The residual philosophy undergrad in me had a minor fit of Ed Sullivan Show–style giddiness at the observation that Tim Scanlon popped up in the comments to my post on happiness (in response to Brad DeLong) below. He made essentially the same point earlier at DeLong’s site, and Will’s obligingly extracted it from the thread there.
Following Scanlon, we can break down the debate over whether “some things are more important than happiness” into two basic areas, and in each we’ve got, as per the last post, (at least) three ways of reading “happiness.” The senses of happiness, implicit in that earlier post, are:
- Hedonic Happiness (HH): Happiness defined as a kind of brute pleasantness of mental states. I achieve Hedonic Happiness by hooking myself up to the Bliss Machine (a thought Nozick entertains before elaborating the better known Experience Machine) and spending my life in some sort of perpetual orgasm.
- Experiential Happiness (EH): Happiness as the subjective sense of affirmation brought about by the belief that one’s desires have been satisfied: I believe that having (or doing) X is worthwhile, and believe that I have (or have done) X. I may increase my experiential happiness by choosing a difficult life as a novelist or painter over that of a feckless playboy, even if (assuming the joys of Borneo distract me from whatever pangs of regret I might feel) the latter life is more joyful, more filled with HH.
- Preference-Realization Happiness (PRH): Happiness defined as that which obtains when which my desires (perhaps even desires regarding events to occur after my death) are actually satisfied in the world, independent of my beliefs about whether that obtains or not. From a first-personal perspective [II(a,b)], of course, PRH and EH will seldom be distinguishable, barring strange sci-fi thought experiments involving Experience Machines or amnesia pills. We generally achieve EH (the satisfaction of having written a good novel) in pursuit of PRH (actually having written that novel). Given the colloquial meaning of “happiness” as something individuals feel, PRH is unusual insofar as it’s realized by states of the world, not subjective mental states.
I entertained some doubt as to whether it really made sense to split out EH as a separate sort of happiness. Isn’t it (one might ask) that we have aims and preferences, whose realization achieves PRH, and that fact is mediated (possibly wrongly) by our beliefs and inferences, bringing about HH? I may feel elation at having done (or thinking I have done) a good painting, but the feeling of satisfaction there is simple HH, isn’t it?
It’s true that EH will often be associated with and hard to distinguish from HH, insofar as it feels nice when we accomplish or experience something we’d wanted to. (It may be a terminological mistfortune, incedentally, that we talk of “preference satisfaction,” in the sense of a set of conditions being met, which drifts a little too easily into talk of “satisfaction” in the Rolling Stones’ sense.) First, as in the example above, someone may knowingly eschew a higher attainable level of HH in pursuit of EH. And it’s tempting to say that they’re just aiming at PRH, so two categories will do. But I think it’s worth carving out as a separate category the cognitive state of approval that triggers the sense of HH—a kind of qualia-free sentiment of approval or affirmation; the belief-plus-normative-assessment package that elicits the emotional buzz. If someone wants to object that the emotional buzz and the cognitive package can’t be wholly disentangled (cf. Nussbaum, for instance: certain qualitatively similar emotional states are distinguishable only by their cognitive content) that’s probably a further point in favor of breaking out the middle category.
There’s the added wrinkle of adjusting for the way preferences change over time. For HH, of course, there’s no trouble: For every given moment (in princple) you can just imagine assigning a value to how “good” someone feels. (This assumes intrapersonal utility comparisons are possible, about which I’m not wholly sure. I may have some mechanism to choose between a night out with the fellas and a night at home reading King Lear; I’m not sure it follows as straightforwardly as is sometimes asserted that I therefore am reducing them to some common denominator of happiness, as opposed to “utility” serving, as David Gauthier insisted, as a theoretical placeholder to signify preference balancings that lacked any independent explanatory value.) It’s trickier with preferences, since we can distinguish (1) the intensity of my desire to bring about X, (2) the strength of my EH, maybe distinguishable from HH, maybe not so easily, when X comes about, and (3) my later view of X.
Imagine, for instance, that religious conviction impels me to go do missionary work, which I view as highly worthy. You look at (1) when seeking to account for my actual decision—note that the actual mechanism here, as opposed to the theoretical shorthand, may not look remotely like a calculation of “expected utility” in any subjective sense, even in the HH-less affirmational sense of EH. Perhaps for some years, I feel strong EH/HH satisfaction at the good work I’ve done. Then, however, I come to believe that my own prior religious convictions are in error, and am dismayed at having wasted so much time on a project I now regard as of little worth. How should we then evaluate, retrospectively, how well my life has gone? Parfit has a good discussion of this I’ll need to return to at some point when I’ve more time, but leave such complications aside for the moment, and let’s review the two areas in which evaluation of happiness will play a role, as per Scanlon:
- I — The personal sphere: (a) Is happiness the most (or the only) important thing for an individual to concern himself with or value or care about? And the related psychological question (b) Is our own happiness the most important or only thing we’re motivated by?
- II — The political sphere: (a) Is the happiness of the persons who constitute a society (or the world) the most important (or only) basis for the formation of just institutions? (b) If it is (in one sense or another of “happiness”) is the concern for happiness to be reflected in a straightforward maximizing way—such that just institutions are those designed to achieve “the greatest happiness of the greatest number”—or are there other constraints that demand a different structure, such as rules designed to create boundaries within which people may pursue happiness, without there being any thought of maximizing “net” or “social” happiness via interpersonal tradeoffs. Or might a just system somehow blend those structures, sometimes enforcing side-constraints, sometimes maximizing.
Note that it’s not necessary that a negative answer for some given conception of “happiness” to I(a) or (especially) I(b) entails a negative answer to II(a). Perhaps [I(b)] you do in fact care that your wife really loves you, rather than that you feel as though she does, or that the novel you’re working on is important and well written, rather than that you merely have that impression. And perhaps [I(a)] you are even right (or, weaker, not irrational) to have that pattern of concern. And yet perhaps, for all that, the god who loves you would give you the false feeling, were the genuine article not attainable, even though you would not have chosen that for yourself if you could somehow create a convincing illusion and then forget you’d done it. Perhaps, in other words, a just state would plug in to the Experience Machine even those who would not opt in, even if once connected they’d forget they were in it.
Now, to return to the discussion from that previous post: Someone might say that [I(b)] there are things she cares about more than happiness, and/or that [I(a)] some things just are more important than happiness. By this, she might simply mean that EH is more important than HH, which I think is the sense in which Will intended it when he started this debate, or (in contemplation of some Experience Machine–like scenario) that PRH is more important than either EH or HH.
Let “A>B” stand for the relation “A is more important than B,” and let H refer to Happiness (particular conception unspecified). DeLong’s Posit (DLP) is that it’s incoherent or tautologically false or otherwise a misuse of language to assert that there exists an X such that X>H. DLP is, I think, fairly clearly wrong as applied to X>HH as evidenced by the various counterexamples above—at minimum for X=PRH or X=EH (or some subset thereof). It may turn out, on reflection, that we think nothing is actually more important than HH, but this would be a substantive conclusion, not a logical or grammatical one. (Also, as concerns I(a), a fully rational person who reached this conclusion would presumably make PRH=EH=HH.) Even if X>HH for some X happens to be wrong, it is at least not self-contradictory.
I think further, though DeLong may disagree, that DLP is wrong for X>EH, where X is some subset of PRH, and that the Experience Machine is a fairly decisive argument for this. So the best reading of DLP is as applied to X>PRH. The thought here is that this generates a contradiction, because any candiate for X can be folded in to PRH. (For a fully self interested person, PRH may just be EH or HH.) In other words, for all X, X is a proper subset of PRH. In that case, X>PRH is tautologically false.
But note that this applies only to I(a), and for pretty much the same reason that there’s a sort of contradiction (though not a strictly logical one) in asserting “X is false, and I believe X.” For the purposes of II(a), from the perspective of a theorist considering the requirements of justice (or from that of “the god who loves you” for an individual), some things might well be more important than any of EH, HH, or PRH. For instance, we might hold what Derek Parfit (and Scanlon) calls an “Objective List Theory” of what makes ones life go well, such that it’s objectively better for certain Xes to be realized, whether or not anyone in the society cares about them (EH/PRH) or is made to feel good by them (HH). For the purposes of II(b), there might be values that don’t trump H in the sense of being alternative goods in competition with H, but nevertheless structure the way H is to be realized. (Think of the way Rawls conceived of the value of justice and the priority of the right over the good.) The only social desideratum from the perspective of justice might be some H, but a principle of fairness might constrain the way H is allocated, such that it’s not required to feed everyone to the Utility Monster, if there are any in the society.
To bring this all back around to DLP: There is a sense in which it’s tautological that “nothing is more important than happiness.” But, as is usual with tautologies, I think it’s a fairly limited and uninteresting sense. All it really means is that what people value most is what they value most, totally independently of any consideration of how much subjective pleasure they get out of those things. This provides no guidance for I(a) since, as Scanlon notes, we do not decide to pursue X because it is a value to us (PRH/EH); it becomes a value to us because we’ve decided for independent reasons to pursue it. And it also provides no guidance for II(a,b). DLP, in other words, is not useful for any of the purposes Layard (critiqued by Will in that original post) was after when he asserted the supreme importance of happiness. As usual, purely lingustic truths make ingredients for weak normative stew.