Working memory capacity and speech production in L2: evidences from a picture description task

Finardi, Kyria; Prebianca, Gicele Vergine Vieira · 2006 · Crossref

DOI: 10.17851/2237-2083.14.1.231-260

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the relationship between working memory capacity (WMC) and second language (L2) speech production, specifically examining how limited cognitive resources influence fluency, accuracy, complexity, and lexical density. Motivated by the view that L2 speaking is a complex cognitive task constrained by working memory, the authors aim to determine if WMC predicts speech performance and whether WMC is a domain-free or task-specific construct. The research seeks to resolve conflicting evidence from prior studies regarding the predictive power of different working memory span tests on L2 oral output. The experimental design involved twelve intermediate-level English as a Foreign Language (EFL) graduate students. Participants completed a picture description task to elicit speech production, which was analyzed for four metrics: speech rate (fluency), errors per 100 words (accuracy), dependent clauses per minute (complexity), and weighted lexical density. To assess WMC, participants took two adapted tests: the Speaking Span Test (SST), considered task-specific, and the Operation Word Span Test (OWST), considered domain-free. The SST required participants to read words and produce sentences aloud, while the OWST involved solving arithmetic operations while memorizing words. Statistical analysis, including Pearson Product Moment Correlation, was used to examine relationships between WMC scores and speech production measures. The results revealed a statistically significant positive correlation between SST scores and speech rate, indicating that participants with higher working memory capacity, as measured by the task-specific SST, produced speech more fluently. No significant correlations were found between the OWST and any speech production measures, nor between either WMC test and accuracy, complexity, or lexical density. Additionally, a significant positive correlation existed between speech rate and complexity, while a significant negative correlation was found between speech rate and weighted lexical density. These findings suggest that participants who spoke faster and used more complex structures tended to use more familiar, high-frequency words, thereby reducing lexical density. The study concludes that working memory capacity appears to be task-specific in the context of L2 speech production, supporting the hypothesis that efficiency in specific processing tasks drives performance. The findings corroborate the existence of trade-off effects among speech production variables; under processing pressure, L2 speakers prioritize certain aspects, such as fluency and complexity, at the expense of others, like lexical density. The authors attribute the lack of broader correlations to methodological limitations, including a small sample size that limited variance in WMC scores and the specific properties of the picture used, which may not have elicited sufficiently complex or lexically dense speech. The study highlights the need for future research to control for non-linguistic variables and utilize diverse elicitation methods to better understand the cognitive constraints of L2 speaking.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-25; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• working memory