LINGUISTIC COMPLEXITY - LOCALITY OF SYNTACTIC DEPENDENCIES

This paper proposes a new theory of the relationship between the sente nce processing mechanism and the available computational resources. Th is theory - the Syntactic Prediction Locality Theory (SPLT) - has two components: an integration cost component and a component for the memo ry cost associated with keeping track of obligatory syntactic requirem ents. Memory cost is hypothesized to be quantified in terms of the num ber of syntactic categories that are necessary to complete the current input string as a grammatical sentence. Furthermore, in accordance wi th results from the working memory literature both memory cost and int egration cost are hypothesized to be heavily influenced by locality (1 ) the longer a predicted category must be kept in memory before the pr ediction is satisfied, the greater is the cost for maintaining that pr ediction; and (2) the greater the distance between an incoming word an d the most local head or dependent to which it attaches, the greater t he integration cost. The SPLT is shown to explain a wide range of proc essing complexity phenomena not previously accounted for under a singl e theory, including (1) the lower complexity of subject-extracted rela tive clauses compared to object-extracted relative clauses, (2) numero us processing overload effects across languages, including the unaccep tability of multiply center-embedded structures, (3) the lower complex ity of cross-serial dependencies relative to center-embedded dependenc ies, (4) heaviness effects, such that sentences are easier to understa nd when larger phrases are placed later and (5) numerous ambiguity eff ects, such as those which have been argued to be evidence for the Acti ve Filler Hypothesis. (C) 1998 Elsevier Science B.V. All rights reserv ed.