Diogo Santos

LANCOG, University of Lisbon

Amending Assessment-Sensitivity

22 March 2019, 16:00

Faculdade de Letras de Lisboa

Sala Mattos Romão (Departamento de Filosofia)

Abstract: According to Dummett (1978) to understand the point of truth is to understand the normative role it plays in the governing of our asserting practices. Dummett’s approach has influenced Assessment-sensitivity views (AS) (e.g. Egan, 2007; MacFarlane, 2011, 2014). AS holds that truth and, hence, the correctness of making and withdrawing assertions is assessment-sensitive. What practically distinguishes this theory from its rivals is its claim about the normative role of truth in the withdrawal of assertions. According to the view, an agent in C2 is obliged to retract an (unretracted) assertion that p made in context C1 if p is not true as originally used (in C1) and assessed from C2. Crucially, the retraction rule renders that an agent is sometimes obliged to retract an assertion that was correct for her to make. Recent experimental data (Dinges & Zakkou, Fintel & Gillies, Kneer, Knobe & Yalcin, Marques) on discourse about personal taste and epistemic modals show that AS’s retraction predictions are in conflict with ordinary speakers’ intuitions. This greatly undermines the purported empirical support for AS. The experimental findings indicate that there is no empirical support for a retraction rule for assertions and that retraction and truth come apart. In this paper I diagnose why AS’s predictions conflict with the empirical data and explain what is wrong with the theory’s depiction of the normative role of truth in the withdrawal of assertions. The diagnosis importantly relies on the claim that retraction is not the only exercitive that agents may use to withdraw the assertoric commitments undertaken by the original assertion – something that those involved in the debate have overlooked.