The role of working memory loads on immediate and long-term sentence recall

ABSTRACT

It is well-established that both phonological and semantic knowledge influence verbal working memory. However, the focus has primarily been on understanding phonological effects despite evidence of semantic influences. Articulatory suppression is a well-established task for preventing phonological processing. Methods to prevent semantic processing have rarely been used in the past, highlighting a need for developing a semantic interference task. We, therefore, conceptualised two novel tasks – an animacy categorisation and semantic relatedness judgement task. This study explored the impact of phonological (articulatory suppression) and semantic loads (animacy categorisation and semantic relatedness judgement) on immediate and delayed sentence recall. Additionally, sentence concreteness (concrete vs. abstract sentences) indexed semantic knowledge in verbal working memory. Across two studies, immediate recall revealed that articulatory suppression (preventing phonological processing) increased the size of the concreteness effect, while the novel semantic tasks (preventing semantic processing) reduced it suggesting that our semantic tasks were indeed imposing a semantic load. Further, relative long-term performance showed that more new words were remembered in articulatory suppression, whereas recall was disproportionately impaired in the semantic relatedness task. Our experimental paradigm offers phonological and semantic suppression tasks that can be used in parallel to investigate the interactions between working memory and language.

KEYWORDS:

• Working memory
• sentence recall
• semantic knowledge
• concreteness effect
• articulatory suppression

Acknowledgements

This work was completed as part of the first author’s PhD thesis. We would like to thank Katie Flannery, Areej Malik, Nivedita Varagunan, and Long Pham for their assistance with data collection, transcription, and scoring.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The material and data that support the findings of this study are available in Open Science Framework at https://osf.io/5a2p6/

Notes

1 Note that although some researchers distinguish between concreteness (experienced by the senses) and imageability (arouses mental imagery), these two concepts are frequently used interchangeably because of their high correlation (e.g., Binder et al., 2005; Fliessbach et al., 2006; Sabsevitz et al., 2005).

2 We thank Jed A. Meltzer for sharing his stimuli with us.

3 An additional analysis was conducted to compare working memory loads across the two experiments. There was no difference between the articulatory suppression and animacy categorization tasks across experiments, BF₀₁ = 23.086 and BF₀₁ = 1.34, respectively. However, the control condition led to better immediate recall than finger-tapping, BF₁₀ = 112.57.

Additional information

Funding

This work was supported by Natural Sciences and Engineering Research Council of Canada [grant number RGPIN-2014-03982] to LA; Ontario Graduate Scholarship to TP; Social Sciences and Humanities Research Council of Canada to TP.