Elsevier

Cognition

Volume 30, Issue 3, December 1988, Pages 191-238
Cognition

Interaction with context during human sentence processing

https://doi.org/10.1016/0010-0277(88)90020-0Get rights and content

Abstract

Psychological theories of natural language processing have usually assumed that the sentence processor resolves local syntactic ambiguities by selecting a single analysis on the basis of structural criteria such as Frazier's (1978) “minimal attachment.” According to such theories, alternative analyses will only be attempted if the initial analysis subsequently proves inconsistent with the context. (See also Ferreira & Clifton, 1986; Ford, Bresnan, & Kaplan, 1982; Rayner, Carlson, & Frazier, 1983). An alternative hypothesis exists, however: If sentences are understood incrementally, more or less word-by-word (Marlsen-Wilson, 1973, 1975), then syntactic processing can in principle exploit the fact that interpretations are available, using them “interactively” to select among alternative syntactic analyses on the basis of their plausibility with respect to the context. The present paper considers possible architectures for such incremental and interactive sentence processors, and argues for an architecture

Résumé

Les théries psychologiques de traitement des langues naturelles ont habituellement supposé que le processeur de phrases résolvait les ambiguités syntaxiques locales en sélectionnant une seule analyse sur la base de critéres structurels comme le principe de l'“attachement minimal” de Frazier (1978). D'aprés ces théories, les analyses alternatives seront seulement envisagées si l'analyse initiale se révèle être inconsistante avec le contexte. (voir aussi …). Cependant, une autre hypothèse est possible: si les phrases sont comprises de façon progressive, plus ou moins mot à mot (Marlsen-Wilson 1973, 1975), alors le traitement syntaxique peut en principe exploiter le fait que les interprétations sont disponibles, et les utiliser de façon “interactive” pour sélectionner parmis les différentes analyses syntaxiques en fonction de leur plausibilité par rapport au contexte. Cet article considère les architectures possibles pour de tels processeurs de phrases interactifs et progressifs, et argumente en faveur d'une architecture telle que les différentes analyses sont offertes en parallèle, et sont distinguées par un appel immédiat au processus de compréhension, selon une interaction sélective ou “faible”, par opposition à l'interaction directive ou “forte”. Nous notons qu'une telle architecture ne compromet en aucune façon l'hypothése de modularité de Fodor (1983). Nous faisons la revue les données expérimentales présentées comme suggérant que le système de traitement des phrases humain était non-interactif et reposait sur des critères purement structurels. Nous présentons de nouveaux résultats qui semblent incompatibles avec la proposition structurelle, et qui soutiennent l'hypothèse interactive. Nous suggérons des raisons qui permettent d'écarter les résultats contraires obtenus auparavant, et concluons que le mécanisme de traitement des phrases humain résoud les ambiguités de type modifieur-attachment en ayant recours á des informations contextuelles et réferentielles de plus haut niveau sous l'intéraction faible.

References (59)

  • T. Winograd

    Understanding natural language

    (1972)
  • A. Ades et al.

    On the order of words

    Linguistics and Philosophy

    (1982)
  • K. Ajdukiewicz

    Die syntaktische Konnexitat

    Studio Philosophica

    (1935)
  • G.T.M. Altmann

    The resolution of local syntactic ambiguity by the human sentence processing mechanism

    Proceedings of the Second Conference of the European Chapter of the Association for Computational Linguistics

    (1985)
  • G.T.M. Altmann

    Reference and the resolution of local syntactic ambiguity: The effect of context in human sentence processing

    (1986)
  • G.T.M. Altmann

    Modularity and interaction in sentence processing

  • G.T.M. Altmann

    Ambiguity, parsing strategies, and computational models

    Language and Cognitive Processes

    (1988)
  • R.C. Berwick et al.

    The role of grammars in models of language use

    Cognition

    (1983)
  • T.G. Bever

    The cognitive basis for linguistic structures

  • P.A. Carpenter et al.

    What your eyes do while your mind is reading

  • W.L. Chafe

    Discourse structure and human knowledge

  • N. Chomsky

    Language and mind

    (1968)
  • H.H. Clark

    The language-as-fixed-effect fallacy: A critique of language statistics in psychological research

    Journal of Verbal Learning and Verbal Behavior

    (1973)
  • S. Crain

    Pragmatic constraints on sentence comprehension

    (1980)
  • S. Crain et al.

    On not being led up the garden path: The use of context by the psychological parser

  • K. Ehrlich et al.

    Pronoun assignment and semantic integration during reading: Eye movements and immediacy of processing

    Journal of Verbal Learning and Verbal Behavior

    (1983)
  • J.A. Fodor et al.

    The psychology of language

    (1974)
  • M. Ford et al.

    A competence-based theory of syntactic closure

  • Cited by (590)

    • Referencing context in sentence processing: A failure to replicate the strong interactive mental models hypothesis

      2022, Journal of Memory and Language
      Citation Excerpt :

      1b) A psychologist told the woman that he was having trouble with to leave. Referential theory, as proposed by Crain and Steedman (1985), Altmann and Steedman (1988), and Ni and Crain (1989), explains this apparent structural preference for resolutions like 1a compared with resolutions like 1b via the principle of parsimony, or that the easiest interpretation given the current discourse state should be chosen among competitors upon encountering ambiguity. For example, when readers encounter ‘the woman’ in a null context (i.e. no preceding discourse information), the subsequent ‘that’ is less likely to be interpreted as a relativizer because this would imply a competitor set of ‘woman’ referents.

    View all citing articles on Scopus

    The authors acknowledge the support of a Science and Engineering Research Council research studentship (awarded to the first author) at the Centre for Cognitive Science, University of Edinburgh; an ESPRIT grant (project 393) to the Centre for Cognitive Science; NSF grant IRI 10413 AO2 to the Department of Computer and Information Science, University of Pennsylvania; SERC grant D/29628 to the Centre for Speech Technology Research, University of Edinburgh; and also the Max Planck Institute for Psycholinguistics, Nijmegen, at which portions of this paper were written while the first author was a visiting scientist there. The authors thank Ellen Gurman Bard, Stephen Crain, Alan Garnham, Nick Haddock, Stephen Isard, Jane Oakhill, Richard Shillcock, Mike Tanenhaus, and Henry Thompson for advice and encouragement, as well as Janet Fodor and Ken Forster for their comments on earlier versions of this paper.

    View full text