Hijacking limitations of working memory load to test for composition in language.

Area of Science:

• Psycholinguistics
• Cognitive Science
• Linguistics

Background:

• The storage versus decomposition debate in language processing remains unresolved.
• Working memory limitations offer a novel approach to investigate linguistic composition.
• English inflectional morphology provides a well-studied paradigm for this investigation.

Purpose of the Study:

• To test the hypothesis that decomposed linguistic forms impose a greater working memory load.
• To investigate whether compositional (regular) English inflections are harder to maintain in working memory than non-compositional (irregular) inflections.
• To identify a diagnostic tool for linguistic composition.

Main Methods:

• A 3-back production task was employed to assess working memory load.
• English inflectional morphology (regular vs. irregular verbs) was the focus.
• Potential confounding factors including frequency, phonology, and orthography were controlled.

Main Results:

• Regular inflections resulted in significantly more working memory errors than irregular inflections.
• Regular forms showed more bare-stem and stem-affixation errors, indicating decomposition.
• Irregular forms showed more past-tense-form errors, suggesting whole-word processing.

Conclusions:

• The findings support the decomposition of regular inflections and whole-word storage of irregulars.
• The regular-irregular difference was modulated by phonological consistency and participant sex, particularly for low-frequency forms.
• Working memory load serves as a robust indicator of linguistic composition.