The Path of Presupposition Projection in Processing - The case of Conditionals

Abstract

While presupposition projection has been explored in depth in theoretical terms, not much is known about possible cognitive processes related to it. We investigate the question of whether projection, which can be seen as a mismatch between the location of where material is introduced syntactically and the level at which it is interpreted, comes with a processing cost that requires additional processing time. In an eye tracking reading study using conditional sentences with German wieder (‘again’) in the consequent and a preceding context sentence, we varied the relative location in which the presupposition of wieder was supported. We argue that the pattern of results we get reflects a processing cost for projection that directly correlates with the length of the projection path as it would be measured in Discourse Representation Theory (DRT). This perspective on the data crucially relies on the representational mechanisms involved in projection, which implies that a non-representational theory such as dynamic semantics cannot offer the same account. A simple processing hypothesis based on distance from presuppositional support measured in clauses also does not suffice to explain the data. The paper thus offers experimental evidence bearing on fine-grained theoretical choices in presupposition theory. It closes by relating our results to some other recent experimental work on presupposition interpretation.