One of the most famous slogans in philosophy is Descartes‘ "I think, therefore I am." This sums up the conclusion that Descartes (pronounced DAY-cart in English or day-CART in French) comes to at the end of the second Meditation. Descartes has been searching for a truth that can‘t be rationally doubted and that he can use to climb out of the well of skepticism that he has dug in the First Meditation. He decides that one thing he cannot doubt is that he exists, because even in the act of doubting, he is doing the doubting.
Derek Parfit in effect denies Descartes‘s conclusion. He would insist that the conclusion is not nearly as innocent and straightforward as it seems. When Descartes says that he exists, he means that there is a continuing subject of experiences or self or ego, or whatever word you prefer, that exists. Parfit denies that there really is any such thing, and therefore denies that we can know that there is.
Of course, Parfit does not make any reference to Descartes in the essay we read. He starts from a very different point: the so-called "split brain" experiments performed by psychologists on patients whose two brain hemispheres have been severed (because the corpus collosum has been cut - a treatment that apparently has been used in some severe cases of epilepsy.) The sorts of verbal responses these patients present suggests an inner life (or lives?) markedly different from our own. As most of you probably know, the right hemisphere normally controls and process information from the left side of the body, and the left hemisphere controls and processes information from the right side of the body. In the simplified experiment that Parfit describes, a patient looks at the center of a screen that is red on one side and blue on the other. On each side is written "How many colors do you see?" The patient responds by writing an answer with each hand. And each answer is the same from each hand: one color. But if we then ask "What is the one color that you see?" the two hands answer differently. If the left side of the screen is red, the two hads anser differently: one writes "red," the other writes "blue."
This is not what you or I would do. We would see both colors in an integrated experience, and no matter which hand we wrote with, we would write "two" in answer to the first question.
Parfit concludes that such patients have two streams of consciousness, each stream unaware of the other. Some have objected to this on the grounds that the sub-dominant hemisphere (typically the right hemisphere in right-handed people and the left on in left-handed people) is not conscious. But Parfit notes that if a patient‘s dominant hemisphere is destroyed, the patient will still behave like a conscious individual, albeit one whose linguistic capacities are highly underdeveloped.
It seems to me that there is probably a lot that we don‘t know about these cases, and may never know. But let‘s accept Parfit‘s basic view here -- that both hemispheres are conscious so long as both are intact -- and see where he leads us.
He notes that one could insist: there is not just one person in these cases; there are two people in the same body. Some psychologists argue that there is just one person; Parfit argues that in a sense, there are none.
If this is right, then the issue of reincarnation, or survival of death in any form takes on a very different complexion. Parfit cannot deny that there is some sense in which there are persons who continue from moment to moment. But what he maintains is that this is a shallow sense and owes its existence to the way we talk rather than to any deep facts about the way things are. Here a comparison might help. Later on, Parfit talks about clubs (organizations) to illustrate a point. We can use the same idea here. Does the Maryland Philosophy Club exist as a thing in its own right? The answer would seem to be no. For there to be a Maryland philosophy club is nothing more than for there to be certain people who have decided to associate with one another for certain purposes and talk about that association in a particular way. The club has no individuality or integrity in its own right. Or again: at the moment, there is on my desk at home a rather outdated and bulky modem, on top of which is a bottle of vitamin E, an aluminum can and a bowl with a spoon and a bit of ice cream residue in it. Suppose I give this collection of objects a name, calling it "Modemess." Does "Modemess" exists? The answer is on the one hand obviously it does. There is the modem and the junk on top of it; no doubt about that. But does it exist as a genuine, unified object? Its existence is surely nominal -- that is, a result of nothing more than the separate facts about the individual objects together with the fact that I have decide to give the lot of them a single name.
Perhaps this is the mode of existence of all (divisible) things, but let that pass for now. It is not how we normally think. In particular, when it comes to ourselves, we normally think that we are unified beings with a certain sort of cohesion, psychological and otherwise. We thin that we exist not merely nominally; we think that our existence as unified entities is not a mere result of somebody deciding to give us a name.
To be fair to recent cultural history, the view that things have a real as opposed to a nominal unity is something that Deconstructionism and related currents of thought have challenged. But whatever we should ultimately think of that, it is not what common sense or common belief tells us. But it is what Parfit is telling us about "ourselves."
Parfit divides theories of personal identity into two broad classes: Ego Theories and Bundle Theories. Ego theories come in different flavors, but all of them claim that what it is to be the same person over time is for some single ego, substance, subject of experiences or what have you to persist over time. The metaphysical ego is primary. The existence of the ego explains the existence of the person and not vice-versa; the ego‘s existence is not just the result of our having given a name to a certain collection of things or processes.
Bundle theories take a quite different approach. Hume was perhaps the first bundle theorist in the Western philosophical tradition (though my uneducated guess is that certain medieval thinkers may have held similar views) but Parfit points out that the Buddha also held a bundle theory. Here is a passage that Parfit quotes from a Buddhist text:
This bundle of elements is void of Self. In it there is no sentient being. Just as a set of wooden parts receives the name of carriage, so do we give to elements the name of fancied being.
Buddha has spoken thus:..."There is no Individual. it is only a name given to a set of elements."
Parfit puts it this way. We can‘t explain the unity of consciousness or the unity of a life by saying that there is some underlying person to whom all the components of the consciousness or the life belong. In particular, there is not some underlying soul or ego or what have you that holds us together, so to speak. Rather, there are two things we should say: (1) what we call a person is in effect a long series of experiences, thoughts, feelings etc. -- "a series of mental states and events" -- and (2) each of these series that we associate with a single person‘s life is tied together by various causal relations, memory being one of the more important ones. (NOTE: Parfit is not saying that each stage in the series can recall all the previous stages.
Does Parfit simply deny that there are persons? As we have already seen, the answer is no -- no more than the Buddhist quoted above denies that there are carriages. What he denies is that persons have any special or fundamental existence. They are assemblages of a certain sort.
At this point, I would like to step outside of the expository framework we have remained in up to now and ask a question or two. (Bear in mind that Parfit has written about all this at much greater length in his book Reasons and Persons.) Think for a moment about three objects: "Modemess," the assemblage on my desk described earlier, the computer on which I am typing these words, and my neighbor‘s dog Bubba. "Modemess" has no integrity of its own. Its claim to being a single object rests entirely on my having given the collection of objects that composes it a name. My computer, to be sure (like the carriage described by the Buddhist) is an assemblage as well, but it is a much more coherent and integrated one. It seems rather less conventional to call it a unified object. Still, its principle of organization is imposed from the outside. But my neighbor‘s dog Bubba is an organism -- a highly integrated system. And, as Locke would likely point out, its principle of organization or activating force is, so to speak, internal to it. We do make a distinction, both in common sense and in science, between mere assemblages and things that have a coherence or unity. On the atomic scale, this is particularly clear. An atom is in one sense an assemblage of electrons, protons and neutrons. But in another sense, it is a highly coherent unity. And collections of molecules can likewise have coherence and unity. Crystals are a particularly clear example. So are molecules of chemical compounds.
What is the point? From the mere fact that a thing is compound, it simply does not follow that its existence as a thing is merely nominal -- merely the result of our attaching a name to it. And thus far, nothing that Parfit has said gives us any reason to think that persons are somehow merely nominal.
Why might he think otherwise?
To get a sense of this, we need to turn to the sections labelled "What We Believe Ourselves to Be" and "How We Are Not What We Believe." Here Parfit uses certain imaginary situations -- though experiments to tease out what he takes our beliefs to be, and then he asks some pointed questions that call those beliefs into serious doubt.
The first cases is familiar to watchers of Star Trek. You enter a chamber. A scanner records all the facts about your brain state, and then sends the information at light speed to another location -- perhaps on another planet -- while destroying your body (painlessly) in the process. On the other end, where the information arrives, a Replica -- a perfect Replica -- of you is instantly assembled. The Replica will be psychologically just like you, will remember what you remembered, and will think that s/he is you.
We might ask: is this really you? Or is it a mere deluded duplicate? As Parfit points out, if the copy were made without destroying you, it couldn‘t be you, since you are just one person. This may lead you to think that it couldn‘t be you in any case, and that in the Star-Trek style case in which your body is destroyed, teleporting would amount to committing suicide.
The next thought experiment involves a range of cases in which a certain percentage of your body and brain is replaced with duplicate matter. Presumably if, say, 2% of your body and brain were replaced, you would have no doubt that the survivor would be you. Parfit points out: we naturally assume that there is a clear answer here: either you survive or you don‘t. And we also tend to assume that there can be no question of partial survival. If, say, replacing 90% of your matter with duplicate matter was not good enough for your survival, it would also not be good enough for your partial survival. And in the case where your body is entirely replaced with duplicate matter, we are essentially in the teleportation case. So whatever we say about the one, we should presumably say about the other.
In fact, this seems a bit quick. If the matter were replaced gradually, over a period of weeks or months or years, we might have a very different view than if it were simply replaced in one fell swoop. After all, most of our cells are replaced over the course of our lives. And while this doesn‘t‘ apply to the brain, I doubt that any of us think that this is crucial to our continued survival. If ordinary metabolic processes eliminated old brain cells and replaced them with relevantly similar new ones, I doubt this would undermine our confidence that we survive throughout. But Parfit points out: if you think that in the artificial replacement case, there is some threshold beyond which we don‘t survive, then you have a problem: how could there be any threshold 50% -- 80% -- 90% that was anything less than completely arbitrary? Suppose you think that unless at least 50% of your matter survives, you don‘t survive. Suppose that exactly 50% of you has been replaced, but that you the real, full you -- still exists. Are you really willing to say that replacing one more molecule means that you die and are replaced by a deluded duplicate who merely thinks she is really you?
Now we may be able to see why Parfit prefers the bundle theory. On the bundle theory, there are facts about the bundle of feelings, thoughts, sensations, etc. and the causal relationships among them. Beyond that, however, nothing more need be said. The bundle theorist simply rejects as confused questions like "would I really survive teleportation?"
So much for what the bundle theory says. Why should we believe it? Hume gave reasons for doubting that there is a metaphysical ego -- an unchanging entity different from all our perceptions in which all of those perceptions subsist. If this truly leaves the bundle theory as the only alternative, then that might be reason enough. But as we have seen, it is one thing to deny that there is a soul or an ego or a substratum. It is another thing to say that the self has no more than a conventional or nominal reality. And that is the view that Parfit wants to urge. So again we ask: what are his reasons? In the article that we are considering, he turns to the split-brain cases. Parfit draws our attention to two features of these cases. One is an element of unity and the other is an element of disunity. But the awareness that (part of) the patient has of seeing only red is unified with an experience of moving his hand and writing "I see only red." This is a special case of a perfectly general phenomenon. I see the keyboard; I hear the click of the keys; I feel the sensation of fullness in my stomach from just having eaten dinner. These are not just discrete, atomic experiences; they are unified into a single awareness. So it is with the split-brain patient‘s experience of seeing the redness and writing "I see red." So far, then, we might insist: the split brain patient is just like you or me; the experiences are had by one single person. But this ignores the disunity that is also part of the story. Just as the experience of red is unified with the wiring of the word "red," so the experience of blue is united with the experience of writing "blue." But these two complex experiences are not unified with each other. The experience supported by one hemisphere appears to be entirely distinct from the experience supported by the other. This is not what we would expect if a single person were having all the experiences.
Is Parfit saying that there are two persons here? The answer is no. He is saying that we can‘t explain all the facts about the patient‘s experiences by saying that the experiences are all had by one person. But what Parfit wants to say is that there are two streams of experience within the patient.
Why not say that these two streams of experience amount to two different subjects or two different egos? The best I can do here is quote Parfit:
While [that is, when] we could assume that "subject of experiences," or "Ego" simply meant "person," it was easy to believe that there are subjects of experiences. But if there can be subjects of experiences that are not persons, and if in the life of a split-brain patient there are at any time two different subjects of experiences -- two different Egos -- why should we believe that there really are such things? This does not amount to a refutation, but it seems to me a strong argument against the Ego Theory.
It seems to me, however, that this is a bit cheap as a way of supporting the bundle theory. I need not assume that there is a single mysterious entity that underlies a biological organism, making it a single organism. I need not, in particular, adopt a substance metaphysics to justify talk of a single dog or flower or (considered as an organism) woman or man. But rejecting the substance view is one thing; accepting the view that the parts of an organism are only conventionally or nominally knit together into a single entity is quite another. In short, what I am suggesting is this: the argument that Parfit is presenting us with is a sort of argument by elimination: it has the form:
Either P or Q
Not-P
Therefore, Q
In particular,
Either the Ego Theory or the Bundle Theory is correct.
The Ego Theory is not correct.
Therefore, the Bundle theory is correct.
The argument is logically correct; it does not commit any fallacies. But that is only one question that needs to be asked about an argument. We also need to ask: are the premises true? If by the "Ego Theory," Parfit means a theory that posits some special metaphysical entity that knits the parts of the person together into a unity -- a soul or mental substance or some such thing -- then we might grant, at least for the sake of the argument, that there is no such thing. But Parfit presents only one other alternative: the Bundle Theory is true, and the the unity of a person is only a nominal or verbal unity. And it isn‘t at all clear that this is the only alternative.
But perhaps we have Parfit wrong. What does he say in response to the problem of explaining the unity of experience? He says
It is simply a fact that ordinary people are, at any time, aware of having several different experiences... Just as there can be a single memory of having had several experiences, such as having heard a bell strike three times, there can be a single state of awareness both of hearing the fourth striking of this bell, and of seeing, at the same time, ravens flying past the bell-tower.
Notice that Parfit is trying to account for two things at once here: the unity of momentary experiences and the unity of experiences over time. The unity of experiences within a single moment is just a fact about experience. Experiences from different modalities (sight, hearing...) combine into single experiences. And these experiences can be linked to earlier experiences in the stream by means of memory.
Is this a genuine unity? Or is it merely a nominal unity?
Parfit gives us not the slightest reason to think that the unity of a single awareness or experience is merely a matter of our calling it an experience -- is merely nominal. The unity seems quite genuine. The question of unity over time is more complicated and less clear, but again, nothing that Parfit says counts against the genuine unity of the elements of the series that constitutes a life. After all, the elements are bound together by causal relations, and not just by any causal relations at that. Certain causal relations would not do. If you tell me about what happened to you yesterday and I remember it later, the two events -- your original experience of what happened to you and my memory of your account of it -- don‘t fit into a single life, even though the original experience that you had is part of the cause of my present memory. The process that led from your experience to my memory is not of the right sort to bring two events into the series that composes a life. And this is not just a matter of convention or description -- at least not obviously.
But what of the thought experiment cases that Parfit considers? What of teleportation or replacement of body and brain parts with duplicates? One can suspect that we don‘t know enough to know what we should make of such cases. Furthermore, it isn‘t clear that our intuitions are worth very much in sorting it out. Would the process of teleportation provide the sort of link between stages of lives that can knit them together into a single life? One wonders if we have the foggiest clue; we don‘t have really have any idea what such a process would actually involve; for that matter, there is nothing like consensus about the solution to how conscious experience arises and how experiences, momentary or otherwise -- or normally linked into a unity. If anyone insisted that we don‘t understand all the relevant causal relations well enough to judge such wild hypotheticals, it is hard to see how anyone could seriously insist otherwise.
What has all this to do with survival? Back to the teleportation case again. Parfit imagines the beam aimed at Mars. He writes:
You do not merely want there to be psychological continuity between you and some person on Mars. You want to be this future person. On the Bundle Theory, there is no such further fact. What you fear will not happen in this imagined case, never happens. You want the person on Mars to be you in a specially intimate way in which no future person will ever be you. This means that, judged from the standpoint of your natural beliefs, even ordinary survival is abut as bad as teleportation. Ordinary survival is about as bad as being destroyed and having a Replica.
Or might we turn it around? Might it be that being destroyed and having a Replica is about as good as ordinary survival? To repeat: not just any relation among distinct experiences can link them together as parts of a life. Psychological continuity is arguably a very strong relation, and may well involve much more than mere duplication would. Suppose that by sheer cosmic fluke, a duplicate of you comes into existence. There is no causal connection between you and it at all; no continuity of process. Must I believe that this fluke duplicate and one that owes its existence to very particular causal processes originating in me are really equally intimately related to me? That I have as much interest in the suvival of one as in the survival of the other? Even if the best answer turns out to be "yes," this is surely not just obvious.
What I am suggesting is that there seems to me to be space between the Ego theory and the Bundle theory. Examples such as the brain bisection cases would no doubt not have straightforward descriptions on such intermediate views. But why should they? And why should the fact that there are messy situations that are hard to describe lead us to a sort of conventionalism about cases in which things proceed more or less "normally?" My point is not that Parfit is wrong but that -- at least in this essay -- he does not make his case.
In any case, if we take the point of view that whatever personhood is, it does not involve a metaphysical Ego or Self, what we can say is that the possibilities for some sort of survival are opened up rather than closed down. Whether any of those possibilities are true, of course is another matter. But that question will be addressed in Raymond Martin‘s article.