Reasons for Persons

Suppose you are replicated on Mars, and the copy of you on Earth is killed ten minutes later. Most people feel like there is some definite answer to whether the Martian is they or someone else. Not an answer got from merely defining ‘me’ to exclude alien clones or not, but some real me-ness which persists or doesn’t, even if they don’t know which. In Reasons and Persons, Derek Parfit argues that there is no such thing. Personal identity consists of physical facts such as how well I remember being a ten year old and how much my personality is similar to that girl’s. There is nothing more to say about whether we are the same person than things like this, plus pragmatic definitional judgements, such as that a label should only apply to one person at a given time. He claims that such continuity of memories and other psychological features is what matters to us, so as long as that continuity exists it shouldn’t matter whether we decide to call someone ‘me’ or ‘my clone’.

I agree with him for the most part. But he is claiming that most people are very wrong about something they are very familiar with. So the big question must be why everyone is so wrong, and why they feel so sure of it. I have had many a discussion where my conversational partner insists that if they were frozen and revived, or a perfect replica were made of them, or whatever, it would not be them. 

To be clear, what exactly is this fallacious notion of personal identity that people have?

  • – each human has one and only one, which lasts with them their entire life
  • – If you cease to have it you are dead, because you are it
  • – it doesn’t wax or wane, it can only be present or absent.
  • – it is undetectable (except arguably from the inside)
  • – two people can’t have the same one, even if they both split from the same previous person somehow.
  • – They are unique even if they have the same characteristics – if I were you and you were me, our identities would be the other way around from how they are, and that would be different from the present situation.

So basically, they are like unique labels for each human which label all parts of that human and distinguish it from all other humans. Except they are not labels, they are really there, characterising each creature as a particular person.

I suspect then the use of such a notion is a basic part of conducting social relationships. Suppose you want to have nuanced relationships, with things like reciprocation and threats and loyalty, with a large number of other monkeys. Then you should be interested in things like which monkey today is the one who remembers that you helped them yesterday, or which is the one who you have previously observed get angry easily.

This seems pretty obvious, but that’s because you are so well programmed to do it.There are actually a lot of more obvious surface characteristics you could pay attention to when categorising monkeys for the purpose of guessing how they will behave: where they are, whether they are smiling, eating, asleep. But these are pretty useless next to apparently insignificant details such as that they have large eyes and a hairier than average nose, which are important because they are signs of psychological continuity. So you have to learn to categorize monkeys, unlike other things, by tiny clues to some hidden continuity inside them. There is no need for us to think of ourselves as tracking anything complicated, like a complex arrangement of consistent behaviours that are useful to us, so we just think of what we care about in others as an invisible thing which is throughout a single person at all times and never in any other people.

The clues might differ over time. The clues that told you which monkey was Bruce ten years ago might be quite different from the ones that tell you that now. Yet you will do best to steadfastly believe in a continuing Bruceness inside all those creatures. Which is because even if he changes from an idealistic young monkey to a cynical old monkey, he still remembers that he is your friend, and all the nuances of your relationship, which is what you want keep track of. So you think of his identity as stretching through an entire life, and of not getting stronger or weaker according to his physical details.

One very simple heuristic for keeping track of these invisible things is that there is only ever one instantiation of each identity at a given time. If the monkey in the tree is Mavis, then the monkey on the ground isn’t. Even if they are identical twins, and you can’t tell them apart at all, the one you are friends with will behave differently to you than the one whose nuts you stole, so you’d better be sure to conceptualise them as different monkeys, even if they seem physically identical.

Parfit argues that what really matters – even if we don’t appreciate it because we are wrong about personal identity – is something like psychological or physical continuity. He favours psychological if I recall. However if the main point of this deeply held belief in personal identity is to keep track of relationships and behavioural patterns, that suggests that what really matters to us in that vicinity is more limited than psychological continuity. A lot of psychological continuity is irrelevant for tracking relationships. For instance if you change your tastes in food, or have a terrible memory for places, or change over many years from being reserved to being outgoing, people will not feel that you are losing who you are. However if you change your loyalties, or become unable to recognise your friends, or have fast unpredictable shifts in your behaviour I think people will.

Which is not to say I think you should care about these kinds of continuity when you decide whether an imperfect upload would still be you. I’m just hypothesising that these are the things that will make people feel like ‘what matters’ in personal identity has been maintained, should they stop thinking what matters is invisible temporal string. Of course what you should call yourself, for the purpose of caring disproportionately about it and protecting its life is a matter of choice, and I’m not sure any of these criteria is the best basis for it. Maybe you should just identify with everyone and avoid dying until the human race ends.

15 responses to “Reasons for Persons

  1. > So the big question must be why everyone is so wrong

    If they are wrong, it’s because the simplified version that we all use as a set of rules is (a) almost always right in the evolutionary environment we evolved in (b) a very QUICK heuristic, (c) a very calorically economical heuristic (because processing = calories).

    It’s like asking why our intuitions about physics are closer to Newtonian than relativistic. Even putting aside the evolutionary argument, Occam’s razor says that until we get more complicated data, we SHOULD go with Newtonianism or single-individual-indentityism.

  2. Consider some sequence prediction algorithm (perhaps Solomonoff induction approximation) that takes your past sense inputs, prior to the scanning and transmission, as data and predicts the next item in the sequence. What does the algorithm predict next, inputs consistent with Mars or with Earth?

    It seems that many simple prediction algorithms will not confidently predict one or another, an uncertainty that maps to human uncertainty about which copy will be the “real you.” The sequence prediction algorithm might generate equivalents to our third-person physics and neuroscience in the course of doing prediction, but that won’t change its basic structure: it will still predict a “next experience.”

    If we can get phenomena that resemble confusion over personal identity (and other psychophysical laws/dualism) in some of the very simplest models of intelligence that we have, why go in search of a convoluted evolutionary psychology argument?

    • Because we are not talking about indexical uncertainty here, but about ‘identity’. If the Earth version were destroyed, there is no indexical uncertainty, but the problem of ‘identity’ still persists.

  3. I suppose that heuristically, we have a very, very robust cluster of “me” in thingspace. So we reason as if the distinction between self and non-self were something absolute, rather than part of an empirical explanation of our sequence of experiences. In the case of being frozen, a phenomenon such as continuity in time gets firmly entrenched as a property of “me”. In the case of the Martian copy, it is continuity in space, which apparently is a little less robust because people seem to find it more plausible that the copy is they, but still we do not question the more basic absolute distinction between self and non-self, instead preferring to profess uncertainty as to where the self is.

  4. Parfit is simply defending, and reasoning on the basis of, one of the scientifically motivated delusions of this epoch.

    There are many points of tension between the view of reality coming from natural science and how reality actually appears to us. One of these points of tension is the relationship between physical atomism and the existence of persons / a self / consciousness. Perhaps a majority of human beings are still substance dualists, believing in a soul-substance other than the matter-substance of their bodies, while the materialist minority mostly believe that “I am my body” or “I am my brain”.

    The even smaller minority who try to have a reasoned, systematic opinion about how the existence of a Turing-equivalent neuro-pudding in a dark bony cavity for several decades could possibly be *the same thing* as the existence of a person who lived out twenty thousand days of life outside that bony cavity … generally appeal to some notion of “computation”. The sensory, emotional, and intellectual experiences whose union in the flow of consciousness make up the whole of a person’s knowledge of reality, are supposed to be the same thing as: lots of ions shuttling back and forth across cell membranes, for several decades, somewhere inside that skull-pudding.

    That this posited identity, between the whole of the life you have known, and the back-and-forth motion of ions in a very small space, has a very tight grip on the imagination of anyone raised to believe in science, should not blind you to the extraordinary intrinsic dissimilarity between the two things that are said to be the same thing. In fact, the dissimilarity is such that, as things stand, this “identity” must actually be treated as a new form of dualism. There are the experiences that we have and know about; there are the ions we believe to be moving around in our skulls; and neurocomputational materialists believe that one accompanies the other in a systematic fashion. But it’s only the very sharp materialists, like David Chalmers, who can bring themselves to notice the dissimilarity, and will actually call themselves dualists – property dualists. On the topic of consciousness, most contemporary “materialists” consider themselves monists while entertaining a belief system which is de facto a property dualism.

    One alternative to espousing neo-dualism is to simply deny the existence of anything which fails to match the scientific picture of matter and the brain. (I will pass over various intermediate stratagems and coping strategies, in which people deny that neo-dualism is a dualism or otherwise hide this from themselves.) The most notorious example of this is found in the debate over qualia. But denial of a robustly and objectively existing person that persists over time is just another example of this.

    My considered opinion is that this particular denial is just another bizarre mistake characteristic of our time, that will eventually be thrown in the dustbin of history by a nonatomistic conception of quantum ontology, applied to living matter. That is, we are going to find out that there is a sharply bounded, intrinsically unified physical entity in the brain (I usually suggest that it will consist of entangled electrons in special functional organelles across a special subpopulation of neurons), and that *is* the self, it is what you are, and your existence is coextensive with its existence. But we don’t have that sort of science yet, and so for now we live in a culture trying to deal with the contradiction between its incomplete science and the facts of subjectivity.

  5. Pingback: The Spamlist! » Reasons for Persons

  6. As you saw when you tried to untangle the quantum lottery, what counts as a single individual is tricky business. Yes, it’s caught up in the pragmatics of contextually cued disambiguation. But I think the problem of reidentification is distinct from the question of personal survival. The former is a third person question, while the latter is a first-person question. They needn’t have the same answer.

    For example: here’s person X. Some time later, in different circumstances, here’s person Y. Is person Y the same as person X? That is, was person Y once the same as person X? On the criterion of psychological continuity, the answer could be yes, but it’s possible that there is a person Z who is also person X in the same way, even though person neither person Y nor Z are psychologically continuous with each other, and thus person Y is not person Z, nor vice versa.

    Specifically: Katja-Earth steps into the matter replicator. A few minutes later, Katja-Mars steps out of the replicator on Mars, and Katja-Venus steps out of the replicator on Venus. Is Katja-Mars the same person as Katja-Earth? Yes. Is Katja-Venus the same person as Katja-Earth? Yes. Is Katja-Mars the same person as Katja-Venus? No. They are distinct branches of Katja-Earth, sharing her common past. Katja-Earth shares a future with neither of them. She steps back out of the replicator on Earth some minutes later; she may not survive the next 10 minutes, or she may continue to live out a long and healthy life. She is a third person distinct from Katja-Mars and Katja-Venus, whose personal future may very well not include junkets to any of the inner planets.

    So: identity of persons based on a criterion of psychological continuity is not a reflexive relation. Clearly, this kind of continuity is only past-directed, not future-directed. Thus it cannot answer the question of personal survival. I don’t know that that question is untangled yet. If I wake up tomorrow, I-then and his then-companions have ways to identify my then-self as identical to my now-self, but I-now know of no way to identify my now-self with my then-self other than habit. That’s something only my then-self can do, only in the reverse direction.

    Without a more carefully constructed example, I’m inclined to answer the question this way: No, I will not survive my encounter with the matter replicator. Parfit’s argument, if I recall correctly, is that *even if* the answer is No, that’s not what matters anyway. Personal survival doesn’t matter, only personal identity based on the criterion of psychological continuity matters. Since there would be someone stepping out of the replicator who is identical to me on the psychological criterion, it doesn’t matter that I won’t survive to see it. Parfit convinces me I can be happy that there would be someone identical to me who gets to see Mars (or travel to other galaxies, etc.), but I think it’s still reasonable to feel sad that it won’t be me.

    • I think you have nailed most of it. I’d quibble about your argument that Katja-Mars and Katja-Venus not being the same person but it’s irrelevant because you have nailed the continuity of it. Even a split second after stepping from the replicator on Mars, Venus and Earth, none of the Katja’s are the same as each other any longer. Katja-Earth is still *legally* Katja. The other two are… what? I’d say they are effectively the equivalent of an interesting specimen of identical triplets and should be treated so. They are of course far more intimately linked than triplets because of the shared early memories. A further interesting question is this: take the putative Christian soul: you die, you are buried, your body rots in the Earth and Judgment day comes. If you are a “good person” then your body is destructively scanned and replaced by a “spirit body”. Is that “spirit body” you? I’d argue that it’s not you in exactly the same way that the Katja-Venus and Katja-Mars are not the same person as Katja-Earth a few minutes (or less!) after leaving their respective replicating machines on Venus and Earth. The only way I can argue for continuity depends on the slowness of replication. I’d say there is a qualitative difference in having parts of your brain/body replaced slowly over time and having your brain/body destructively scanned in a short period of time. I guess I’m arguing that a “person” is the combination of brain/body *and* memories and that memories by themselves just don’t cut it.

  7. Ah, something just clicked for me. I completely believe that if I am cloned and my “old” self is killed, then I’ll still be alive. The continuity will remain. But it never *felt* like it. It always felt to me like something important died in that scenario. And I finally realized that I feel what a rationalist feels in a haunted house. ( He doesn’t believe in ghosts, but he feels fear. Likewise, I believe my clone is me, but I feel afraid/uncertain.

    • Sorry I can’t make the link. To me a person is a combination of their memories and their body/brain. Making an identical copy out of different matter and then letting the copy and the original person run forward in time is pretty clear that they’re not the same person. For me it’s an easy fix. If you destructively scan the first person so that they *disappear* then the copy *thinks* it’s the same person but it’s *not*. It’s a twin with the original person’s memories and *it’s* twin has just *died*.

      Here’s a counter example that illustrates the point: organ transplants.

      Let’s say you get a transplanted liver put in you because you partied excessively. Is that you? Most people would argue naively yes.
      In reality you are now a hybrid but still *mostly* you even though part of you has died. The hybrid, however has continuity with the past.

      So let’s roll this forward, replacing every single organ in the body but *stopping* at the brain. You have a continuum of beliefs about whether that’s *really* the same person and the truth is it isn’t the same person it’s a *hybrid*. Likely, however, the hybrid person would still believe they have continuity and they still have the original brain so they could make an argument for being the same person if not the entire body and that’s the crux of the matter: the seat of identity is a combination of the brain and the memories.

      Now here’s the crux: if the brain is now replaced with the donor’s memories are the the same person? Definitely not. Not even a hybrid.

      Are they the same person, however, with the donor’s brain but copies of the original memories put in? Definitely not but they would believe that they were.

      If you take the track that a copy of a person destructively scanned is the same person as the original then I equate that to saying that two twins are the same person but yet they are *not*. There is a clear difference in my mind to being destructively scanned and being replaced slowly piece by piece. One is the process of life (i.e. I am not composed of exactly the same cells as I was when I was a child but *some* of me is and I have the same memories) and the other is *killing me* and making a copy of me that thinks it’s me.

      • I think you are confused about this. Your philosophy sounds nice, but you are missing the definition of a person. I define “me” as a model that takes input and gives output. I don’t care how that model is instantiated: flesh, wires, atoms, etc… So a copy of me is still “me” because it’s the same model. Given any input it would give the same output “I” would give.
        It seems to me you are too attached to your body (this particular instantiation) as being the definition of “you”.

        • It is you that is confused. A copy of you is the same model but it’s not you in exactly the same way that there are tens of thousands of copies of a 2011 mustang out there. Each one of them is unique. It’s not the *same*.
          You are confusing verbs with nouns.

  8. A lot of psychological continuity is irrelevant for tracking relationships. For instance if you change your tastes in food, or have a terrible memory for places, or change over many years from being reserved to being outgoing, people will not feel that you are losing who you are. However if you change your loyalties, or become unable to recognise your friends, or have fast unpredictable shifts in your behaviour I think people will.

    You seem to be talking about qualitative identity here, as opposed to numerical identity.

  9. “Suppose you want to have nuanced relationships, with things like reciprocation and threats and loyalty, with a large number of other monkeys.”

    We are apes, actually.


Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.