Monday, August 10, 2009

The Importance of Implications

A recent discussion of wellbeing over at Go Grue reminded me of how easily philosophical debates can degenerate into mere terminological disputes and "concealed tautologies" (as Sidgwick and Parfit warn against) if we're not careful. For example, a hedonist might be tempted to define 'wellbeing' as 'the felt quality of a life'. But then they are not actually saying anything of substance when they assert that 'wellbeing is the felt quality of a life'. Given what they mean by the former term, this claim becomes logically equivalent to the empty tautology, 'The felt quality of life is the felt quality of life,' which is hardly a claim worth making.

In order to make our affirmation of a particular theory of wellbeing substantive, we must have some independent grasp of the meaning of the term. For example, we might analyze S's interests in terms of what's desirable for S's sake. But even if we lack a full-blown analysis of the concept, we should at least have some sense of its conceptual role, i.e. what implications it has for other philosophically important concepts.

I think it's very dangerous to just test one's linguistic intuitions in isolation. Often, people seem be drawn to hedonic (or 'mental state') theories of wellbeing simply because they think it sounds wrong to say that someone was "harmed" by an event that they never learned of (or that otherwise impacted their subjective experiences). This does not strike me as a philosophically interesting or relevant consideration. Instead of assessing what combination of words sounds natural to our ears (as if we were mere grammarians), we should be assessing the real content of the claim, as revealed by its philosophical (e.g. normative) implications.

The concept of 'wellbeing' seems important only insofar as it has implications regarding what we have normative reason to want and to do when we care about a person (including ourselves). If we do not think that a loving parent would (qua loving parent; bracketing all other motivations) want to plug their child into a Nozickian experience machine to live a meaningless life of pleasant delusion, or that a prudent person would want this for themselves, then we shouldn't be hedonists about wellbeing. It's that simple.


  1. "If we do not think that a loving parent would (qua loving parent; bracketing all other motivations) want to plug their child into a Nozickian experience machine to live a meaningless life of pleasant delusion, or that a prudent person would want this for themselves, then we shouldn't be hedonists about wellbeing. It's that simple."

    Isn't that a little quick? I know a lot of ethicists might go for this, but a consequentialist? Try this:

    If we think that a loving parent, qua loving parent, wouldn't kill one of his children to harvest his organs for the sake of five of his other children (provided he wouldn't get caught, there wasn't another option, bla bla bla), then we shouldn't be utilitarians. It's that simple.

    We really need to know what other considerations are at stake and whether these judgments can be given satisfactory justifications before we make such conclusions. I'm not saying you've never done that, but it isn't that simple.

  2. I've always thought that, used in this context, "sounds wrong" sounds ambiguous. On the one hand, it could be a comment on the feel a particular expression or expressions give us. On the other, it could be used to mean something like "sounds in a way suggesting the falsehood of what is asserted".

    Moreover, it's unclear what weight the feel of the expression carries as far as our evaluation of its truth is concerned. If it is a good indicator that the assertion is false, then it's misleading to suggest that we should assess the content "instead of" assessing the feel .

  3. Pavel - perhaps I could say that we should assess the content by way of assessing its implications, rather than by way of assessing the feel (in isolation).

    The basic worry here is that most people don't really have a very firm grasp of what a word like 'harm' means. Maybe in ordinary language it's even ambiguous or indeterminate between multiple possible interpretations (some of which build in the requirement that a harm must be experienced -- many people report this linguistic intuition, at least). But there's only one sense of harm that is philosophically interesting, which is the interpretation I've pinned down here with implications for what's desirable for an individual's sake.

    So I want to encourage others to discuss this philosophically interesting question, rather than using 'harm' in some other sense that leads people to merely affirm concealed tautologies and other boring claims. By encouraging people to think about the implications of the term, we encourage them to resolve any ambiguity of content in favour of the more philosophically interesting interpretations.

  4. Nick - I'm not seeing the analogy. Obviously all sorts of other considerations come into play in the utilitarian case, since a loving parent will be reluctant to harm one of their children, even for the sake of a greater benefit to another. But there's nothing like that going on in my case. (If you disagree, I hope you'll explain further.)

    N.B. It doesn't seem part of the nature of (multiply-instantiated) love that you'll necessarily be motivated to trade off the interests of one loved one against another when appropriate. So the actions of a loving parent don't seem particularly relevant to assessing utilitarianism in the first place. Your imagined parallel argument is a non-starter.

    On the other hand, it does seem in the nature of love that one must desire the increased well-being of each loved one, considered individually. So if we don't think that love should move us to plug a child into an experience machine, even when all else is equal (i.e. we are only concerned with the interests of this one child, and no other considerations are in play), then that would seem to pretty conclusively establish that the interests of the child aren't served by plugging them into the machine.

    If you think that some other considerations are somehow clouding this judgment, you'll have to say what they are. But I haven't seen any reason yet to doubt that this is a perfectly straightforward case.

  5. I didn't mean anything very sophisticated by my analogy. Here's all I was really going for. Although it is true that mental state welfarism has this deeply counterintuitive consequence in the experience machine case, it would be a mistake to say that it must be rejected on that ground, without carefully considering arguments in its favor. This is much as it is with utilitarianism and the transplant objection. Just as you shouldn't say "no one believes organ harvesting is okay, so utilitarianism is not true", you shouldn't say "no one believes living in experience machines is good for people so mental state welfarism is not true". Things are not that simple. That's all I was going for.

    I'll pass on advancing arguments for mental state welfarism here. Suffice it to say that they exist and deserve careful treatment. Perhaps you just think that the arguments for mental state welfarism are just terrible, and don't require response?

  6. I see what you mean now; you've misunderstood me. You're worried about mere prima facie intuitions, whereas I'm appealing to considered judgments here. That is, my target is the "hedonist" who doesn't take their view to have any implications about what people have reason to want or to do. Hence my claim: if you don't think that caring about someone requires you to plug them into an experience machine (all else equal), then you shouldn't be a hedonist.

    I mean the whole conditional to be the "simple" claim. Even a hedonist could endorse that, if they reject the antecedent. But (absent some super-compelling argument of which I'm unaware) most people probably won't want to do that. So they shouldn't find hedonism so appealing.

  7. (Incidentally, I think this whole "won't somebody think of the implications?" approach is pretty vital for utilitarians. As you note, utilitarianism sounds crazy when you're just asked to make superficial judgments about what seems "okay" or not. But again, that's partly a linguistic thing; once we start thinking about the fundamental philosophical implications, it doesn't seem nearly so crazy to say that we have more reason to kill one to save five, etc., while noting that for familiar 'indirect utilitarian' reasons the ideal code for society would not recommend such a decision procedure, etc. etc.)

  8. Do you find the idea that we are living in a simulation as uncomfortable as the idea of being in Nozick's machine? It strikes me that the reason we tend to reject the experience machine is that pleasure alone isn't sufficient to have a "good" life. But we can still reject that proposition and maintain that a person's wellbeing depends on their experiences.

    I have trouble understanding how you could have a reason to want something for someone who can't experience anything. Why not also be concerned with the harm that comes to fictional characters or people who never existed? If experience isn't a relevant characteristic of harm, then how could you justify only being concerned with people who actually existed?

  9. Hi Eric - note that experiences aren't the only things that non-existent people lack! (They also lack desires, goals, achievements and, you know, lives.) No harm comes to fictional characters; only fictional harm. So your concluding questions are senseless.

    "I have trouble understanding how you could have a reason to want something for someone who can't experience anything."

    That not what I claimed. The dispute is about what we have reason to want, not who we have reason to want it for. I agree that non-conscious zombies aren't subjects of welfare (because I think only conscious beings can have mental states - incl. genuine desires or goals - at all). But I think we should want more for them than just certain kinds of experiences. We should want for people to achieve their goals in fact.

    "Do you find the idea that we are living in a simulation as uncomfortable as the idea of being in Nozick's machine?"

    I'm not sure what you're getting at here. (Nozick's machine offers whatever experiences you like, not just pleasant ones. It still sucks.) I think it matters that we're interacting with other real people; I don't think it matters what the medium is, so a shared simulation would be fine.

    P.S. Everyone agrees that wellbeing depends in part on the quality of one's experiences. (That's one thing we care about for a person's sake.) The question is whether it depends on other (more objective) qualities of one's life in addition.

  10. Richard,

    Thanks for the clarification! I need to find some way of familiarizing myself with this terminology without going through an undergrad course in philosophy.

    I am still skeptical that there are non-experiential qualities that are important for wellbeing, but I think I understand the gist of your position now.

  11. First, I'm not convinced that the concept of well-being is only important in terms of its normative implications. It partly implies descriptive facts about the world that might be of interest in their own right.

    Second, just as some hedonists build their view into the meaning of the term, you're doing the same. You write that well-being is:
    "what's desirable for an individual's sake."
    "what we have normative reason to want and to do when we care about a person"

    The most natural way to read these claims is that you've made well-being parasitic on normative facts about other peoples' desires. That looks pretty weird to me, since one would think that the explanation goes the other way. Regardless, it looks like a substantive claim. Perhaps caring about someone only involves attending to certain aspects of their overall well-being, for instance. (Perhaps being independent is an aspect of well-being, and this is stifled by people caring about you too much.) I don't think this is true, but I don't think it's incoherent, which it would be if your definition were correct.

    Or one could read your definitions in another way, a way that makes them completely uninformative, and leaves just as much room for the hedonist to read in their definition. What is someone's "sake"? What is it to care "about" someone? It's tempting to understand someone's sake as their well-being, and to care about someone is to care about their well-being. But if your definition is this uninformative, it doesn't prevent the hedonist from pulling the same move again. They can claim that "someone's 'sake'" just means "their felt quality of life".

    I don't think hedonists understand "well-being" to mean "felt quality of life" merely as a way to cheat in arguments. It's because it's not clear that there is any way to define it that is informative, theory neutral, and not dependent on linguistic intuitions. That project seems to me to be doomed from the start. (I don't think this is merely a fact about the concept of well-being in particular: the same seems to apply more generally.)

    (I believe Guy Fletcher has just submitted a thesis on what well-being is. As far as I know, this is an underexplored topic: most writers just give a list of things that contribute to well-being, but don't tell you what the thing is to which they contribute.)

  12. Alex, I'm not following you.

    The whole point of my post is that there's a difference between substantive normative claims and mere stipulations.

    If a theorist is "read[ing] in their definition", or claiming that some part of the LHS just means the RHS, then they are affirming a concealed tautology, and so saying nothing at all. There is literally no content to the claim in question. They might as well say "pleasant experiences are pleasant experiences".

    But if we have an independent notion of reasons/desirability, then we can make substantive (non-tautologous) claims about what things we have reasons to want for a person. The hedonist will answer that we ought to want pleasant experiences for them; I will add further things in addition; but in any case, this is a substantive answer to the question, and not simply a stipulation of its meaning.

    Since my basic analysis of 'welfare' is perfectly compatible with any substantive view about welfare (including hedonism), it clearly does not "build [my] view into the meaning of the term". It simply serves to make it clearer what is at stake. Welfare is a kind of value, and value is (or, even for Mooreans, at least has implications for) desirability. Even hedonists should agree with this (or something like this), if they want their affirmation of hedonism to actually be saying anything at all.

    P.S. What descriptive facts are in dispute when ethicists argue about whether hedonism is true?

  13. Sorry Richard, I'll try again. Perhaps I've misunderstood what you're trying to do.

    The general point is the one in the final paragraph: You are trying to give an account of the conceptual role of well-being which meets three criteria: (a) It is not based merely on what "sounds wrong" (i.e. linguistic intuition), (b) it is not already a move towards any substantive theory, and (c) it must say something, otherwise there's no point in doing this at all. But I don't think that anything could possibly meet all three criteria.

    So at least one of those criteria ought to be relaxed. I'm not sure which one. But B is a possibility. We might think that some concealed tautologies are nonetheless informative. (That just is the paradox of analysis.)

    And I think this prolem is demonstrated by your own effort. By defining "X's welfare" as "desirable for X's sake", you have two serious problems.

    First, your claim is substantive because some people do not think that there is a straightforward link between what is desirable and what is valuable. Perhaps, for instance, desirability is a matter of reasons, and reasons do not always straightforwardly link up to values. So some would deny that everything that makes someone well-off is something that is desirable. For example, perhaps it is not desirable that wrongdoers are happy, even though this definitely contributes to their well-being.

    Perhaps you could ammend the view in this respect. Perhaps you could instead say that "X's welfare" means "What is valuable for X's sake". But that is itself totally uninformative. What is someone's "sake"? My grip on "sake" is worse than my grip on "welfare", and the only way I know how to think of the former is in terms of the latter. But then what we have is "X's welfare" means "what is valuable in terms of X's welfare", and that's not what you were after at all.

    As I say, perhaps I've misunderstood what you're up to. But *if* you are trying to explain the conceptual role of well-being in a non-linguistic, theory-neutral and interesting way, I don't think it's possible.

    (On the p.s., I should have phrased that a little but more tentatively, but I had in mind things like "Jane committed suicide because her life was going so badly". The truth of that descriptive claim might depend on what theory of well-being is true.)

  14. Hi Alex, I think I see your worry now. We need to be careful with the informativeness [c] criterion. When one offers a theory of welfare, or makes claims about what things are harms or benefits, this is clearly aimed at being substantive, and hence ought to "say something" non-trivial.

    But it's perfectly okay to give a trivial conceptual analysis beforehand, insofar as this will help make clearer what you are claiming when you go on to make your substantive claims about welfare.

    You may complain that a theory-neutral conceptual analysis won't be "interesting" to anyone who's already thinking clearly. But the problem is precisely that, in philosophy, people often aren't thinking particularly clearly. So it sometimes helps to state the obvious.

    (For example, by rewording our ordinary notion of welfare in terms of "what's desirable for a person's sake", we can remind people of the philosophically important aspects of the concept, and hence encourage them to hold these implications in mind when formulating their substantive claims.)

  15. Richard, I think we're getting there, but I'm not sure you've quite seen the force of the worry.

    My point isn't just that anything that is theory-neutral will be uninformative.

    It's also partly that the explanation you give of the conceptual role of well-being had better be something that we can grasp without recourse to the concept of well-being itself. Otherwise it's not merely trivial to give such an anaylsis, it's completely pointless. I think that your view fails on this count, since "someone's sake" is completely opaque unless we make sense of it in relation to well-being.

    (I presume the point about desirability was clear: Saying that welfare is "what's desirable..." is not merely to "remind" someone of something they already knew. It's a substantive claim, one that some, me included, are inclined to think is false.)

  16. No, I've explained elsewhere that being 'desirable for the sake of S' does not mean being desirable, tout court. So it's perfectly compatible with thinking that vicious people ought to suffer.

    Anyway, I'm not too concerned about my particular proposal here. (Another possibility would be to understand the well-off life as one that is "desirable" in the sense that holding fixed what lives there are, you have [prudential] reason to want the high-welfare life to be yours rather than someone else's.)

    My point is just that an interesting notion of welfare had better be connected in some such way to reasons and desirability (of a sort, or in certain circumstances). This couldn't possibly be "completely pointless", because you'll notice I was responding to someone who failed to acknowledge even this much. He really did seem to be talking about 'wellbeing' in a way that had no normative implications whatsoever -- just asserting hedonism as a concealed tautology. If a little bit of (even "opaque" or circular) rewording will help people avoid this black fate, then that's "point" enough for me.

  17. Final go, then I'll perhaps post on this at my place sometime.

    I repeat that the basic point about desirability is this: Well-being is a kind of value, and since its an open question as to just how desirability relates to the evaluative and the deontic, it's equally an open question as to how well-being relates to desirability. If something is desirable only if there's some reason to bring it about, then the well-being of wrong-doers may not be desirable *at all* (by which I don't merely mean that it may not be desirable all-things-considered/tout court).

    And this is relevant on the broader point, since it's perfectly coherent to think that well-being doesn't relate in *any way* to reasons or desirability. Perhaps it's a kind of value that fails to have any deontic implications. That is wildly implausible, but not incoherent.

    You might say: "Fine, but at least we have to bear in mind that well-being has normative implications, even if it's an open question as to whether or not it has deontic implications." That seems correct, except that having normative implications is only necessary, not sufficient, for some concept being the concept of well-being. And you're getting the sufficient condition only by bringing in the notion of a "sake", which hinders more than it helps. I have no idea what that word means except a slightly fancy word for well-being. Someone confused about what an analysis of well-being was an analysis of would not be helped by talking about "sakes" instead. So it seems to me, at any rate.

  18. Yes, again, it's perfectly compatible with my view to say that you have (in the actual circumstances) no reason at all to desire the welfare of the vicious. Seriously. Follow the link.

    For the record, it does strike me as completely incoherent to think that some genuine (kind of) value might not be related in any way to (some kind of) desirability. Until they explained more, I would no longer have the faintest clue what such a person meant by "value".

    You seem to be getting stuck on the idea that one particular kind of value might not correspond to one (different) kind of desirability, but of course that's precisely what we'd expect given that there are various kinds of both. It's a 1:1 correspondence, not an all:all correspondence. (Note that one common way to express the idea that we don't have reason to desire the welfare of wrongdoers is to say that the welfare of wrongdoers is no longer of [impersonal] value. We presumably shouldn't conclude from this that welfare has no implications for value.)

    But this is all a bit besides the point. I don't doubt that someone completely clueless about wellbeing would remain so even after I tried to cue their responses with talk of "sakes" and such. But they are not my target audience here. So you're bringing up a completely different dialectic.

    Again, my target is the person who [has the capacity to grasp the concept of wellbeing perfectly well, but] unreflectively settled on a more or less stipulated theory of wellbeing by consulting their linguisitic intuitions. And my aim is to get them to think more about the substantive philosophical implications of the various theories instead. In order to do this, I'm presupposing that on some level they already understand perfectly well what all my talk of "what's desirable for a person's sake" (i.e. wellbeing) is about. I just think that wording it like this can help jolt their memory, by making especially salient some of the relevant normative implications that they'd previously been neglecting.

    Bearing this dialectic in mind, I could see relevant objections being directed at either my goal or my methods. For example, do you think my target here is better off sticking with their concealed tautologies? (That seems crazy.) Or, granting that they'd do better to attend to the philosophical implications and make substantive claims, do you have a pragmatic suggestion for how to better elicit such a response?


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.