Research Group
in Analytic Philosophy

Abstract

According to an influential proposal by Williamson (1996), assertions are subject to a single rule of the form "one must: assert p only if p has C”. This elicits the question: what is property C? The most prominent responses are “factive”: they specify C so that an assertion is permissible only if it is true (e.g., assert p only if p is true, assert p only if you know p). Factive accounts are opposed to non-factive ones. Non-factive accounts maintain that some untrue assertions are permissible, and endorse a weaker standard for asserting (e.g. assert p only if you reasonably believe that p).  The main rationale behind non-factive accounts are “getterised” false assertions: false assertions that the speaker has the best reasons to believe to be true. These assertions seem permissible qua assertions, against the predictions of factive accounts.  In this paper, I will attempt a defence of non-factive accounts. To do so, I will focus on two putative reasons to resist them. First, that gettierised false assertions may be explainable within a factivist framework. Second, that non-factive accounts have no resources to explain the incorrectness of untrue assertions.  After rejecting the first claim, I will propose a solution to address the second: pairing non-factive accounts with the old-fashioned idea that assertions aim at truth. This provides non-factive accounts with the resources to explain the ‘incorrectness’ of untrue assertions: as a failure to fulfil the purported aim of assertion, rather than as a failure to fulfil its norm.