<p>In <i>Reasoning with Attitude</i>, Luca Incurvati and Julian Schlöder defend what they call inferential expressivism: a proposed synthesis of inferentialism and expressivism, according to which semantic values are given by inferential relations between attitude expressions. In this contribution, I examine how far inferential expressivism departs from more traditional forms of inferentialism and, insofar as it does diverge, put pressure on the claim that inferentialists must become inferential expressivists in order to explain a wide range of puzzling target notions. I do so by focusing on what I call the ‘Many Meanings Problem’: the challenge of accounting for semantic distinctions both across and within domains. Oversimplifying greatly, this problem generates the following dilemma. Either, inferential expressivism explains all these many meanings by appealing to an expansive repertoire of speech acts, or it keeps that repertoire sparse and lets other inferential resources do the fine-grained explanatory work. The first horn earns inferential expressivism its keep, but risks bringing it into conflict with its own methodological commitments. The second generates no such conflict, but threatens to undermine the motivation for adopting inferential expressivism in the first place. Still, my conclusion is conciliatory rather than negative: I suggest that inferential expressivism is best understood not as a rival to traditional inferentialism, but as a close collaborator engaged in advancing a shared inferentialist project.</p>