Shallow processing and attention capture in written and spoken discourse

Processing of discourse seems to be far from uniform with much evidence indicating that it can be quite shallow. The question is then what modulates depth of processing? A range of discourse devices exist that we believe may lead to more detailed processing of language input (Attention Capturers), thus serving as modulators of processing enabling some parts of discourse to be processed more than others. We list some of these and describe two in detail. We introduce the text-change procedure in which texts are presented twice in succession with a possible change to a word on the second presentation. By analogy with visual change-blindness, text change provides a method of tracking attention capture and depth of processing in text. In Experiment 1, the effect of italicization in writing was investigated. Italicization led to an increase in change detection. In Experiment 2, we investigated focus-driven word stress in speaking. Once again, focus-driven stress led to increases in change detection. These examples show how word stress increases depth of processing. We discuss the findings in terms of the Granularity account of focus, in which deeper processing is characterized as a more detailed semantic specification