In other words, the typology of “causes” reveals that two people can choose the same vacation (or whatever) but for different reasons. Ah, but this isn’t published…. Erm, to resort to the place of last resort: doesn’t Distinction include a discussion of how upper and middle [sic] class Frenchies interpret photographs differently? I seem to remember that the photographs of potholes and garbage were “ugly” to the working class folks, and “beautiful” to the elites.

Sorry I’m incoherent. More when I can think straight.

As far as I can tell from rational choice literature, the distinction they draw on is fuzzy, and seems to be along the lines that preferences are very close to choice itself (say, a preference for _Adaptation_ versus _Atonement_) and tastes are closer to commodities that one can think of as being terms in a utility function (say, a taste for quirky). But this seems extremely loose as far as I can tell and I’m not sure I have it right. More to the point, it’s clear that from the different posts above that these terms have really different meanings to different people, so my confusion about trying to talk them in an essay seems about right.

Unfortunately for my essay I only have 7500 or so words. I’m caught up mostly on the question of how you can talk about preferences as an explanation for action if you don’t think preferences actually characterize something specifically real inside the head. That is, there’s no reason to think that two people who have the same preference have something similar going on in their brains, beyond just that this holistic characterization can be made. This gets you very close to a position that preferences are just that which is ‘revealed’ in choices, but if you go there, then you mostly lose the prospect of preferences being explanatory, at least if by explanatory we mean by reference to causes. Instead, Gintis has a recent essay in which he says basically that preferences (along with beliefs and constraints) are an apparatus for re-describing choices; to me that position is a dead-end for explanation, since redescription may be useful but it isn’t explanation.

I’m with you, though, that the prefs-as-contingent and taste-as-fixed is kind of shaky, theoretically. I also think, with regards preferences, it runs the dramatic risk of being tautological. Assumedly, preferences are ‘revealed’ through action – a quick econlit search shows recent research like Udo Ebert’s “Revealed Preference and Household Production” (which abstract I’ve reproduced below). Since the strong assumption is that preferences lead to action, if we can determine household production functions, and they are consistent, then whammo – we can go backwards to derive the preference function.

See how nicely that works? We can’t know preferences, but we can know outcomes (what people spend money on). Bit if the outcomes are stable and consistent, we can assume that they reflect preferences.

So IMHO, we’re having two conversations here. The first one is the one that Jeremy left us with – the role of preferences in determining social action. My guess is that this is meant as an address to the economic assumption that straightforwardly, preferences->action. The second conversation is more derived from the Bourdieu-like concerns over where taste comes from. And in that conversation, there is lots of room for debate still, about structural and situational ways that preferantastes come about and affect cultural capital.

My own interest is in the second with regards my research on art markets and whatnot (but really I would love to knock out of the box the assumption that preferences are stable and lead to action).

Abstract from Ebert’s “Revealed Preference and Household Production”. March 2007 Journal of Environmental Economics and Management.

The paper deals with the possibilities of recovering the underlying preference ordering from observed behavior when nonmarket goods are employed in household production. The problem is relevant for the evaluation of environmental goods and for the measurement of welfare in environmental policy. It is shown that preferences can be recovered if and only if a corresponding (mixed) demand system can be integrated. This system can be derived from observable behavior and the household production functions imposed. The conditions for its integrability are presented and can be checked. Therefore the approach suggested is operational and allows to decide whether the behavior observed and the household production functions chosen (as maintained hypothesis) are consistent. This result is important since the evaluation of nonmarket goods in this framework crucially depends on the choice of the household production functions.

I think that’s right, and perhaps they are not so different. I think there’s something to be found in the fact that taste tends to emerge with respect to cultural choices – taste in food, art, literature, music, fashion, dress, architecture, and perhaps people (friendship and sexual partners). It’s still easy to evaluate, and hence moralize, tastes though I don’t think we put much stock on basing morality on a ‘faculty’ of taste. With respect to preferences, these are connected to a broader range of situations involving social choice, not just cultural ones. A preference simply implies that we have the capacity to observe ourselves making meaningful choices for not entirely arbitrary reasons.

I’m a little resistant – though perhaps for no good reason – to think of tastes as fixed and preferences as flexible, though I understand the argument. How do we account for changes and shifts in tastes, which I do think happen? Of course, acknowledging shifting or changing tastes doesn’t necessarily sever the connection between taste and social class since, as in the case of fashion, one of the markers of ‘appropriate’ behavior for certain groups is not to stick with your tastes for too long lest you be seen as passe, out of touch. Still, if we think of changes in taste as a driving motor for changes in preferences at point-of-sale (e.g., as when we buy music), we’d need to account for how taste contributes to contingency and fixity at the same time rather than making tastes more fixed and preferences more contingent. What I want today is not necessarily what I wanted yesterday, and this isn’t because I’m just willy-nilly inconsistent, but because I’ve developed ‘better’ or ‘different’ tastes. Taste allows you to claim a ‘learning process’ for your preferences, perhaps – whereas preferences, because they are highly situational, just change depending on context.

Well, maybe this is wrong. I don’t know – hard one!

Just a summary of my take, which actually changes every time I think about the two concepts…I just had a feeling that there’s a kind of first-order/second-order differentiation between preferences and tastes. Not to say that preferences aren’t ’socially constructed,’ but out of a concept of taste we can generate (using familiar distinctions such as good/bad, strong/weak, refined/unrefined, idiosyncratic/usual, etc.) certain kinds second-order descriptions of social behavior that we can’t do as readily with preferences. Perhaps, glibly, preferences lend themselves to aggregation and tastes to evaluation and discrimination? In any event, tastes seemed to me a kind of stylization of preferences, a linking of preferences into patterns or as JLena says cultivated habits, in ways that allow choices to become socially-recognizable statements of a kind: my preferences express who I am, but don’t determine the whole of me (since any combination of links is contingent). That allows us to find one another appealing/disgusting, intriguing/boring, surprising/ordinary. Maybe we can still arrive at these kinds of feelings and attitudes with the concept of preferences, but it still seems easier to reach at them through a concept of taste.

http://unionstreet.wordpress.com/2007/12/12/taste-for-taste/

The impetus for our conversation was that preferences are assumed to be so fixed, but the content of those preferences might matter a lot. Theoretically, can I today have a preference for Chinese food, and tomorrow a preference for Italian food? Or do my preferences have to be fixed and fixated on one or the other? And if I have a preference for heterogeneity (this was the issue – that Amazon always assumes I’m going to continue to want/be more of what I already have/am), how might the preference->action link look differently?

These might all be settled issues, and maybe Jeremy would know – or maybe you can call these preferences or tastes and it doesn’t matter much..