Construal presents a new theory of sentence processing, one that allows a limited type of underspecification in the syntactic analysis of sentences. It extends what has arguably been the dominant theory of parsing (the garden-path theory developed by Lyn Frazier and colleagues) through the 1980s into new and previously unexplored domains, and greatly advances the potential for insights into how meaning is both made and understood.
Frazier and Clifton, both pioneers in parsing theory, present new psycholinguistic theory and experimentation concerning how "nonprimary" phrases are analyzed in sentence comprehension. They define a process of "construal" and show how it accounts for cases in which the parser does not fully determine structure during the course of ordinary comprehension.
The idea of construal arises in part through the authors' critical review of the challenges to their established framework for research on structural parsing. While they demonstrate that the principles of parsing theory remain valid for a wide variety of languages and grammatical constructions, they go beyond them to clearly identify those types of constructions built by the process of construal. Frazier and Clifton show that construal follows distinct principles, and they flesh out their hypothesis with previously unexamined evidence and new empirical tests.
“For the past 15 years research on human sentence parsing has been dominated by the work of Frazier, Clifton and their colleagues at UMass, Amherst. This stimulating monograph provides a wealth of new evidence and argument—all used persuasively to refine and update their theoretical position. It is a tour de force which will ensure that their research remains the focus of continuing debate for another 15 years.”
—Don Mitchell, University of Exeter
“This is required reading for anyone interested in how we understand sentences we read or hear, and how we decide among the alternative interpretations of ambiguous sentences.”
—Stephen Crain, University of Maryland at College Park