In my previous posts on Brandom, I may have come across as unduly critical, or as dismissive. That couldn’t be further from the truth. There is much that I find in Brandom’s work that is important, and that I will continue to draw from. In my previous posts I have been homing in on the differend as I see it between the positions I am working through and those developed by Brandom. Those differences have only been put forward in a rough and ready manner – this is, after all, just a working blog – and thus I’ve been thankful for comments from Pete of the deontologistics blog (his comments can be found here). They have forced me to clarify some of my points further. Before wrapping up on Brandom I want to list a few more thoughts that seem to be in need of further development. I’ll begin with where I most agree with Brandom.
By far the area where I find Brandom developing his thoughts in a trajectory I can follow is in his efforts to develop a nonnatural and nonreductive semantics, or theory of concepts. That this is what Brandom is up to has been recognized by many readers of Brandom. Daniel Dennett, for example (in a review of Making it Explicit that can be found here), takes Brandom to task for precisely this reason. He sees aspects of Brandom’s approach as entailing an ‘unacceptable flight from naturalism,’ and all in order to avoid the ‘collapse,’ as Dennett puts it, of the normative into the nonnormative (he cites as further evidence for this aversion the complete absence of references to ‘evolution’ in Making it Explicit). As Brandom sees it, however, Dennett is continuing to follow the either/or mandate that can be traced to the tradition from Hume through Quine, which ‘offers the friends of modality a stark choice: either show how to explain modalities in nonmodal terms or learn to live without them.’ Dennett opts for the former whereas Brandom sees the entire mandate as a false dilemma. As Brandom sees it, our use of modality in discourse is not reducible to the form of natural-scientific predictive utility, understood here, as Ruth Millikan does for instance, as a consequence of the norms of proper functioning of reproducing biological systems that have evolved by means of natural selection; rather, and as has been discussed in previous posts, our ability to use normative vocabulary derives from the community. Dennett finds this ‘bootstrapping’ approach incapable of resolving the regress of how we come to use normative vocabulary. As Dennett argues, ‘Community is Brandom’s skyhook and he can’t have it. He knows this, but he prefers not to dwell on what it would take to secure community as his base of operations.’ This is much in line with my critique of Brandom’s reliance upon a Hegelian social holism as the transcendental condition (see earlier posts); however, rather than return to a naturalist, reductive account of the normative as Dennett does, I will follow Brandom and his effort to avoid the Hume/Quine either/or mandate (that is on the traditional reading of Hume which I don’t follow). In particular, if one were to follow Dennett, Millikan, et. al., then semantics would then become a specialized form of natural science, or yet another instance where a philosophical problematic spins off a scientific discipline. For Brandom, however,
…semantics is not just one more special science. It is (also) a philosophical discipline…[which entails] understanding and if need be [the] criticism of the concepts employed by other disciplines [including the natural sciences]. (in “Modality, Normativity, and Intentionality,” p. 599).
Brandom is correct to push this distinction; however, I’m not sure that the normative direction is the best way to go to cash out the promise I think upholding this distinction holds. To situate this discussion in terms I used in an earlier post where I discuss Nagarjuna and the concept of emptiness (sunyata), I agree that concepts should not be reduced to being nothing other than their inferential relations to one another, for this would undermine the object-ivity of the concept. This is why I think Brandom calls upon the object-ivity of the Community and the holism this entails in order to provide the objective basis that would be lacking if concepts were nothing other than their inferential connections to other concepts. Similarly, normativity cannot be reduced to natural scientific relations, relations that map the predictive utility of an autonomous, in-itself reality, for here too the normative “oughts,” the potential for being mistaken and hence for being other, becomes reduced to being simply a state of affairs of what is. Rather, concepts, and the normativity they entail, are inseparable from states of affairs; they are, moreover, the conditions for identifying problems for which there can be mistaken or faulty solutions (Dennett uses the term faulty to provide for a naturalist account of normativity). But concepts are not to be confused with these states of affairs, or with the predictive functions that map relations between states of affairs. On this point, I largely follow Deleuze and Guattari’s critique of Chomskian grammar:
When linguists (following Chomsky) rise to the idea of a purely language-based abstract machine, our immediate objection is that their machine, far from being too abstract, is not abstract enough because it is limited to the form of expression and to alleged universals that presuppose language…The abstract machine is pure Matter-Function–a diagram independent of the forms and substances, expressions and contents it will distribute. (A Thousand Plateaus, p. 156).
When concepts are reduced to being functional mappings of the identifiable states of affairs in the world–much as Chomskian grammar functions as an abstract mapping of the identifiable languages one finds spoken in the world–then concepts are not abstract enough. They are predetermined by the objective states of affairs, whether the ‘alleged universals that presuppose language,’ or the objectivity of the Community. Concepts are thus not reducible to the abstract mappings of science, and hence the importance of Brandom’s effort to develop a nonreductive semantics. At the same time, however, I find the normative semantics Brandom develops to be susceptible to the Deleuzo-Guattarian critique just cited, and believe that a problematizing semantics may be better able to cash out much of what Brandom seeks to do. A problematizing semantics is problematizing precisely because it resists codification in terms of a predetermining, already established identity, whether in terms of already established languages for Chomsky or the Community for Brandom. For Deleuze what we have instead is the event, or multiplicities. From Deleuze’s perspective, a semantics needs to be understood not in terms of established uses but rather in terms of the faculties that engender established uses, and more importantly in terms of the problems that force the faculties to employ the diagrams cited in the previous citation. As Deleuze puts this in Difference and Repetition,
Take, for example, the linguistic multiplicity, regarded as a virtual system of reciprocal connections between ‘phonemes’ which is incarnated in the actual terms and relations of diverse languages: such a multiplicity renders possible speech as a faculty as well as the transcendent object of that speech, that ‘metalanguage’ which cannot be spoken in the empirical usage of a given language, but must be spoken and can be spoken only in the poetic usage of speech coextensive with virtuality. Difference and Repetition, p. 193.
The phonemes are certainly real – hence the reality of the virtual as Deleuze discusses it – but their reality is hidden (or withdrawn to adopt Levi Bryant’s term) within the actualization of the ‘terms and relations of diverse languages’ as these are actually spoken, and empirically available. Again the phonemes are themselves empirically available as well, as any linguist who studies them will tell you, but it is precisely the relationship between these phonemes as a ‘virtual system’ and their actualization in a language that accounts for their being a multiplicity that is irreducible to the actual terms and relations of a given language. In short, the ‘virtual system’ of phonemes is irreducible to an empirical, natural presentation of the actual languages (and a similar virtual system would be at work with respect to the biological organism’s capacities to utter the sounds that it can). At the same time there is the ‘transcendent object,’ the object that is irreducible to the relations between terms, both the virtual system of relations between phonemes and the actual relations of the words of a language. This transcendent object cannot be spoken within the empirical usage of the language – it is incapable of a reductive, naturalist account – but can only be spoken, as Deleuze puts it, ‘in the poetic usage of speech.’ This poetic usage of speech is the problematizing semantics that unsettles established usage, and yet it is precisely the poetic usage, and the transcendent object inseparable from established usage, that allows for the possibility of the norms and standards that come to be identified with the ‘social achievements’ (as Brandom puts it) of the community. To sum this all up as pithily as I can: aesthetics is first philosophy.