Use of Verb Information in Syntactic Parsing: Evidence From Eye Movements and Word-by-Word Self-Pace...
Download
1 / 17

Ferreira and Henderson (1990) - PowerPoint PPT Presentation


  • 87 Views
  • Uploaded on
  • Presentation posted in: General

Use of Verb Information in Syntactic Parsing: Evidence From Eye Movements and Word-by-Word Self-Paced Reading. Ferreira and Henderson (1990). Sentence processing models.

loader
I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
capcha

Download Presentation

Ferreira and Henderson (1990)

An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

Presentation Transcript


Use of Verb Information in Syntactic Parsing: Evidence From Eye Movements and Word-by-Word Self-Paced Reading

Ferreira and Henderson (1990)


Sentence processing models

  • Interactive model: All sorts of information (lexical, syntactic, semantic) communicate in an unconstrained fashion to produce the most plausible reading at the earliest stages of sentence comprehension


Sentence Processing Models cont.

  • Garden Path Model: Distinct modules within the language system. Communication occurs, but in a constrained fashion.

  • Minimal Attachment Principle: Sentences initially receive the simplest syntactic analysis possible


Research Questions

  • Does verb information aid readers in their initial parsing of temporarily ambiguous sentences?

  • Can verb subcategorization override minimal attachment?


Methodology

  • Experiment 1: Eyetracking

  • Experiment 2: noncumulative word-by-word self-paced paradigm

  • Experiment 3: cumulative word-by-word self-paced paradigm


Experiment 1

  • Subjects: 12 members of University of Massachusetts community

  • Materials: four lists of 80 experimental items and 72 fillers

  • Factors: verb bias, presence of complementizer


Verb Types

  • Minimal Attachment Verbs: Strongly biased for a minimal attachment of the ambiguous noun phrase, e.g. guess

  • Nonminimal Attachment Verbs: Strongly biased for a nonminimal interpretation of the ambiguous noun phrase, e.g. brag


Sentence Frames


Results-First Fixation Duration


Total Reading Time


Regressions

  • Verb bias had effect on regressions to the disambiguating region.

  • Reanalysis appeared to be easier with nonminimal attachment verbs

    (4 vs. 21 regressions)


Experiment 2

  • Self-paced reading task (non-cumulative)

  • Subjects: 24 University of Alberta undergraduates

  • Materials: Same as Experiment 1


Results- Mean Reading Times


Experiment 3

  • Same as Experiment 2, but with a cumulative self-paced reading task

    (allows reader to reread)


Results –Mean Total Reading Time


Discussion

  • Expt. 1 : Verb bias did not prevent misanalysis, but seemed to facilitate reanalysis

  • Expt. 2: Robust effect of verb bias on postdisambiguating region (less total reading times) → verb bias aids reanalysis

  • Expt. 3: weaker effects (methodological issues)


SO…

  • The fact that sentences lacking a complementizer are harder to process regardless of verb bias supports the garden-path model!


ad
  • Login