Robbins (2000) raises several objections to the characterisation of nonconceptual content in "The Paradox of Self-Consciousness" (Bermudez 1998). He attacks what I called the Priority Principle (PP), namely, that mastery of a language is a necessary condition of concept mastery. In this reply I explain why I remain unconvinced by his objections and offer a further argument in support of PP.
2. Robbins (2000) characterises my argument in support of the Priority Principle as resting upon the following premise:
P2. The capacity for rational judgement requires both (a) the capacity to form warranted beliefs and (b) the capacity to identify and articulate warrants for at least some of those beliefs.
It is because, on the one hand, mastery of a concept requires the ability to employ that concept in rational judgments, and hence the capacity to identify warrants for judgements, and, on the other hand, the capacity to identify warrants requires linguistic competence, that I think that concept possession is constitutively linked to language mastery.
3. Robbins objects that P2 is too strong. Why should a concept- possessor be able to identify warrants for beliefs involving that concept? Why can we not appeal to some form of epistemological externalism (such as reliabilism, for example) that would allow a belief to be warranted even when the believer is completely unaware of what makes it warranted? Robbins points out that there is no reason why a believer should always be able to identify whether and in virtue of what a particular belief is warranted. Why should we insist on it even in a single case?.
4. I am fully in agreement with Robbins that there is no reason for endorsing either of the following two principles: (A) If a believer has a justified belief then he must believe that his belief is justified. (B) If a believer has a justified belief then he must be able to articulate what makes his belief justified (by producing some sort of argument leading up to the conclusion that his belief is justified). So why then do I accept (P2)?
5. It is, I think, a basic fact that possessing a concept requires appreciating when it is appropriate to apply a concept and what inferences it is appropriate to draw from that concept. It is not enough simply to be disposed to apply the concept in the right circumstances, or to draw the right sort of inferences. A concept-user must apply the concept in the right circumstances because he appreciates that those are the right circumstances - and he must draw the appropriate inferences because he appreciates that they are appropriate. It is not enough for reasoners to be credited with mastery of the concept IF. . . THEN, that they should simply possess a brute disposition to conclude Q from the premises P and IF P THEN Q. The reasoner must make the inference because they perceive it as a valid inference [NOTE 1].
6. It is open to Robbins (or anyone else!) to object that imposing this requirement on concept possession begs the question. Perhaps it does. It seems to me, however, that there is a clear distinction between two different types of thinking. There is, on the one hand, thinking that involves a sense of the normative commitments incurred by making certain kinds of judgement and a sense of the appropriate grounds for making those inferences and, on the other thinking that does not. It is only the first of these that I would describe as conceptual. But disputes over this may well just be terminological. The important point is that we recognise the difference, not how we label it [NOTE 2].
7. Robbins appeals to considerations of parsimony in support of the Conceptual Requirement Principle. We should not multiply types of thinking beyond necessity, so let us stick with the principle that all thought is conceptual as the default option. The general thought that we ought to be parsimonious is unimpeachable. But the problem is that we cannot be as parsimonious as Robbins would like us to be. Quite apart from the general distinction made in the previous paragraph we also need to distinguish between thinking that is properly systematic and generative (that meets what Gareth Evans called the Generality Constraint) and thinking that involves only a limited degree of recombinability. Most theories of domain-specific modules dedicated, for example, to social cognition or intuitive physics fall squarely into the second group. It is, of course, precisely to modular thinking of this kind that appeal is most often made in characterising the types of thinking engaged in by non-linguistic animals and prelinguistic infants.
8. Let me end by sketching out a further line of argument in support of the Priority Principle, one that did not feature in the book. Once again, the premise of the argument is what Robbins calls Principle P2, namely, that concept possession requires the ability to identify and articulate warrants for at least some beliefs. Articulating such warrants will involve making second-order judgements about the inferential relations between thoughts and other mental states (such as perceptions) as well as between thoughts tout court. That it is to say, it will involve intentional ascent. Intentional ascent, however, requires holding a thought in mind in a way that reveals its structure (otherwise it would be impossible to evaluate its inferential connections with other thoughts). This, it seems to me, is possible only if the thought is linguistically vehicled. At least, we have no grasp on any other way in which a thought can be held in mind in such a way that it can be the object of further thoughts. The argument, therefore, is that intentional ascent requires semantic ascent.
9. If this argument from intentional ascent to semantic ascent is sound then it provides a further way of understanding the differences between linguistic thought and non-linguistic thought, as there will be certain types of thinking that are in principle unavailable to non-linguistic creatures, namely, all those involving intentional ascent, or what is sometimes called metarepresentation. This category will include thoughts about modality (taking modal operators to apply to sentences) and second-order desires, as well as certain types of belief revision. But the most interesting type of thinking that seems to require language (because it requires intentional ascent) is the ascription to other subjects of mental states. This raises the question of how the complicated forms of psychological awareness and psychological self-awareness discussed in Chapter 9 of Bermudez (1998) can be possible at the nonlinguistic and non-conceptual level. I will address this question in my forthcoming reply to Gary Fuller and Carol Slater.
 For further discussion see Bermudez (1999).
 I don't actually think that the issue is simply terminological. The most plausible accounts we have of the nature of concepts analyse them in terms of their role in the first type of thinking. See Peacocke (1992), for example.
Bermudez, J. L. (1998) The Paradox of Self-Consciousness. Cambridge MA. MIT Press.
Bermudez, J. L. (1999a) Precis of "The Paradox of Self-Consciousness". PSYCOLOQUY 10(035) ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1999.volume.10/ psyc.99.10.035.self-consciousness.1.bermudez http://www.cogsci.soton.ac.uk/cgi/psyc/newpsy?10.035
Bermudez, J. L. (1999b) 'Naturalism and Conceptual Norms', Philosophical Quarterly 49, 77-85
Peacocke, C. (1992) A Study of Concepts. Cambridge MA. MIT Press.
Robbins, P. (2000) Paradox Twice Lost. PSYCOLOQUY 11(057) ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/2000.volume.11/ psyc.00.11.057.self-consciousness.8.robbins http://www.cogsci.soton.ac.uk/cgi/psyc/newpsy?11.057