In Reasons and Persons, Parfit invites us to imagine a character called 'Proximus' who cares more about the nearer future. This 'bias towards the near' means that he would wholeheartedly prefer to undergo intense pain later rather than a mild pain now. This seems irrational. We may advocate temporal neutrality in general, and so think that "mere differences in timing... cannot have rational significance." I agree with this, so in this post I want to address a couple of objections Parfit makes to this view.
Parfit's basic strategy is to compare the bias towards the near with two other temporal biases we have. The bias towards the future is seen in our tendency to be relieved when bad things are past. And the bias towards the present is our tendency to care especially about the pain we are now experiencing. Parfit claims that temporal neutrality requires us to denounce both these tendencies, but such denunciation seems crazy. So we should conclude that mere differences in timing can be significant after all. His most compelling example is of "the mounting excitement that we feel as some good event apporaches the present -- as in the moment in the theatre when the house-lights dim." Such excitement seems perfectly reasonable, yet - Parfit claims - it is an instance of the bias towards the near. Shouldn't temporal neutrality require us to be just as excited about distant pleasures?
Well, no. We should be more careful in understanding the scope of the temporal neutralist's claims. Obviously the claim is not that timing never matters for anything: if a bomb is about to explode, better to start running sooner rather than later! For similar reasons, it makes more sense to fear the explosion when the risk is in the future rather than the past. It's perfectly reasonable to attend more to present events, and to be excited about those that will very soon be present, etc. The neutralist need not deny any of this. Their claim is simply that one's preferences should not involve any temporal bias, or time-inconsistency. Note that temporally responsive feelings are perfectly endorsable from a timeless perspective. I do not later regret feeling excitement before the show, but I would regret choosing a lesser nearby good over a greater distant one. That's a revealing difference.
If we recast Parfit's proposed biases in terms of preferences (e.g. for the bias towards the present, say you would forsake greater future benefits in order to obtain a small boost of pleasure right now), they no longer seem any more defensible than Proximus' bias towards the near. So I don't think Parfit has any good objection to temporal neutrality after all.
Monday, September 24, 2007
19 comments:
Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)
Note: only a member of this blog may post a comment.
Subscribe to:
Post Comments (Atom)
Posts
I think your distinction between feelings and preferences is a good tack for the temporal neutralist to take, at least in response to the sorts of cases Parfit raises.
ReplyDeleteI remain, though, unconvinced by the T.N. position. I think this is in part because I am drawn to a philosophical version of the ‘what doesn't kill us makes us stronger’ cliche: It seems for at least some of us that we are able to 'recontextualize,' one might call it, past suffering because of how we later come to know that it fits into the narrative of our lives. And that narrative is arguably one that can only be constructed retrospectively. I think this has some bearing on the issue at hand. Let me try to explain with an example.
I have been fortunate to have been blessed with a life relatively free of suffering and pain, certainly compared to people who have not had the good health and financial advantages that I've had. But with that caveat against self-indulgence, I would say that my time of most intense and prolonged suffering was in high school. Let's just say that suffering was of strength X. Suppose I am asked now whether I would prefer that X be in my past or in my future. T.N. would suggest that since both X’s are of equal painfulness, I have no reason to prefer past X to future X. But it seems to me that I do have reason to prefer past X. If X is in my past, I have the advantage of possibly being able to fit it into a story of my life (perhaps a story that explained how that suffering was important to what I have become). That's of course something I can only build after the fact. If X is in my future, and that’s all I now of my future, then I am not able to construct such a narrative, deprived of a richly colored (and yet-to-come) backdrop with which the future suffering could be interwoven.
Although the phenomenological feel of the past X and future X are, by stipulation, the same at the respective times they are happening, there is the potential that, by a change in surrounding circumstances, our post-future X self could not hold the same, shall we say, ‘life-affirming’ attitude toward future X, once it is in his past, as my present self can hold toward past X. (Post-Future X self might, for example, die as a result of X, or soon thereafter for some unrelated reason, hit by a bus, say, immediately after he suffers X.) Since we don’t know at the present what those circumstances will be, we’re left with a kind of epistemic blindspot that we don’t have when we’re looking back on our past suffering.
A key issue, I think, is the following: It seems that what might matter more to us is not the phenomenal feel of the suffering as such, but rather our overall attitude to the suffering as refracted through some kind of transfiguring narrative in the context of our life as a whole, as the result of which it can come, ex post facto, to seem less bad, perhaps even necessary or beneficial. The trouble with the T.N.'s implied hedonic calculus is that it seems to be comparing coffee grounds to coffee grounds when we are really comparing coffee grounds to coffee.
The sorts of cases that Parfit gives are highly artificial, and that may seem a merit in the way that it focuses our intuitions. But by abstracting suffering from any sort of context, I worry that Parfit skews our intuitions. In normal cases, in which our lives consist of more than simply 2 instances of suffering, that suffering has the potential to fit into a larger pattern. And Parfit’s cases don’t reflect that.
I suppose that it could just be stipulated that our X suffering is so fruitless that it cannot be fit into any narrative at all. Yet even if this is right, it would only show that sometimes we do not have reason to prefer past X over future X. For it seems that at least sometimes past X can prove fruitful, whereas the difficulty is that we don’t know whether future X will pan out in the same way. Hence, we would seem to have reason to prefer past X in such cases, even if the X’s are, phenomonologically-speaking, the same.
Likewise, suppose we could have the benefit of a crystal ball, telling us the future, so that we could fit future X into the story of our life. Of course, even if that’s possible in some cases, or by supposition in a thought-experiment, that’s not true in all cases. It would seem than in those cases when that’s not true, everything else being equal, it would be reasonable to prefer past X.
(It's probably worth pointing out that my intuitions may be odd here. I think that suffering is not incidental or instrumental to the value of a life, but in part constitutive of that value, provided it is incorporated into a compelling narrative. That may be an uncommon view. )
-Andrew
Andrew - I like that a lot. Parfit's discussion was oversimplified by the assumption of a false theory of welfare (simple hedonism). I recently argued that pains are not so straightforwardly aggregative, and you expand that point even further by bringing in issues of broader context, which I absolutely agree with you about. (I'm a big fan of 'narrative' talk.)
ReplyDeleteBut this is consistent with temporal neutrality as I am understanding it. If considerations of temporal context mean that the long past pain is actually better for me (overall, considering my life as a whole) than the short future pain, then the temporally neutral injunction to promote general welfare tells me to prefer the past pain in such a case. Crucially, there is no time-inconsistency here: even in the past, I should prefer to be suffering the pain then rather than later, in the knowledge that this way will give me the best life overall. This is a decision I can endorse from a timeless perspective, and from all particular times. It is not biased in favour of my future self, or anything like that. It is what is best for my life overall.
It's worth noting though, that in the odd artificial case involving crystal balls and everything we need to ensure that 'all else is equal', we can tease out the isolated issue of whether temporality has any intrinsic (or fundamental) importance. From these rare cases, we may conclude that it does not. We should be temporally neutral in this limited sense. But you're quite right that in most cases it will be instrumentally significant, and the neutralist will not want to deny this. Often, we should care about differences in timing, because these go hand in hand with other differences which matter in themselves.
If I identify more strongly with near future selves than far future selves and thus would prefer to move pain to far future selves, is this still compatible with temporal neutrality? It seems like it should be, in that it's not the time of suffering that I care about but rather the emotional distance I have to the sufferer.
ReplyDeletePeople typically do value the near future much more than the far future. However if they knew that they were to be frozen and reanimated in 50 years, they would value that far future more. This implies it's not the absolute time that is the deciding factor.
Yeah, that sounds right. We may think that persistence through time is a matter of degree, so that closer future selves are "more me" than distant ones (but due to qualitative rather than temporal relations).
ReplyDeleteI don't mean to drag things on a tangent, but you say that you thought his theatre example was most compelling.
ReplyDeleteI personally found the My Past or Future Operations (p165) example far more compelling. Do you think this example is faulty somehow? It refers to preferences, not feelings.
I think that's a similar case. It refers to "preferences" only in the affective sense of being "relieved" to hear that the operation is past. If one could deliberately choose between the options (say by time travel, or praying to God, or whatever), then it strikes me as irrational to choose the option that is worse for you, even if the pain is in the past.
ReplyDelete(There is the added complication, as discussed above, that it is not always clear in these cases which pain is really worse for you. My earlier post discussed the second 'operation' case in more detail.)
Richard,
ReplyDeleteThank you—I am starting to see the appeal more of the timeless perspective, when before I could not understand why it would be appealing. I’m glad you’re interested in the narrative-style stuff. We’ll have to talk about that more. I still do remain unconvinced, though about T.N.
My worry is mainly an epistemic one, or more precisely, a worry about how knowledge of own our epistemic limitations bears on what it is rational to care about.
You give a case in which we take a kind of sub specie aeternitatis perspective on our own lives, viewing painful events in their rich contexts, such that we could compare a past pain to a future pain, each modified by its relevant context. In such a case, I would agree that considerations of temporality should not play a role in our decision-making: It doesn’t make sense here to prefer that a pain be in the past.
But I don’t think that this sort of thought experiment shows that a temporal bias is unjustified. Here’s why:
My worry is that that such a comparison will never be epistemtically possible from our perspective as temporally-situated agents; we simply can’t look at our lives from this perspective, since we do not know how the future will turn out. We have, one might say, an epistemic disability, in that we cannot know how future events might be redeemed by their context, whereas we can know that about past events.
Your suggestion appears to be, “Ok, discounting the disability and assuming full knowledge of how a future painful event is modified by a rich context, what would we do then?” But is this a fair move? I worry that it’s like telling an ugly person, “Suppose you don’t have this disability of being ugly. What would you do then? That’s what’s really best for you,” the result of which is that he asks a beautiful girl out, who spurns him cruelly. What it’s rational for him to do is not what he would do, were he not to have this disability. It’s what he would do given that he has this disability. I think that when we take our actual epistemic disabilities into account, it is justified to prefer that a pain be in our past.
I agree with Andrew.
ReplyDeleteIf for no other reason (and I think there are other reasons), I think we are justified in discounting future costs and benefits because of our compounding uncertainty about the future.
Assuming that uncertainty away may be interesting theoretically, but it doesn't seem to be of any benefit to real people making real decisions with real imperfect knowledge of the future
Sure, I grant the practical point; but I'm more interested in the theoretical fundamentals.
ReplyDeleteRichard,
ReplyDeleteI took "relieved" to be cognitive, not affective. I take Parfit to mean it in the sense of something along the lines of "seeing reason to not treat the past pain as being (as) bad". But I fear at a basic level, I simply don't share your intuitions on this case. If I personally could change that situation (by praying to God or whatever), I would be tempted to have the larger pain in my past instead of the smaller pain in the future.
Andrew and Gil,
Think of it this way - If Richard is wrong, and we should not be temporally neutral, then the following is true:
There isn't just reason to discount future pleasures and pains because they're less likely. There's also a further reason to discount them: simply because they occur in the future. That is, we ought to discount future pleasures and pains more than in mere correlation with probability.
That is the view Richard is arguing against. If that view were true, it would be a surprising and important result - for practical purposes too.
We all agree that you ought to discount future events in correlation to their probability. But should we discount them more than this?
Yes, I think we should discount them more.
ReplyDeleteOne reason, I think, is that the effects linger.
So, if I move a pleasure from time T2 to T1 (or a pain from T1 to T2) I am improving my life during the interval between T1 and T2.
One way this might happen is that my attitude is improved (or not diminished) during that time and I am thus in a better position to enjoy other pleasures, overcome other challenges, etc.
But that's merely another instrumental reason for discounting. We can imagine a case where it doesn't hold. The question is whether, all those contingencies aside, simply being located in the near future is itself an additional reason to favour an option, regardless of the benefits that might accrue to one's life as a whole (which can be appreciated from a timeless perspective, after all).
ReplyDeleteRichard,
ReplyDeleteIt sounds like you're saying something like: "If we ignore any reason why we should prefer A to B, is there any reason to prefer A to B?"
If you're not saying this, could you give me an idea of what an argument you would accept might look like?
Would it have to be something like a claim that the near future is always more important than the far future?
Would it be relevant to be more concerned with my near future self (because he's closer to being me) than my far future self?
Richard,
ReplyDeleteYou say: "Sure, I grant the practical point; but I'm more interested in the theoretical fundamentals."
I think the practical point is a deep one. In fact, I don't see the appeal of the theoretical project without annexing it to the practical project. What is the point of investigating those "theoretical fundamentals" supposed to be?
I take the T.N. theorist to be setting the stage for a claim that could in some sense be action-guiding, or perhaps more properly, attitude-shaping: It does not make sense, he wants us to think, to prefer that a given pain be in one's past to its being in one's future. If it is to have any point at all, it seems to me, his theory ought to be trying to refine the patterns of our care.
On your view, T.N. seems to be saying something like this: "Here's my theory of what a rational pattern of care is. Now, I grant that as an agent, acting with your epistemic limitations in the real world, it would never be practical for you to adopt it, since you can't ever know your future. But still, this is the rational pattern of care. After all, I showed it was so in my thought experiment!"
I guess I just don't see why the theoretical view would even be all that interesting, even purely theoretically. What's the point of having a theory of rational care that it would be foolish (dare I say irrational?) for any temporally-situated agent to adopt as a practical guide to care?
-Andrew
Gil - I would accept what Alex is trying to do, i.e. present a case where a mere difference in timing (taking care to hold fixed all other variables to avoid confounding) makes an intuitive difference as to which option is preferable.
ReplyDelete"Would it be relevant to be more concerned with my near future self (because he's closer to being me) than my far future self?"
Depends what you mean by 'closer'. If you mean that you will become him sooner, then yes, that's exactly the property I want to evaluate. If you mean because he is more similar to you, then no, that is not relevant, since it isn't anything intrinsically to do with time.
Andrew - I don't think philosophy needs defending. But I'm happy to share what it is that I personally find so engaging about such theoretical issues. First and foremost, there is the sheer wonder of studying the fundamental structure of (normative) reality. Aren't you just a little bit curious about what really matters, at the most fundamental level? I'm curious about whether temporal relations are among those things, or if their import is merely derivative/instrumental. This is an intrinsically fascinating question - it need not serve any practical "point". Such inquiry is the deepest kind of value; our other practices ought to be serving it!
Secondly, one could add that understanding fundamental reality tends to have practical benefits. In this case, we may note that the practically advisable level of discounting will be larger if there is intrinsic reason to discount in addition to all the instrumental ones. But I don't so much care about that. I want to know what's true, not just what's useful.
Richard,
ReplyDeleteI hope you're not tagging me with a form of pragmatism/ anti-philosophy, because that's not what I was intending to advance at all. Such a tack would be a way of circumventing argument, and that's not my interest here.
You say: "Aren't you just a little bit curious about what really matters, at the most fundamental level? I'm curious about whether temporal relations are among those things, or if their import is merely derivative/instrumental."
Of course. That question is a supremely interesting one. That's why I'm engaging in this debate with you. My concern is whether a thought experiment that discounts the epistemic limitations of actual agents (specifically their inability to know the future) is going to deliver a true answer about what really matters. I don't think it will; I think it will deliver an answer to the question about what would matter, if the future were determined and we were omniscient. But since we're not that way, I don't think it gives us a true answer to the question about what does matter. I think that question can only be answered by taking certain practical factors into consideration--most especially, our inability to know the future. My point was not that the overarching theoretical question itself was uninteresting, but that it becomes less interesting by divorcing itself from these practical factors. I suppose that the questions what matters given our limited human faculties and what would matter, were our faculties vastly different, have answers that are both part of the fabric of reality. I'm just much more drawn to the former theoretical question, both for its intrinsic interest and because I think it's the view that ought to be attitude-shaping (and hence the view that has practical relevance). But it is not as if my ultimate appeal is to a theory's "usefulness"; as you are, I am interested in its truth.
I remain unconvinced by your view about what really matters. But I am totally in agreement with you that it matters (above and beyond what's "useful"--I term I never used, by the way) to have a have a true view about what really matters.
-Andrew
Now I'm confused. You asked, "What's the point of having a theory of rational care that it would be foolish... for any temporally-situated agent to adopt as a practical guide to care?"
ReplyDeleteHow is this not pragmatism?
"My concern is whether a thought experiment that discounts the epistemic limitations of actual agents (specifically their inability to know the future) is going to deliver a true answer about what really matters."
How else can we distinguish intrinsic (fundamental) from instrumental (derivative) import?
My view is that - based on your earlier arguments - you should hold not that time fundamentally matters, but that narrative incorporation fundamentally matters (and that as a contingent matter timing will often affect this). Right?
There would be a few ways of construing pragmatism. One view, which I've never been able to make sense of, is that X is true if X works. That just seems crazy. Another view is don't worry about what's true and do what works. Rorty sometimes says things like this. (Although I am a fan of Rorty's, I like him qua other things, not qua pragmatist.) Maybe there's another construal of pragmatism that I'm missing, but at least on either of these views, I don't see myself as taking a pragmatist line.
ReplyDeleteI think I may see our disconnect: When I say "practical," you are perhaps taking that to mean pragmatism / useful. I didn't mean it in that sense at all. I was thinking more of the Critique of Practical Reason sense of "practical," i.e., reasons directed toward action. My worry was that the thought experiment would suggest results that illicitly undermine these practical reasons, and that our patterns of care would fail to be practical in that sense.
You say: "My view is that - based on your earlier arguments - you should hold not that time fundamentally matters, but that narrative incorporation fundamentally matters (and that as a contingent matter timing will often affect this). Right?"
--Yes, that seems right. You've put it much better (and more perspicuously) than I did! Although I am not sure whether I would go so far as to say that narrativity was fundamental exactly; I think it may just be a contigent fact about (many) modern humans that we care about it. But you're right that I am certainly not endorsing the view that a temporal bias is in itself of fundamental importance. Our need for narrative incorporation warrants this temporal bias. So were there no need for narrative incorporation, we might have reason to be temporally neutral. I agree with you there. (Was this your point?--I'm sorry if I have misunderstood it).
My worry was that this would end up being a Pyrrhic victory for the T.N. theorist. For from our perspective as practical agents (in the sense of acting/caring on the basis of reasons, not in the crass sense of just doing what works), given our epistemic limitations, and given that we care about the narrative shape of our lives (as I take it most of us do), we will always have reason for that temporal bias. Of course, one can explain away these contingent facts about us (that we have epistemic limations, care about narrativty, etc.). But I guess I would be suspicious that in doing so, a thought experiment really yielded up what would be MY rational patterns of care, or give ME a reason to adopt some alternative, temporally-neutral pattern of care. When I said that such a theory wouldn't be practical, I meant that it would not provide US (given the sort of beings we are) with genuine reasons for action/ care/ concern, etc. That was all. I am not disputing the theoretical point that if we abstracted from our circumstances we might have reason to be temporally neutral. So perhaps we are not in disagreement?
Thanks for carrying on the debate: I know this is well-trodden ground for you, and I am new to the territory and perhaps making quite elementary mistakes! (I have read Parfit on personal identity, but not so much on these issues.) But, in any event, I do appreciate your patience and indulgence.
-Andrew
Oh, no need for that - I've enjoyed the discussion too! You've shifted it in new and interesting directions.
ReplyDeleteIn light of your penultimate paragraph, I think we're actually in agreement on the original issue. I'm curious about your two senses of "practical", though. When will they come apart? How does timing provide us with reasons, if not that it is advantageous for us to attend to? Perhaps you merely mean that it is reliably correlated with other reason-giving factors (e.g. narrative incorporation). But that is different from having timing itself provide us with reasons.
Note that T.N. does not claim that time has no (even instrumental) significance. The claim is just that time has no fundamental significance; any import it has must be derived from other factors.
Perhaps you are a pragmatist in a third sense, such that you are interested only in what reasons we have, but not so much in the structure of these reasons -- whether they are instrumental or fundamental, etc.?