We use words to communicate, enact laws, and make promises, but we also insult, discriminate, and subordinate gender, ethnic, and sexual minorities. Recently, Rae Langton and Mary Kate McGowan, appealing to speech act theory have argued for the "Constitutes Thesis"--our words do not merely cause the harms of oppression, but are in and of themselves acts of oppression. Opponents of the thesis criticize it as "philosophical sleight of hand," "metaphorical," and "strikingly implausible" because utterances are dissimilar to paradigmatic acts of oppression: words neither lynch or exploit, nor prevent the development of one's rational capacities. In How We Do Bad Things with Words, I address these concerns and advance my own account of the Constitutes Thesis, appealing to the interplay between speech acts, social norms, and institutions. Our words oppress by engaging the norms of oppressive social practices.

I begin by investigating whether exercitive models of oppressive speech are adequate for explaining how some speech constitutes oppression. I argue that exercitive models fail to characterize the full range of oppressive speech because these models focus on one speech act type as the source of oppressive speech and privilege the enactment of oppressive policies (Chapter 1). Then, I argue for my account, the 3Cs model, which characterizes oppressive speech in terms of a wide range--ceremonious, communicative, and collateral--speech acts (Chapter 2). However, speech act theory alone is not sufficient for defending the Constitutes Thesis. The critics of the thesis are right in this regard, so I provide a framework, the Constitutes-Iteration analysis, that demonstrates which theoretical resources are required for explaining how speech constitutes oppression (Chapter 3). We need a theory of oppression and a theory of social practices. I then offer a sketch of such a theory, maintaining that an action is a contribution to a social practice when it either accords with the conventions or social norms of the practice, or legitimizes these conventions and norms (Chapter 4). Finally, I argue that some utterances oppress by engaging these norms and conventions, and defend my account against objections (Chapter 5).