LILA LOVES LAGAVULIN. It is her favorite whisky. Aatif likes watching college basketball but has little time for baseball. Taniquill prefers the feel of flannel pajamas to cotton. Pei Lin enjoys being in rooms that are painted eggshell blue more than those painted canary yellow. Alejandro is more in the mood to listen to jazz than classical music. Tyus totally goes for spicy food. As a result, each has a reason to go in for what they favor over what they disprefer. (1) In such matters of mere taste one has a reason to choose what one favors or prefers. (2) In such matters, one's "stance" or favoring can play a role in grounding reasons, at least if we suppose that the attitude is based on an accurate descriptive understanding of what one's options are really like. (3) These reasons need not be decisive, obviously, but they carry some pro tanto weight, at least in many contexts. I will call this the "Modest Claim." (4) The astute reader will spot that my choosing that label for the view reveals that I did not expect this view to be highly contentious.
But it is. A wide range of influential philosophers, including T. M. Scanlon, Michael Smith, Ralph Wedgwood, Richard Arneson, Roger Crisp, and Richard Kraut, maintain that an individual's favorings or stance never play a normative role in grounding reasons. (5) Too often it is a bit obscure why people deny the Modest Claim. Indeed, there has been real confusion about the best formulation of the claim. In this paper I will try to clarify the central claim and articulate the considerations that seem to motivate people to resist it. I will argue that these considerations are unpersuasive and that we should accept the Modest Claim.
If one were ever going to grant normative authority to contingent attitudes it would surely be in the context of matters of mere taste. (6) Thus the most plausible and coherent views that deny the Modest Claim embrace what I will call broad normative stance-independence. Shafer-Landau, in the context of characterizing a type of moral realism, explicated this notion of stance-independence. Proponents of normative stance-independence maintain that truths in the relevant normative domain, in our case reasons for action, obtain "independently of any preferred perspective" and are "not made true by virtue of their ratification from within any given actual or hypothetical perspective." (7)
As the above examples made clear, a great variety of attitudes are covered by the relevant notion of an agent's "stance," including, among others, loving, liking, wanting, desiring, craving, valuing, and preferring. Further, there are different levels of stance, such as when one wants to love Radiohead more than KC and the Sunshine Band. It is an advantage for the friend of the Modest Claim to have so many options. I will not champion here the normative relevance of a particular stance. I do think the above examples offered of favoring attitudes are all tempting stances for the defender of the Modest Claim to point to. It may be that more than one such favoring attitude grounds or partially grounds reasons. (8) However, the friend of the Modest Claim need only assert that at least one such stance grounds reasons, and they need not claim that it does so in all contexts. I will use "favoring attitude" or "stance" as the generic and "preference" or "desire" as the favored example of a particular stance. I will use "stance-independence" to refer to full stance-independence and "stance-dependence" to refer to at least partial stance-dependence.
Three ambiguities in understanding the most useful and important usage of stance-independence are worth considering before we proceed. First, consider a view that says that there are completely objective, stance-independent criteria for what is beautiful and that, while it is valuable to interact with the beautiful, appreciating the beautiful is even more valuable. Such a view might say that the relevant sort of appreciation is conative--being moved by beauty or loving it, for example. Such a view claims that one's normatively favored conative reactions are the key to this extra value. Should this be thought of as a fully stance-independent view or not?
In this case, the appreciation is thought to be warranted by stance-independent norms and only warranted reactions are thought to be of value. (9) Such a view will say that conative reactions ground reasons, but only if those conative reactions are themselves warranted by the object of the attitude. The most important divide is between views that maintain that conative attitudes can play a role in grounding value even if the object of the attitude does not, by itself, justify or merit the attitude, and views that deny this. The latter sort of view still seems to me to side with Socrates in the Euthyphro question of where value originates. The attitudes are, on such a view, still normatively slaves of stance-independent values. Only by properly responding to what is stance-independently valuable can they generate value.
The friend of the Modest Claim, as I will understand it, maintains that, even in contexts in which none of the options commands or warrants the relevant favoring attitude, nonetheless where the attitude happens to go still plays a role in grounding reasons. So we will understand the relevant sort of stance-dependent theorist as claiming that, at least in some cases, one's stance plays a grounding role even when that stance is not itself normatively required or favored by the stance-independent value of the object. Our question is whether the stances one has no stance-independent reason to have can ground normativity. (10)
The second ambiguity concerning stance-independence is what a stance is in the relevant sense. Dale Dorsey has shown how one can focus on contingent cognitive attitudes such as beliefs about what has the relevant sort of value within a recognizably subjectivist framework. Ruth Chang has suggested a voluntarist view according to which stipulating that one has a reason can, in some contexts, create a reason. (11) What is crucial to both views, I take it, is the thought that even if one's cognitive attitude or stipulation hits on something that is not stance-independently favored, it still has direct normative upshot. Both of these views still grant authority to an agent's contingent stance, even if not her conative stance. (12) As I understand the Modest Claim, it maintains that some such "favoring stance" can create reasons in such contexts and need not insist that the relevant favoring attitude is a conative state. However, I think there are strong reasons to incline toward a conative version of the view and I will assume such a version here for the sake of simplicity.
The third ambiguity concerns the level at which one's attitudes must endorse an option to count as authoritative. Some argue against the Modest Claim in this way. They maintain that, while one must favor a sensation for it to give one a reason in matters of mere taste, still, they maintain, one need not have a higher-order favoring attitude toward that pleasure in order for it to be reason-giving. Thus, they conclude, the reason here is stance-independent.
I think this argument mistaken. To see why, consider the full-on subjectivist who thinks that a particular favoring attitude grounds all of an agent's reasons. Now this alleged fact, that those attitudes ground reasons, is, according to the subjectivist, not itself hostage to anyone's further favoring attitudes. Subjectivists maintain that favoring attitudes at some level or other ground one's reasons. They need not maintain, and have not tended to maintain, that for the attitudes at level N to ground reasons, there must be a further favoring attitude at level N+1 or higher toward the N-level attitude. If maintaining that one's favoring attitudes at a specific level ground reasons that are not themselves in need of ratification from some further favoring attitude toward it was enough to make one a fully stance-independent theorist, then most full-on subjectivists do not accept any stance-dependence. If most subjectivists do not count as embracing stance-dependence on a construal of what makes a view stance-dependent, then so much the worse for the usefulness of that construal. Stance-dependence in that sense has rarely been endorsed, even by subjectivists.
Some influential stance-dependent views look to higher-order attitudes, such as what one's idealized self wants one's ordinary self to want. But a view that claimed normative authority for all (informed) attitudes at all levels, regardless of higher-order ratification, would clearly remain stance-dependent. Further, the higher-order stances that have been purported to have normative upshot were not claimed to be made reason-giving or well-being grounding only if there was some higher-order favoring attitude toward the lower-order stance. (13)
To see in action what I regard here as the mistake I am warning against, consider an argument from Guy Fletcher to the effect that even hedonistic views that take pleasure to be a sensation one intrinsically wants for its intrinsic phenomenological properties, and maintain that pleasure necessarily benefits one, do not count as relevantly attitude-dependent. He argues that "on the hedonistic theory, pleasure is good for you even if you have no pro-attitude toward it." (14) And that is true. If you bundle the pro-attitude into a state, as Fletcher does with pleasure, the hedonist does not claim you need an additional desire toward the bundled state for pleasure to benefit. But you might just as well bundle together the favoring attitude and the object of that attitude, call that a desire satisfaction, and say that views that claim that desire satisfaction benefits whether one has an additional desire for desire satisfaction or not are not subjectivist. (15) Such maneuvers will implausibly result in having to say that...