Abstract

Since people can often successfully interpret utterances that flout or ignore conventions, Davidson concludes that shared conventions are neither necessary nor sufficient for linguistic interpretation. This conclusion is based on an overly narrow conception of what it is to know, and to share, a language. Rather than, as Davidson argues, simply interpreting the meaning the speaker intends their words to be interpreted as having (and their words’ truth conditions), successful interpretation requires interpreting the illocutionary act the speaker intends to be interpreted as performing (and the act’s felicity conditions). This change in focus highlights the need for many types of shared conventions, beyond the conventional meanings of words that Davidson considers and dismisses as unnecessary. When any one convention is ignored or flouted, interpretation is possible because the apparently unconventional utterance nonetheless conforms to a host of other shared conventions. Conventions are necessary for linguistic interpretation.