Past Desires

Consider the following example from Parfit’s Reasons and Persons:

When I was young what I most wanted was to be a poet. This desire was not conditional on its own persistence. I did not want to be a poet only if this would later still be what I wanted. Now that I am older, I have lost this desire. I have changed my mind in the more restricted sense that I have changed my intentions. But I have not decided that poetry is in any way less important or worthwhile. Does my past desire give me a reason to try to write poems now, though I now have no desire to do so? (1984, 157)

The answer is clearly “no.” Parfit’s past desire to become a poet doesn’t give him any reason to write poems now. And, from this, we can conclude that the fulfillment of his past desire to be a poet is of no benefit to him, for it is plausible to suppose that if the fulfillment of his past desire were beneficial, he would have a reason to write poems. But can we further conclude that, in general, the fulfillment of a past desire is not, itself, beneficial? -- see note 1 below.

As Steve Luper rightly points out in his forthcoming paper “Past Desires and the Dead,” the answer is “no,” for there is more than one possible explanation for why Parfit wouldn’t benefit from writing poetry now that he has no desire to do so. One explanation is that it is the “pastness” of his desire that undermines what prudential value there might otherwise have been in satisfying it. But another explanation points to the process by which Parfit’s desire has come to past and whether that process is something that Parfit would have opposed. Luper claims that although a desire doesn’t always lose its prudential significance when it is removed, it does always lose its prudential significance when it is given up or removed voluntarily (i.e., given up or removed by a process that one would not oppose) -- see note 2 below. So Luper believes that what explains the fact that Parfit would not benefit from satisfying his past desire is that his desire has been voluntarily given up. Furthermore, Luper claims that the non-fulfillment of a past desire that has been involuntarily removed can be harmful. By contrast, I believe that the non-fulfillment of a past desire is never, itself, harmful and that the best explanation for why Parfit would not benefit from satisfying his past desire by writing poetry is that his desire has come to past, not that it has come to past by a process that Parfit would not oppose. So who’s right, me or Luper?

To determine which of us has the better explanation for why Parfit doesn’t harm himself in choosing not to write poetry, let’s consider a variant of Parfit’s original case. Suppose that, unbeknownst to Parfit, Dr. Evil has been slipping a drug into Parfit’s morning tea and that this drug has caused Parfit’s interest in poetry to wane, allowing his interest in philosophy to wax. Eventually, as a result of the drug, Parfit has lost all desire to be a poet. And let’s assume that Parfit would oppose this type of external manipulation of his desires. Now the question is: Is the fulfillment of his past desire to be a poet of any prudential importance? In other words, does his failure to write poetry now harm him by making it the case that his past desire to be a poet will never be fulfilled? My inclination is to say “no.” Of course, I might want to say that Dr. Evil harmed Parfit in giving him the drug, for Dr. Evil thereby violated his autonomy. And I might want to say that if Parfit would have been happier being a poet as opposed to the philosopher that he is now, then Dr. Evil harmed him by causing him to be less happy than he would have otherwise been. But I would deny that Parfit is harmed by the non-fulfillment of his past desire to be a poet, despite the fact that his desire was removed by involuntary external manipulation. Unfortunately, I have no argument for this. It’s just a brute intuition. So I’m curious whether others share my intuition. For if my intuition is correct, then we should, contrary to Luper, hold that the non-fulfillment of a past desire is of no prudential importance whether or not it has come to past as a result of a process that one would oppose.

Notes:

Note 1: It could be that doing what will satisfy a past desire is beneficial, as where the fulfillment of one’s present desires is beneficial and one presently desires to do something that will satisfy this past desire. But this is not a case where the fulfillment of the past desire is, itself, beneficial. It is the fulfillment of the present desire that is beneficial.

Note 2: As Luper points out, we don’t disapprove of our desires changing, say, as a result of rational deliberation or as a result of an unmanipulated shift in our concerns from one valuable thing to another. But we do disapprove of our desires changing as a result of involuntary external manipulation (e.g., brainwashing). And when it comes to certain desires (e.g., one’s desire to do philosophy as opposed to one’s desire to smoke cigarettes), many of us would oppose someone else giving us a drug to remove the desire.

Comments

[Scott Wilson]

Doug,

I think I am on your side on this one. I am not sure if my reasons for this view are sound or not, but I am sure someone will let me know if they are not.

Desires are the sorts of things that come into existence, perdure for a certain amount of time, and go out of existence. If you believe that the satisfaction of a desire can make a person better off, then it would seem that the thing that is to be satisfied must exist at the time of its satisfaction to have this effect. If the desire does not exist anymore, than there is nothing left to satisfy.

As Luper rightly points out, there are several ways for a desire to cease to exist: I can voluntarily allow it to be replaced by other desires, external manipulation might rid me of it, dementia, and death, to name a few. However, in each of these cases, the desires cease to exist, after which there is nothing left to be satisfied or frustrated.

The alternative view, that desires can be satisfied after a person's death, makes the ontological status of desires rather mysterious to me. Just what is being satisfied in these cases: where do the desires of the dead reside? Of what do they consist? The view I am advocating doesn't have to answer these questions. Desires are mental states, and when a person dies, and has no mental states, she she no desires (and hence nothing to be fulfilled or frustrated).

[Doug Portmore]

Scott,

I think that you’re right: the desire and the desire-thwarting state of affairs must be cotemporaneous. Let “P” stand for some person, “S” for some state of affairs, and “tn” for some moment in time. With these abbreviations, we can represent an unfulfilled desire as follows: P desires, at tx, that S obtains at ty but S does not obtain at ty. There are at least two possibilities concerning tx and ty. One possibility is that tx refers to a time prior to ty. If so, we’re dealing with an unfulfilled past desire, as is the case in Parfit’s example. Another possibility is that tx and ty both refer to the present time. In that case, we’re dealing with an unfulfilled present desire. For instance, I presently desire a cold beverage but don’t have one. Thus my present desire is unfulfilled.

Now assuming that you and I are right that an unfulfilled past desire doesn’t constitute a harm because the desire and the desire-thwarting state of affairs must be cotemporaneous, there is only one move available to the desire-fulfillment theorist who wants to countenance a harm in the revised version Parfit Poet Case (the case where Dr. Evil causes his desire to be removed): re-describe the intentional object of the relevant desire so as to convert it from a past desire into a present desire. Thus “P desires, at t1, that {S obtains at t2}” becomes “P desires, at t1, that {it is true now, at t1, that S will obtain at t2}” -- the intentional objects are demarcated by braces. To illustrate, let t1 be the time when Parfit was young and wanted to be a poet, let t2 be the present, and let S stand for the state of affairs where Parfit is a poet (i.e., someone who writes poetry). Thus Parfit’s desire, at t1, that {S will obtain at t2} gets re-described as the desire, at t1, that {it is true to say, now at t1, that S will obtain at t2}. In other words, Luper can claim that what Parfit desired when he was young (t1) is that it was true say then (at t1) that he will a poet at t2. In doing this, the desire-fulfillment theorist is able to account for Parfit’s being harmed by not being a poet now in terms of what was at the time an unfulfilled present desire, thereby avoiding the implausible view that an unfulfilled past desire constitutes a harm. The problem, though, is that this move is entirely ad hoc. What reason is there to think that this is the relevant description of the intentional object of Parfit’s desire? Moreover, if we ought to re-describe the intentional objects of such desires in order to countenance such harms, then, to be consistent, we ought to re-describe Parfit’s desire in the original version of Parfit’s Poet Case. That is, we should say that what Parfit desired when he was young was that it was true back then that he would become a poet when he grew up, which was, at the time, an unfulfilled present desire. But, in that case, Luper would need to claim that Parfit was harmed back when he was young whether or not his desire was subsequently removed voluntarily or involuntarily.

[Campbell]

Here's a proposal, which I take to be in the general spirit of Luper's position (as you describe it). First, define the notion of endorsing a past desire as follows. An agent endorses a past desire just in case, if it were up to her, she would still have the desire. The idea is that, in order to see whether an agent endorses a particular past desire, we "go" to the nearest possible world at which her having or not having the desire is within her control; if she's got the desire at that world, then she endorses it (at the actual world), otherwise she doesn't endorse it. On my proposal, the satisfaction of a past desire benefits an agent only if she endorses the desire, in this sense.

Now, consider Parfit's original case. Here, the nearest possible world at which Parfit has control over the relevant desire is the actual world, and in the actual world he doesn't have the desire any more. So, he would not be benefited by the satisfaction of the past desire. Consider next your modified case, with Dr Evil. Here, Parfit's lacks control in the actual world, since Dr Evil's manipulation has taken control away from him. So we need to look at the nearest possible world where there's no Dr Evil. Presumably, at such a world, Parfit will again have lost the relevant desire. So, again, the past desire isn't endorsed, and its satisfaction isn't beneficial. Consider finally a case involving past desires of the dead. Suppose Alanis, who died last year, desired consistently throughtout her life that the Cleveland Browns win the Superbowl in 2005. Clearly it's no longer up to her whether or not she has that desire (she doesn't have the desire and there's nothing she can do about it, because she's dead). But, we may suppose, at the nearest possible world where she's not dead, she still has the desire. Hence, the proposal allows that Alanis would be benefited by the Cleveland Browns winning the Superbowl next year.

[Doug Portmore]

Campbell,

Your proposal is interesting, and I believe that it is more plausible than the position that I've been attributing to Luper in that it gets the intuitively correct result in my revised version of Parfit's Poet Case: the result that Parfit would not benefit from writing poems now and has no reason to do so, despite the fact that his desire to write poetry was involuntarily removed by Dr. Evil. However, I find it highly counter-intuitive to suppose that Alanis is harmed if the Cleveland Browns don't win the Superbowl in 2005. But, in any case, I think that we should, for now, avoid cases involving posthumous harm, because people's intuitions may be swayed by the thought that Alanis can't be harmed after her death.

So let's stick to a Parfit-like case. So suppose that there's a man named Schmarfit, who is now a philosopher and who, when young, desired to become a poet when he grew up. Furthermore, let's suppose that Dr. Evil removed Schmarfit's desire in the same way that Dr. Evil removed Parfit's desire in my revised version of Parfit's Poet Case. Now let's suppose that in the nearest possible world where there's no Dr. Evil Schmarfit, unlike Parfit, continues to desire to be a poet and indeed eventually becomes a poet. Nevertheless, in the actual world, Schmarfit no longer desires to be a poet and has in fact become a famous Oxford philosopher. Does Schmarfit harm himself if he fails to do what is necessary to become a poet (i.e., write poems)? Your proposal would imply that he would. I find this highly counter-intuitive and would therefore reject your proposal.

You might respond that overall he is better off not writing poems in that the harm that would result if he doesn't write poems (the harm in his past desire being unfulfilled) is outweighed by the harm that would result if he writes poems: the frustration of those present desires that will be fulfilled only if he doesn't spend time writing poems. But it seems to me that it's not just that he has no net reason to write poems; it's that he no reason at all (not even a pro tanto reason) to write poems now that he has no desire to do so -- at least, the fact that writing poems now would fulfill his past desire doesn't give him any reason to write poems now. Or so it seems to me.

[Campbell]

With regard to any desire of an agent we can distinguish: (1) the times at which the desire is held by the agent; (2) the times at which the desire is satisfied; and (3) the times at which the agent is benefited by the satisfaction of the desire.

(As an aside, the second of these seems problematic. Consider Alanis's desire the Browns win the Superbowl in 2005, and suppose that the Browns do, in fact, win. At what time, or times, is the desire satisfied? Just at the precise time when the game ends? Or just the time when the Browns are awarded the trophy -- some kind of bowl, presumably? Or at every time from the moment of victory onwards? Or at every time in the history of universe? It seems hard to say. But let's not get into that right now ...)

Now, we may define the following two principles:

(P1) For any times b and s, if an agent is benefited, at b, by the satisfaction of a desire, at s, then the agent must hold the desire at b.

(P2) For any times b and s, if an agent is benefited, at b, by the satisfaction of a desire, at s, then the agent must hold the desire at s.

(N.B. in both cases s is intended to refer to the time at which the desire is satisfied. These things are a little hard to express in nice English.)

The difference is that, whereas P1 requires that the desire be held at the time of benefit, P2 requires that it be held at the time of satisfaction. Of course, if we think that these two times must be the same, then the two principles will, for us, come to the same thing. But if these times are allowed to differ, then the two principles have interestingly different implications. Consider the Parfit case: in his youth, Parfit desires to write poetry, but later in life he loses that desire. And suppose that, for some reason, Parfit does write poetry later in life, even though he has no desire to do so. Parfit's writing poetry satisfies his earlier desire. Does he benefit from this? P2 implies that he does not benefit, because he doesn't hold the desire at the time of its satisfaction (later in life). But P1 allows that he does benefit; however, if he does benefit it must be at a time prior to the time at which the desire is satisfied.

Now, suppose that we accept P1 but not P2. Then, on our view, there's a sense in which we can benefit the dead; for there are things we can do now that will be of benefit to those who are now dead. But there's also a sense in which we cannot benefit the dead; for these people will not be benefited now, when they're dead, but earlier, when they're alive. A person cannot be benefited while dead, but she can be benefited (while alive) by events that do not occur until after she's dead.

Dunno whether that's a plausible view. But it is interesting, I think.