There is, then, reason to think that the relatively recent and conspicuous interpretive turn in anthropology, with its strong political character, has not confronted directly the shadowy but suffuse presumption of empiricism that has characterized ethnographic anthropology's rise as a social ‘science’ and has helped to eclipse from view the absolutely critical extent to which ethnographic inquiry is also and always an exercise in metaphysics.2 It is notable that the ontological implications of Foucault's notion of power were not seen as such by anthropologists. (Given his anti-metaphysical proclivity, though, the great French thinker must take some blame for this oversight.) In conceiving of power in terms of production as well as repression, he had in mind the creation of the real. Although his work tended to concentrate on the constraining force of the reality thus produced, he also took this constituting function to mean that power must be seen in a positive as well as negative light. This supplementary understanding of power points directly to the importance of ontology for anthropology, to the way in which humans in their relations to things and to one another (including the ethnographic interaction) participate in the creation of reality. Because it seems not simply to blur but virtually to eradicate the distinction between ethics and power, I am loath to use the term ‘power’ in this universalistic (Nietzschean) way. Nevertheless, the usage plainly and forcefully suggests that at the end of the day, anthropology is—whether its practitioners know it or not and despite its quite proper credentials as social science— ontology, and that therefore ontological preconceptions, both those of the studied and the student, should be an explicit and pivotal concern of anthropological inquiry.
As they are constructed through experience, these preconceptions plainly are historical. Yet precisely because they are preconceptions, they serve also to constitute reality. As synthetic a priori, they mark a zone of ambiguity between theory and practice, or between mental act and bodily action, and therefore, under most quotidian circumstances, their hosts are in no position to tell them from reality, including the reality of the hosts themselves. They are innocently enacted in the ‘natural’ course of everyday life. If this is correct, then it suggests that it is to our great advantage to seek to isolate and identify these ontological preconceptions, which betray themselves in their own existential and discursive practice (where ‘betray’ means both, on the one hand, ‘deliver’ or ‘construct’, and, on the other, ‘disrupt’). And since the anthropologist's onto- logical preconceptions are critical to his professional inquiry, it is in his direct, professional interest to do the same for his own, taking advantage of the disruption offered by ethnographic confrontation to jar his reflexive insight and rethink reality. It is a central contention of this book that of anthropology's synthetic a priori, dualism remains one of the most, if not the most, stubborn and comprehensive, and that it has worked and continues to work to restrict profoundly—at the heart of the discipline's defining purpose—the anthropologist's ability to plumb the reality of other cultures.
Obviously, an anthropology rooted in ontological dualism is at an elemental disadvantage when facing an alien culture in which the real is projected as basically ambiguous, for such a reality cannot be neatly factored into things that simply stand outside to one another. Instead, an ambiguous reality presents entities as concretely participant of one another and therefore only relatively self-contained or identifiable. In order to capture the characteristic human dynamic of such a reality, I deploy the notion of ‘primordial choice’ (cf. Evens 1995). This notion is predicated directly on a reality the most diacritical feature of which is ambiguity. Under this ontological condition, reality may be seen as finally somehow giving rise to itself, presenting an open-ended causality not subject to analysis that pursues a cause-effect regression. This is so because basic ambiguity does not finally allow for things that are utterly separate and distinct, things that thereby lend themselves to the logical canons of identity, non-contradiction, and the excluded middle. Instead, basic ambiguity amounts to imperfect holism, and imperfect holism entails a causal dynamic in which, paralogically, the effect both issues from and is continuous with the cause. Inasmuch as cause and effect participate in each other, it cannot be otherwise.
Such causality amounts to a kind of self-generation, wherein what is both does and does not give rise to itself or is at once self-contained as well as other to itself, and therewith open to its other. Among humans, this fundamentally opaque causal process is so pronounced that it effects a reflexivity indistinguishable from what we, in the West, are accustomed to call ‘choosing’. The openness of ontological ambiguity manifests itself consummately as, to make here a fruitful distinction, not natural selection but moral selection. Whereas in natural selection there is indeterminacy but no witting agent to speak of, in moral selection the self that is distinguished, not as a given but as a phenomenal function of the process of selection itself, continues to select with at least tacit intentionality. Still, the evident agency of moral selection is fundamentally imperfect. In the West, we are inclined to epitomize choice in terms of explicit intentionality and wholly autonomous selection. In fact, though, notwithstanding Kant's metaphysic of morals, choice as such is always necessarily to some critical extent heteronomous as well as autonomous. Put another way, to emphasize the paralogical nature of moral selection, heteronomy serves as an enabling condition of autonomous choice.3
By ‘primordial choice’, then, I do not intend choice that is perfectly witting, individual, and free, as if it springs from a self-transparent or noumenal or transcendental or absolute self. Still, as befits self-generation, to a notable extent a primordial choice also resists reduction simply to determination by what is other. It is characterized thus by a fundamental indeterminacy and is therefore also creative. Indeed, such a choice is so wonderfully creative that it tends to found a self-identity as well as the particular social and cultural world—the ‘second nature’—that inevitably corresponds to any such purposeful identity. Hence, for example, as one differentiates oneself as either male or female, one distinguishes the world as dualistically gendered; or as one delineates oneself as a cogito, an ‘I think’, one identifies the world as divided pristinely between the subjective and the objective. The capacity of such choices to create worlds is relative and limited to be sure, but it is also plain and consequential.
Although primordial choice always presupposes essential ambiguity, paradoxically such a choice, if it is sufficiently oblivious to its own constant heteronomy, can project the self as complete unto itself and the corresponding world as given to immaculate boundaries between one thing and another. This picture of things is paradoxical because if immaculate rather than fuzzy boundaries are the rule, the primordial choice that issued in these boundaries becomes logically inconceivable. Primordial choice is intelligible only in a nondualistic world. Of course, if one takes for granted formal, classical logic and its law of bivalence, then by definition primordial choice is paradoxical in its own right. But precisely because nondualism does not take such logic for granted, proceeding instead according to a ‘logic’ of ambiguity, primordial choice remains a conceivable proposition. A logic of ambiguity allows for the possibility and functionally specific use of formal logic but ultimately does not admit of an absolute boundary between logic and practice.
Indeed, if choice is to be meaningful as such, which is to say, creative or originary,