Assertion and Retraction

Sometimes our speech actsSpeech acts go really well. They’re in good normative order when we make them, and our relationship with them remains happy ever after. We continue to endorse them, continue to be content to stand by them, continue to endorse our conversation’s being in whatever state they put it in to, and continue to be content to bear whatever commitments we undertook by making them. But sometimes not. Sometimes we want to disavow a previous speech actSpeech acts. Maybe because we’ve learned that it wasn’t in good normative order when we made it. Maybe because we no longer endorse the change in conversational state that it brought about, or because we want to get out from under some commitment that the speech actSpeech acts saddled us with. When this happens, we’re liable to retract the speech actSpeech acts. We’ll use one of a variety of locutions—“never mind”, “scratch that”, “I take it back”, “I guess I was wrong”, “wait, that’s not right”, for example—to, somehow or other, do something that’s aptly described as retracting or taking back or disavowing our previous speech actSpeech acts. Retraction is something that definitely happens, and the process(es) by which it happens are something that any completed account of linguistic communication will need to include an account of. (And that any partial account will need to leave room for). What’s not a sure thing is the correct theoretical characterization of what we’re doing when we retract. In this paper I’ll take up three questions about retraction: The first is the question of reductionism vs. anti-reductionism about retraction—is there a special, sui generis speech actSpeech acts of retraction, or does retraction always proceed by way of some speech actSpeech acts of a familiar type? (Asserting the negation, apologizing, etc.) The second is the question of what kind of theoretical framework for thinking about speech actsSpeech acts provides the most natural home for anti-reductionism about retraction—a framework that puts the speaker-side norms governing the use of speech actsSpeech acts on the theoretical ground floor, or one that puts audience- or context-side effects on the ground floor? (Spoiler: Effects-first views are better.) Finally, I argue that thinking about retraction gives us some helpful leverage in thinking about the effects of assertion, motivating a view according to which assertions are best thought of as endorsements of forward-looking constraints on the conversational context, and that this view of assertion is independently attractive.