Psychology Reader

Good-Enough Representations in Language Comprehension

Fernanda Ferreira, Karl G.D. Bailey, and Vittoria Ferraro · 2002 · Current Directions in Psychological Science

Role
Foundational argument · Critical reappraisal
Difficulty
Accessible
Full work length
5 pages
Type
Journal article

Access

Psychology Reader links to legitimate sources and does not host article PDFs.

Reading focus

Syllabus: How Do We Understand and Use Language?

Question what it means to say that a reader has understood a sentence. Separate a plausible overall impression from an accurate account of its grammatical relationships.

A question to consider. What pair of comprehension questions could show that someone has repaired a sentence's structure while retaining an incorrect earlier meaning?

Why this reading

An interpretation can feel coherent while getting who did what wrong. Ferreira and colleagues question whether successful sentence processing always yields a detailed, accurate representation. Their argument separates repairing grammatical structure from correcting the meaning a reader retains.

Before you read

Syntax concerns how words are organized; compositional interpretation uses their meanings and relationships. A garden-path sentence initially invites an analysis that later words undermine. A heuristic is a shortcut; reanalysis revises an earlier interpretation.

Read for

  1. Distinguish evidence that a reader noticed and repaired a syntactic problem from evidence that its initial meaning disappeared.
  2. Compare the roles of plausibility, word order, and grammatical structure in the examples.
  3. Ask what would distinguish incomplete encoding from later forgetting or difficulty answering the comprehension question.

Consider afterwards

  1. Design two questions about one ambiguous sentence: one checking the eventual grammatical analysis and one checking whether an earlier mistaken inference persists.

Editorial context

This short conceptual review proposes a research approach rather than a complete processing model. Its laboratory examples do not show that people usually misunderstand conversation. ‘Good enough’ describes possible limits of representation, not advice to accept careless interpretation.