Reviews the history of verbal data in a variety of fields, places protocol analysis in its historical context, and examines more recent claims and criticisms regarding protocol analysis. Concludes that protocol analysis, when conducted according to certain principles, can be an important tool for researching the composing process. (MG)

Three- to nine-year-olds' (N=80) post-exposure production of regular and irregular suffixes indicated that subjects found it easier to learn a regular suffix when they heard it used with phonetically similar base forms. Subjects were more likely to overgeneralize the regular suffix to irregular forms when they had heard it used in conjunction with…

Hudson addressed two issues on second-language acquisition: the validity of the multidimensional model and the application of developmental sequence findings to assessment and pedagogy. This response suggests that his lines of argument are based on serious conceptual flaws. (Contains 37 references.) (LB)

This study of sequential memory in 16 children with language impairment found that list-final suffix effect was substantially larger than in control children, even though other aspects of their recall were normal. Children with language impairment were more dependent upon unanalyzed acoustic and phonetic representations of speech. Response…

The Shortlist model is presented, which incorporates the desirable properties of earlier models of back-propagation networks with recurrent connections that successfully model many aspects of human spoken word recognition. The new model is entirely bottom-up and can readily perform simulations with vocabularies of tens of thousands of words. (DR)

Finds that temporal markers modify the availability of preceding words: segmentation markers like "around two o'clock" and "then" reduce this availability, whereas continuity markers like "and" improve this availability. Supports the hypothesis that segmentation markers lead readers not to integrate new information…

Examines the role of a "local cue"--the grammatical aspect expressed by the verb "ru/-ta"--in Japanese temporal clauses. Ninety-six children were examined. Results show that the success of verb forms as clues requires that temporal relations be clearly expressed. This supports the claim that overt surface clues facilitate…

Connectionist (parallel distributed processing) modeling provides a new way to approach the neurological study of language. This method focuses on the interplay between a computational model and the appropriate neurological, neuropsychological, and speech and language data, couched in connectionist mechanisms that map naturally to what is known of…

Outlines some theoretical differences in the interpretation of nonword repetition and discusses how best these differences could be resolved. (eight references) (GLR)

A theoretical integration of the storage and processing functions of working memory in language comprehension is presented. A computational theory is described in which storage and processing are fueled by activation. Capacity is expressed as the maximum amount of activation available in working memory for either storage or processing. (SLD)

Four experiments demonstrated that the comprehension of novel compound nouns involves the recognition of a general relationship between two categories. They reinforce earlier findings that the process of creating meanings may be equally as efficient as the process of selecting meanings. (33 references) (Author/LB)

Recent investigations of filler-gap dependencies in sentence processing have assumed that the parser must compute an antecedent-trace relationship in which the trace site is identical to the canonical position of the moved phrase. Pickering and Barry's challenge to this view is refuted and a "direct association hypothesis" is suggested.…

Pickering and Barry's recent argument against the existence of empty categories (ECs) in human sentence processing is disputed. It is argued here that ECs may still play a linking role between thematic role assigners and wh-phrases. One possible parsing algorithm is given that accounts for Pickering and Barry's data. (28 references) (Author/LB)

Papers by Gorrell and by Gibson and Hickok question Pickering and Barry's (PB) arguments against empty categories in sentence processing. This reply disputes Gorrell's claims that PB's interpretation of the data is inadequate and, in agreement with Gibson and Hickok, reinforces the arguments that the gap location is irrelevant to the formation of…

Several different conceptualizations of applied linguistics are evaluated, ranging from "applications of linguistic theory" to alternative models for studying language that extend and complement generative grammar as a theory of language. It is shown that they imply substantive differences in goals, methods, and priorities of language study. (30…