Epistemic norms governing assertion

Suppose a friend asserts to you that an event you were planning to attend together has just been cancelled due to an asbestos hazard.You accordingly cancel your taxi, change into your pyjamas and stay in for the night.The next day, however, you come to find out that the event wasn't actu­ally cancelled, after all! It turns out you friend had simply made it up, asbestos and all, because he was feeling too lazy to make the effort.

There are many ways in which your friend's assertion is criticisable. One of them is morally. From a perspective from which what matters are things like being a good person and doing the right thing, your friend shouldn't have asserted what he did. But let's bracket the moral shortfalls of your friend's assertion and home in on an entirely different problem, one that has to do with (put roughly) your friend's take on the accuracy of what he was saying.

According to one very popular view in speech act theory, in the philosophy of language, as well as in social epistemology, your friend's assertion—one that is not only false but based on no evidence whatsoever—is going to count as defective for reasons that are entirely epistemic.² On this way of thinking, assertion, as a kind of speech act—viz., a speech act by which a sentence is uttered in the indicative or declarative mood—is normatively constrained in the sense that one should assert something,A, only if one possesses some (to-be-specified) epistemic credential (e.g., true belief, justified belief, knowledge), with respect to A.The matter of what this creden­tial is supposed to be is controversial. But to the extent that this general idea is on the right track, an implication is that your friend may be understood as lacking the epistemic authority to assert what he did (fabricating on the basis of no evidence whatsoever is, after all, not asserting on the basis of a very good epistemic credential).

One of the first philosophers to appreciate that assertions are subject to this kind of criticism was G.E. Moore (1944, pp. 524—5), who registered that it seems paradoxical to assert a sentence such as the following:

(1) It is raining, but I don't believe that it is raining.

Asserting (1) would certainly sound odd. But here we need to be careful. The most obvious way that an assertion can be paradoxical is if what is asserted is (like the liar sentence—i.e.,“This sentence is false”) is semantically paradoxical, viz., if it is true only if false.³ But notice that the content of (1) isn't paradoxical in the sense that the liar sentence is.After all, suppose that, when S asserts (1), both conjuncts are true: it is raining and you don't believe that it is raining (as might be the case if one asserts (1) when inside a windowless room, when it is in fact raining outside). In such a context of utterance, what one asserts when one utters the sentence in (1) would be a true proposition! So, in what sense did Moore think asserting something like (1) is paradoxical?

In order to appreciate what's amiss with asserting a proposition like (1), it's important to think not only about semantics of assertion but also the pragmatics.⁴ Put another way, we need to look not only at what is explicitly said when one asserts something, but also at what is presupposed and implied.

For illustrative purposes, consider by way of comparison the speech act of questioning. You might ask:

(2) Is it raining?

In uttering a sentence like (2) in the interrogative mood, it's of course possible that you already know the answer to what you are asking (you might be pretending not to know, or just ask­ing to see whether someone else knows). However, in normal circumstances, by uttering (2), you are thereby representing yourself as not knowing whether it is raining, or more generally as being in some way epistemically impoverished or ignorant with respect to the matter of whether it's raining; at least, you represent yourself as in need more information to settle the matter for yourself.

Asserting is, as a speech act, a kind of‘epistemological mirror image' of questioning, at least, in a mirror that reverses how it is you represent yourself, epistemically, with respect to the content of these respective speech acts. On one simple and popular way of thinking: ignorance is the norm of questioning just as knowledge is the norm of asserting; and so asserting stands to questioning as knowledge to ignorance.⁶

Williamson and others have attempted to get a lot of mileage out of this simple idea. For one thing, it looks as though if by asserting you represent yourself as knowing, then there's a simple story for why (1) sounds so odd. Knowledge asymmetrically entails belief.And so, if knowledge is the norm of assertion, then by asserting you represent yourself (via asserting the first conjunct) as believing something that you then immediately explicitly commit yourself to not believing by (via asserting the second conjunct) claiming that you know it.

Here, though, a clarification is needed. If the apparent paradoxicality of ‘Moore-paradoxical' sentences like (1) were all that needed expla'ining by considerations to do with how we rep­resent ourselves, when asserting, then we might be inclined toward something weaker than the Williamsonian line that knowledge is the norm of assertion. After all, if the idea that belief is the norm of assertion were true, then this would suffice to explain why (1) seems paradoxical.

But as Moore (1962, p. 277) had also observed—and as Williamson (1996) has drawn par­ticular attention to—sentences like (2) also sound paradoxical, and not just sentences like (1).⁷

(3) It is raining, but I don't know that it is raining.

Moreover, challenges like “You didn’t know that!” (and not merely:“You don’t believe that!”) sound perfectly reasonable when used as challenges to assertions made in ignorance.The best explana­tion for this and other data about our patterns of using ‘know'8 (and more generally, of asserting as a communicative practice) recommend, according to Williamson (2000, pp. 253—5), identify­ing the kind of epistemic credential that's needed to warrant assertion as not merely belief, or true belief,⁹ or justified belief,¹⁰ but knowledge. Expressed, as it often is, as a necessary condition on epistemically permissible assertion, the idea is as follows:

Knowledge Norm of Assertion (KNA-Nec): one is properly epistemically posi­tioned to assert that p only if one knows that p.¹¹

But there might be more to the idea that knowledge is norm of assertion. As Mona Simion (2016) puts it:

[...] it looks as if a knowledgeable speaker need not do more — regarding his warrant for p — in order to be in a good enough epistemic position to transmit testimonial knowl­edge that p to her hearer.Thus, for the characteristic epistemic purpose associated with the practice of assertion, knowledge seems to be both necessary and good enough.

The idea that speakers who assert knowledgeably are beyond epistemic reproach has led Simion, along with DeRose (2002) and Hawthorne (2004) to embrace, along with KNA-Nec, a further sufficiency thesis according to which:

Knowledge Norm of Assertion (KNA-Suff): one is properly epistemically posi­tioned to assert that p if one knows that p.¹²

Here is not the place to adjudicate whether knowledge is either necessary or sufficient for epis- temically proper assertion.¹³ Rather, the aim has been to just briefly outline both why assertion is thought to be normatively constrained, epistemically and then to register ‘knowledge' as popular way to think substantively about such an epistemic norm.This will provide us with the relevant background to see how intellectual humility and assertion might interact in interesting ways.

29.4