The Apparent Rationality of Prudential Concern
Consider the following apparently straightforward inference:
I do not expect to die soon. Therefore I expect to be alive in the future. I expect I will have experiences in the future. I anticipate having experiences in the future. Because experiences can be pleasant or unpleasant, I have reason to care about the quality of those experiences.
Notice the flow of argument: from a straightforward prediction of fact—that my death is not imminent—and the seemingly innocuous observations that persons persist through time, that persons have experiences, and that experiences vary in quality, to the conclusion that I have a reason to care about the quality of my future experiences. The steps in the argument seem innocent and deeply familiar. These ideas are so closely linked as to seem inseparable.
I suggest they seem inseparable because the core concept of a person is that of a subject of experience that persists through time. Because experiences can be pleasant or unpleasant, we think a subject has reason to care about their quality in the future.
By “subject of experience” I just mean whatever has experiences. I am not claiming that the concept of a person is that of a Cartesian ego, or a spiritual substance, or a biological organism, or its brain, or any sort of psychological entity. I claim only that the concept of a person is of something that has experiences. This claim is uncontroversial.
The above argument illustrates how a motivational idea—having a reason to care about something—can be embedded in what appears to be a straightforward factual description.
If we accept the argument, and conclude that I have reason to care about the quality of my experiences, it is hard to avoid the added conclusion that I do not have the same kind of reason to care about the experiences of persons who are not myself. The inference seems to show that I have a special kind of reason to care about my own future experiences that I do not have to care about other people’s experiences. Reasons of this kind are called “prudential.”
The ideas of prudence and imprudence are tightly linked to rationality. Derek Parfit exposed this link in Reasons and Persons:
…’imprudent’ expresses an objection about rationality. This is shown by the more common words with which people are criticized for acting imprudently. When someone does what he knows will be worse for himself, he would be called by many ‘stupid’, ‘an idiot’, ‘a mug’, or ‘a fool’. This shows that this objection is about irrationality. If we believe that an imprudent act is not irrational, the charge ‘imprudent’ will cease, for many people, to be a criticism. [Parfit, 1984, p 318]
What Matters in Survival
Now consider what matters in survival. One thing we care about is that we continue to have experiences.
If I continue to have experiences, I continue to exist. And although I continue to exist through periods of unconsciousness, an end to my conscious experience is an end to what matters in my survival. Clearly, continuing to have experiences is a large part of what matters to people about their survival. Many would say that to continue to have experiences is necessary for everything that matters to them about their survival.
Parfit said personal identity is not what matters in survival, a proposal that met with resistance from many philosophers. David Lewis, after agreeing with Parfit’s view that what matters in survival is psychological continuity and connectedness, went on to say:
What matters in survival is survival. If I wonder whether I will survive, what I mostly care about is quite simple. When it’s all over, will I myself—the very same person now thinking these thoughts and writing these words—still exist? [Lewis, 1983, p 145]
That is, what matters in survival must coincide with personal identity because “survival” means “continued existence.”
It is easy to become confused by Parfit’s view that, in some cases, what matters in survival is preserved, but personal identity is lost. If I accepted the little argument with which I began this post, I would also say:
If I ceased to exist, I would cease to have experiences. Much of what matters to me about my survival depends on continuing to have experiences. I cannot make sense of the statement that I might cease to exist but nonetheless, in every way that matters, continue to have experiences.
Confusion is a test for meaning. The confusion expressed in the above passage is understandable if the core meaning of the word “person” is the idea of a persisting subject of experience.
In Reasons and Persons, Parfit argued that personal identity consists in psychological continuity and connectedness, with any cause, and without branching. Even if we allow that psychological continuity and connectedness is a good criterion for judging the truth or falsehood of personal identity statements, it is clear that “same person” does not semantically imply “is psychologically continuous with and connected to.” We know it does not because its denial is not confusing.
Many people believe they would survive transformations in which psychological continuity and connectedness are lost. Bernard Williams described such transformations, involving torture and drug-induced delusions, which he believed there was at least “a risk” he would survive.[Williams, 1970, p 91] But examples of this belief are easy to find without resorting to exotic thought experiments. They appear in widespread fears of senile dementia and similar debilitating diseases. If senile dementia ran in my family, I might easily succumb to anticipatory dread of the confusion and anxiety that dementia patients suffer after losing the mooring of their memories. I might dread this experience despite the fact that, when I reach that stage, my psychological connections with myself at present will be very weak.
A defender of the psychological criterion might reply that my anticipatory dread of this fate is rationally justified because the onset of dementia is gradual. Because psychological changes from each day to the next are not great, psychological continuity is maintained even though connectedness, over the entire span of the disease, is extremely weak. The transformation of dementia is, in this way, like the transformation of a young child into an adult with very different psychological attributes from the child he was.
On this reply, whether or not I can rationally anticipate having the experiences of the human being who will survive my dementia depends on whether the changes are sudden or gradual. Granted, if I were diagnosed with dementia, I would be concerned about the pace of degradation. Among other things, I would want to know how much time I had left in which I would remain competent to accomplish certain tasks. But I would not think of the pace of the transformation as the difference between surviving and not surviving. The severity of the changes matters to me much more than their pace.
If I accept the psychological criterion, I should regard rapid-onset dementia with catastrophic memory loss as equivalent to death. Diagnosed with such a condition, I should not anticipate having the experiences of the survivor. Yet I know I would find it difficult not to anticipate having those experiences. I would be inclined to fear the experience of confusion and anxiety which dementia patients suffer. It is no comfort to to told that the survivor of my dementia will not remember my life, and will have a completely different personality. Those facts, along with the fact of suffering, make the prospect of dementia seem worse than death.
If instead of common senile dementia, I were diagnosed with a virulent encephalitis which would destroy my medial temporal lobes overnight, wiping out all my memories, I might still expect to survive that transformation. I might anticipate having the experiences of the survivor. If the psychological criterion were correct, then I would be mistaken to anticipate having those experiences. But I would not be making a logical mistake. I would not be confused, or incoherent.
Further evidence that the meaning of “same person” is more closely tied to the idea of continuing to have experiences than to other kinds of continuity is found in the ways people talk about life after death. Any child understands the question, “If you could choose, would you rather come back as a cat or a horse?” The question is about reincarnation. It does not carry the suggestion of any measurable psychological continuity, such as memories or personality being carried over. The question is simply, “If you could now choose what life to lead after your human life is over, and you had the choice between the experiences of an ordinary cat and those of an ordinary horse, which would you prefer?”
Those who have thought hard about personal identity may find the question deeply flawed, even incoherent. Arguably, it rests on a disreputable presupposition of the existence of something resembling a spiritual substance. Such criticisms do not affect my point, which is a psychological fact: children (who are not metaphysicians) are not confused by the question, and exhibit strong preferences in their answers: one would rather be a horse because a horse is fast and strong; another would prefer to be a cat because cats (the child believes) live coddled, luxurious lives.
Most people, including those who deny the possibility of reincarnation, are not confused by questions like, “What animal would you rather come back as?” This is despite the fact that they do not suppose that reincarnation involves psychological continuity and connectedness. If I expect to be reincarnated as an otter, it is not an otter with memories of being human. What I expect is that after my death I will have an otter’s experiences.
In contrast, most people would be confused to be told that they can rationally expect to have experiences following an event in which they cease to exist. To accept it is to abandon the core idea of a person which we imbibed with our mothers’ milk—that of a continuing subject of experience.
To mitigate the confusion we need conceptual reform—the process of building a new concept or concepts to do the work of the old one that has been shown to be inadequate. Discovering that a concept which we use every day is inadequate is like discovering structural weakness in aging infrastructure, such as in a pillar supporting a bridge. We cannot safely use the bridge until we replace the pillar with something solid.
Consider information-based teleportation. I am scanned in Vancouver; my information is transmitted to Sydney, Australia; a copy of me is constructed there; the original in Vancouver is destroyed. If I am trying to decide whether to be teleported, I want to know whether I should anticipate having the experiences of my copy in Australia.
Parfit argued that if my copy is a faithful replica of me, then what matters in my survival will have been preserved. I accept his conclusion, and will not review his arguments here. If we accept his conclusion, we must say that having a replica is about as good as ordinary survival. (Alternatively, “ordinary survival is about as bad as being destroyed and having a replica.” [Parfit, 1984, p 280]) And if that’s true, then I have reason to anticipate having the experiences of my replica if, and only if, I have reason to anticipate having future experiences in ordinary survival. If there were a difference in this respect between teleportation and ordinary survival, it would be an important one.
The Extreme Claim
In Reasons and Persons, Parfit considered the view that we have no special reason to care about our future experiences—that concern for our own future is not rationally required—a view which he called The Extreme Claim. After considering arguments pro and con, he did not conclude either that the Extreme Claim is true or that it is false. “Though the Extreme Claim is defensible, it can also be defensibly denied.” [Parfit, 1984, p 312]
Anticipation of experience is part of the complex emotional structure that I, following Ray Martin, call ‘self-concern.’ [Martin, 1998] Anticipating having experiences in the future is, for most of us, inseparable from believing we have reason to care about the quality of those experiences. It is also, for most of us, inseparable from actually caring about the quality of those experiences.
Bernard Williams described a range of cases which demonstrate plasticity in self-concern. Describing the same set of facts in different ways can induce normal, thoughtful human subjects either to extend self-concern to the survivor of a transformation, or to withhold it. That different descriptions tend to influence human subjects in this way is a psychological fact about human beings.
Moreover, a single description can have different effects on different persons, inducing some to extend, and others to withhold, self-concern. Information-based teleportation can be used to demonstrate this plasticity. If it is described this way:
You will be scanned in Vancouver, and the file containing your information will be transmitted over the internet to Australia. A living copy of you will be constructed there. The original in Vancouver will be destroyed.
many people would regard it as a death sentence. They would not anticipate having the experiences of the replica in Australia, and would refuse to be teleported. That is a psychological fact about those people. But I, and Parfit, and others (including several non-philosophers I’ve talked to), would readily anticipate having the experiences of our replicas. That is a psychological fact about us.
In asserting or denying that to be teleported is to have reason to anticipate having the experiences of the survivor, neither group is making a factual mistake. There are no facts about the process of information-based teleportation that settle the question which group is right.
Neither the Teleportation Users Group or the Anti-Teleportation League is making a mistake
Someone who thinks that psychological continuity and connectedness is what matters in survival might object that people who regard ordinary teleportation as death are making a mistake, because teleportation preserves psychological continuity with strong connectedness, and ordinary teleportation is non-branching. This objection misses the point. A person who regards teleportation as death may be fully aware that teleportation preserves psychological continuity. Although he may acknowledge that his replica in Sydney is a good replica, he may deny that his replica is himself. He can do so coherently because the traditional concept of a person is not based on psychological continuity and connectedness.
In describing his denial as coherent, I am making a claim about the meaning of words. “Is the same person as” does not mean “is psychologically continuous and connected with, without branching.” It means “is the same subject of experience as.” Psychological continuity and connectedness is at best a plausible criterion for personal identity. Members of the Anti-Teleportation League reject teleportation because they reject this criterion.
I claim that the traditional meaning of the word “person” includes, as a core element, the idea of a continuing subject of experience—an entity that persists through time and has experiences. It might be objected that this idea is too vague, because there is no clear criterion for applying it. It might be thought that, in devising a plausible criterion for personal identity, Parfit has clarified and elucidated an indispensible, though muddy, concept, rather than recommended conceptual reform.
I agree with this objection in part. There is no clear criterion for deciding the truth or falsehood of statements of the form, “a is the same subject of experience as b,” unless such statements are embedded in a scientific, or at least empirical, account of what such subjects are. And to come up with an empirical account of subjects of experience, or persons, is the main aim of the contemporary philosophical debate about personal identity.
Nonetheless, the objection fails because it does not explain the continuing unease we feel when considering transformations in which psychological continuity and connectedness is not preserved. If “is the same person as” meant “is psychologically continuous and connected with,” no one would have anticipatory fear of living for years, in an institution, as the survivor of rapid-onset dementia with complete memory loss and personality changes. But people have such fears.
Why the Extreme Claim is True
Given that there is no fact of the matter whether the Teleportation Users Group or the Anti-Teleportation League is correct, we can legitimately conclude that no fact provides a reason for anticipating future experience. If there is no fact of the matter in the case of teleportation, there is no fact in the ordinary case either. Nothing in nature affords a reason for anticipating future experiences. There is no natural relation that justifies anticipation. Anticipation is just something we do—a feature of human psychology.
Anticipation is a psychological attitude that is strongly motivational. To anticipate having experiences is to give imagined future experiences a similar motivational effect to actual experiences in the present. If I feel the intense heat of a stovetop, I pull my fingers away. If I see a red-hot iron approaching my skin, I anticipate a burning pain and immediately move out of the way. If I cannot move because I am shackled to a torture rack, I suffer as I witness the approach of the hot iron, much as I suffer later when it touches my skin. These are psychological facts about me.
There is a scientific explanation of these psychological facts. Part of the explanation is about natural selection. Motivational anticipation of experience has survival value. This is true for human beings and for many other animals.
The survival value of anticipation is greatly amplified when combined with the ability to imagine the non-immediate future in detail. I will not say this is an exclusively human characteristic, but it is marked in human beings. A dog suffers when witnessing the approach of a drunken master who is in the same foul mood he was when he kicked the dog yesterday and last week. But as far as we know, dogs do not worry in summer, when food is plentiful and life is pleasant, about how they will get through the harsh conditions of the winter to come. Human beings excel at that sort of worrying. Because we care so much about the quality of our future experiences, we devote a lot of energy to making detailed plans for the future; and we are very good at it. As increasing wealth has banished the wolf from our doorsteps, consigning him to outer fields and the woods beyond, more and more of our energy is being directed further and further out.
Information-based teleportation is a transformation which anticipation of experience can cross. I know that, because I would readily anticipate having the experiences of my replica in Australia. That is another psychological fact about me. And I don’t think there is anything unusual about me, other than the fact that I have thought about teleportation, imagining in detail what the world would be like if information-based teleportation became a reality. The process is physically possible, and likely, I think, to become technically possible this century. Because of its many benefits, it is likely to be commercialized, and to be adopted by many people. Although some will hesitate to use it at first—fearing that the process is equivalent to death—most will be persuaded when they encounter the replicas of their friends and loved ones and see that there is no discernible loss. People who hold to the belief that teleportation is a death sentence will start to seem backward and deluded, as a geocentrist today would seem backward and deluded to almost everyone, and as those who deny evolution seem backward and deluded to many of us. It will become obvious to most people that nothing important is lost in the process of human replication.
Replication preserves similarity, nothing more. When I accept that replication preserves everything important in survival, I am confronted by the question: how similar must my replica be to myself in order to preserve what is important in my survival? And, if I have not yet been convinced that anticipation is irrational, I am confronted by the more specific question: how similar must my replica be in order for me to rationally anticipate having my replica’s experiences?
One difficulty is that similarity comes in degrees, covering a continuous spectrum of cases ranging from no detectable difference to differences as great as the differences between any two persons. Parfit’s enlightening discussion of similar spectrum cases shows that there is no place to draw the line. [Parfit, 1984, pp 231-243] If I try to draw it anywhere on the spectrum, and say that I can reasonably anticipate having the experiences of replicas on one side of the line but not the other, I am confounded by the fact that the differences between two replicas on opposite sides, but both very close to the line, are very slight. Those differences cannot justify anticipating having one replica’s experiences and not the other’s.
Anticipation of experience seems not to admit of degrees. If I expect to undergo some transformation—whether teleportation, or ordinary aging, or a career change, or catastrophic rapid-onset dementia with loss of all my memories—I either anticipate having the experiences of the survivor, or I do not. I don’t know how to anticipate on a sliding scale, extending more concern for some survivors than for others. Whether or not I will have the survivor’s experiences seems to me to be an all-or-nothing matter.
I can easily describe a spectrum of transformations in which the survivors bear many different degrees of similarity to me now. Since I am unable to come up with a clear idea of having my survivor’s experiences which can be applied strongly or weakly, so as to match this spectrum, I am led to doubt that there is any real difference between having and not having my survivor’s experiences. I am led to doubt that there is any reason to anticipate future experiences, either in exotic cases like teleportation or in the usual ways that we survive.
Spectrum arguments show that no theory that claims attributes are all that is important about persons has room for a reason to anticipate future experiences. Since I recommend such a theory, I conclude that there is no such reason. Anticipation of experience is not rationally required.
Conceptual Reform – A Change of Attitude
It does not immediately follow that anticipation of experience is irrational. Anticipation of experience might still be rationally permissible.
It is difficult to show that anticipation is not rationally permissible. To anticipate having an experience is to expect that the experience will occur, and to have a distinctive emotional attitude towards that experience. The attitude of anticipation is motivational. It is a disposition to try to bring about pleasant experiences and to prevent unpleasant ones. How could it be shown that this attitude is irrational, that is, not rationally permissible?
One way would be to show that anticipation of experience cannot be separated from the belief that there is a reason for it. If I were to anticipate having some experience, and I stopped believing that anticipation is rationally required, would what remains of my attitude be anticipation, or some other attitude?
Anticipation never feels like an attitude we can rationally choose to withhold. There is a compulsion about it; we feel we ignore it at our peril. Clearly, the belief that anticipation is rationally required adds to its motivational force. If the belief were to die, what remains would be a much weaker motivator.
I cannot form a clear idea of an attitude of anticipation unaccompanied by the belief that some relation provides a reason for anticipation. That is evidence that an attitude is not anticipation unless it is bolstered by the belief. But it is not conclusive evidence; perhaps it is just a failure of my imagination.
Contrast anticipation of experience with another attitude: sympathy for another person. Sympathy is not usually thought to be rationally required. Either to be sympathetic or to be unsympathetic is not normally considered irrational. (Some people consider it irrational to be moved by sympathy, but I will ignore their position until a later post.)
Although sympathy may move people involuntarily, it is widely recognized to be somewhat tractable to choice. Sympathy is not just felt; it can be given or withheld. The idea that sympathy is partly voluntary is expressed in descriptions like, “He hardened his heart, but she opened hers.”
Parfit claims that psychological continuity and connectedness, without branching, is the only plausible criterion for personhood, or at least the best one. He may wish to make a stronger claim: that what we should mean by “being the same person” is “being psychologically continuous with and connected to, without branching.”
If he does so, he has moved beyond conceptual analysis into the territory of conceptual reform. Conceptual reform is a legitimate aim for philosophy. But philosophers who engage in conceptual reform should make it clear what they are doing. To recommend reform of the concept of a person is to recommend that we change the meaning of the word “person.”
The concept of a person is one of the most important concepts we have. We cannot abandon it: society would collapse. For many of its uses—property ownership and contractual obligation, to take two hard-boiled examples—we have no ready substitute. Even if the concept of a person is deeply incoherent or based on false assumptions, we cannot simply stop using it, because we depend on it too deeply. Our only hope is to reform it: to recommend a new, more robust meaning for the word “person” that is free from incoherence and false assumptions, and that will do the work we need it to do.
The traditional account of the reason we all have for anticipating our future experiences is breathtakingly simple. Persons persist; they have different experiences at different times; experiences differ in quality. If there is no fact that rationally justifies anticipation of future experience, what part of the simple account should we give up?
I recommend giving up anticipation itself. We should continue to accept that persons persist, and have experiences, and that some experiences are much worse than others. Rejecting Open Individualism, we should also accept that there are differences between persons—that I am not you. Accepting all this, there is room to deny that I have a reason, which you do not, to anticipate having some experience, for example, the delightful experience of eating leftover tongue in madeira sauce for lunch.
My recommendation may seem unclear. You might think that giving up anticipation has no bearing whatever on the truth or falsehood of the very simple account I gave of the reason for anticipation.
My recommendation may also seem too daunting. How would we get by without anticipation? Anticipation of experience evolved because it has survival value. It goads us into action just as effectively as, or even more effectively than, an actual goad. It goads us to do things that we think will benefit ourselves in the future. Quite often, our plans work out and we actually do benefit as a result.
I will address the dauntingness first. In recommending that we give up anticipation, I do not recommend that we stop caring about our futures. Anticipation of experience is not the only motivating attitude available to us. Sympathetic concern is another motivating attitude. If we abandon anticipation, it is open to us to be motivated by sympathetic concern for ourselves in the future.
Sympathetic concern is, of course, an attitude which we can take towards others. It may seem odd to have sympathetic concern for oneself. But I think it is odd only because the sympathetic concern we might feel for ourselves is usually eclipsed by the more salient and compelling concern that accompanies anticipation. When I manage to weaken, or stun, my inclination to anticipate having experiences in the future, by persuading myself that I have no reason to do so, then I actually find it not difficult to have sympathetic concern for my future self.
I think sympathetic concern is enough to motivate people to look out for their own interests. Anticipation is overkill. Anticipation may have been well suited to an earlier time, before our species dominated the planet. It is a very strong motivator. The belief that it is rationally required adds to its strength. The fear, and driving appetite for personal gain which accompany anticipation of experience may have been necessary for the survival of early hominids struggling to compete in an unforgiving African jungle. I don’t think they are necessary today.
These days, the strong emotions of anticipation primarily motivate human beings to compete with each another, more than with other species. Especially in industrialized nations, where human beings collectively control more resources than is needed for the survival and comfort of everyone, these strong emotions have become dysfunctional. Instead of improving the chances of individual survival, they threaten the survival of our species and the livability of the planet.
If we were to root out anticipation of experience, banishing it from our thoughts, the role of sympathetic concern could expand to fill the motivational void. Sympathy has much to recommend it. Because it is an attitude which we have towards others as well as one we can have towards ourselves, it tends to favour the claims of morality over the claims of self-interest when those come into conflict. Any reason I have for sympathetic concern for my future self is the same kind of reason I have for sympathetic concern for other people.
Compared to anticipation, sympathy also is a much better fit to a reformed concept of persons. That is true of Parfit’s recommended reform, that non-branching psychological continuity and connectedness should be considered the unifying idea of persons. It is also true of my recommendation that what unifies a person over time is informational content, or (more simply) similarity. Parfit’s recommendation is a restricted case of my recommendation.
One reason sympathy is a better fit is that it admits of degrees. Psychological connectedness—which is just psychological similarity with (according to Parfit) any cause—admits of degrees, as does similarity in general. We have different degrees of similarity to different people, and we extend different degrees of sympathy to different people. We tend to be most sympathetic to those who resemble us in certain respects.
I hope these remarks succeed in making the proposal to give up anticipation somewhat less daunting. I will now address the charge that my proposal is unclear.
To summarize my position: I continue to believe that I persist through time and have experiences, some more painful than others. But I deny this gives me a reason to anticipate having my future experiences. I deny it because no one has been able to find a relation in nature that rationally justifies anticipation of experience. The puzzle cases provide good evidence that there is no fact of the matter as to whether or not anticipation is justified. I conclude that it is not justified. That much seems clear enough.
The truth or falsity of my view, so outlined, does not depend on whether or not it is psychologically possible to give up the emotion-laden attitude of anticipation. However, I have also voiced my suspicion that anticipation depends on the belief that it is rationally required. If that suspicion is true, and I continue to anticipate having experiences in the future, then I do not fully believe my view. Anticipation of experience requires believing that there is a reason for it—believing that anticipation is rationally required by some real relation between me, now, and in the future. Believing that there is such a reason is what gives anticipation its compelling edge. If that compelling edge were removed, anticipation would cease to be. What would remain would be a different attitude that is more weakly motivating, one resembling—perhaps indistinguishable from—sympathetic concern.
If that’s right, and I am to believe the truth about persons, I must stop anticipating having experiences in the future.
On the Possibility of Believing the Extreme Claim
When contemplating the simple facts that I persist through time and have experiences, some pleasant and some painful, I find it hard to believe that I have no reason to anticipate my future experiences. If I had no alternative view, I might be unable to believe it.
But I have developed an alternative view. I continue to believe that I persist through time and have experiences, some more painful than others, but I don’t believe this gives me a reason to anticipate having my future experiences. Because I don’t believe that, I no longer anticipate them. I have an alternative attitude, which is sympathetic concern for my own future. Having that alternative allows me to care for my own interests in the future. It makes it psychologically possible for me to stop anticipating my experiences.
In this way I find it possible to believe the Extreme Claim, and the truth about persons.
Lewis, David (1983) “Survival and Identity”, in Martin and Barresi, ed. (2003) Personal Identity, Blackwell.
Martin, Raymond (1998), Self-Concern, Cambridge Studies in Philosophy.
Parfit, Derek (1984), Reasons and Persons, Oxford University Press.
Williams, Bernard (1970), “The Self and the Future,” in Martin and Barresi, ed. (2003) Personal Identity, Blackwell.