Despite abundant research on embodied processing in L1, it remains unclear about the temporal characteristics of embodiment effects, particularly regarding the comparison of time courses between L1 and L2. Using conceptual modality switch costs (costs incurred from processing words involving different perceptual modalities) as the modality-switch-effect indicator, this study compared ERP brainwaves between modality-switch and -nonswitch conditions as Chinese-English bilinguals performed property verification tasks in both languages. The results showed that (1) across languages, modality switch effects were observed no later than 160 ms on N1 latency and also on N400 amplitude, but were not observed on the LPC component; (2) the modality switch effect on N1 amplitude was significant in L1, but it did not reach significance in L2. These results highlight the early involvement of embodied mechanisms in meaning retrieval, and reveal different activation levels between L1 and L2 embodied simulations at the initial stage of language comprehension.
While behavioural evidence has demonstrated that stems can be recognised irrespective of position and affixes are subject to more nuanced positional constraints, the present study employed EEG to observe this aspect of visual word recognition. Morphemically transposed compound nonwords (dreamday) elicited significantly reduced N250 and N400 ERP components relative to their controls (shadeday), as well as enhanced P600 activity. Conversely, transposed suffixed (fulplay) and prefixed (portex) nonwords did not elicit morpheme transposition effects relative to their controls (fulprofit, dockex), producing equivalent N250/N400/P600 components across conditions. Finally, robust morpheme transposition effects were found for semantically transparent conditions (fallwater vs. sellwater) across the N250, N400, and P600 components, whereas interference effects only emerged in the P600 time window for semantically opaque conditions (moonhoney vs. snowhoney). These findings offer insights into the morpho-orthographic and semantic processing that underlies morpheme transposition effects, further demonstrating the distinct positional encoding mechanisms of English stems and affixes.
Reading three-character Chinese compound words was studied in two experiments using the lexical decision task and the eye-tracking technology to examine the extent to which they are recognised via their components. To this end, the frequency of the first character was manipulated in a continuous manner separately for two kinds of word structure: those where the first character formed an independent constituent (A-BC words) and those where it was a part of an embedded compound word (AB-C words). The study found evidence for both compositional and holistic processing. Moreover, A-BC words were read somewhat faster than AB-C words in the sentence context but not in isolation. Word-internal structure also affected the way attention was allocated within words. The pattern of results is consistent with parallel dual route models of compound word reading. It is concluded that Chinese readers demonstrate flexibility in using compositional and holistic processing in reading three-character words.
Previous studies established that handwriting production involves central and peripheral processes, but the transition mechanism between central and peripheral processes remains unclear. To elucidate this transition, we employed a priming paradigm integrating writing video primes with picture naming writing tasks while recording electroencephalogram (EEG) during Chinese character writing. Orthographic processing represents a critical stage of central processing immediately preceding peripheral processing, while peripheral processing involves motor processing. We manipulated orthographic processing by varying the orthographic relatedness between prime and target characters, and motor processing by altering writing direction. Cluster-based permutation analyses revealed a main effect of orthographic relatedness at 320-390 ms, an interaction with writing direction at 474-494 ms, and main effects of writing direction at 516-536 ms and 712-734 ms. These findings empirically support a sequential central-to-peripheral transition during Chinese character handwriting, revealing three successive stages, namely orthographic processing, orthographic-motor interaction, and independent motor processing.
A concurrent visuospatial working memory (vsWM) task has been shown to reduce predictive gaze behaviour in several visual world eye-tracking experiments. This reduction could be due to general aspects of prediction or an effect contingent on the additional demand of visually co-present objects. In an N400 event-related potentials study, L2 English participants heard the same sentences and completed the same vsWM task from a study which showed a reduction in predictive gaze behaviour, but without visually co-present objects. We observed robust effects of predictability but evidence against an influence of the vsWM task. These results suggest that vsWM does not interfere with the pre-activation of a target word without a co-present visual array. We conclude that previously observed effects of vsWM on prediction are instead likely specific to gaze behaviour and situations in which spoken language refers to a co-present relevant visual context.
We examine the processing of Tense Harmony, a semantic constraint requiring temporal compatibility between matrix and embedded clauses in English it-clefts (It was the gardener that was calling/*will call). Critically, Tense Harmony applies to Connected Clauses but not to string-identical clefted Relative Clauses. An acceptability judgment task shows that Tense Harmony violations receive significantly lower ratings and slower judgments compared to simple tense mismatches, despite being fully grammatical as a Relative Clause. A Grammatical Maze experiment reveals increased reading latencies immediately at the disambiguating region for these violations. Results demonstrate that Tense Harmony is computed incrementally, that clefted Relatives elicit garden-path effects, and that semantic constraints guide structural expectations. These findings support hybrid models of sentence processing incorporating not only retrieval but also predictive mechanisms in dependency formation, extending our understanding of how semantic and structural information interact in real-time sentence processing.
Previous studies revealed that grammatical gender may shape perception and categorisation, yet the extent of this influence remains unclear, specifically in simultaneous bilinguals with two gendered languages. This pre-registered ERP study investigates how two partially contrasting grammatical gender systems modulate perception in Ukrainian-Russian bilinguals. Twenty-six participants completed a non-verbal categorisation task assessing associations between primes (depicted nouns with matching or mismatching grammatical genders across L1s) and target (male/female faces). Behavioural results (response types and reaction times) showed that bilinguals were not affected by the prime-target gender congruency for matching primes or by more dominant/proficient L1 for mismatched primes. ERP analyses showed no significant modulations of predicted components (N1, P2/VPP, N300) by grammatical gender for either type of primes. This suggests that grammatical gender alone may not independently modulate categorisation processes in bilinguals with two gendered L1s, especially when explicit conceptual or semantic activation is not required.
Based on Complementary Learning Systems theory, this study used a semantic priming lexical decision task following recent or remote learning in single or diverse contexts to examine novel word semantic integration via ERPs and representational similarity analysis (RSA). Behavioral results showed stronger priming for diverse contexts at recent testing, an effect absent after consolidation. Target-word (existing words) N400 revealed semantic priming at recent testing for diverse contexts but only at remote testing for single-contexts. Prime (novel words)-elicited N400 was more negative at recent than remote testing, especially for single-contexts. RSA showed higher representational similarity at remote testing in early time windows (140-198 ms and 224-246 ms), indicating offline consolidation enhanced early neural stability. In contrast, diverse contexts increased similarity in later time windows (236-294 ms), reflecting richer semantic associations. Findings suggest contextual diversity and offline consolidation support semantic integration through temporally distinct but complementary mechanisms.
While the comprehension of sentential negation has been widely studies, the production of negation have received far less attention. This study examined the cognitive and neurophysiological processes underlying negative responses. Using a two-phase verification paradigm, participants first read brief narratives describing the daily activities of two characters, followed by a verification task requiring a Yes/No response through button presses. ERP analyses showed significant differences between Yes and No responses across both early and late time windows. Source estimations revealed a distinct two-step pattern: Yes responses involved early activation in posterior-temporal regions, while No responses showed late greater activation in anterior regions. Lateralised Readiness Potential (LRP) analyses further supported this two-stage process, indicating an initial bias toward affirmation followed by inhibition. These findings suggest that the production of negation engages a two-stage processing mechanism, likely involving inhibition circuitry during the second step.
Unlike information-seeking questions (ISQs), rhetorical questions (RQs) do not ask for information but make a point. Behavioural data showed that this difference in illocution type is mainly signalled by intonation, with RQs requiring more specific tonal patterns than ISQs. This EEG study investigated the processing of tonally marked ISQs and RQs, when the illocution type was congruently or incongruently cued by a preceding mnemonic visual stimulus. We predicted a prosodic expectancy positivity (PEP) in incongruent vs. congruent conditions and earlier detection of incongruence in tonally marked ISQs. ERP results and an exploratory analysis of oscillatory brain activity showed congruency effects for tonally marked ISQs and RQs and different timing: PEP-like effects and alpha suppression occurred earlier in tonal ISQs than RQs. Beta-band suppression emerged at the utterance's end for both. The results show that tonally marked RQs require a more nuanced prosodic analysis of the entire utterance for interpretation.
Previous research has demonstrated grammatical constraints under which phonologically-related words help to resolve tip-of-the-tongue (TOT) states: Words in a different grammatical class from the TOT increase resolution while words in the same grammatical class do not. The present research investigated whether grammatical class effects on TOT resolution result from nonpresented neighbours, which can be activated via nonword primes' suffixes. During TOTs, participants completed an auditory lexical decision task where one item was a first-syllable nonword prime or unrelated nonword. Suffixes of nonword primes, which lack lexical representations, indicated a part of speech different from the target, same as the target, or ambiguous in part of speech. Different part-of-speech nonword primes and ambiguous part-of-speech nonword primes increased TOT resolution, whereas same part-of-speech nonword primes did not. These results refine earlier theories of grammatical class constraints on phonological priming of TOT resolution by documenting a novel influence of a TOT's first-syllable neighbours.
We report three experiments using a novel variation of the word flanker paradigm to test whether readers process articles and nouns together, taking advantage of the fact that, in Spanish, there must be gender agreement (concordance) between articles and nouns. In Experiment 1, we asked Spanish native speakers to make grammatical gender judgments about centrally presented Spanish nouns (e.g. "mesa", "table"). The central nouns could be presented with concordant article flankers ("la mesa la"), discordant article flankers ("el mesa el") or no flankers. We found that responses were slower and less accurate in the discordant condition. In Experiment 2, a lexical decision task, this effect was weaker than in Experiment 1. In Experiment 3, a semantic categorisation task, we found no evidence of the article flanker effect. This suggests that articles and nouns are processed jointly, but the impact of this joint processing depends on the task.
Mild traumatic brain injury (mTBI) often leads to subtle cognitive-communication difficulties. This study combined behavioural measures, eye-tracking, and cognitive tasks to investigate sentence processing in individuals with mTBI. Twenty-five participants with mTBI and 30 neurotypical controls completed a sentence-picture matching task featuring canonical and noncanonical sentence structures while eye movements were recorded. Cognitive function was assessed using working memory and attention tasks. Although aggregate group differences were minimal, eye-tracking measures revealed group differences in attentional allocation during sentence processing, suggesting altered coordination of cognitive resources following mTBI. These findings are most consistent with resource reduction accounts of sentence comprehension, which posit that comprehension difficulties arise from reduced efficiency in deploying cognitive resources as processing demands increase. These results additionally highlight the value of multimodal approaches for detecting subtle language processing differences following mTBI.
This study explored how bilingual language experience modulates the lexical properties of words produced in an English verbal fluency task. Ninety-nine French-English bilinguals completed letter and semantic fluency trials, and the resulting words were analysed for "cognateness", phonological neighbourhood density, and lexical frequency. We examined how these properties varied based on bilinguals' L2 proficiency, immersion, and age of onset. Results showed that more proficient bilinguals and those with longer immersion produced lower-frequency words with denser phonological neighbourhoods, suggesting a larger and more integrated L2 lexicon. Earlier onset bilinguals, in contrast, produced words with higher lexical frequency than later onset bilinguals. Bilingualism effects were more pronounced in the letter condition, likely due to its broader lexical search space and higher cognitive control demands. These findings highlight how bilingual lexical activation is shaped not only by language experience and acquisition profile, but also by the cognitive and linguistic constraints of the task.
