One way to approach the puzzle of deontic constraint is to ask whether rational action necessarily has a consequentialist structure, or whether it can incorporate nonconsequential considerations.
Unfortunately, many theorists (philosophers and social scientists) have been misled into believing that the technical apparatus of rational choice theory, introduced in order to handle the complications of probabilistic reasoning, is also one that prohibits the introduction of nonconsequential considerations into the agent’s practical deliberations. In other words, it is sometimes thought that decision theorists are necessarily committed to consequentialism, or that consequentialism is simply the expression of Bayesian reasoning, when applied to practical affairs. Deontic constraint, or rule-following behavior, according to this view, is either not mathematically tractable, or else violates some elementary canon of logical consistency.
There is absolutely no reason that a rational choice theorist cannot incorporate deontic constraints—or any other type of rule-following behavior—into a formal model of rational action as utility-maximization (although, in so doing, it would perhaps be prudent to shift away from the vocabulary of utility-maximization toward that of value-maximization, given the close connection in many people’s minds between utility theory and consequentialism). The commitment to consequentialism on the part of many rational choice theorists is the result of a straightforward oversight that arose in the transition from decision theory (which deals with rational choice in nonsocial contexts) to game theory (which deals with social interaction). Early decision theorists adopted a consequentialist vocabulary, but did so in a way that made consequentialism trivially true, and thus theoretically innocuous.
Since I am inclined to put rules on the “preference” rather than the “belief” side of the preference-belief distinction, what really needs to be shown is that the preference through which an agent’s commitment to a rule is expressed may also be rational. In order to do so, it is necessary to challenge the prevailing noncognitivism about preferences, or the view that desires are somewhat less susceptible to rational reevaluation than beliefs.
My goal is to take what I consider to be some of the best thinking done in the past couple of decades in epistemology and philosophy of language, and show how it “ﬁts” with some of the most important work being done in evolutionary theory, in order to reveal the deep internal connection between rationality and rule-following. One of the major forces aiding and abetting the noncognitive conception of preference, for well over three centuries, has been a commitment to representationalism in the philosophy of mind (i.e., the view that “representation” constitutes a central explanatory concept when it comes to understanding the contentfulness of our mental states).
The alternative strategy, which has recently been developed with considerable sophistication by pragmatist theorists like Robert Brandom, is to start with a set of concepts that are tailor-made for the explanation of human action, and then extend these to explain belief and representation. This is based on the plausible intuition that human action in the world is more fundamental than human thought about the world.
This analysis serves as the basis for my defense of what I call “the transcendental necessity of morality.”
Reading the philosophical literature, it has come to my attention that “Kantian evolutionary naturalism” is not a particularly well-represented position in the debates over the foundations of human morality. This is a deﬁciency I hope to remedy. The basic Kantian claim, with respect to moral motivation, is that there is an internal connection between following the rules of morality and being a rational agent.
I would like to defend the rationality of deontic constraints at the level of action, but am not committed to defending “deontology” as a theory of justiﬁcation.
There is also an inclination among moral philosophers to draw a sharp distinction between “moral” and what are called “conventional” obligations, such as rules of etiquette, or “social norms” more generally. I reject this distinction, not because I think morality is conventional, but rather because I follow Emile Durkheim in thinking that all social norms (or “conventions” in this way of speaking) have an implicitly moral dimension.