Date: 13 July, 1993


In 1993, I submitted a reworked portion of Chapter 5 "II. Thought and Conventional Language" of my dissertation to the Australasian Journal of Philosophy. It was rejected on the basis of the following report by an anonymous referee. The report raises a number of issues which need further discussion. One of these is whether it is possible to interpret Sellars as holding a, more or less, unified view; or whether we should view Sellars as undergoing a change of mind.

The central claim of the paper is that many critics of Sellars have misinterpreted his thesis (N) that "all awareness . . . is a linguistic affair." Contrary to what these critics believe, Sellars did not intend to deny the existence of a "prelinguistic awareness and thought" (p. 3). This misinterpretation, according to the author, results from failing to take notice of Sellars' commitment to a distinction between conceptual and nonconceptual thinking and awareness, a distinction implicit in a number of passages cited by the author and explained in Sellars' late paper "Mental Events" (1981) in terms of the distinction between a Language and a Representational System. In view of this distinction, it is plausible to suppose that in asserting (N) Sellars merely wished to deny that conceptual awareness can exist prior to language. Hence, the author concludes, there is no reason to suppose that Sellars is committed to "linguistic idealism", the view that all awareness is linguistically mediated.

There are three main problems with the paper:

(1) It misplaces the real point of the controversy about (N). The controversial nature of (N) is not alleviated by recognizing that for Sellars there can be nonconceptual awareness prior to (independent of) language. In the context of the argument in EPM against "the myth of the given", nonconceptual awareness for Sellars amounts to sensing a nonintentional phenomenon (to be sharply distinguished from thinking); and to say that sensing is language independent is a rather uninteresting claim. (N), a thesis which Sellars labels a 'psychological nominalism', remains fully controversial in the eyes of many critics even if it is restricted to conceptual thinking or awareness. Theorists of empirical leanings (including constructionists like Piaget or connectionists like Smolensky) have strenuously denied that the use of concepts is language dependent.

(2) The author does not seem to appreciate the difference between the ontogenetic claim that there is thought (awareness) prior to the acquisition of language, and the metaphysical claim that thought is essentially language-like (structurally, functionally, etc.), i.e. that there is a language of thought (in Fodor's sense). Thus he takes the claim (b) "there is a 'language of thought' [in Fodor's sense] prior to the language of speech" to be a corollary of the claim (a) "awareness and thought [are] genetically prior to [the acquisition of] language." But (b) is neither entailed nor presupposed by (a). Thought may have a language-like structure (e.g. compositionality, recursiveness, etc.) whether or not, ontogenetically, it develops in the child prior to the acquisition of a conventional language. Sellars' acceptance of the LOT hypothesis (b) is part and parcel of his conception of thought (conceptual awareness) as analogous to speech (thought is a sort of 'silent speech', and speech is 'thought-out-loud'). But Sellars' acceptance of (a) is only relative to nonconceptual 'thought', a sense of 'thought' for which there is no language of thought.

(3) The author tends to read out of context some passages Sellars cited in support of his interpretation. As mentioned, (N) occurs in EPM (1956); to support an interpretation is at least uncautious without a careful assessment of whether Sellars' views have changed, or whether his later statements merely clarify or extend his earlier views, or indeed, whether they are consistent with those views. Take for example the passage from Sellars (1975) cited on p. 9, where the issue of subconceptual thinking (in babies and animals) comes up. It is not at all clear that it supports the author's interpretation of (N). Is a "'rudimentary' form of conceptual thinking" still conceptual thinking in a sense that seriously challenges the critics' interpretation of (N), given especially the context in which (N) occurs? Sellars' statement "We interpret their [babies' and animals'] behavior using conceptual thinking as a model but qualify this model in ad hoc and unsystematic ways which really amounts to the introduction of a new notion which is nevertheless labeled 'thinking'" does not seem to support the claim that Sellars did not (in 1975, much less in 1956) hold the view that all thinking, all awareness in the proper and interesting (intentional) sense of the word, is a linguistic affair. Nor does the quotation from "Mental Events" (1981) cited on p. 5 support the author's interpretation. The point there is simply that (conceptual) thoughts are not literally linguistic events though all (nonparroting) linguistic events are thoughts (i.e., thoughts-out-loud). This is consistent with saying that all (silent) thoughts are language-like, i.e. occur in a medium (mentalese) which is analogous to speech.

[Return to Forum]