Pleasure itself is good. In particular, (a) if X has a pleasant or enjoyable experience, this is inherently good for X in that it is part of (not just something that promotes) X’s well-being and (b) X’s pleasant or enjoyable experience is inherently or non-instrumentally good or valuable to X (the experience is non-instrumentally agent-relatively valuable). Though we do not typically make the distinction between [a] and [b] clear, it can be teased out from a competent language-user.

I want to focus on [b], setting aside [a] (and the idea of the pleasure of any human, or any sentient being, being non-instrumentally good or valuable simpliciter or agent-neutrally). (Some, pointing out that we mainly or entirely use the language of good-for in a normative sense in ordinary language, balk at the category of agent-relative value, treat welfare-good-for as inherently normative and call it a day. If you share this perspective, feel free to treat the category of agent-relative value as a technical stipulation. Or perhaps substitute welfare-good-for language as you like.) In particular, I’m interested in how a fitting attitude account (FA) of the inherent agent-relative value here would have to go. On one general formulation of FA, the value of that-P is nothing more than the second-order property of that-P having some first-order property feature in virtue of which it is fitting to value that-P. 

But it is not plausible to apply such a formula in this case. The agent-relative, non-instrumental value for X in X’s experiencing pleasure is not a function of some further property of X’s pleasure-experience; rather, it seems to be a function simply of the pleasure (of the experience being a pleasure-experience). Because of this, the standard FA formula just mentioned (and any other FA formula that implies the same explanatory structure) seems wrong.

What seems to be happening here is this: though what usually makes a response to instances of a type of state of affairs fitting is the token having some feature distinct from its type (so that what we are responding to, in the most specific sense, is the further feature or the state of affairs having the further feature, though it is the state of affairs or object/entity not the further feature itself that would be the object of any attitude-type response). And so, once the relevant type is specified, we have all the information about what the “normative stimulus” is. Formally that would seem to be what is going on (or what the FA theorist should say is going on). Plausibly, this structure of fitting-response would seem to be unique to the (non-instrumental, agent-relative) value in pleasure (but maybe not!).

We could make the more-standard Scanlon-type (or similar) FA-formula fit, but it would sound odd and not provide any additional information: when it is fitting for a person to non-instrumentally value an experience, this is in virtue of that experience being a pleasure-experience for that agent. (Set aside whether that statement is true. The point here is largely formal, though I’m assuming we could formulate something along these lines that is extensionally adequate if this is not.)

Perhaps my point here is simply a technical point about how to formulate FA for the special case of the value to X of X’s pleasure (of the being-pleasant or taking-joy-in aspect of an experience that X is having) – and a way of making sure the FA theorist is not tripping over her own feet in explaining this case.

Maybe, but the point here seems relevant to a special role in the constitution of value that pleasure seems to play. In particular: it is because pleasure itself is valuable (non-instrumentally, to the agent) that associated experiences and activities are valuable (non-instrumentally, to the agent). Or, in FA terms: it is fitting to non-instrumentally value pleasant experiences and activities just in case (and maybe because?) it is fitting to non-instrumentally value pleasure itself. However, I have neither argued for nor explained this apparent dependency. It seems likely that just how this would go will depend on one’s theory of value (and that the FA approach would yield a distinctive pattern of explanation that perhaps has not been explored).

(One might also take this case to be evidence that FA fails – perhaps that sui generis non-instrumental, agent-relative value supervenes on pleasure, with our fitting responses being responses to the value, not the pleasure. I don’t think that is right, but it is worth considering in the larger context of the virtues and vices of having a general FA-style approach to some or all types of value or evaluative features.)