = C, then necessarily B = C (566). However, the transitive property is present in his
theory of identity by the uniqueness condition, which is why I explain his conclusion using the
uniqueness condition.
me). Understanding the phrase this way, Parfit’s claim that relation R is what matters may be

expressed as follows:

(T) When I care about whether I survive (i.e. whether I will be the same person as some

person at a later time), I fundamentally care about whether there will be at least one

person in the future who is R-related to me.

According to (T), in the case of my fission into A and B, I do not (or at least, rationally should

not) care that I don’t survive the fission, because there exists at least one person (both A and B)

who are R-related to me, which is what I actually care about and fundamentally want.

Parfit’s argument for what matters in survival

I have reconstructed Parfit’s argument for (T) as follows, using a scenario in which person A

divides into person B and person C, and person D becomes person E just by the passage of time

(just a case of ordinary survival, where D = E) (566-567).

(i) Survival (i.e. persistence of personal identity into the future) has value. That is, if we

were presented with two options, our survival and our non-survival, we would

preferentially choose the former.

(ii) Personal identity is simply the R relation with the uniqueness condition: PI = R + U.

(iii) If we pretended person C didn’t exist, then we would say that A survives as B.

(iv) This relation between A and B, ignoring C, is the same as the relation between D and E.

(v) By acknowledging the existence of C, the only thing that changes about the relation

between A and B is that it doesn’t satisfy the uniqueness condition.

(vi) So the relation between A and B is the R relation. (From (ii), (iv), (v))
(vii) From the point of view of A, although the loss of uniqueness entails lack of survival, it is

not as bad as death. (As Parfit says: “Double survival is not the same as ordinary survival.

But this does not make it death” (566).)

(viii) So the value of uniqueness is negligible compared to the value of R. (From (vi), (vii))

(ix) So the R relation contributes most to the intrinsic value of PI. (From (viii), (ii))

(x) So the R relation is what fundamentally matters in personal identity. (From ix)

Is uniqueness really negligible?

Parfit’s point that the uniqueness condition makes little or no difference to R is somewhat

undefended. Consequently, one way to attack his argument is to argue that U makes a big

difference sometimes. Is it possible to have a case in which I want to be psychologically

connected (R-related) with exactly one person?

Suppose my magical people copier is faulty. Suppose today I want to make N copies of

myself. The copier destroys the original body (an intentional effect), then creates N copies, each

of which are living, thinking beings who think they are me and have my psychology; i.e. each of

these copies is R-related to me. However, due to a glitch, N – 1 of the copies will feel

excruciating, intolerable, incurable pain, and the remaining copy will feel fine. Obviously this is

not the same as death, but might we say this is almost as bad or possibly even worse than death?

Let us consider the related case of permanent illness. Suppose that, after my genes are

sequenced, the doctors tell me that I have a neural disorder that will cause me excruciating,

intolerable pain for the rest of my life. Suppose my two options are (1) to continue living and

endure the pain, or (2) death by a suitable means. (I will not attempt to argue for either option

here.) If you think (2) is the better option, then you may be inclined to believe that it is probably

better for me if the extra N – 1 copies not to have existed at all, i.e. you want N = 1, presumably
so that the people R-related to me will not experience as much pain. That is to say, the person

that steps into the copier wants the R-relation to be unique to one other person; U adds a lot of

value to R because it makes the difference between no pain and lots of pain. If you believe

option (2) is better, then you don't care just about R when we say we care about identity in

survival. On the other hand, if you think that any amount of pain is preferable to death, then

option (1) is the preferable option, and you don’t mind the creation of an arbitrarily large number

of copies of me experiencing pain. The bottom line is that whether you accept Parfit’s argument

should inform your prioritization of options (1) and (2) and vice versa.

Works Cited

Parfit, Derek. “Personal Identity.” The Norton Introduction to Philosophy. Ed. Gideon Rosen,

Alex Byrne, Joshua Cohen, and Seana Shiffrin. New York: Norton, 2015. 558-569. Print.

I pledge my honor that this paper represents my own work in accordance with University

regulations.

91003