Back

Dissociable roles of neural pattern reactivation and transformation during recognition of words read aloud and silently: An MVPA study of the production effect

Bailey, L. M.; Matheson, H. E.; Fawcett, J. M.; Bodner, G. E.; Newman, A. J.

2024-02-23 neuroscience
10.1101/2024.02.20.581164 bioRxiv
Show abstract

Recent work surrounding the neural correlates of episodic memory retrieval has focussed on the decodability of neural activation patterns elicited by unique stimuli. Research in this area has revealed two distinct phenomena: (i) neural pattern reactivation, which describes the fidelity of activation patterns between encoding and retrieval; (ii) neural pattern transformation, which describes systematic changes to these patterns. This study used fMRI to investigate the roles of these two processes in the context of the production effect, which is a robust episodic memory advantage for words read aloud compared to words read silently. Twenty-five participants read words either aloud or silently, and later performed old-new recognition judgements on all previously seen words. We applied multivariate analysis to compare measures of reactivation and transformation between the two conditions. We found that, compared with silent words, successful recognition of aloud words was associated with reactivation in the left insula and transformation in the left precuneus. By contrast, recognising silent words (compared to aloud) was associated with relatively more extensive reactivation, predominantly in left ventral temporal and prefrontal areas. We suggest that recognition of aloud words might depend on retrieval and metacognitive evaluation of speech-related information that was elicited during the initial encoding experience, while recognition of silent words is more dependent on reinstatement of visual-orthographic information. Overall, our results demonstrate that different encoding conditions may give rise to dissociable neural mechanisms supporting single word recognition.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.