Tuesday, September 21, 2010

Natural Arbitrariness

One intuitive basis for considering Future Tuesday Indifference to be irrational is that it involves drawing "arbitrary distinctions" within one's preference set. (I take this sort of line to be most famously associated with Michael Smith.) Derek Parfit, however, denies that this line is available to proponents of a merely 'procedural' conception of rationality. Instead, he claims:
Our preferences draw arbitrary distinctions when, and because, what we prefer is in no way preferable. It is arbitrary to prefer one of two things if there are no facts about these things that give us any reason to have this preference. [On What Matters, Chp 3.]

In order to explain why basing a meta-hedonic preference on the day of the week is arbitrary, whereas basing it on the felt quality of the experience is not arbitrary, Parfit thinks we must appeal to the substantive normative fact that only differences of the latter sort are reason-giving. I think this is not quite right.

We can specify a non-normative sense of 'arbitrariness' in terms of how differences in one's preferences match up with differences in natural properties. Since "being Tuesday" is a merely conventional (rather than natural) property, there's a straightforward sense in which the Future Tuesday Indifferent agent is being arbitrary: they are not treating (naturally) like cases (normatively) alike.

Still, I don't think this non-normative sense of arbitrariness is enough to get the proceduralist out of trouble. For we can imagine a variation on the case where, instead of their preferences featuring Tuesday (as such), the agent's degree of meta-hedonic concern varies according to the waxing and waning of the moon (or some similarly natural variation). Suppose that they have full ordinary concern about future pains experienced when the moon is full, but only half as much concern for agony felt under a half-moon, and no concern at all to avoid agony felt under a new moon (regardless of whether they'd recognize it as such at the time). In this case, there doesn't seem to be any straightforwardly procedural grounds for criticizing the agent -- their preferences are not arbitrary in the non-normative sense -- but it still seems objectively crazy.


  1. Your moon case shows that desires can track the natural facts and be crazy. It's equally true that tracking some conventional facts is not crazy: My desire to shake hands with my *right* hand is based only on conventional facts, but the desire seems none the worse for it. As you suggest, the natural/conventional distinction just seems totally independent of the rationality of desires.

  2. Well, it would be crazy if you had an ultimate (non-instrumental) desire to shake hands with your right hand.

  3. Richard,

    I largely agree with you, but I think there is something more to arbitrariness. You say that FTI is a failure to consider “(naturally) like cases (normatively) alike.” But that understanding still requires some idea of what kinds of natural similarities are relevant and irrelevant. No two events are naturally identical if they occur at different times or in different spaces. They can be similar in the sense of lots of other properties, but there will always be differences, including differences in space and time. The waning of the moon is one example, as is the pinkness of chalk (as I tried to argue in seminar yesterday). So we aren’t looking for identical natural facts; we are looking for similar natural facts, and even “similarity” is insufficient. We need them to be similar in the relevant ways.

    The challenge is for the Subjectivist to explain the idea of relevant similarity without reliance on object-given reasons. Here are a few ways in which the Subjectivist might proceed:

    1) The Subjectivist could admit that there is no such thing as all-out arbitrariness, but only arbitrariness from some point of view or in terms of some set of rules. Future Half-Moon Indifference (with Present Half-Moon Horror) might be prudentially arbitrary because it fails to consider our future telic desires, of which prudence requires consideration.

    2) The Subjectivist might use an informed-desire view to isolate the relevant kinds of properties. For meta-hedonic desires, these properties will turn out to be hedonic likings and dislikings for any rational, self-interested agent. In terms of Future Half-Moon Indifference, we might restrict our knowledge of future moon stages, say, by living underground. Upon learning that the pain we experienced on some evening occurred on a half-moon, our rational response would be to admit that the moon stage is not a relevant property in our desires about pain. This is not because we have object-given reasons to have the meta-hedonic aversion to pain at all times, but rather because our meta-hedonic desires track hedonic likings and dislikings because those psychological facts provide motivation. I suppose this is a matter of reflective rejection.

    3) The Subjectivist might say a distinction is arbitrary if there is no further explanation for them. A legitimate explanation should be comprised of non-disjunctive, general rules. I think this thought might captures the intuitive understanding of arbitrariness as drawing distinctions based on a whim. So it should pick out a coherent property about the case, and we should have the same reaction to that property in cases other than that particular one. It seems that when we have this kind of explanation, our distinction is non-arbitrary. So there has to be something about future half-moons that affects our desires other than the meta-hedonic aversion to pain. That property could not be half-moonness itself if we have Present Half-Moon Horror, because we would then be making an exception to the non-disjunctive, general rule regarding half-moonness. It would have to be something about the future in general, but then our preferences on future full moons would be irrational, because we would be making exceptions to the non-disjunctive, general rule regarding future pain.

  4. Hi Jake, can you say a bit more about how you're imagining the subjectivist could justify rejecting as (necessarily) 'arbitrary' a meta-hedonic desire set according to which one's strength of aversion to an episode of pain is consistently proportional to the fullness (or lack thereof) of the moon at that point of the lunar cycle?

    (1) "Relative" arbitrariness doesn't count. Sure, it's arbitrary relative to the standards of temporally neutral prudence. But, equally, temporal neutrality would qualify as 'arbitrary' relative to the rules of lunarcy (excuse the pun). This does nothing to break the normative symmetry that is intuitively so objectionable about subjectivism.

    (2) Why think that (merely procedural) "informed reflection" will necessarily lead to ordinary human-like desires? It's true that we don't consider lunar properties relevant to our meta-hedonic desires. But another agent might -- that's the whole point of the example. If you restrict their knowledge of the lunar cycle, the lunartic agent will (between screams) patiently explain to you that they are not in a position to assess the degree to which their current agony violates their reflective preferences -- it essentially depends on this extrinsic relation, after all.

    (3) This anti-disjunctive move is closest to my proposed understanding of arbitrariness. It doesn't seem to rule out the consistent lunar-tic.


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.)