A second way in which Kant's synthetic a priori promoted rather than undermined dualism is this. Correlative to his preservation of an absolute distinction between subject and object, that is, between an inside or mental world and an outside or physical one, is the consideration that his argument continues to predicate a reason that is autonomous and pure. Although by definition it is not analytic, the synthetic a priori is nonetheless conceptual before it is experiential. Indeed, even while it moves to criticize reason by referring it to the evidence of experience, Kant's synthetic a priori proposes that that evidence is not exactly given but intellectually constituted. As Burke puts it (1969: 191), for Kant “experience derives its appearance from the nature of consciousness (the ‘I think, or ‘transcendental synthesis of apperception’).” That is to say, instead of starting with human experience (in the mediatory middle, so to speak), he starts with reason qua reason (transcendental or not), arguing that the human world is begotten conceptually (categorically). Thus, he ends up with such universals as ‘cause’ and ‘object’, that is, with relative but determinate notions that he universalizes, rather than with ‘universals’ that are basically—and paradoxically—conditioned or not specifiable outside of their particular manifestations.3
At this juncture, I can make explicit that the philosophical notion of the a priori is connectable to anthropology at the very core of the latter's scholarly enterprise. As the notion of the a priori directly implicates the idea of universals, so it bears squarely on the question of cultural relativism. Indeed, the deployment of this philosophical notion has powerful implications for anthropological inquiry, from—to sum up broadly the direction of this inquiry—nineteenth-century evolutionism to twentieth-century relativism.
Since empirical statements can always be refuted by further observation, their truth (what indeed it is usual to call synthetic truth) is essentially contingent. But in a manifest sense, so is the truth of analytic statements, since it is dependent on the truth of the other such statements that go to make up the particular system of logic by which truth of this nominal—or, if you like, pure—kind is defined. Moreover, the truth of analytic statements may be regarded as finally necessary or a priori rather than contingent or a posteriori only when the system of logic from which they derive is itself necessarily universal rather than particular (as Kant [1963], in line with his transcendental idealism, professed is in fact the case with classical Western logic [see also Evens 1983]).
Obviously, then, since neither synthetic nor analytic knowledge in itself can support a claim to the epistemic superiority of one culture over another, both are compatible with cultural relativism as opposed to (the Enlightenment idea of) progressive cultural evolutionism. Kant's synthetic a priori, however, which, by virtue of the so-called transcendental deduction, makes some specific concepts both necessary and universal, certainly can offer such support. This helps explain why Kant paid virtually no attention to the role of culture and society in his philosophy of reason and knowledge. To be sure, in picturing consciousness as active rather than passive, Kant made room for cultural activity. By insisting, though, that the synthetic a priori is universal in the received sense (the sense defined by dualism or absolute division between the universal and the particular), and hence according to the epistemological principle of certainty rather than that of basic ambiguity, he could not take advantage of this ethnological opportunity. On the contrary, by defining humanity primarily in terms of reason—a reason whose purity had been critically cut, but only to be further and sublimely rarefied by transcendental distillation—Kant made it possible to conclude (and here I anticipate the interpretation of the Holocaust in chapter 4), whatever his own opinions in such matters, that any peoples who could be shown to fall short of such reason were backward, inferior, or less than human.
Toward True Synthetic a Priori: Wittgenstein and Merleau-Ponty
Wittgenstein
It seems ironic that in his critique of metaphysics, Kant, in the thrall of reason, failed to follow to its logical conclusion the implication of his own Copernican revolutionary thesis, for if it is the case that all knowledge presupposes a point of view, then his ‘transcendental logic’ cannot escape this circumstance. Even if Kant's ‘categories’ serve as preconditions of empirical knowledge, they too are a form of knowledge and therefore cannot but entail a point of view. The truth of Kant's usage of ‘transcendental’ (that is, its metalogical status) notwithstanding, unless Kant is claiming that this point of view is no less than God's, it must be characterized, as with all points of view, as particular and therefore as less than autonomous.
During his lifetime, Kant did not escape criticism of this kind. Most notably, Johann Georg Hamann, a compatriot of Kant and fellow inhabitant of the city of Königsberg, developed a substantial critique in reaction to Kant's critique of reason, one that had an enormous influence on both eighteenth- and nineteenth-century thinkers (Beiser 1987: chap. 1). But I wish to focus on the twentieth century, when the idea of the synthetic a priori was subjected to further extensive revision, entailing fundamental rather than formal confusion of the a priori and the synthetic. Thus, for example, Saul Kripke (1980), in his brilliant Naming and Necessity, differentiates necessity from analyticity by arguing that referents are fixed by names rather than descriptions. In other words, for Kripke, a spade is a spade because it has been so called (or, as he says, “rigidly designated”), this name being passed on from one speaker to another, rather than because a spade fits a certain criterial description (the description could be variable or even wrong). Here, as in the biblical book of Genesis, the identity of a thing originates with its name, resulting in a truth that is a posteriori but also quite necessary (a spade is indeed a spade, no matter if, say, under changed circumstances, it is no longer found to be black). If a truth is both necessary and dependent upon concrete evidence for its warrant, a matter of sensory perception and yet inevitable, it would seem to play loose with the difference between the a priori and the a posteriori.
Again, for example, Harvard philosopher W V. O. Quine (1953) argues, in his well-known essay, “Two Dogmas of Empiricism,” for a gradualistic rather than dualistic distinction between the analytic and the synthetic. As he sees it, from the point of view of pragmatism, every system of logic in the end must give way at its edges to the lessons of experience, making its analyticity distinctly relative. If Quine is right, then there is—in the final instance—no logically necessary a priori but only a synthetic one.
In another place (Evens 1983), writing on the efficacy of the Nuer incest prohibition, I drew on Quine's argument about epistemological gradualism as between logic and experience, in order to propose a nondualist solution to the so-called anthropological problem of primitive mentality. My solution centered on a notion of half-logic, predicated on the ontological thesis of basic ambiguity, and capacitated therefore to endorse, in a subtle but forthright way, the apparently magical possibility of a rule or convention that enjoys the force of nature. Here, however, in order to ground my ideas directly in a phenomenological theory of perception, it is fitting to appeal to the thought of M. Merleau-Ponty and Ludwig Wittgenstein, both giants of twentieth-century philosophy. Their work readily lends itself to discussion in terms that leave no doubt as to the essential relevance of the idea of the a priori to professional anthropology.
At one point in his career, Ludwig Wittgenstein asked his student and friend, M. O'C. Drury, to read to him from J. G. Frazer's The Golden Bough. They read only from the first volume, not getting very far because of the profusion of Wittgenstein's critical remarks. His commentary, soon developed by him in writing, was eventually published as “Remarks on Frazer's Golden Bough” In this terse essay, Wittgenstein (1971) offers scathing criticism of Frazer's anthropological understanding of magical and religious practices among so-called primitives, and in doing so he sets out his own conception of how such practices might best be understood. His