Norms of Assertion and Constitutive Rules

Kent Bach quoted here the following passage from Williamson that I have long found quite perplexing:

That assertion has such [constitutive] rules is by no means obvious; perhaps assertion is more like a natural phenomenon than it seems. One way to find out is by supposing that it has such rules, in order to see where the hypothesis leads and what it explains. That will be done here. [pp. 238-9]

It’s, of course, from Knowledge and Its Limits. What puzzles me about it is the contrast between what has constitutive rules and natural phenomena. So the idea, I take it is that natural phenomena, like lightning and thunder, perhaps, have no associated norms, but other kinds of things might have constitutive norms. I’ve always wondered why Tim only cites these two options. Why can’t a phenomenon have associated norms, but no constitutive norms? Moreover, is this passage meant to endorse the view that anything normative is, for that very reason, non-natural? That doesn’t strike me as a compelling view, and as I recall, Williamson never explains this distinction or the point of the contrast. So, is the contrast just between the normative and the natural, and should we think that everything normative is subject to constitutive rules?


Norms of Assertion and Constitutive Rules — 4 Comments

  1. I too was puzzled by Williamson’s contrast between what has constitutive rules and natural phenomena. Prior to the passage you quote, which occurs at the beginning of his chapter on assertion (Ch. 11), Tim likens speech acts to games and contrasts games with the act of jumping. This example illustrates your point, for clearly there can be norms for acts of jumping that are not constitutive (leaving aside the special and obviously game-like cases of high jumping and long jumping competitions). Now Tim is careful to distinguish norms in general from constitutive rules (he doesn’t seem to distinguish rules from norms — I do, in section 2 of a very old paper, ‘Analytic Social Philosophy: Basic Concepts,’ to which there is a link here), and he points out various ways in which assertions may be praised or criticized (I would add that most ordinary speech acts can be praised or criticized in the same or similar ways). He observes that these norms are not specific to assertion, and goes on to suggest that there is one norm that is.

    In the earlier thread that you linked to above, I expressed my doubts that there are any constitutive rules for ordinary speech acts, such as assertions, as opposed to the obvious norms of sincerity, relevance, and the like) for ordinary speech acts in general. At the end (to date) of that thread, you asked about the distinction between norms and constitutive rules of assertion. This was in connection with my rather weak Belief Account of assertion — assert only what you believe — which imposes a mere sincerity condition on assertion, one that is not specific to assertion. The idea was to combine that with an account of belief that would explain the very things that the Knowledge Account of assertion (and other stronger accounts) aims to explain, but without invoking any constitutive rules (sincerity is not limited to assertions).

    Anyhow, you asked me to say more about the distinction between norms and constitutive rules of assertion. I suspect that you’re really wondering why I think constitutive rules are not needed for ordinary speech acts. This question is addressed in section 7.3 of my book with Mike Harnish, Linguistic Communication and Speech Acts (MIT 1979), where we argue, largely on the basis of the theory developed in earlier chapters, against John Searle’s reliance on constitutive rules. The gist is this. The various conditions that Searle plausibly identifies as singly necessary and jointly sufficient for the performance of speech acts of various types can be fully accounted for on a Gricean intentional/inferential model of communication. Moreover, Searle’s account is so formulated that it could only work for speech acts where the so-called force indicating device is present.

    Rather than go into more detail, I’ll just mention two relevant features of our Gricean account. First, we individuate types of speech acts in terms of types of attitudes expressed by the speaker. Second, the key notion of expressing an attitude comes in (here we borrow Austin’s terminology) at the illocutionary, not perlocutionary level. It is one thing to express an attitude, such as a belief, and another thing actually to have it, i.e., to be sincere — much less to get the listener to believe what one believes, which in the best case is to convey knowledge. I’m sure this is too compressed to make a lot of sense, but if you’re interested and don’t have that long out-of-print book handy, you might check out the first few sections of a handbook article, ‘Speech Acts and Pragmatics,’ posted here.

  2. Kent, this is very helpful. Do you think the notion of constitutive rules is the same in Searle and Williamson? I wonder since the natural account of a constitutive rule is that you aren’t engaging in the speech act in question if you violate the rule, and that’s not Williamson’s notion. After reading the first couple sections of the piece you referenced, I think the way you use the term can be the same as Williamson’s.

    I’m sure you’ve thought of this already, but given Williamson’s view that knowledge is a mental state, there’s a question as to whether the illocutionary act of making a statement expresses the attitude of belief or of knowledge. Of course, the question only arises, I think, if Williamson’s argument succeeds in showing that knowing is a state of mind…

  3. Do Searle and Williamson operate with the same notion of constitutive rules? (You could also ask whether either of their notions is the same as Rawls’s practice conception of rules — see “Two Concepts of Rules,” Phil. Rev. 1955.) Probably not, in which case the argument I mentioned in my previous comment may not apply to Williamson’s notion. For Searle constitutive rules take the form ‘X counts as Y in C’ (C is a social or institutional context), and these aren’t normative — they don’t specify conditions in which one must Y (or may Y) but only how to Y. So it’s not clear that it makes sense to speak of violating such a rule. For Williamson the constitutive rule for assertion takes the form ‘One must: assert p only if p has C’ (C the property specified by the correct account of assertion), and this is clearly normative. Obviously such rules can be violated. Indeed, as Williamson says, “breaches … may even be common” (p. 240). He goes on to say that “some sensitivity to the difference … between conforming to the rule and breaking it presumably is a necessary condition of playing the game, speaking the language, or performing the speech act,” but whatever exactly he means here, I don’t think that his constitutive rules are the same as Searle’s counts-as rules. He may not even mean the same thing by ‘constitutive’.

    As to whether the illocutionary act of making a statement expresses the attitude of belief or of knowledge, in the sense of ‘express’ that Harnish and I operate with (p. 15), one can succeed in stating that p even if one does not thereby give the listener even prima facie reason to think that one knows that p. But even a blatant liar gives the listener at least prima facie reason to think that he believes that p. Otherwise, he could not succeed in telling the lie (of course, the listener can understand the lie without accepting it or even being at all inclined to accept it).

  4. Pingback: Certain Doubts » Stats for May

Leave a Reply

Your email address will not be published. Required fields are marked *