Learning phonotactic-like regularities in immediate serial recall.

Area of Science:

• Cognitive Psychology
• Psycholinguistics
• Speech Science

Background:

• Learning a spoken language involves acquiring phonological rules, including phonotactics (permissible sound sequences).
• Previous research shows speech errors respect artificial phonotactic constraints, indicating learning in the phonological domain.

Purpose of the Study:

• To investigate if errors in nonlinguistic domains, specifically immediate serial recall, reveal similar learning patterns.
• To explore domain-general versus language-specific constraints on phonotactic learning.

Main Methods:

• Participants performed verbal immediate serial recall of 6-item sequences, divided into two 3-item subsequences.
• Some items were restricted to specific positions, paralleling artificial phonotactic constraints in speech learning experiments.
• Recall errors were analyzed for sensitivity to absolute and probabilistic constraints.

Main Results:

• Recall errors demonstrated sensitivity to experimental statistical regularities, mirroring findings from speech error studies.
• Both absolute and probabilistic constraints influenced recall errors.
• Learning effects were less pronounced than in phonological learning and less influenced by hierarchical structure.

Conclusions:

• Errors in immediate serial recall can reflect learning of statistical regularities, suggesting domain-general learning mechanisms.
• Phonotactic learning involves both domain-general constraints and language-specific factors.