J. L. Austin, How to Do Things with Words
Notes on sections IV - VII

Curtis Brown

I. Constatives vs. Performatives Revisited

In these sections, Austin begins accumulating reasons for doubt about his original constative/performative distinction. Let's just review a few of the reasons for doubt.

Recall the main distinctions between constatives and performatives:

constatives performatives
(1) are true or false (2) are felicitous or infelicitous
(3) saying something (4) doing something

(1) Constatives as well as performatives seem to be subject to something like felicity conditions. That is, there are various ways they can "go wrong" even if they aren't false.

Austin's examples: (a) saying something with a false "implication." Saying something "implies" that you believe it, so there's something fishy about saying something you don't believe. But this is not logical implication, so the "something fishy" here is not self-contradiction.

Similarly, if I say "The cat is on the mat, but I do not believe that it is," I have not literally contradicted myself (since what I've said is entirely possible). But I have committed what G. E. Moore called a "contradiction in use." Similarly, "I do not exist" expresses a proposition that is not self-contradictory, but it does involve a contradiction in use, in the sense that I can't say it truly even though it could be true.

So if I say something without believing it, what is wrong with my utterance? Austin: it's insincere. But this seems very similar to the insincerity of promising something you don't intend to do. What I've said is infelicitous even though it may not be false.

(b) Saying something with a false "presupposition." Consider the following (improved!) variant on Austin's example: "All of my articles in the Journal of Philosophy have been praised by world-famous philosophers." Or: "Every time I have given a talk at Harvard University, I have gotten a standing ovation." These sentences do not logically imply that I have published articles in the Journal of Philosophy or given talks at Harvard, but they do presuppose these things. There's something very fishy about saying these things if I haven't published in the Journal of Philosophy or spoken at Harvard.

Austin suggests that this is similar to the way performatives are infelicitous if conditions for their success are not met, e.g. saying "I insult you" when there is no practice of insulting someone by saying this.

An even better example of type (b) might be this. I ask a friend what his son's name is. Just as I ask the question, he has a horrible attack of heartburn, and blurts out "Urgh!" I mistakenly think that his son's name is Urgh. Later I tell my wife, "Little Urgh is really a charming guy." My sentence presupposes that 'Urgh' is the name of a person; if it isn't, then I don't succeed in saying anything, just as I don't succeed in insulting you if the right practice doesn't exist.

(2) Performatives as well as constatives seem to be subject to something very like truth or falsity. This is related to the fact that some sentences could be used either descriptively or performatively, or possibly both at the same time. "I am sorry" -- performative or descriptive? "I approve" -- do I thereby give my approval, or merely make a statement about my attitudes? Etc. (See interesting chart on p. 79.)

It seems that if I say "I apologize" and successfully apologize, then the statement that I apologize must be true. (Austin also notes the related 1st/3rd person asymmetry here: "He apologized" is descriptive; for that matter "I apologized" (past tense) is descriptive, not performative.)

Anyway, the point is that if I say "I apologize" and thereby apologize, then it is true that I apologize. Is this that much different from saying "Grass is green" and its being true that grass is green? (Searle: the difference is in "direction of fit.")

(3) Constatives are a matter of saying something, not doing something. But isn't saying something in fact a way of doing something? Moreover, there are lots of utterances that seem not only to state something but to have a performative structure:

I warn you that the bull is loose

I declare that you are the orneriest child I've ever seen

I conclude that your argument is worthless

(Compare Austin on "expositives" at p. 85.)

(4) And of course, although a performative may be a matter of doing something, it also involves saying something.

Austin also tries but fails in lecture V to find a grammatical criterion for performative utterances (best attempt: the first person singular present indicative active).

So at the end of our reading for today, he decides that it is time for a "fresh start" on the problem. (p. 91.)

Last update: March 30, 2010
Curtis Brown  |  Philosophy of Language   |  Philosophy Department  |   Trinity University