How Language is Used to Oppress (HaLO)

Oppressive speech is speech that harms and disempowers its targets. It includes hate speech (slurs “Yid”, “Spic”), derogatory attitudes (“migrant vermin”), and negative moral talk (“gypsies are feckless”). Such speech doesn’t just offend but seeks to harm the target during the conversation, so that they feel humiliated, attacked, denigrated, and silenced. This harm is not confined to conversation. Utterances shift social norms by changing people’s opinions about how targets ought to be treated. Oppressive speech uses this effect to establish and maintain oppressive norms. Recent rhetoric in politics is a vivid example of this. This is a significant cause of economic and social harm to many groups. Oppressive speech is thus one of the most urgent social and political issues of our time.

There is currently no simple, unifying framework that models oppressive speech or explains its many effects.  This project seeks to address this gap through a multidisciplinary approach that combines the complementary strengths of game theory, and theories of social norms and social injustice. Such an approach is essential because philosophers of language have typically modelled communication as the exchange and evaluation of propositions, without paying sufficient attention to: (a) what the utterances are used to do in the conversation; (b) what the utterances are used to do socially; and (c) how the social structures affect both these processes. The proposed approach will provide a unified account of these phenomena for oppressive speech. Without theorising about what utterances are used to do, rather than just what they mean, we cannot explain the social effects highlighted above.

The core insight of the project is that there is a two-level hierarchy of games: conversational games are embedded within a larger social game. Oppressive actors in conversational games seek to acquire power in the larger social game to achieve later payoffs (e.g. an unfair distribution of economic resources). I hypothesise that this hierarchical game can explain: (i) the effects of oppressive speech within a conversational game; (ii) the way speech can shift norms that govern the social game; and (iii) how the desire to maintain social injustice motivates oppressive speech. 

The objectives of the project are three interlinked parts of a new model of oppressive speech:

  • 1) Deliver an initial model of oppressive speech as a conversational game in which speakers make utterances that change the rules of the game. These changes are intended to shift power from targets to the speaker. This deals with the lowest level game.
  • 2) Explain the harmful social effects on targets, by explaining how oppressive speech sets up unjust social norms. This explains how the conversational game embeds in the social game and how its effects leak out into the social game. 
  • 3) Explain how unjust social structures affect speech, what motivates agents in a group to oppressive acts, and what is needed to change harmful norms. This explains how payoffs in the social game motivate oppressive actions in the conversational game and how unjust social structures prime the conversational game to enable these actions.


MSCA Fellowship


1.4.2019 - 30.03.2021


Dr. Mihaela Popa-Wyatt