I applaud the arguments in Lepore and Stone (2015) that Gricean, Neo-Gricean, and Relevance theories of conversational implicature and utterance interpretation are deeply flawed because the additional meanings speakers convey when using sentences are conventional rather than calculable. I then go on to rebut several conclusions Lepore and Stone endorse that do not follow: that there is no such thing as conversational implicature; that in figurative speech speakers do not mean anything beyond what the sentences they utter mean; that anything a speaker means is something the speaker directly intends and says; and that any meanings conveyed conventionally are given by the grammar or semantics of the language. Along the way, I argue that conventions are constituted by certain causal processes, not mutual expectations, and I distinguish two types of speaker meaning.