In this thesis, I examine a particular kind of breakdown that occurs when a person with relatively little social power is unable to do ordinary things with their words. Specifically, I look at a model for understanding unjust pragmatic dysfunction, which is a kind of wrong that belongs to a wider class of injustices that are experienced through or in virtue of, language. This wider class of breakdowns I will call “linguistic injustice”. I will suggest that judgements about what a speaker is trying to do with their words can be undermined by culturally shared values and norms of behavior. I start with the work of Rae Langton, who puts forward a schema for which to understand silencing qua speech acts (the actions performed through and by speech), that builds upon J.L. Austin’s 1962 work How To Do Things with Words. I move on to present an altered account of these instances that attempts to describe how addressees are unable to accurately understand the speech acts of people from traditionally oppressed groups. Finally, I bring in sociolinguistic tools in order to help understand the mechanics of those failures I will call “illocutionary deafening”. This will allow for the careful examination of the way in which gender, race and class (among other factors) can lead to addressees’ failure to understand restrained black men, in the back of police cars—who are saying “I can’t breathe”—as genuine pleas for medical assistance.