دانشگاه اصفهان نشریه پژوهش های زبان شناسی 2322-3413 18 1 2026 03 21 Cross-Domain Structural Priming of Action Sequences and Relative Clause Constructions: Tracing the Embodied Bases of Syntactic Structures آماده‌سازی ساختاری میان‌حوزه‌ای توالی کنش‌ها و ساخت‌های موصولی: در جستجوی پایه‌های بدن‌مند ساختارهای نحوی 71 91 29371 10.22108/jrl.2025.143908.1888 FA امیرعباس رفیعی فاضل دانش‌آموختۀ دکتری علوم شناختی-زبان‌شناسی، گروه زبان‌شناسی، دانشکدۀ علوم انسانی، دانشگاه تربیت مدرس، تهران، ایران ارسلان گلفام دانشیار گروه زبان‌شناسی، دانشکدۀ علوم انسانی، دانشگاه تربیت مدرس، تهران، ایران حیات عامری دانشیار گروه زبان‌شناسی، دانشکدۀ علوم انسانی، دانشگاه تربیت مدرس، تهران، ایران سحر بهرامی خورشید دانشیار گروه زبان‌شناسی، دانشکدۀ علوم انسانی، دانشگاه تربیت مدرس، تهران، ایران Journal Article 2025 01 05 Abstract One of the controversial questions in neuropsychological investigations of language is whether syntactic mechanisms interact with a domain-general hierarchical processor or operate independently from other cognitive domains. Inspired by theoretical conjectures regarding structural parallels between language syntax and action sequences, this study focused on the correspondence between sentences featuring extraposed and center-embedded subject Relative Clauses (RCs) and complex actions with linear and nonlinear sequences, respectively. To explore this relationship, we conducted a series of cross-domain structural priming experiments to determine whether the abstract structures of actions influenced participants’ processing speed in grammaticality judgments of RCs. Thirty-one native Persian speakers aged 18-32 years voluntarily participated in the study. We manipulated the participants' exposure to complex actions with linear and nonlinear sequences across three experimental conditions: execution, observation, and reading of means-end sentences. The results from pooled data across all experimental conditions suggested a structural interaction between linguistic and nonlinguistic domains. However, we found no significant evidence of priming effects within each experimental condition. Using linear mixed-effects models, the overall analysis indicated a main effect of the experimental condition. Keywords: Structural Priming, Embodiment of Syntax, Relative Clause (RC) Constructions, Action Sequences, Persian. Introduction Several scholars have proposed theoretical frameworks to explore the connection between linguistic syntax and the brain action systems (e.g., Glenberg & Gallese, 2012; Knott, 2012; Pulvermüller & Fadiga, 2010). To investigate how bodily representations of syntax are integrated within the sensorimotor ("mirror") system, Pulvermüller and Fadiga (2010) suggested that the hierarchical structure of a center-embedded Relative Clause (RC) (e.g., “the man [whom the dog chased] ran away”) resembles a superordinate sequence, where one action is nested between two others (e.g., open door [switch on light] close door). In recent years, a growing body of empirical research has provided evidence for shared structural processing mechanisms between linguistic and nonlinguistic domains (e.g., Casado et al., 2018; Roy et al., 2013; Van de Cavey & Hartsuiker, 2016). These cross-domain structural interactions suggest that syntax processing may be less language-specific than traditionally believed. This study investigated whether prior exposure to linear and nonlinear action sequences could influence Response Time (RT) patterns in grammaticality judgments of extraposed and center-embedded subject RCs. We also explored whether the occurrence and strength of priming effects varied based on the modality, in which the prime stimuli were presented: execution (motor), observation (sensory), and reading of means-end sentences (verbal). Materials & Methods We employed a 2 × 2 × 2 × 3 design incorporating 4 within-participant variables: prime type (linear sequence vs. nonlinear sequence), sentence correctness (correct vs. incorrect), trial congruency (congruent vs. incongruent), and prime condition (execution, observation, and reading). These variables served as our independent factors. We measured RTs as an indicator of how quickly participants judged the grammaticality of sentences in terms of subject-verb agreement, which was our primary dependent variable. RTs were calculated as the duration from the presentation of a question mark immediately following the last word (the verb) to the moment participants responded by pressing the appropriate keystroke (the right shift key for ‘True’ and the left shift key for ‘False’). Following the methodology of Allen et al. (2010), we formulated 32 means-end sentences as prime items with half structured as linear sequences (e.g., “Pouya picked up the hand shovel, dug a hole, and sowed the beans”) and the other half as nonlinear sequences (e.g., “Pouya turned on the tap, wore the dishwashing gloves, and washed the spoon”). For target items, we selected two types of subject RCs distinguished by their embedding locations: 8 extraposed RCs (e.g., “Parastār āmpul-i-rā zad-Ø [ke mard-rā bihuš kard-Ø]”/“The nurse gave the injection [that anesthetized the man]”) and 8 center-embedded RCs (e.g., “Mard-i [ke tabar-rā tiz kard-Ø] deraxt-rā qat’ kard-Ø”/“the man [that sharpened the ax] cut down the tree”). Half of these sentences included a subject–verb morphosyntactic anomaly. The (un)grammaticality of sentences was not revealed until the last word, which was always the 8th word in the two RC types. Additionally, we constructed a set of filler materials, including 12 conjoined sentences and 12 two-action means-end sentences, ensuring at least one filler item intervened between each prime-target sequence. Each experimental session was divided into 4 blocks, each presenting a different combination of prime pairs and target sentences to create structurally congruent and incongruent trials. We utilized a cross-domain structural priming method, where participants read, observed, or executed prime action sequences before judging the grammaticality of sentences in a self-paced reading task. The study involved three visits spaced approximately two weeks apart. All stimulus presentations and data collection were managed using DMDX software (version 6.0.1.0). Discussion of Results & Conclusion By pooling the RTs of extraposed RCs across all conditions, we observed a statistically significant difference between congruent and incongruent trials (V=5100, p=.04). For center-embedded RCs, the Wilcoxon signed-rank test indicated a marginally significant difference (V=5821, p=.09). Additionally, we analyzed the overall data using linear mixed-effects models in R (lme4 package; Bates et al., 2015). The final best-fitting model included the main effects of condition, sentence correctness, and congruency; however, no significant effect of prime type was found. Our findings provided suggestive, yet inconclusive, evidence for cross-domain priming in structural information processing. The limited number of participants in each condition, as well as the relatively small number of trials eligible for analysis, might partly account for these results. Another consideration was that the prime stimuli were constructed from a linguistic perspective, focusing on the general goals of actions rather than the specific movements involved (i.e., at a detailed movement level). (K. Pastra, personal communication, February 18, 2022, and February 21, 2022). This approach might have impacted the effectiveness of the primes. Furthermore, the lack of comparable mean RTs across the tested modalities suggested that they might differ in terms of timing, manner, and the amount of information they conveyed (see Fischer & Zwaan, 2008). یکی از مباحث بحث‌‌انگیز در تحقیقات عصب-روان‌شناسی زبان پاسخ به این پرسش است که آیا سازوکارهای نحوی با یک پردازشگر سلسله‌مراتبی حوزه‌عام در تعامل هستند یا عملکردی مستقل از سایر حوزه‌های شناختی دارند. بر پایۀ نظرپردازی‌های صورت‌گرفته مبنی بر وجود تناظرهای ساختاری بین نحو زبان و توالی کنش‌ها، در این پژوهش جملات موصولی فاعلی پسایندی و درونه‌ای‌میانی را به ترتیب در تطابق نسبی با توالی خطی و غیرخطی کنش‌های غیرساده قرار دادیم. از این ‌رو، در مجموعه‌ای از آزمایش‌های آماده‌سازی ساختاری میان‌حوزه‌ای بررسی کردیم که آیا ساختار انتزاعی کنش‌ها می‌تواند سرعت پردازش آزمودنی‌ها را در قضاوت دستوری جملات موصولی تحت تأثیر قرار دهد. بدین‌منظور، 31 آزمودنی فارسی‌زبان در دامنۀ سنی بین 18 تا 32 سال به‌طور داوطلبانه در این مطالعه شرکت کردند. متعاقباً بسامد رویارویی آنها را‌ نسبت به کنش‌هایی با توالی خطی و غیرخطی در سه شرایط آزمایشی اجرا، مشاهده و خواندن جمله‌های وسیله-هدف تغییر دادیم. نتایج حاصل از تجمیع داده‌ها در شرایط آزمایشی مختلف بیانگر نوعی تعامل ساختاری بین حوزه‌های زبانی و غیرزبانی بود. اما شواهد درخور توجهی برای اثر آماده‌سازی در هر یک از شرایط آزمایشی به‌طور جداگانه یافت نشد. نتایج تحلیل کلی داده‌های به‌دست‌آمده از طریق مدل‌های اثرات آمیختۀ خطی نیز مؤید اثر عمدۀ شرایط آزمایشی بود.

https://jrl.ui.ac.ir/article_29371_7464a84a417a214ff7fb176357da187d.pdf