C. examines a class of sentences (viz. 'let's'-sentences) which appear to be excellent candidates for a Gricean analysis (in terms of conversational implicature), but he shows that such an analysis would be incorrect, which suggests some problems for the applicability of the Gricean approach.