The Dependency Locality Theory: A Distance-Based Theory of Linguistic Complexity

A major issue in understanding how language is implemented in the brain involves understanding the use of language in language comprehension and production. However, before we look to the brain to see what areas are associated with language processing phenomena, it is necessary to have good psychological theories of the relevant behavioral phenomena. Recent results have suggested that constructing an interpretation for a sentence involves the moment-by-moment integration of a variety of different information sources, constrained by the available computational resources (see, e.g., Ford, Bresnan, and Kaplan 1982; MacDonald, Pearlmutter, and Seidenberg 1994; Trueswell, Tanenhaus, and Garnsey 1994; Trueswell1996; Tyler and MarslenWilson 1977; McClelland, St. John, and Taraban 1989; Pearlmutter and MacDonald 1992; Crain and Steedman 1985; Altmann and Steedman, 1988; Ni , Crain, and Shankweiler 1996; see Gibson and Pearlmutter 1998 and Tanenhaus and Trueswell 1995 for summaries). This chapter presents evidence for one theory of resource use in sentence co"mprehension: the dependency locality theory (DLT ). If the evidence for a theory such as this one accumulates, it will then make sense to look for neural correlates of the theory (see Kaan et ale 1998; Harris 1998, for some initial attempts to find event-related potential measurements of brain activity corresponding to the components of the DLT ). An important part of a theory of sentence comprehension is a theory of how sentence structures are assembledsentence parsingas words are input one at a time. Two important components of sentence parsing consume computational resources: