Conclusions

• Parsing is highly incremental and sensitive to memory load

• Disambiguation:

  • syntactic strategies seem central, but
  • modularity is controversial

• Reanalysis

  • Serial and repair (Lewis, Fodor & Inoue)
  • Monotonic (Sturt & Crocker)
  • Parallel (Gibson)

• Psychological reality and representations:

  • Current syntactic theories do not easily integrate with processing models
    • non-incremental, difficult reanalysis
  • Psychological evidence is often most easily described in terms of ‘dependency-based’ representations (e.g. theta-attachment, thematic monotonicity)

• Graded preferences, interacting constraints, and experience?

Previous slide

Back to first slide

View graphic version