Game theory, cheap talk and post‐truth politics: David Lewis vs. John Searle on reasons for truth‐telling

Journal for the Theory of Social Behaviour 48 (3):306-329 (2018)
  Copy   BIBTEX


I offer two potential diagnoses of the behavioral norms governing post‐truth politics by comparing the view of language, communication, and truth‐telling put forward by David Lewis (extended by game theorists), and John Searle. My first goal is to specify the different ways in which Lewis, and game theorists more generally, in contrast to Searle (in the company of Paul Grice and Jurgen Habermas), go about explaining the normativity of truthfulness within a linguistic community. The main difference is that for Lewis and game theorists, “truthful” signaling follows from an align- ment of interests, and deception follows from mixed motives leading to the calculation that sending false information is better for oneself. Following in the Enlightenment tradition, Searle argues that practical reasoning, which involves mastery of at least one language, requires that actors intend to communicate. This intention includes constraining the content of statements to uphold veracity conditions. After distinguishing between these two accounts, I will artic- ulate the implications for explaining, and even informing actions, constitutive of post‐truth politics. I argue that the strategic view of communication is suffi- cient neither to model everyday conversation nor to reflect a public sphere useful for democratic govern- ment. Both the pedagogy of strategic communication as cheap talk, and its concordance with new digital information technologies, challenge norms of truthful- ness that underlie modern institutions essential to an effective public sphere.

Author's Profile

S. M. Amadae
Massachusetts Institute of Technology


Added to PP

521 (#27,285)

6 months
223 (#8,528)

Historical graph of downloads since first upload
This graph includes both downloads from PhilArchive and clicks on external links on PhilPapers.
How can I increase my downloads?