Consensus and Excellence of Reasons

It is plausible to suppose that the normativity of evaluative (e.g., moral and epistemic) judgments arises out of and is, in some sense, dependent on our actual evaluative practice. At the same time, though, it seems likely that the correctness of evaluative judgments is not merely a matter of what the underlying practice endorses and condemns; denial of this leads one into a rather objectionable form of relativism. In this paper, I will explore a social practice account of normativity according to which normativity is grounded in our actual social practice of evaluation. I will show how this account allows normativity to be dependent on our actual evaluative practice, while allowing the correctness of evaluative judgments to be independent of this practice in important ways, and how the resulting temporal logic of reasons gives us a conception of morality and other sorts of evaluative discourse that is not historically local. here are different accounts of what it is for a perceived normative requirement to be correct. These accounts can be seen as lying on a continuum, from those divorcing correctness completely from what the community actually does to those that completely reduce moral correctness to what the community does. The former group may be called transcendentalists1; they think that what is in fact, for example, morally correct in no way depends on what the community takes to be morally correct; questions of the correctness of normative judgments (moral, epistemic, semantic, etc.) are completely divorced from the actual practice of the community. Under the heading of “transcendentalist” fall Platonists and other assorted thirdrealmers. At the other end of the spectrum we have attributivists, those who think that the correctness of a moral judgment is merely a matter of what a particular T 86 JEREMY RANDEL KOONS community thinks, or of what a particular person thinks. That is, normative correctness is reducible to some community’s practice, or to some person’s psychological states. Into this camp fall relativists such as Ruth Benedict, along with subjectivists, various emotivists, and others of their ilk. It is plausible to suppose that the truth of the matter lies somewhere in the middle of the continuum: that is, the correctness of evaluative judgments (e.g., questions of moral rightness and wrongness, epistemic justification, etc.) is in some sense dependent on our actual evaluative practice. At the same time, though, this dependence is not a tight one: the correctness of evaluative judgments is not merely a matter of what the underlying practice endorses and condemns; denial of this leads one into a rather objectionable form of relativism. The question then becomes, What are the details of this via media? How do we sketch the relation of dependence between the correctness of evaluative judgments and our actual evaluative practice in a way that does justice to both of these intuitions? In this paper, I will explore a social practice account of normativity—an account according to which normativity is not transcendent in origin, but is instead grounded in our actual social practice of evaluation. I will show how this social practice account of normativity does justice to both of the intuitions I have described—that is, how this account allows normativity to be dependent on our actual evaluative practice, while independent of it in important ways—and how the resulting temporal logic of reasons gives us a conception of morality and other sorts of evaluative discourse that is not historically local. In other words, a social practice account of normative judgment can avoid both attributivism and historical localism. I. TWO CONCEPTIONS OF NORMATIVITY Traditional theories of normativity divide into two competing camps. A social practice view of normative utterances is most plausible if we view it as arising in the context of these two views, and as responding to difficulties the two traditional views cannot address. The first camp is the attributivist camp. On this conception of normativity, to say that one ought to do x is to say that the doing of x follows from some set of rules, or that the community endorses the performance of x, or that the performance of x follows from one’s beliefs (or would after suitable reflection). Thus, the attributivist conception is essentially a descriptive notion of normativity. To say that one ought to do x is to describe something, either that person’s belief set, or the practices of her community, or a set of rules, or some such. It is to say that the endorsement of x follows as a matter of fact from this practice or set of beliefs or rules. The other traditional conception of normativity is the transcendental conception. Far from asserting that particular endorsements follow from a community’s practice or an individual’s belief set, this view claims that CONSENSUS AND EXCELLENCE OF REASONS 87 normative claims aim to correspond to community-independent norms. The correctness of a normative judgment is then entirely independent of what anyone in the community thinks or does. These are only rough sketches of the two positions, but more is not needed, for certain examples clearly confound both conceptions, no matter how these conceptions are ultimately fleshed out. A good example is the introduction of the “Wilt Chamberlain” rules in basketball. These were a set of rules introduced for the simple, albeit theoretically puzzling, reason that he was too good a player. His ability to swat balls away from the basket led to both the rule against goaltending and the extension of the key. His ability to leap from behind the foul line and stuff the ball led to a revision in the rules regarding foul shots.2 This example brings out the inadequacy of the attributivist conception of norms: [I]f claiming that an act is permissible in a game is just saying that this entitlement follows from some fixed set of rules definitive of that game, then it would seem we must conclude that a different game was being played post-Wilt from the one played pre-Wilt. But is this a reasonable conclusion? If so, then we must say that every record book is mistaken since they all take statistics before and after Wilt to have been set in the same game. Everyone who talks about basketball would be subject to a crucial ambiguity of which they are totally unaware. Presumably all sorts of legal issues concerning, say, television rights over basketball games would have to be reconsidered.3 Nor can it be said that these revisions were uniquely determined by the state of the game pre-Wilt. Not only is it highly unlikely that it occurred to anyone that such a good player might come along, but these particular solutions to the Wilt problem were underdetermined by the practice of basketball pre-Wilt. In other words, “[t]here were many ways the game could have been revised to retain its interest. (The baskets could have been raised, players over 7 feet tall could have been banned, etc.).”4 This “normative underdetermination of the future emendations of the rules by the present needs and past practices demonstrates not only the inadequacy of the attributive conception, but that of the transcendental conception as well. There cannot be a single basketball game out there waiting for us in Plato’s heaven since there are several—indeed, probably infinitely many—coherent games we might have opted for.”5 Thus, the transcendentalist and the attributivist are making opposite errors. The transcendentalist divorces normativity too far from social practice, claiming that our practice has no bearing on normative truth. The attributivist, on the other hand, makes the connection too tight, so that any change in our actual practice makes it the case that we are simply engaging 88 JEREMY RANDEL KOONS in another activity. This suggests that the correct path is a sort of via media, on which normativity arises out of our social practice, but according to which the correctness of a normative claim isn’t just a matter of whatever I (or we) say it is, or of whatever follows from my belief set, etc.6 In what follows, I will demonstrate how a social practice account can avoid both attributivism and historical localism. This is an important result, because it will show that although crude social practice accounts may run afoul of one or both of these problems, social practice accounts of normativity need not be seen as inevitably succumbing to them. Before continuing, let me make some brief remarks about social practices themselves. I think we all have a rough, intuitive idea of what a social practice is. It is, for example, what anthropologists posit to explain certain regularities in the behavior of a community. Here is one example: “[C]onsider the practice in basketball of not stepping outside the sidelines when one is in possession of the ball. It is clearly useful as part of a systematic account of the ‘dance of the basketball players’ to take them to be committed to following such a rule, even though they do not always follow it, nor does any penalty for non-compliance universally follow.”7 We attribute commitments such as these to communities whose behavior we seek to explain. It is the fact that a bit of behavior (staying inside the lines, stopping at stop signs) is caught up in such a web of commitments that makes possible its classification as an action (and its performer as an agent), as opposed to mere behavior. In interpreting a community in terms of social practices, one takes the members of the community to be committed to, and bound by, certain appropriatenesses. The normativity is only shallow, however, as positing such practices does not involve committing oneself to following them.8 One advantage of explaining norm-governed behavior in terms of social practices is that such an explanation allows us to account for normativity without positing spooky non-natural properties or objects; we need only posit social practices, which are sets of commitments to practical appropriatenesses (mainly implicit), commitments we posit to explain the behavior of a community. Acting according to one of these appropriatenesses is acting for a reason. And so reasons are, on this account, products of social practices and the appropriatenesses implicit in these practices. Despite the advantages of a social practice account of normativity, it seems as though there is a significant disadvantage, however: it seems as though this account leads us into a rather nasty version of relativism. After all, if it is our community’s practice that fixes the correctness of an action, then it seems straightforwardly contradictory to question whether an action that accords with our community’s practices is correct. We will see, over the course of this paper, how a sophisticated social practice account can evade this worry. CONSENSUS AND EXCELLENCE OF REASONS 89