Expressivism, inferentialism, and the status of attitudes

The aim of this paper is to show that expressivism about attitudes is not a tenable position. Although this claim has been often made in the literature, traditional arguments against attitudinal expressivism assumed a dated form of expressivism. In order to show that ascriptions of attitudes cannot be seen as expressive in either of them, the paper discusses two more recent versions of expressivism: quasi-realism and inferentialist expressivism. Quasi-realism escapes traditional arguments against attitudinal expressivism because it allows expressive statements to be true in a deflationary sense. Nonetheless, it is implausible to adopt quasi-realist expressivism with regard to attitudes because this position cannot avoid substantial psychological commitments, despite Toppinen's recent arguments to the contrary. This conclusion also pertains to quasi-realist hybrid expressivist conceptions. The recently developed conception of inferentialist expressivism arguably creates the most accommodating theory for attitudinal expressivism as it does not make any substantial psychological commitments. Still, the criteria of bifurcation assumed in this framework does not support the claim that attitudes play an expressive role.