Contextual effects in understanding indirect requests∗

Subjects were presented with sentences like "Must you open the window?" either embedded within a story context or as an isolated sentence. Within the story context, the sentence could have its literal meaning, "Is it necessary that you open the window?" or its indirect or conveyed meaning, "Do not open the window." Subjects gave comprehension responses and paraphrase judgments for the sentences. In a story context subjects take longer to understand literal sentences than conveyed ones. Without context the conveyed requests take longer than the literal sentences. A second experiment found that in a story context subjects take as long to understand a direct request, "Do not open the window," as to verify an indirect request, "Must you open the window." The data from these experiments suggest that a person understanding an indirect request in context need not construct the literal interpretation before deriving the conveyed request. Notes This research was supported by National Institute of Education Grant G‐77–0015 to David E. Rumelhart and Public Health Service Grant MH 15828–09 to the Center for Human Information Processing. I wish to thank David Rumelhart and the entire LNR Research Group for their support and guidance in all phases of this work. Additional thanks to Robert Olds for his assistance in running the experiment and to Robert Glushko, Jeff Miller, James McClelland, Donald Norman, and Arthur Samuel for their helpful comments on earlier drafts of this paper. Reprint requests may be sent to Raymond W. Gibbs, Department of Psychology C‐009, University of California at San Diego, La Jolla, California 92093.