If I say that a state of affairs, S, is good (good simpliciter or agent-neutrally valuable or good), I usually mean something like this: S is better than some implied comparison class of other states of affairs. And so, in this usage, ‘good’ is like ‘tall’: it is a disguised comparative feature, not a non-comparative feature. This kind of value, then, is properly analyzed in terms of better-than. It is for this kind of reason that many theorists take value simpliciter (or other sorts of value or all sorts of value) to be properly analyzed in terms of better-than. Here’s why that seems wrong to me.
Consider the proposition that pleasure is good (again, good simpliciter). To keep things simple (and because it is likely correct), suppose we can state the same fact by saying that someone experiencing pleasure is good (the implication being that the verbal form of attribution to an object versus that of attribution to a state of affairs does not have any metaphysically important upshot). In this case, unlike the other, the goodness that we are attributing is genuinely non-comparative. We are not saying that someone experiencing pleasure is better than some reference class of relevantly similar states of affairs.
Plausibly, the goodness feature here is more like height than it is like tall in the tall/taller/tallest case. It is something like a fundamental feature or dimension that exists in some quantity, in terms of which (and only in terms of which) the comparative properties exist and make sense. Thus, as we explain taller-than in terms of quantity of height here compared to quantity of height there, so we explain better-than (and hence merely-nominally non-comparative goodness) in terms of quantity of (genuinely non-comparative) goodness here (in this object, in this state of affairs) compared to quantity of (genuinely non-comparative) goodness there (in that object, in that state of affairs).
We might highlight the distinction here by considering two terrible states of affairs, A and B, where A is better (less terrible) than B, but A is also good relative to some (also terrible) comparison class. In this case, A is both good (in the merely nominally non-comparative sense) and not good (in the genuinely non-comparative sense). Overall, there is no net quantity of goodness in A (and we might even suppose that there is no quantity of goodness in A, at all, in any respect). This shows us that we can, and sometimes need to, talk about how much goodness there is in a thing as well as about one thing being better than another. Though there are probably contexts in which one or the other use of ‘good’ is dominant, I think we can always sensibly talk about both things: (i) the goodness and how much of it is here or there (like having-height) and (ii) one thing, in respect of how much goodness it has, being good relative to a reference class (like being-tall).
Something similar is probably true for agent-relative value or goodness. Maybe, though this is less clear to me, for good-for (if this is distinct from agent-relative goodness, which I think it is) and attributive goodness (e.g., something being a good chair) as well.
Does this seem more or less right?
NOTES
1 This discussion is inspired by parts of Mark Schroeder’s recently-updated (March, 2023) Stanford Encyclopedia of Philosophy entry on value theory. A difficult, somewhat opinionated survey, but highly recommended.
2 Thanks to Derek Bowman for a very helpful discussion.
Policy of Truth 
It seems to me that you have to decide a prior issue before addressing this one: is “goodness” only derivatively attributed to states of affairs, or is that attribution primary and free-standing?
You get the comparative analysis of goodness if you think of goodness as attributable primarily to courses of action. If the agent faces alternatives A1 and A2 at t, and has to choose between them, then if we adopt some normative standard, one is better than the other. Given certain background conditions, “goodness” then becomes the property common to every action that’s better than the alternatives. States of affairs are only good insofar as they are states of affairs brought about by good actions. The experience of pleasure is good only because the experience is aimed at, and (by some standard) consistently trumps the other alternatives.
It’s only if you detach goodness from actions, and attribute it directly to states of affairs abstracted from the sets of actions that produce them, that you get a rationale for the non-comparative conception of goodness you have in mind. I’m not sympathetic to that move, but if you grant it, I guess your view has plausibility.
You might find a different example clarifying. Take the concept of “clean” or “cleanliness.” In one sense, “clean” is a comparative aesthetic concept: X is cleaner than Y when X perceptually appears to be less dirty than Y. For instance, if X and Y are rooms, X could look cleaner than Y if X “sparkled” more than Y, but all the dirt was in some hidden corner.
In another sense, “clean” is a comparative functional concept: X is cleaner than Y when X actually lacks less of whatever counts as “dirt” given the function of X’s and Y’s as organs, organisms, or artifacts. So if X and Y are operating rooms, X is cleaner than Y if X is less prone to hospital-borne infections than Y, regardless of how outwardly “shiny” or “dull” it looks.
In a third (notional) sense, we could imagine an “absolute” conception of cleanliness: X is absolutely clean if X contains no dirt by any conception, aesthetic or functional, of dirt. In other words, it would, quantitatively, contain 0 units of dirt. (This is sort of analogous to “absolute 0” heat on the Kelvin scale.) I guess I can (sort of) make sense of that as an abstract thought, but I’m not sure it has any real-world application.
Same with goodness, mutatis mutandis. The example you give of our “needing” a conception of absolute goodness, is totally abstract. It’s not clear to me it has any real-world application, any more than an absolute conception of cleanliness does.
LikeLiked by 1 person
Thanks, Irfan. Quite helpful.
(1) The possibility that you bring up, I take it, goes like this: goodness is primarily attributed to actions in a comparative sense and so its secondary attribution to states of affairs (whatever its verbal form) is comparative as well. This tracks some Objectivist-y metaethical themes. I’m not sure why, though, on this kind of picture, why the actions (and the standard adopted for the purposes of choosing between alternatives) would not be normative rather than evaluative (more reason or normative pressure to perform A1 than there is to perform A2). The question of whether to perform A1 or perform A2 seems like a paradigmatically normative question to me. By contrast, if the context is not that of deciding what to do (say, perhaps, when I look back on my options and my decision), then evaluative language seems to make some sense (but is most naturally interpreted as getting at a slightly different normative/evaluative feature): choosing A1 was better than choosing A2. However, if a bit forced, I don’t think it does violence to evaluative language to describe one having more reason to choose A1 than to choose A2 in terms of A1 being a better choice than A2. I do think evaluative language can be “stretched” in this way. The question is why we have the different concepts and whether they get at importantly different properties. Hard questions. I think that the view you articulate here would count as either a reduction or elimination of the normative in favor of the evaluative. But, despite all this, I can certainly make good notional sense of your point. My question, I guess, would be what there is in the way of self-contained, strong evidence for this kind of view.
(2) I didn’t give good, concrete examples. That is bad. I’ll make up for that now, at least a bit (but I’ll also shift the topic some). My very general case of non-comparative goodness was that of the goodness simpliciter in a person (or sentient being) experiencing some degree of pleasure. But I think it will help, in the case of pleasure, to shift from agent-neutral (good simpliciter) to agent-relative value. So we have: it being good relative to X (or valuable to X) for X to experience (some degree of) pleasure. (I did not fill-out the pleasure case in terms of agent-relative value because most people do not separate welfare-good-for from agent-relative value in a general way. I do. I also take agent-relative value to be normative or action-guiding, while good-for — controversially, including welfare good-for — not to be.) Is the ‘good’ or ‘valuable to’ concept here non-comparative? This much, I think, is obvious: we don’t have in mind some kind of comparison class, as we do when we say ‘that NFL quarterback is tall’. Though it is natural to assume that the property is non-comparative if the concept is, I don’t think we can rule out some analysis according to which the apparently non-comparative feature is actually comparative in some way. Note: non-comparative need not mean ‘absolute’ in your sense. In many cases, a concept for there being literally none of the underlying quantity or stuff (dirt, length, non-comparative goodness) is pretty useless (if often entirely coherent). I’ll readily admit that my argument here is mostly just an expression of linguistic intuition, with accompanying presumptive metaphysical intuition (how I take us to use the relevant bits of language, what they seem to mean or refer to).
(3) Your treatment of the same (or a very similar) case is interesting. Applying the general view that you articulate, you say: “[t]he experience of pleasure is good only because the experience is aimed at, and (by some standard) consistently trumps the other alternatives.” I take this statement to be about agent-relative value (that is also an element in the agent’s welfare). More explicitly (and using my terminology): “X-experiencing-pleasure being intrinsically or non-instrumentally valuable to X comes to it being better for X to choose options that lead to her experiencing pleasure than it is for X to choose options that do not lead to her experiencing pleasure (according to some standard that X bears the right relationship to).” A lot is omitted in leaving a blank spot here for the standard and the relevant sort of agential relationship to the standard. One worry I have here is this: unless these elements are filled-in just so, what is generated is a kind of instrumental value.
(4) However, once this work is done (and assuming it can be done…), your view constitutes a way to make this kind of attribution of goodness to pleasure comparative (in a non-obvious way) — but not on the tall/taller/tallest (height) model. All of this, I will say, is quite distant from our language use (and hence from a more “conceptual” style of analysis). Though I don’t rule out non-conceptual or purely metaphysical-explanatory analyses out, I think we have to start from our linguistic-cum-metaphysical intuitions and, if we do, the evidential hurdle is pretty high for treating the relevant properties here as disguised comparative properties.
LikeLike