The research presented here aimed at studying the processing of syntactically ambiguous Prepositional Phrases (PP) following a Noun Phrase Object (NPO). The ambiguity lies in the fact that the PP could be attached either to the verb (VP-attachment, ''The boy broke the window with a rock'') or to the NPO (NP-attachment, ''The boy broke the window with the curtain''). The conditions on the NPO were the following. First, the NPO always played the role of Result of the predicate, a condition which was supposed to favor NP-attachment. Second, the NPO began either with an indefinite article, a definite article, or a demonstrative adjective. In the first case, NP-attachment was assumed to be more natural and less costly than VP-attachment because of the presence of expectations and the absence of inferences. In the other two cases, NP-attachment was assumed to be less natural and more costly than in the first case (particularly in the demonstrative condition) because, in these cases, it was required to infer a set of referents. An off-line choice task (Experiment 1) and a self-paced reading task (Experiment 2) both produced results in accordance with the predictions.
Schelstraete, M.-A. (1996). Definiteness and prepositional phrase attachment in sentence processing. Cahiers de Psychologie Cognitive, 15(5), 463-486. https://hdl.handle.net/2078.5/68348 (Original work published 1996)