The process of sentence parsing is clearly influenced by the detailed information recovered from the lexicon for each successive word that appears in the sentence. This information may be used to guide or direct the assembly of new syntactic structures during the initial phases of parsing. Alternatively, it may be used to check or monitor the internal consistency of syntactic frameworks constructed on some other basis. This paper first describes a range of different models in each of these two categories and then reviews the empirical evidence bearing on the alternative models. Overall, it is concluded that current findings are most compatible with structure-checking models. The data are discussed in terms of a model based on a symbol-processing formalism (unification grammar) and the outline of a parser based on a connectionist processing system.
Stored only when both papers are in the corpus — 86 citations exist upstream.
Imported from OpenAlex (CC0) — nothing here is generated.