Judging Plausible Deniability

Plausible deniability (PD) can be used strategically to convey content without dealing with the repercussions of asserting it explicitly.

For instance, one might say “I admire your courage for wearing this dress” to convey an insult without incurring the social costs of explicit aggression. PD underlies such important phenomena as politeness, covert threats, political dog whistles, and flirting. Despite this, PD is understudied and as of yet imperfectly understood.

Arguably, PD cannot be reduced to the mere probability of the relevant interpretation.

Rather, it also relies on subtle social structures, such as power dynamics, attribution of responsibility and commitment, and their interactions with the linguistic expression of content.

This makes PD an extremely complex and fascinating phenomenon from the point of view of linguistics, philosophy of language and, more generally, the cognitive and social sciences.

The PlauDe (Plausible Deniability) project aims to develop a unified account of PD that untangles its conceptual components and unifies its various guises under one theoretical umbrella.

PlauDe will develop a formal computational model of PD to achieve a high degree of clarity and precision and generate quantitative empirical predictions that will be tested experimentally.

The model will improve over previous accounts by analyzing PD as a compromise between conveying content to another language user and avoiding judgment by a judge or bystander, be it concrete or imagined.

This model will be used to systematically analyze real-world cases of manipulative language use with the aim of disseminating the judge-based account, which is not only a technical innovation but will also form the basis for a non-technical explanation of the phenomenon, resulting in a workshop to help entities vulnerable to PD-based linguistic strategies.