Lexical Access in Comprehension vs. Production: Spatiotemporal Localization of Semantic Facilitation and Interference

Abstract

Humans understand words faster when they are preceded by semantically related words. This facilitation is thought to result from spreading activation between words with similar meanings. Interestingly, in language production, semantic relatedness often has the opposite effect: in object naming for example, a related prior word delays the naming time of the current object. This could be due to competition during conceptual search or later interference at the motor preparation stage. However, no study has systematically compared the facilitory and inhibitory effects and thus their neurobiological relationship is unknown. We contrasted maximally parallel production and comprehension tasks during magnetoencephalography and found that in comprehension (specifically word reading), semantic relatedness modulated activity in the left middle STG at 180-335ms, consistent with prior findings on the spatiotemporal localization of lexical access. In contrast, a semantic interference pattern for the production task (object naming) occurred in a postlexical time-window at 395-485ms in left posterior insular cortex, consistent with postlexical motor preparation. Thus, our data show that semantic priming during comprehension and interference during production are not two sides of the same coin but rather they clearly dissociate in space and time, consistent with a lexical account for comprehension and a post-lexical one for production. Significance statementThe processing of semantically related words has been a central tool for understanding the organization of the mental lexicon. One striking observation is that semantic relatedness tends to be facilitory in comprehension but inhibitory in language production, perhaps because only production involves a conceptual search through semantically related candidates. The neurobiology of this contrast is not understood. Our magnetoencephalography results demonstrate that the facilitory pattern is first observed in classic left temporal lexical access regions at ~200ms, whereas the inhibitory pattern occurs later and in the insular cortex. These findings show that the two effects do not co-localize in space or time and suggest that the inhibitory effects in production stem from a late motor preparation stage.

Spatiotemporal properties of common semantic categories for words and pictures

“…Another recent MEG study (Dirani and Pylkkänen, 2023) examining generalization between written words and pictures using picture naming and word reading tasks showed different results. Significant decoding (animal vs. tool) activates surprisingly early around 75 ms for pictures, and 95 ms for words.…”

“…Furthermore, the temporal dynamic of object recognition may also be influenced by the specific experimental task employed in the study, such as a one-back task, category judgment, or detection task. As such the precise timing of semantic category activation remains an open question (Giari et al, 2020;Kaiser et al, 2016;Dirani and Pylkkänen, 2023;Miozzo et al, 2015;MacGregor et al, 2012;Dirani and Pylkkänen, 2020;Simanova et al, 2010).…”

“…This includes testing the hub-and-spoke theory of semantic representations where the representation first arises in a modality-specific way and later transfers to the amodal hub (Ralph et al, 2017;Humphreys et al, 2015;Pobric et al, 2007). There is limited research done to investigate the time of cross-modal generalization using M/EEG and the reported time dynamic of cross-modal representation varies (Leonardelli et al, 2019;Dirani and Pylkkänen, 2023;Iamshchinina et al, 2022;Simanova et al, 2010).…”

“…The different time courses of the shared semantic representation between (Dirani and Pylkkänen, 2023) study and (Leonardelli et al, 2019) study could be attributed to variations in participant tasks and experimental paradigms. Importantly, in (Dirani and Pylkkänen, 2023) a block presentation of categories was utilized to demonstrate cross-modal categorization. This might result in anticipatory effects making the semantic categorization occur earlier.…”

Behavioural and EEG evidence for inter-individual variability in late encoding stages of word production

“…Greenham et al (2000) found a similar pattern, specifically larger amplitudes for semantically incongruent than congruent words, but with an N450 component. Using MEG (which allows for spatial precision as well as temporal precision), Dirani & Pylkkänen (2018) found the semantic interference effect at around 395-485 ms in frontal motor regions, and take this result to suggest that this effect occurs during motor planning. On the other hand, Dell'Acqua et al (2010) found evidence of the semantic interference effect at a much earlier time window of around 100 ms after stimulus onset, mostly in left frontal and temporal electrodes.…”