Published online by Cambridge University Press: 04 September 2014
Expressives like damn convey a negative attitude toward an entity or toward a situation. What is particularly interesting about such expressions is the looseness of the relation between their syntax, which is the syntax of normal attribute adjectives, and their interpretation (Potts 2005, 2007). An experiment on various negative expressives manipulated the placement of the expressive as a prior utterance, or inside the subject or inside an object of the verb or preposition. Experimental participants were asked what the speaker was most likely to have a negative attitude towards − the subject, the object, or the entire situation. The test items were of two types, ‘non-causal’ and ‘causal’, exemplified by The holiday is on the damn weekend and The dog is on the damn couch. In the non-causal items, the subject (holiday) cannot plausibly be taken as being responsible for the state of affairs described. However, in the causal items, the subject might be responsible for the state of affairs described. The same range of interpretations was observed for all placements of damn. The prior utterance condition (Damn. The dog is on the couch.) yielded more entire situation interpretations than the sentence-internal damn items. Overall, subject damn items yielded more subject interpretations than object damn items. However, as predicted by the hypothesis that blame would devolve on a potentially responsible agent (the culprit hypothesis), there were more subject interpretations in the causal items than in the non-causal items. The results suggest that considerable pragmatic inferencing is involved in the interpretation of expressives, consistent with a proposal that an expressive constitutes a separate speech act.
This work was supported by Grant HD18708 from the National Institutes of Health to the University of Massachusetts. Thanks to Shayne Sloggett, Josh Levy, Matt du Pont, Adina Galili, Marysa Mezzetti, Jennifer Dimiyan, and Chidima Oranekwu for help collecting and analyzing the data. We are very grateful to Chris Potts, Bernard Fradin, and Kyle Rawlins for discussion of the issues here, and to Chris Potts and an anonymous reviewer for comments on the manuscript.