In spite of the discomfort that cultural anthropologists experience whenever anyone mentions the "R" word (rationality; not relativism), RCT is actually quite complementary to anthropological theories of cultural production. Indeed, far from being the predatory, colonizing force it is commonly perceived to be—the monster that ate the social sciences and then went on to snack on the humanities—on close inspection, RCT is downright humble and restrained in its claims. For one, it does not presume to say anything about the desires, goals, sentiments, values, and beliefs that shape individual preferences. All of these are defined as exogenous to rational choice model. Thus, it makes no threatening advances on cultural theory. Neither does it posit anything about the social institutions and social relations in which individual rational actors are embedded, leaving the field clear for sociology. As it appears to have been convincingly demonstrated by Tversky, Kahneman, and Thaler that rational choice models are psychologically unrealistic, psychology can rest easy. So can history, since RCT's formal mapping of desired ends floats above historical events and transformations.
Given these extremely limited claims, one might wonder what all the fuss is about.
Regenia Gagnier's paper suggests that part of the fuss is about the ahistoricity of RCT models, which renders them incapable of charting how choice develops through time in specific historical circumstances. Unlike literary modes of thought, which include diachronic as well as synchronic elements, RCT is resolutely synchronic. Choice—and, therefore, assessment and intentionality—takes place in a slice-of-time, much like "culture" used to be located in an "ethnographic present" in anthropological studies. Both are obviously narrative conventions that carry with them likely distortions. The modern meanings of choice and preference, Gagnier points out, are themselves historical; and—if I may take the liberty of an anthropological reading of her discussion—these meanings, which are constitutive elements in a folk theory of social agency, are simultaneously constitutive forces in the production of choices and preferences. In other words, we are back to Geertz's (1973) definition of culture as "a model of and a model for" behavior. If, as Gagnier convincingly argues, the meaning of choice changes over time and with it the ways in which people evaluate and enact their choices, then the neat boundaries between the endogenous and exogenous in RCT models blur.
Hazel Markus's paper suggests that another part of the fuss is about the cultural specificity of the model of agency that is assumed by RCT. Since actors situated in specific sociocultural contexts employ different models of agency, we need different psychological models to explain their preferences and choices. This is a point that has been strongly argued by cultural anthropologists interested in understanding different modalities of agency (see, for example, Mahmood 2004) and their constitution of different kinds of identities and subjectivities. I will return to this point shortly.
Together, Gagnier's and Markus's papers raise at least two serious challenges to RCT. The first points to RCT's failure to adequately conceptualize processes of subject formation. If, as Foucault (1980) has argued, the subject does not precede power relations (or social relations more generally) but rather is produced by these relations, then the capacities that constitute a subject's modes of agency are themselves the products of these relations. In addition, power—in which I would include both the enforcement of cultural norms and the incitements of cultural desires—is not merely a constraining force (as it is in more conventional theories of social action) but a productive one as well. Once cultural norms and beliefs are conceptualized as crucial to the production of desiring subjects and their modes of agency, it follows that culture cannot be exogenous to models of social action, including rational choice theory.
A second and closely related challenge raised by Markus's and Gagnier's papers is to the static, nonprocessual character of the relations between subject and action in RCT. The desiring subject—the author of the preferences that are revealed by the modeling of choice—is presumed by RCT to be a stable entity. But, if choice as a form of social action is constitutive of social relations, and these social relations in turn produce desires and capacities, it follows that desiring subjects do not precede choice, but are constituted by them. This means that subjects are—even in the presence of strongly repressive regimes—always unstable and contingent formations.
Rather than view RCT as an alien force threatening to take over the social sciences and humanities, it seems more accurate to recognize it as a domestic one that has been with us—by which I mean the humanities and the humanistic social sciences—all along. Its core premises have long been pervasive in political science, economics, and sociology. Even in the most humanistic of the social sciences, anthropology, its premises have been accepted by many anthropologists. As Ken Arrow pointed out in the discussion, we use the language of intentionality and choice all the time. Although I take it that Arrow's point here is that the pervasiveness of these concepts attests to their universality, an alternative interpretation is that they have become sufficiently hegemonic in academic and popular discourse as to have erased the history of their production.
The dualisms that Julie Nelson argues reflect a cognitive habit permeating economics mirrors Talcott Parsons' (Parsons and Bales 1955) dichotomy between "instrumentality" and "affect"—the first a passionless interest fashioned by reason and pursued through calculated rational action, the second uncalculated and irrational. The instrumental/affective dichotomy is central to the Parsonian theory of social action and underpins the concept of "interest" as a motivating force in "economic action" social action. In defining instrumentality as rooted purely in objective interest, the unpredictability and unmanageability associated with interest as purged from models of economic action and relegated to other institutional domains in the Parsonian model of modern society. Family and religion might be governed by affect, but the economy is governed by instrumental action.
Parsons' notion of instrumentality, in turn, is rooted in Weber's (1978) definition of economic action as the quintessential pursuit of interest by means of rational technique—in other words, by technique which is consciously and systematically oriented to the experience and reflection of the actor (Weber 1978, 65). Weber’s definition of “economic action” rests on the distinction between action oriented toward the satisfaction of a desire for utilities and action oriented toward the satisfaction of other desires (1978, 68). He defines utilities as “the specific and concrete, real or imagined, advantages of opportunities for present or future use as they are estimated and made an object of specific provision by one or more economically-acting individuals” (1978, 68). In addition to “goods” (non-human objects which are the sources of potential utilities) and “services” (utilities derived from a human source, so far as this source consists of active conduct”), “social relationships which are valued as a potential source of present or future disposal over utilities are . . . also objects of economic provision” (1978, 69). This definition of utilities opens up his definition of “economic action” to a much broader range of actions that might be initially surmised. Indeed, it makes it difficult to distinguish “economic action” from other social actions, including those oriented toward establishing and maintaining kinship relations and friendships. Thus, what seems at first glance to be a rigorous and narrow definition of “economic action” turns out, on closer scrutiny, to rest on concepts that blur the boundaries of the definition. The problem, however, is not that Weber’s definitions are not sufficiently rigorous. Rather, the problem is that his analytic strategy of differentiating “economic action” from other social actions is ill-conceived. Few, if any, social actions are so singularly oriented as to be easily classified according to such a scheme. Indeed, the question we might well pursue is how such a scheme of desire, intention, and choice became hegemonic and what modes of agency and subjectivity it has produced.
 Markus’s paper is unavailable for publication, but the key ideas are to be found in Hazel Markus and Shinobu Kitagawa, “Models of Agency: Sociocultural Diversity and the Constitution of Action,” in Cross-cultural Differences in Perspectives on the Self, ed. Virginia Murphy-Berman and John J. Berman, Nebraska Symposium on Motivation 49 (Lincoln: University of Nebraska Press), 1-57.
Foucault, Michel. 1980. Power/knowledge: Selected interviews and other writings, 1972-1977. Brighton, England: Harvester Press.
Geertz, Clifford. 1973. The interpretation of cultures: Selected essays. New York: Basic Books.
Mahmood, Saba. 2004. Politics of piety: The Islamic revival and the feminist subject. Princeton, NJ: Princeton University Press.
Parsons, Talcott, and Robert F. Bales. 1955. Family, socialization, and interaction process. New York: Free Press.
Weber, Max. 1978. Economy and society: An outline of interpretive sociology. Vol. 1. Ed. Guenther Roth and Claus Wittich, trans. Ephraim Fischoff et al. Berkeley: University of California Press.