THIS MACHINE KILLS FASCISTS: DETECTING PROPAGANDA WITH FORMAL MODELS OF MASS DISCOURSE
I bring work in formal pragmatics to bear on mass discourse analysis. It's a common observation that mass media messaging has profound effect on the formation of public opinion and the shaping of political behavior. Among the ways we might conceive of mass media's effects is one on which such messaging effectively issues a "deliberative script" vis-à-vis issues of public importance. One insight of this project is that the script it provides is a function not only of the lexical items selected and individual speech acts performed, but also of the structure of those discourses proffered for mass consumption. Models which construe information structure in terms of sub- and super-questions nicely capture the contours of a deliberative script and so render tractable the project of high-resolution comparison between rival scripts. In many instances of heated public debate, we see opponents not merely giving different answers to particular questions, but adopting very different deliberative strategies; addressing the question of appropriate health care policy, the liberal will raise the question of whether there exists a positive right to such care, while the conservative will raise the question of how much coercive force the state has any business imposing. For a variety of reasons, including the vast existing empirical work on the corrosive effects of extreme political polarization, and theoretical considerations around the source of legal or governmental authority (e.g. the Rawlsian notion of public justification), I argue that divergent deliberative scripts should be of interest to the political theorist interested in the health and legitimacy of democracies. I then develop a model of mass discourse which makes use of foregoing work on the information structure of non-mass discourses, and in terms of which mass discourse pathologies related to the above notion of highly divergent deliberative scripts are precisified.
A SELECTION OF BISCUITS: CONDITIONALS AND RELEVANCE
January 25, 2025
While some conditionals assert a relationship between their antecedents and consequents (e.g. If it's sunny tomorrow, I'll go to park), others, so-called 'Biscuit Conditionals', have been analysed as establishing a relationship between the truth of the antecedent and the act of asserting the consequent (classically: There are biscuits in the kitchen if you're hungry). Recent treatments of biscuit conditionals have explained the existence of such a speech act in terms of relevance preservation. I draw attention to the fact that these conditionals seem utterable even under conditions where relevance is highly unconstrained (as in discourse-initial utterances). I argue that, in such cases, it is the information-structural update which is rendered conditional on the truth of the antecedent.