The weakness of must: In defense of a Mantra

Abstract

<p class="p1">Many semanticists have claimed that must’s meaning is weaker than</p><p class="p1">epistemic necessity, a claim that <span class="s1">von Fintel & Gillies </span>(<span class="s1">2010</span>) dub “The Mantra”.</p><p class="p1">Recently <span class="s1">von Fintel & Gillies </span>have argued in an influential paper that the Mantra is</p><p class="p1">false, and that the intuitions that have driven it can be accounted for by appealing to</p><p class="p1">evidential meaning. I show that <span class="s1">von Fintel & Gillies </span>do not provide a compelling</p><p class="p1">argument against the Mantra, and that their theory of evidential meaning, while</p><p class="p1">promising in certain respects, also has serious empirical and conceptual problems.</p><p class="p1">In addition, a variety of corpus examples indicate that speakers who assert must</p><p class="p1">p are not always maximally confident in the truth of p. As an alternative, I reimplement</p><p class="p1"><span class="s1">von Fintel & Gillies</span>’ theory of indirect evidentiality in a probabilistic,</p><p class="p1">Mantra-compatible framework. Ultimately, both sides of the debate are partly right:</p><p class="p1">must is weak in several respects, but it also encodes an indirect evidential meaning.</p>