Content area
Syntactic priming, wherein prior exposure to a linguistic structure influences subsequent processing, has been less extensively explored for ditransitive structures (double-object and prepositional-object) among L1-English L2-Chinese learners. This study investigates cross-linguistic priming in 57 native English-speaking L2 learners of Chinese, varying in proficiency, using a translation selection task (comprehension) and a picture description task (production). The results reveal significant cross-linguistic priming effects in both comprehension and production tasks. Notably, in Task 2, the production rates of both syntactic structures under priming conditions were significantly higher than those observed under non-priming conditions. In addition, advanced learners exhibited stronger priming effects than beginners. These findings tentatively support the developed version of the bilingual lexical-syntactic representation model and the error-based implicit learning model, indicating that cross-linguistic priming is modulated by learners’ proficiency, where increased exposure and error-driven adjustments strengthen the connection between L1 and L2 syntactic representations. Furthermore, there is also an indication that modality might influence structural preferences, with prepositional-object structures possibly being favored in comprehension and double-object structures potentially more common in production, although these observations remain tentative. Overall, the insights offer valuable implications for optimizing Teaching Chinese as a Second Language strategies, particularly through tailored comprehension and production activities to enhance syntactic acquisition.
Introduction
Syntax poses a significant challenge in teaching Chinese as a second language (TCSL), particularly for the second language (L2) learners with English as their first language (L1). One common source of difficulty is the negative transfer from English syntactic patterns to Chinese, which often leads to ungrammatical constructions. For example, learners may produce sentences like 妈妈做饭给他 (Māma zuòfàn gěi tā, Mom cooks for him), which, although structurally acceptable in English, violate Chinese syntactic norms. Among the most problematic structures are ditransitive structures, including both double-object (DO) and prepositional-object (PO) structures, which frequently lead to errors in both comprehension and production. Despite their ubiquity in everyday communication, these structures remain difficult for learners to master, pointing to deeper challenges in cross-linguistic syntactic processing.
Cross-linguistic syntactic priming refers to the facilitative effect whereby exposure to a syntactic structure in one language aids in the processing or production of the same structure in another language. This phenomenon offers a valuable framework for investigating bilingual syntactic development (Hartsuiker et al., 2004). While previous research has examined DO and PO structures primarily in the Chinese-to-English direction (e.g., Wang & Liu, 2018), relatively little is known about how syntactic priming operates in the reverse direction, that is, how L1-English L2-Chinese learners process and produce these structures. Moreover, few studies have systematically examined whether and how priming effects vary across proficiency levels—a gap that limits our understanding of the underlying psycholinguistic mechanisms and hinders the development of evidence-based pedagogical strategies for L2 instruction.
To address this gap, the present study aims to investigate cross-linguistic syntactic priming of DO and PO structures in L1-English L2-Chinese learners, focusing on the role of Chinese proficiency in modulating priming effects during sentence comprehension and production. Drawing on Bernolet’s (2008) developed model of bilingual lexical-syntactic representation, and the error-based implicit learning model (Chang et al., 2006), a translation selection task (for comprehension) and a picture description task (for production) were employed to address the following questions: (1) Does cross-linguistic syntactic priming occur between English and Chinese for DO and PO structures? (2) How does Chinese proficiency influence these priming effects? By analyzing priming effects and acquisition patterns across proficiency levels, this study seeks to provide insights into the psychological mechanisms underlying L2 syntactic processing and offer practical strategies for TCSL, ultimately reducing negative L1 transfer and enhancing L2 learners’ mastery of Chinese ditransitive structures.
Background literature
Chinese ditransitive structures in L2 acquisition
Ditransitive structures in Chinese typically fall into two types: double-object (DO) and prepositional-object (PO) structures (Dryer, 1986; Heine & König, 2010). In DO structures, the verb is typically followed by an indirect object (usually a recipient) and a direct object (usually a theme), as in 我送给他一个礼物 (Wǒ sònggěi tā yīgè lǐwù, I gave him a gift). In contrast, PO structures use a preposition, most commonly 给 (gěi), to introduce the indirect object, as in 我把那个礼物送给他 (Wǒ bǎ nàgè lǐwù sòng gěi tā, I gave the gift to him).
Although DO and PO structures appear regularly in native input, they pose distinct challenges for L2 learners due to syntactic mismatches between English and Chinese. DO structures have long been debated in Chinese linguistics across various theoretical frameworks (e.g., Li, 2007; Wang & Ding, 2014; Zhou, 2009) and often involve strict verb-specific constraints, especially with less frequent verbs like 通知 (tōngzhī, inform), and 提醒 (tíxīng, remind), which are not easily mapped onto familiar English patterns. However, their most prototypical subtype—the giving constructions (e.g., 给, gěi, give)—tends to be more accessible for L1-English L2-Chinese learners due to clear structural and semantic parallels with English DO structures (Chang, 2014). PO structures, on the other hand, appear less constrained by verb type but often involve greater syntactic complexity. They frequently appear within bǎ constructions (e.g., 把…送给…, bǎ…sònggěi…), whose rules are less transparent to L2 learners and more context-sensitive. Moreover, English L1 learners may overgeneralize from the uniform use of the English preposition to, resulting in frequent errors in parsing or production in Chinese. As a result, despite their surface simplicity, PO structures may impose a heavier cognitive burden on L2 learners and present substantial learning challenges.
Given these challenges, ditransitive structures offer a particularly rich domain for investigating syntactic priming and structural adaptation in L2 Chinese acquisition. Understanding how L2 learners acquire these structures and whether their processing is modulated by syntactic priming can help inform both theoretical accounts of bilingual syntax and practical approaches to instruction in TCSL.
Cross-linguistic syntactic priming
In linguistic research, priming refers to the influence of language structures or meanings that participants have previously been exposed to on subsequent language processing, which is manifested in language comprehension or language production (McDonough & Trofimovich, 2011). Priming research comprehensively explores diverse forms in the field of linguistics, notably encompassing lexical, semantic, and syntactic structure priming. Among these, syntactic priming—the focus of this study—occurs when individuals replicate syntactic structures recently encountered through reading, listening, or producing language in their subsequent sentence generation (Yang & Zhang, 2007). For example, exposure to a PO structure like The company offered a job to the graduate increases the likelihood of producing a similar structure, such as He sent a gift to her. This reflects the underlying mechanisms of syntactic processing during sentence generation.
Cross-linguistic syntactic priming, particularly relevant to bilingual or L2 learners, involves the influence of a syntactic structure in one language on processing or production in another (Kootstra & Muysken, 2017). For instance, Loebell and Bock (2003) demonstrated that dative structures (double-object and prepositional-object) in German primed similar structures in English among bilinguals, suggesting shared syntactic representations across languages. Similarly, Hartsuiker et al. (2004) found that Spanish passive sentences primed English passives in bilinguals during a dialogue task, indicating cross-linguistic syntactic persistence. In the Chinese context, Zhang and Wang (2012) observed that Indonesian active and passive sentences primed Chinese equivalents in Indonesian-Chinese L2 learners, highlighting the potential for L1-to-L2 priming in Sino-Tibetan languages, though such studies remain scarce. Collectively, these findings highlight the role of L1 in influencing L2 syntactic choices—a phenomenon that is central to understanding L2 acquisition.
Various tasks have been used to investigate cross-linguistic syntactic priming, each suited to specific modalities, with priming effects shown to vary depending on the type of task. Comprehension tasks, such as translation selection, assess how learners process primed structures by presenting them with primed sentences and multiple response options to assess their structural preferences (Li, 2006), while production tasks, such as picture description, require learners to generate sentences based on visual stimuli following a prime (Bock, 1986). Empirical studies have further illustrated how priming patterns differ by task modality. For example, in a comprehension-based paradigm, Thothathiri and Snedeker (2008) used a visual-world eye-tracking task to demonstrate that English DO primes led listeners to focus on potential recipients, while PO primes directed attention to objects, revealing modality-specific processing effects. In contrast, in a production setting, Bock (1986) employed a picture description task and found robust structural priming effects, with exposure to DO or PO structures increasing the likelihood of using matching syntactic forms. These differences inform this study’s use of translation selection (comprehension) and picture description (production) tasks to compare priming effects in L1-English L2-Chinese learners, addressing a gap in modality-specific research for this language pair.
L2 Proficiency dynamics
Proficiency significantly modulates cross-linguistic syntactic priming in L2 learners, another central focus of this study. Several studies have shown that as L2 proficiency increases, learners are more likely to exhibit syntactic priming across languages. For example, Wang (2009) found that higher-proficiency L1-Chinese L2-English learners showed stronger priming of English ditransitive structures than lower-proficiency learners in both written and oral tasks, suggesting that syntactic representations become increasingly shared. Similarly, Bernolet (2008) reported that cross-linguistic priming was observed only in advanced L1-English L2-Dutch learners, indicating that proficiency facilitates structure priming across languages. These findings collectively suggest that syntactic priming may reflect long-term structural learning that strengthens with L2 proficiency.
However, other studies have suggested that cross-linguistic priming may be more robust at lower levels of L2 proficiency. Li (2006), for instance, argued that cross-linguistic priming is a transitional cognitive process primarily observed in early stages of L2 acquisition. His study found that lower-proficiency Chinese-English learners were more likely to rely on L1 word order when translating, leading to greater priming effects, while higher-proficiency learners exhibited reduced reliance on L1 structures. Similarly, Xu (2014) reported that in a translation selection task involving Chinese complex sentences, priming effects were stronger among low-proficiency learners and diminished as L2 proficiency increased, possibly due to suppression of L1 representations. These findings suggest that at early stages, learners may employ a “translation strategy” that increases priming magnitude, whereas more advanced learners exhibit greater syntactic autonomy and flexibility, reducing observable priming. Such proficiency effects appear to be modulated by factors including direction of transfer, task demands, and structural complexity.
The interaction between proficiency and task modality further shapes priming effects. Shin (2010) examined English-Korean bilinguals and found that cross-linguistic structural priming in production occurs only among proficient L2 speakers, suggesting that bilingual syntactic processing in production is initially asymmetrical but develops with increasing proficiency. Xu and Zeng (2024) extended this to L1-English L2-Chinese learners, using a self-paced reading task to assess comprehension. They found that higher-proficiency learners exhibited stronger priming of sentential complement structures, indicating that proficiency enhances the expectation for syntactic repetition in comprehension tasks. In contrast, Ye (2024) investigated L1-Chinese L2-English learners with non-canonical cleft structures, using picture description (production) and description verification (comprehension) tasks. While bidirectional priming was observed, the strength of priming was not sensitive to proficiency, suggesting that for certain complex structures, structural complexity may override proficiency effects in both modalities. Additionally, Lei and Wang (2025) explored cross-domain priming from mathematical processing to L2 English sentence comprehension among Chinese learners, finding that higher proficiency increased priming effects in comprehension, suggesting that proficiency may also facilitate structural processing in related cognitive domains.
These findings reveal a multifaceted landscape of proficiency’s influence on cross-linguistic syntactic priming, where advanced learners leverage integrated syntactic representations to enhance priming in comprehension (e.g., Xu & Zeng, 2024) and production (e.g., Shin, 2010), while early learners rely on L1-driven priming as a scaffold for acquisition (Li, 2006; Xu, 2014). The divergence in priming strength across modalities and structures, heightened in comprehension yet tempered by complexity in production (e.g., Ye, 2024), suggests a dynamic interplay between linguistic proficiency and cognitive adaptation, potentially extending to cross-domain processing (e.g., Lei & Wang, 2025). For TCSL, this implies a transformative pedagogical framework: early learners could benefit from comprehension-focused activities to harness L1 priming, while advanced learners might thrive with production tasks that challenge structural complexity, fostering a deeper, more autonomous syntactic mastery tailored to their evolving linguistic capabilities.
Theoretical models
To explain cross-linguistic syntactic priming, several models have been proposed. The bilingual lexical-syntactic representation model (Schoonbaert et al., 2007) proposes that bilinguals share lemma, syntactic feature, and combinatorial nodes across languages, enabling priming when translation-equivalent verbs (e.g., English give and Chinese 给gěi) activate shared syntactic structures. This model accounts for verb repetition and translation equivalence effects observed in this study’s tasks. Building on this, Bernolet’s (2008) developed model suggests that L2 proficiency determines the degree of syntactic sharing, with advanced learners exhibiting stronger priming due to integrated L1-L2 representations. This model is directly relevant to this study, as it predicts stronger priming in advanced L1-English L2-Chinese learners for DO and PO structures.
In contrast, the error-based implicit language learning model (Chang et al., 2006) offers a dynamic alternative, suggesting that syntactic priming emerges from prediction error-driven adaptation. When a learner’s expected syntactic structure (e.g., a double-object dative) is violated by input (e.g., a prepositional-object dative like give a book to someone), the resulting prediction error triggers an implicit adjustment of syntactic representations, strengthening the primed structure (e.g., the Chinese equivalent 把一本书给某人, bǎ yìběnshū gěi mǒurén). Khoe et al. (2021) further elaborate that the system responds by implicitly strengthening the unexpected but observed structure, increasing its likelihood of future use, with this learning being cumulative and long-lasting, thus unifying syntactic acquisition and priming. Supporting this, van Dijk and Hopp (2025) found that in German-English cross-linguistic priming, prediction errors triggered by verb biases (e.g., favoring DO or PO structures) led to distinct priming patterns (PO in L1-to-L2, DO in L2-to-L1), reflecting the role of L1-driven expectations in shaping implicit adjustments. Similarly, Şafak and Hopp (2025) showed that prediction errors in L1-German L2-English learners, observed via eye-tracking in comprehension tasks, persisted into production, though effects were moderated by verb-specific constraints, such as non-alternating verbs (e.g., donate), which limit structural options and reduce priming magnitude. Recent modeling work by Khoe et al. (2021) further supports this, showing that error-driven learning in a bilingual Dual-path model simulates cross-linguistic priming by adjusting structural expectations based on prediction errors, a process that accumulates over time and applies to both within- and cross-language contexts.
This error-based perspective complements the lexical-syntactic model by emphasizing adaptive, experience-driven changes over static representations. It directly supports the study’s aim to explain cross-linguistic syntactic priming effects in L1-English L2-Chinese learners, suggesting that prediction errors, shaped by L1 expectations and structural constraints, may underlie the observed DO and PO priming across comprehension and production modalities. For TCSL, this model advocates for structured exposure to varied syntactic inputs, inducing prediction errors to enhance implicit learning of Chinese DO and PO structures across proficiency levels.
The study
This study investigates cross-linguistic syntactic priming of DO and PO structures in L1-English L2-Chinese learners, focusing on how Chinese proficiency influences priming effects in comprehension and production. Conducted with native English-speaking international students in Beijing, China, the study refers to these participants as L2 learners of Chinese with L1 English to reflect their status as non-native Chinese learners with varying proficiency levels, aligning with the study’s focus on second language acquisition (SLA). It employs two tasks: a translation selection task (Task 1) to assess comprehension and a picture description task (Task 2) to evaluate production. Targeting the L1-to-L2 direction, the study explores acquisition patterns of DO and PO structures and aims to clarify the psychological mechanisms underlying structural priming by testing the bilingual lexical-syntactic representation model (Bernolet, 2008; Schoonbaert et al., 2007) and the error-based implicit learning model (Chang et al., 2006).
Research questions and hypotheses
To guide the investigation, the study addresses the following research questions (RQs) and corresponding hypotheses, covering both comprehension (Task 1) and production (Task 2):
RQ1:
Does cross-linguistic syntactic priming occur between English and Chinese for DO and PO structures in a translation selection task (comprehension)?
Hypothesis 1a: Exposure to English DO or PO structures will prime the selection of corresponding Chinese DO or PO structures in the translation selection task, as shared syntactic representations facilitate cross-linguistic priming (Loebell & Bock, 2003).
Hypothesis 1b: Priming effects will be stronger for PO structures than DO structures in comprehension task, given the higher frequency of PO structures in English (Bresnan et al., 2007), which may dominate L1-driven syntactic processing (Bao, 2017; Zhang & Lu, 2012).
RQ2:
How does Chinese proficiency modulate cross-linguistic syntactic priming in the translation selection task?
Hypothesis 2a: Advanced Chinese learners will demonstrate stronger cross-linguistic syntactic priming effects compared to beginner and intermediate learners. This is because higher proficiency is expected to enhance the shared syntactic representations between L1 and L2, facilitating priming across languages (Bernolet, 2008).
Hypothesis 2b: In contrast, beginner learners may exhibit weaker or no priming effects specifically for DO structures. This may be due to stronger influence from their L1 (English) syntactic preferences, where PO structures are more commonly used, potentially limiting DO priming at lower proficiency levels (Bresnan et al., 2007).
RQ3:
Does cross-linguistic syntactic priming occur between English and Chinese for DO and PO structures in a picture description task (production)?
Hypothesis 3a: English DO or PO priming sentences will increase the production of corresponding Chinese DO or PO structures in the picture description task, consistent with cross-linguistic syntactic persistence (Hartsuiker et al., 2004).
Hypothesis 3b: Participants will produce more DO structures than PO structures under priming conditions, as DO structures are syntactically simpler and align with the “principle of linguistic economy” (Givón, 1993).
RQ4:
How does Chinese proficiency influence cross-linguistic syntactic priming in the picture description task?
Hypothesis 4a: Advanced learners will show stronger priming effects than beginner and intermediate learners, reflecting greater L2 syntactic autonomy and shared representations (Wang, 2009).
Hypothesis 4b: Beginner learners are expected to produce more DO than PO structures in non-priming conditions, due to the complexity of PO structures (e.g., bǎ constructions).
Task 1: cross-linguistic syntactic priming in English-Chinese translation selection (comprehension)
This task investigates cross-linguistic syntactic priming in comprehension (RQ1) and the influence of Chinese proficiency (RQ2) among L1-English L2-Chinese learners, using a translation selection task. It employs a 2 × 3 mixed factorial design, with sentence structure (DO, PO) and Chinese proficiency (beginner, intermediate, advanced) as independent variables, and the number or percentage of DO and PO structures chosen as the dependent variable.
Methods
Participants
Twenty-seven native English-speaking international students (16 male, 11 female; mean age = 24.7 years) from the UK, USA, and Australia were recruited from Beijing Language and Culture University (BLCU), Tsinghua University, and China University of Geosciences (CUG). Participants had normal or corrected-to-normal vision and were right-handed. Their proficiency levels were assessed via the Hanyu Shuiping Kaoshi (HSK), a standardized test of Chinese language proficiency for non-native speakers, taken within the past six months, along with self-reported study duration and instructor evaluations. Participants were then evenly distributed across three levels: beginner (HSK 3–4, 6 months-1 year; n = 9), intermediate (HSK 5, 1–2 years; n = 9), and advanced (HSK 6, > 2 years; n = 9). Detailed participant information is in Appendix A.
Materials
The task used a questionnaire with 30 multiple-choice items, each presenting an English priming sentence (15 DO, 15 PO, 5–10 words in length) followed by two Chinese translations (DO and PO) differing in syntactic structure but similar in wording. Vocabulary was restricted to HSK Levels 甲 (A) and 乙 (B) in Hanyu Shuiping Cihui yu Hanzi Dengji Dagang (National Chinese Proficiency Test Committee, 2001) for accessibility, verified by professional Chinese teachers. A pilot study with 20 native Chinese speakers confirmed no structural bias (Table 1).
Table 1. Wilcoxon test results for pilot translation sentence selection task
Z-score | p-value | |
|---|---|---|
PO priming | − 0.175 | 0.861 |
DO priming | − 1.133 | 0.257 |
Procedure
Participants received task instructions in their preferred language (English or Chinese) and signed a consent form collecting demographic and proficiency data. To minimize vocabulary interference, participants familiarized themselves with Chinese characters before the formal task. They then completed the questionnaire, selecting the most appropriate Chinese translation for each English priming sentence.
Results
The 27 participants provided 810 valid responses. Prior to analysis, data were tested for normality (Shapiro–Wilk test, p > 0.05) and homogeneity of variance (Levene’s test, p < 0.05). Due to non-homogeneous variances, Welch’s t-test was used for between-group comparisons (e.g., across proficiency levels, df≈16), while paired samples t-tests were used for within-subject comparisons (e.g., DO vs. PO selection within conditions, df = 26).
Overall priming effects
Table 2 summarizes the distribution of DO and PO selections. Under DO priming, participants selected DO structures 28.02% (227/810) and PO structures 21.98% (178/810); under PO priming, PO structures were selected 30.99% (251/810) and DO structures 19.01% (154/810). Paired samples t-tests (n = 27, df = 26) compared DO and PO selections within each condition. Under DO priming, the difference was not significant, t(26) = 1.96, p = 0.061, d = 0.38, suggesting a weak priming effect. Under PO priming, PO selections significantly exceeded DO selections, t(26) = -5.14, p < 0.001, d = 0.99, indicating robust priming. The stronger PO priming effect likely reflects the higher frequency of PO structures in English (Bresnan et al., 2007), which aligns with Hypothesis 1b, suggesting L1-driven expectations influence L2 comprehension.
Table 2. Distribution of sentence structure selections by priming condition (unit: number of sentences)
DO selection | PO selection | |
|---|---|---|
DO priming | 227 (28.02%) | 178 (21.98%) |
PO priming | 154 (19.01%) | 251 (30.99%) |
Bold indicates the highest frequency value within each condition
Proficiency effects
Table 3 and Fig. 1 detail selections by proficiency level. Within-group comparisons used paired samples t-tests (n = 9, df = 8). Beginners showed no DO priming effect, selecting more PO structures (29.26%) than DO (20.74%), t(8) = 6.51, p < 0.001, d = 2.17, indicating L1 dominance. Under PO priming, the difference was not significant (26.67% PO vs. 23.33% DO), t(8) = 1.73, p = 0.121, d = 0.58. Intermediate learners showed a marginal DO priming effect (31.11% DO vs. 18.89% PO), t(8) = 2.07, p = 0.073, d = 0.69, and a strong PO priming effect (30.74% PO vs. 19.26% DO), t(8) = -5.32, p < 0.001, d = 1.77. Advanced learners exhibited significant priming for both: DO priming (32.22% DO vs. 17.78% PO), t(8) = 3.57, p = 0.007, d = 1.19; PO priming (35.56% PO vs. 14.44% DO), t(8) = -5.57, p < 0.001, d = 1.86. These findings suggest increasing syntactic priming with proficiency, aligning with Hypothesis 2a.
Table 3. Distribution of sentence structure selections by Chinese proficiency level and priming condition (unit: number of sentences)
DO selection | PO selection | ||
|---|---|---|---|
Beginner level | DO priming | 56 (20.74%) | 79 (29.26%) |
PO priming | 63 (23.33%) | 72 (26.67%) | |
Intermediate level | DO priming | 84 (31.11%) | 51 (18.89%) |
PO priming | 52 (19.26%) | 83 (30.74%) | |
Advanced level | DO priming | 87 (32.22%) | 48 (17.78%) |
PO priming | 39 (14.44%) | 96 (35.56%) |
Bold indicates the highest frequency value within each condition
[See PDF for image]
Fig. 1
Distribution of sentence structure selections by priming condition and Chinese proficiency level (unit: number of sentences). Note: Fig. 1 illustrates selection patterns across proficiency levels, showing stronger priming in advanced learners
Cross-proficiency comparisons
Between-group comparisons (Table 4) used independent samples t-tests (Welch’s adjustment, df≈16). Under DO priming, intermediate learners selected more DO structures (31.11%) than beginners (20.74%), t(16) = 3.57, p = 0.002, d = 1.78; advanced learners (32.22%) outperformed beginners, t(16) = 7.26, p < 0.001, d = 3.63, but the difference between advanced and intermediate was not significant, t(16) = 0.27, p = 0.788, d = 0.14. Under PO priming, PO selection was significantly higher among intermediate (30.74%) than beginner learners (26.67%), t(16) = 2.35, p = 0.032, d = 1.17; advanced learners (35.56%) outperformed beginners, t(16) = 3.34, p = 0.004, d = 1.67, but the difference between advanced and intermediate was not significant, t(16) = 1.90, p = 0.075, d = 0.95. The linear increase in priming effects with proficiency supports Hypothesis 2a, though beginners’ PO preference reflects L1 reliance (Hypothesis 2b).
Table 4. Distribution of sentence selections by priming condition and Chinese proficiency level (unit: number of sentences)
DO selection | PO selection | ||
|---|---|---|---|
DO priming | Beginner level | 56 (20.74%) | 79 (29.26%) |
Intermediate level | 84 (31.11%) | 51 (18.89%) | |
Advanced level | 87 (32.22%) | 48 (17.78%) | |
PO priming | Beginner level | 63 (23.33%) | 72 (26.67%) |
Intermediate level | 52 (19.26%) | 83 (30.74%) | |
Advanced level | 39 (14.44%) | 96 (35.56%) |
Bold indicates the highest frequency value within each condition
Task 2: cross-linguistic syntactic priming in English-Chinese picture description (production)
This task investigates cross-linguistic syntactic priming in sentence production among L1-English L2-Chinese learners, using a picture description task. It employs a 2 × 3 factorial mixed design, with sentence structure (DO, PO) and Chinese proficiency (beginner, intermediate, advanced) as independent variables, and the number or percentage of DO and PO structures produced as the dependent variable, consistent with Task 1. The task consists of two subtasks: Subtask 1 (non-priming, describing images without priming sentences) and Subtask 2 (priming, describing images after viewing English priming sentences). The task addresses research questions RQ3 (priming in production), and RQ4 (proficiency effects), testing Hypotheses 3a, 3b, 4a, and 4b.
Methods
Participants
Thirty native English-speaking international students (16 male, 14 female; mean age = 21.7 years) were recruited from the same institutions as in Task 1. To minimize potential bias arising from task familiarity, a different cohort of participants was recruited for Task 2. Participants were from the UK, USA, Canada, and Australia, with normal or corrected-to-normal vision and right-handedness. Participants were evenly divided into non-priming (n = 15) and priming (n = 15) groups, with 5 participants per proficiency level in each group. Proficiency levels were defined based on HSK certification, self-reported study duration, and instructor evaluations: beginner (HSK 3–4, 6 months-1 year), intermediate (HSK 5, 1–2 years), and advanced (HSK 6, > 2 years). Detailed participant information is summarized in Appendix B.
Instruments and materials
Both subtasks were conducted using E-Prime software on a Dell mobile PC with a 14-inch display, ensuring precise timing of stimuli presentation. The task materials were displayed on a screen with a resolution of 1366 × 768 pixels, with a 65 cm distance between the screen and participants. The materials were presented in black text on a white background, using 34-point Times New Roman font for English text and Song font for Chinese characters. The primes and images were shown on the screen in randomized order according to the task design. Participants completed a series of practice trials to become familiar with the task, and once they felt comfortable, they were able to begin the formal trials. While E-Prime was used to manage presentation and timing, no reaction time or error rate data were recorded in this study.
Subtask 1 (Non-priming condition) consisted of 20 images depicting everyday actions suitable for either DO or PO sentence constructions (e.g., giving, showing, handing). Participants were instructed to describe each image in natural Chinese, guided by Chinese lexical prompts presented below each image. These prompts were selected from HSK Levels 甲 (A) and 乙 (B), as outlined in Hanyu Shuiping Cihui yu Hanzi Dengji Dagang (National Chinese Proficiency Test Committee, 2001), and were verified by professional Chinese language instructors for appropriateness and accessibility. Each participant completed 20 trials and produced 20 sentences in total. A pilot study with 10 native Chinese speakers confirmed that the images elicited DO and PO structures with equal probability in the absence of priming, ruling out structural bias in the visual stimuli.
Subtask 2 (Priming condition) used the same 20 images, each preceded by an English priming sentence—10 of DO-type (e.g., Mom hands me a bag) and 10 of PO-type (e.g., Mom hands a bag to me). Each priming sentence contained 5–10 words in length and used vocabulary distinct from the image prompts to prevent lexical repetition. Each priming sentence was uniquely and consistently paired with a specific image, but with no semantic connection, thus avoiding conceptual overlap. Following the priming sentence (presented for 3 s), participants viewed the corresponding image and produced a sentence in Chinese. Each participant produced 20 sentences.
Procedure
Participants received task instructions in their preferred language and provided informed consent along with demographic and language background information. Before starting the task, they familiarized themselves with the prompts to avoid issues such as incoherent or incomplete sentences due to unfamiliar vocabulary. The task began with a welcome screen outlining the procedure, followed by a brief practice phase. Participants proceeded to the formal task by pressing a designated key once they were comfortable with the process.
Data analysis indicators
Produced sentences were categorized into three types: DO structures, PO structures, and “other” (e.g., subject-predicate sentences, serial verb constructions, incomplete sentences, or phrases). If a sentence deviated from standard Chinese syntax but remained intelligible, such as 奶奶讲给我们一个故事 (Nǎinai jiǎnggĕi wǒmen yígè gùshì, Grandma tells us a story), it was considered a correct sentence and categorized as a DO structure sentence based on the “understandability principle.” This principle allows for the inclusion of syntactically deviant but semantically clear constructions. Minor errors in word choice, tense, or word order did not affect categorization.
Results
The 30 participants produced 600 valid sentences across Subtask 1 (non-priming, n = 15) and Subtask 2 (priming, n = 15). Results are presented in three parts: (1) priming vs. non-priming conditions, (2) proficiency effects, and (3) cross-proficiency comparisons, addressing RQ3 (priming in production), RQ4 (proficiency effects). Prior to analysis, data were tested for normality using the Shapiro–Wilk test and for homogeneity of variance using Levene’s test. Results indicated that the data approximated normality (p > 0.05), but variance was not homogeneous across groups (p < 0.05). Consequently, Welch’s t-test was used for independent samples t-tests to account for unequal variances. Due to the task design, paired samples t-tests were used for within-subject comparisons in Subtask 2 (e.g., DO vs. PO priming within the same participants), while independent samples t-tests were used for between-subject comparisons (e.g., non-priming vs. priming, and cross-proficiency comparisons).
Priming vs. non-priming conditions
Table 5 summarizes sentence production across conditions. In the non-priming condition (Subtask 1), each participant produced 20 sentences, which were categorized into DO, PO, or Other structures. A paired samples t-test was used to compare the proportion of DO (17.33%) and PO (15%) structures within the same participants (n = 15, df = 14). The difference was not significant, t(14) = 1.81, p = 0.09, d = 0.47, indicating no inherent structural preference in the absence of priming, though a slight trend toward more DO structures was observed.
Table 5. Distribution of sentence production by priming condition (unit: number of sentences)
DO production | PO production | Other production | ||
|---|---|---|---|---|
Subtask 1: Non-priming | 52 (17.33%) | 45 (15%) | 203(67.67%) | |
Subtask 2: Priming | DO priming | 106 (35.33%) | 16 (5.33%) | 28 (9.33%) |
PO priming | 22 (7.33%) | 98 (32.67%) | 30 (10%) | |
Bold indicates the highest frequency value within each condition
Other production includes subject-predicate, incomplete, or non-DO/PO structures
To assess the overall effect of priming, independent samples t-tests (with Welch’s adjustment for unequal variances) were used to compare Subtask 1 (non-priming) and Subtask 2 (priming), which involved different participant groups (n = 15 per subtask, df≈28). Under DO priming, participants produced significantly more DO structures (35.33%) than in the non-priming condition (17.33%), t(28) = 9.84, p < 0.001, d = 3.72. Similarly, under PO priming, PO structures (32.67%) were produced more frequently than in the non-priming condition (15%), t(28) = 9.66, p < 0.001, d = 3.65, demonstrating a robust priming effect.
Within Subtask 2, paired samples t-tests were conducted to compare the production of primed structures (n = 15, df = 14). Under DO priming, participants produced significantly more DO structures (35.33%) than PO structures (5.33%), t(14) = 10.70, p < 0.001, d = 2.76, reflecting a strong priming effect. Similarly, under PO priming, PO structures (32.67%) were produced more frequently than DO structures (7.33%), t(14) = 7.46, p < 0.001, d = 1.93, also indicating a robust priming effect. These findings support Hypothesis 3a by confirming cross-linguistic syntactic priming in production. The stronger DO priming effect (d = 2.76) compared to PO priming (d = 1.93) may be attributable to the syntactic simplicity of DO structures, as predicted by Hypothesis 3b. Table 5 also shows a significant reduction in “Other production” under priming conditions (9.3–10% vs. 67.67% in non-priming), highlighting priming’s role in enhancing target structure use (Hypothesis 3a).
Proficiency effects
Table 6 details production by proficiency level, addressing RQ4. In the non-priming condition (Subtask 1), all proficiency levels produced predominantly “Other” structures (63–74%). Paired samples t-tests within each level (n = 5, df = 4) showed no significant DO-PO differences (beginner: t(4) = 1.50, p = 0.208, d = 0.67; intermediate: t(4) = 0.00, p = 1.000, d = 0.00; advanced: t(4) = -0.38, p = 0.724, d = 0.17). Beginners produced more DO structures (18%) than PO structures (8%), partially aligning with Hypothesis 4b, which argues that beginners favor DO structures due to the complexity of PO structures (e.g., bǎ constructions). Intermediate and advanced learners showed balanced DO-PO production (intermediate: 17% DO vs. 17% PO; advanced: 17% DO vs. 20% PO), suggesting that increased proficiency enables learners to better handle the complexity of PO structures, even without priming.
Table 6. Distribution of sentence production by Chinese proficiency level and priming condition (unit: number of sentences)
DO production | PO production | Other production | |||
|---|---|---|---|---|---|
Beginner level | Non-priming | 18 (18%) | 8 (8%) | 74 (74%) | |
Priming | DO priming | 27 (54%) | 8 (16%) | 15 (30%) | |
PO priming | 11 (22%) | 23 (46%) | 16 (32%) | ||
Intermediate level | Non-priming | 17 (17%) | 17 (17%) | 66 (66%) | |
Priming | DO priming | 35 (70%) | 5 (10%) | 10 (20%) | |
PO priming | 8 (16%) | 32 (64%) | 10 (20%) | ||
Advanced level | Non-priming | 17 (17%) | 20 (20%) | 63 (63%) | |
Priming | DO priming | 44 (88%) | 3 (6%) | 3 (6%) | |
PO priming | 3 (6%) | 43 (86%) | 4 (8%) | ||
Bold indicates the highest frequency value within each condition
Other production includes non-DO/PO structures. Percentages reflect production within each condition
Then independent samples t-tests (Welch’s adjustment) were used to compare non-priming and priming conditions within each level (df≈8). For beginners, DO priming increased DO production to 54% from 18%, t(8) = 10.85, p < 0.001, d = 4.85, and PO priming increased PO production to 46% from 8%, t(8) = 9.00, p < 0.001, d = 4.02. Intermediate learners showed stronger effects, with DO priming increasing DO production to 70% from 17%, t(8) = 13.05, p < 0.001, d = 5.83, and PO priming increasing PO production to 64% from 17%, t(8) = 6.89, p < 0.001, d = 3.08. Advanced learners exhibited the strongest priming, with DO priming increasing DO production to 88% from 17%, t(8) = 24.35, p < 0.001, d = 10.89, and PO priming increasing PO production to 86% from 20%, t(8) = 9.24, p < 0.001, d = 4.13.
Within Subtask 2, paired samples t-tests were employed to examine within-subject differences in DO and PO structure production under each priming condition. As shown in Fig. 2, for beginners, under DO priming, DO structures (54%) significantly outnumbered PO structures (16%), t(4) = 10.16, p < 0.001, d = 4.54, reflecting a strong preference for primed structures. Similarly, under PO priming, PO structures (46%) significantly outnumbered DO structures (22%), t(4) = 9.80, p < 0.001, d = 4.38, indicating robust priming despite limited L2 proficiency. For intermediate learners, DO priming led to significantly more DO structures (70%) than PO structures (10%), t(4) = 9.49, p < 0.001, d = 4.24, and PO priming resulted in more PO structures (64%) than DO structures (16%), t(4) = 8.23, p < 0.001, d = 3.68, suggesting stronger priming with increased proficiency. For advanced learners, DO priming produced significantly more DO structures (88%) than PO structures (6%), t(4) = 14.06, p < 0.001, d = 6.29, and PO priming led to more PO structures (86%) than DO structures (6%), t(4) = 11.31, p < 0.001, d = 5.06, demonstrating near-ceiling priming effects.
[See PDF for image]
Fig. 2
Distribution of sentence production by Chinese proficiency level and priming condition (Unit: %). Note: Fig. 2 illustrates the proportion of DO, PO, and other structures produced across proficiency levels and conditions, highlighting stronger priming effects in advanced learners
These findings, as depicted in Fig. 2 and detailed in Table 5, strongly support Hypothesis 4a, demonstrating that priming effects intensify with increasing proficiency, with effect sizes escalating from beginner (d = 4.02–4.85) to advanced learners (d = 4.13–10.89). The significant priming effects across all levels—beginners showing a 36–38% increase in DO and PO production, intermediates a 47–53% increase, and advanced learners a 66–71% increase—emphasize this trend. Figure 2 visually reinforces these proficiency-driven differences, highlighting advanced learners’ near-ceiling priming performance (88% DO under DO priming, 86% PO under PO priming), which reflects enhanced L2 syntactic priming.
Cross-proficiency comparisons
Table 7 compares sentence production across proficiency levels under the same priming conditions in Subtask 2, addressing RQ4. Since this involves comparisons between different groups (beginner, intermediate, advanced), independent samples t-tests (Welch’s adjustment) were used (n = 5 per group, df≈8). Under DO priming, advanced learners produced significantly more DO structures (88%) than beginners (54%), t(8) = 10.76, p < 0.001, d = 7.61. The difference between intermediate learners (70%) and beginners was significant, t(8) = 5.70, p < 0.001, d = 4.03, while the difference between advanced and intermediate learners was also significant, t(8) = 5.11, p < 0.001, d = 3.61. Under PO priming, PO production increased significantly with proficiency: intermediate learners (64%) outperformed beginners (46%), t(8) = 5.54, p < 0.001, d = 3.92; advanced learners (86%) outperformed intermediates, t(8) = 6.65, p < 0.001, d = 4.70; and advanced learners significantly outperformed beginners, t(8) = 12.01, p < 0.001, d = 8.49. Figure 2 visually depicts these trends, showing advanced learners’ near-ceiling primed structure production (88% DO under DO priming, 86% PO under PO priming). The increasing effect sizes (d = 3.61–8.49) highlight the strengthening of priming effects with proficiency, supporting Hypothesis 4a (Wang, 2009). Table 7 also shows a decline in “Other production” with proficiency (beginner: 30–32%, intermediate: 20%, advanced: 6–8%), reflecting improved L2 syntactic accuracy.
Table 7. Distribution of sentence production by priming condition and Chinese proficiency level (unit: number of sentences)
DO production | PO production | Other production | ||
|---|---|---|---|---|
DO priming | Beginner level | 27 (54%) | 8 (16%) | 15 (30%) |
Intermediate level | 35 (70%) | 5 (10%) | 10 (20%) | |
Advanced level | 44 (88%) | 3 (6%) | 3 (6%) | |
PO priming | Beginner level | 11 (22%) | 23 (46%) | 16 (32%) |
Intermediate level | 8 (16%) | 32 (64%) | 10 (20%) | |
Advanced level | 3 (6%) | 43 (86%) | 4 (8%) |
Bold indicates the highest frequency value within each condition
Discussion
This study explored cross-linguistic syntactic priming of DO and PO structures among L1-English L2-Chinese learners, examining the role of proficiency across comprehension (Task 1) and production (Task 2) tasks. The findings contribute to understanding the cognitive mechanisms underlying syntactic processing and have implications for TCSL. In the following subsections, the results are interpreted in relation to the research questions, situated within the context of prior studies, and discussed in terms of their theoretical and pedagogical implications.
Cross-linguistic syntactic priming in comprehension and production (RQ1 and RQ3)
Participants showed consistent cross-linguistic syntactic priming effects in both comprehension and production tasks, with increased selection and production of target structures following exposure to syntactic primes. These results support Hypotheses 1a and 3a, indicating that prior exposure to L1 structures can facilitate corresponding L2 representations. Extending prior research (e.g., Hartsuiker et al., 2004; Zhang & Wang, 2012), the findings demonstrate that abstract syntactic representations are transferable across typologically distant languages and modalities, pointing to shared processing mechanisms in L2 learners.
These findings align with both the bilingual lexical-syntactic representation model (Schoonbaert et al., 2007) and the error-based implicit learning model (Chang et al., 2006), but more strongly support the latter. While the lexical-syntactic model accounts for cross-linguistic structural priming through the activation of shared syntactic representations across languages, it does not fully explain the real-time, input-sensitive nature of the observed priming effects. Specifically, participants exhibited a marked tendency to reproduce primed structures immediately after exposure, such as a significant increase in PO selections in comprehension (30.99% under PO priming vs. 19.01% under DO priming, t(26) = -5.14, p < 0.001), and increased DO productions in the production task (35.33% under DO priming vs. 5.33% under PO priming, t(14) = 10.70, p < 0.001). These patterns suggest that participants were not merely retrieving pre-existing representations, but were actively adjusting their syntactic expectations in response to recent input—an adaptive mechanism central to the error-based model.
According to this model, priming effects arise when a syntactic structure encountered in the input creates a mismatch with the learner’s existing expectations. This prediction error then triggers implicit learning, resulting in an increased likelihood of using the recently encountered structure. The dynamic, experience-driven adjustment observed across both comprehension and production tasks therefore aligns more closely with the prediction error-based learning mechanism than with a purely activation-based account. While the lexical-syntactic model effectively explains cross-linguistic alignment in terms of shared representations, it does not fully capture the adaptive processes reflected in these real-time syntactic updates. The consistency of these effects across modalities reinforces the idea of a shared syntactic system in bilinguals, while also suggesting that modality-specific factors may shape how prediction errors are generated and resolved.
Influence of Chinese proficiency on priming effects (RQ2 and RQ4)
Cross-linguistic syntactic priming varied significantly by proficiency level: advanced learners exhibited robust priming effects in both tasks, while beginners showed weaker or absent effects, particularly for less familiar structures such as DO. This supports Hypotheses 2a and 4a and highlights the modulatory role of linguistic experience in syntactic processing.
The modulation of cross-linguistic priming by Chinese proficiency highlights the role of linguistic experience in shaping syntactic processing in L2. Advanced learners’ stronger priming effects suggest more entrenched L2 syntactic representations that are increasingly dissociable from L1 influence. This aligns with models proposing that as L2 proficiency develops, learners rely less on surface-level translation strategies and more on abstract structural mappings shared across languages (e.g., Bernolet et al., 2007). In contrast, beginners’ reduced or absent priming, particularly for DO structures, may reflect a lack of fully established L2 representations, or persistent activation of L1-preferred structures (e.g., PO in English), which limits cross-linguistic priming.
Interestingly, this proficiency-based pattern contrasts with findings from studies such as Li (2006) and Xu (2014), which reported stronger priming among lower-proficiency learners. One possible explanation is that those studies involved more syntactically complex sentence types, encouraging reliance on direct translation strategies that inadvertently increased priming. This discrepancy highlights that priming strength is not solely determined by proficiency but is also shaped by structural complexity and task characteristics. Such variation reinforces the idea that structural priming is not merely a reflection of repetition or strategy use, but serves as a window into the developmental status of interlingual syntactic links. Furthermore, it highlights the interactive influence of linguistic experience, task demands, and structure-specific frequency in shaping the strength and direction of priming effects, thereby offering valuable insights into the evolving architecture of L2 syntactic representations.
Speculative observations on structural preferences across modalities
An exploratory observation from this study is the potential difference in syntactic structure preferences (DO vs. PO) between comprehension (Task 1) and production (Task 2) tasks across proficiency levels. In the comprehension task, participants showed a preference for PO structures across all proficiency levels, possibly because these structures align more closely with L1 (English) expectations, where PO structures are more frequent (Bresnan et al., 2007). This tendency might reflect learners’ uncertainty in L2 processing, particularly at lower proficiency levels, leading them to favor structures perceived as more complex and complete (Xu, 2014). In contrast, the production task revealed a dominance of DO structures, likely due to their syntactic simplicity and alignment with the “principle of linguistic economy” (Givón, 1993), especially among beginner learners. Advanced learners, however, displayed more balanced performance across both tasks, suggesting greater flexibility in adapting to task-specific demands as proficiency increases.
These observations suggest that structural preferences may be influenced by both modality-specific cognitive demands and learners’ proficiency, with comprehension tasks potentially amplifying L1-driven expectations, particularly among lower-proficiency learners who rely more on familiar L1 patterns, while production tasks favor simpler L2 structures such as DO, especially for beginners with limited syntactic control. However, due to the limited sample size, these patterns should be interpreted cautiously, and future research with larger samples is needed to confirm these modality-driven differences in syntactic preferences and their interaction with proficiency levels.
Theoretical implications
The findings provide preliminary support for Bernolet’s (2008) developmental model of lexical-syntactic representation, which argues that syntactic representations are initially language-specific but become increasingly shared between L1 and L2 with rising proficiency. Advanced learners’ robust priming effects (e.g., 88% DO production under DO priming) suggest strong L1–L2 priming, allowing for efficient priming of abstract syntactic frames like ditransitives (Schoonbaert et al., 2007). In contrast, the weak or absent priming observed among beginners, particularly for DO structures in Task 1, aligns with the model’s prediction that early-stage learners lack shared L1–L2 syntactic representations. At this stage, the syntactic systems of the two languages remain largely independent, limiting the extent to which cross-linguistic priming can occur, especially between typologically distant language pairs such as English and Chinese.
Complementing this view, the error-based implicit learning model (Chang et al., 2006) provides a dynamic account of how such representations may evolve. From this perspective, structural priming reflects prediction error: the discrepancy between expected and encountered syntactic forms drives adaptation. Beginners, whose PO expectations are grounded in English L1 preferences (Bao, 2017; Zhang & Lu, 2012), may experience fewer prediction errors when encountering PO primes, leading to limited adaptation. By contrast, advanced learners, being more attuned to L2-specific patterns, may register larger prediction errors when encountering dispreferred structures, thus facilitating greater learning and stronger priming effects. This model also explains modality differences: comprehension priming may rely on predictive processing, where learners anticipate structures based on primes (Xu & Zeng, 2024), while production priming involves output facilitation, where primed structures are more readily retrieved (Shin, 2010). Supporting this, van Dijk and Hopp (2025) found that prediction errors in German-English priming were modulated by verb biases, with L1-to-L2 priming favoring PO structures, similar to the current study’s findings. However, verb-specific constraints, such as non-alternating verbs (e.g., donate), may reduce priming magnitude, as noted by Şafak and Hopp (2025), suggesting a need to explore verb effects in future English-Chinese studies.
While both models explain key aspects of the data, they differ in explanatory emphasis. Bernolet (2008) focuses on the developmental emergence and consolidation of shared structures, whereas Chang et al. (2006) emphasize real-time updating driven by processing experience. The findings of the present study suggest that these two mechanisms likely operate in tandem. Early-stage L2 processing may rely heavily on L1-based representations, but with increasing exposure and prediction error, learners progressively build shared syntactic networks through implicit learning.
Furthermore, the results challenge the common assumption in shared representation models that typological similarity is a prerequisite for cross-linguistic priming (e.g., Hartsuiker et al., 2004). Despite substantial structural differences between English and Chinese, learners showed clear priming effects, particularly for DO structures. This suggests that abstract syntactic convergence can occur even in typologically distant language pairs, thereby broadening the applicability of these models beyond closely related languages.
Pedagogical implications for TCSL
The findings offer actionable insights for TCSL, particularly when tailored to learners’ proficiency levels and task modalities. For beginners, who exhibit limited cross-linguistic priming and rely heavily on L1 structures, comprehension-focused activities can leverage L1-to-L2 priming to build initial L2 representations. For example, teachers might design a 20-min translation activity where students first encounter English PO primes (e.g., She gave a book to him) and then select the corresponding Chinese translation (e.g., 她把书给他, Tā bǎ shū gěi tā) from two options (DO and PO). This activity, supported by visual aids like sentence diagrams, can highlight structural differences, reducing negative L1 transfer. Contrastive teaching—comparing English and Chinese ditransitive structures—can further facilitate learning. For instance, teachers might use a bilingual handout to illustrate the optional use of bǎ in Chinese PO structures (e.g., 她把书给他 vs. 她给他书), emphasizing syntactic flexibility. To assess effectiveness, teachers could track students’ accuracy in translation tasks over a semester, monitoring reductions in L1-driven errors.
For advanced learners, production tasks can foster syntactic autonomy. A 30-min picture description activity, where students describe images (e.g., a person giving a gift) after exposure to varied DO and PO primes, can encourage alternation between structures, enhancing flexibility, as evidenced by their near-ceiling priming performance. These activities could be integrated into a communicative language teaching framework, using real-life scenarios (e.g., role-playing gift-giving situations) to promote naturalistic use. However, challenges remain. For example, beginners may struggle with complex PO structures, necessitating simplified input (e.g., shorter sentences, familiar vocabulary), while advanced learners may require more immersive contexts, such as group discussions in Chinese, to sustain progress. Digital tools, like language learning apps with priming-based exercises, could further support these strategies, offering scalable practice opportunities. Over the long term, such tailored approaches may enhance learners’ syntactic accuracy and fluency, though their effectiveness depends on consistent implementation and learner engagement, highlighting the need for teacher training in priming-informed pedagogical design.
Limitations and future directions
This study has several limitations that warrant consideration. First, the small sample size (27 in Task 1, 30 in Task 2) and limited item number precluded a power analysis, restricting the generalizability of the findings. A larger sample could enhance statistical power and enable more robust conclusions about priming effects across proficiency levels. Second, the focus on auditory-visual tasks neglected written modalities, which may yield different priming patterns. Incorporating written tasks, such as sentence completion (e.g., Desmet & Declercq, 2006; Pickering & Branigan, 1999; Salamoura & Williams, 2007; Wang, 2009) or written translation, could provide a more comprehensive understanding of cross-modal priming effects.
To address these limitations, future research could explore priming in diverse L1 backgrounds, such as Japanese or Korean learners, whose languages share typological features with Chinese (e.g., use of Chinese characters), potentially revealing how structural similarities influence priming. Additionally, investigating verb-specific effects, particularly with non-alternating verbs like 通知 (tōngzhī, inform), could clarify the role of structural constraints in priming, as highlighted by Şafak and Hopp (2025). Specific research questions to guide these efforts include: How do written priming tasks influence DO/PO preferences in L1-English L2-Chinese learners? How do verb biases modulate cross-linguistic priming across proficiency levels? Addressing these questions with larger, more diverse samples and varied task modalities will provide deeper insights into L2 syntactic priming.
Conclusion
This study provides preliminary evidence of cross-linguistic syntactic priming in L1-English L2-Chinese learners, demonstrating that higher proficiency enhances priming effects, and suggesting that task modality may influence structural preferences. The findings tentatively support the developed bilingual lexical-syntactic representation model and point to the role of error-based implicit learning mechanisms in L2 syntactic processing. By offering insights into the psychological mechanisms of priming and practical strategies for TCSL, this study contributes to the growing field of cross-linguistic syntactic priming research. Looking ahead, future investigations could broaden the theoretical scope of priming research by integrating cross-disciplinary approaches, such as combining psycholinguistic methods with computational modeling to simulate L1-L2 syntactic interactions. Additionally, exploring priming in diverse linguistic and cognitive contexts, such as heritage language learners or multilingual settings, using advanced experimental techniques, could further clarify the universal and language-specific aspects of syntactic processing. These directions will advance our understanding of L2 syntactic acquisition and inform the development of evidence-based, proficiency-sensitive teaching practices, ultimately benefiting language education in diverse second language learning contexts.
Acknowledgements
I would like to express my sincere gratitude to all the student participants for their invaluable contribution to this research. Their time and effort were essential to the success of this study. I am also deeply grateful to the reviewers for their insightful feedback and constructive suggestions. For the purpose of open access, the author has applied a Creative Commons Attribution (CC BY) licence to any Author Accepted Manuscript version arising from this submission.
Author contributions
M.L.: conceptualization; methodology; investigation; formal analysis; writing; editing; data curation; visualization; validation.
Funding
No funding was received.
Availability of data and materials
The study materials are accessible on the Open Science Framework at the link: https://osf.io/m4a35/. The datasets generated and analyzed during the current study are not publicly available, due to the privacy and confidentiality agreements with the participants, but can be obtained from the author upon reasonable request.
Declarations
Ethical approval
The research conducted in this study adhered to the Belmont Report.
Human ethics and consent to participate
The study was reviewed and approved by the Research Ethics Committee at Beijing Language and Culture University. All participants provided informed consent, and their rights and welfare were protected throughout the study.
Competing interests
The authors declare no competing interests.
Abbreviations
Beijing Language and Culture University
China University of Geosciences
Double-object
Hanyu Shuiping Kaoshi
First language
Second language
Prepositional-object
Second Language Acquisition
Teaching Chinese as a Second Language
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
Bao, J. H. (2017). Yingyu—Hanyu kuayu yufa qidong xiaoying shiyan yanjiu [An experimental study on cross-linguistic syntactic priming between English and Chinese] (Master’s thesis, Beijing Foreign Studies University).
Bernolet, S. (2008). Lexical-syntactic representations in bilingual sentence production (Doctoral dissertation, Ghent University).
Bock, JK. Syntactic persistence in language production. Cognitive Psychology; 1986; 18,
Bresnan, J., Cueni, A., Nikitina, T., & Baayen, R. H. (2007). Predicting the dative alternation. In Cognitive foundations of interpretation (pp. 69–94). KNAW.
Chang, F; Dell, GS; Bock, K. Becoming syntactic. Psychological Review; 2006; 113,
Chang, H. Muyu wei Yingyu he Fayu de xuexizhe dui Hanyu shuangbinju ji qi yuge zhuanhuan jiegou de xide yanjiu [A study on the acquisition of Chinese double-object structures and their dative alternations by English and French learners]. Yuyan Wenzi Yingyong; 2014; 2, pp. 96-106.
Desmet, T; Declercq, M. Cross-linguistic priming of syntactic hierarchical configuration information. Journal of Memory and Language; 2006; 54,
Dryer, MS. Primary objects, secondary objects, and antidative. Language; 1986; 62, pp. 808-845. [DOI: https://dx.doi.org/10.2307/415173]
Givón, T. English Grammar: A function-based introductionVolume I; 1993; Amsterdam, John Benjamins Publishing Company: [DOI: https://dx.doi.org/10.1075/z.engram1]
Hartsuiker, RJ; Pickering, MJ; Veltkamp, E. Is syntax separate or shared between languages? Cross-linguistic syntactic priming in Spanish-English bilinguals. Psychological Science; 2004; 15,
Heine, B; König, C. On the linear order of ditransitive objects. Language Sciences; 2010; 32,
Khoe, YH; Tsoukala, C; Kootstra, GJ; Frank, SL. Is structural priming between different languages a learning effect? Modelling priming as error-driven implicit learning. Language, Cognition and Neuroscience; 2021; 38,
Kootstra, GJ; Muysken, P. Cross-linguistic priming in bilinguals: Multidisciplinary perspectives on language processing, acquisition, and change. Bilingualism Language and Cognition; 2017; 20,
Lei, W; Wang, B. Impacts of working memory, L2 proficiency, and cognitive flexibility on cross-domain structural priming. Poznan Studies in Contemporary Linguistics; 2025; 61,
Li, M. Xiandai Hanyu shuangbinju de zai renshi [A re-interpretation of double-object structures in modern Chinese]. Yuyan Jiaoxue Yu Yanjiu; 2007; 6, pp. 38-44.
Li, RB. Kua yuyan jufa qidong ji qi jizhi [Cross-linguistic syntactic priming and its mechanism]. Xiandai Waiyu; 2006; 29,
Loebell, H; Bock, K. Structural priming across languages. Linguistics; 2003; 41,
McDonough, K; Trofimovich, P. Using Priming Methods in Second Language Research; 2011; Routledge: [DOI: https://dx.doi.org/10.4324/9780203880944]
National Chinese Proficiency Test Committee. (2001). Hanyu shuiping cihui yu Hanzi dengji dagang (xiuding ben) [Chinese proficiency vocabulary and character level outline (Revised ed.)]. Economic Science Press. (Original work published in Chinese)
Pickering, MJ; Branigan, HP. Syntactic priming in language production. Trends in Cognitive Sciences; 1999; 3,
Şafak, DF; Hopp, H. Learning L2 grammar from prediction errors? Verb biases in structural priming in comprehension and production. Bilingualism: Language and Cognition; 2025; [DOI: https://dx.doi.org/10.1017/S1366728925000033]
Salamoura, A; Williams, JN. Processing verb argument structure across languages: Evidence for shared representations in the bilingual lexicon. Applied Psycholinguistics; 2007; 28,
Schoonbaert, S; Hartsuiker, RJ; Pickering, MJ. The representation of lexical and syntactic information in bilinguals: Evidence from syntactic priming. Journal of Memory and Language; 2007; 56,
Shin, J-A. Structural priming and L2 proficiency effects on bilingual syntactic processing in production. Korean Journal of English Language and Linguistics; 2010; 10–3, pp. 499-518. [DOI: https://dx.doi.org/10.15738/kjell.10.3.201009.499]
Thothathiri, M; Snedeker, J. Give and take: Syntactic priming during spoken language comprehension. Cognition; 2008; 108,
van Dijk, C; Hopp, H. Structural preferences in language learning across languages: evidence from verb bias effects in cross-linguistic L1-to-L2 and L2-to-L1 priming. Language, Cognition and Neuroscience,; 2025; [DOI: https://dx.doi.org/10.1080/23273798.2025.2476703]
Wang, R., & Liu, Y. (2018, July). Cross-linguistic Study on Visual Syntactic Priming of Imbalanced Chinese-English Bilinguals. In 4th International Conference on Arts, Design and Contemporary Education (ICADCE 2018) (pp. 414–418). Atlantis Press. https://doi.org/10.2991/icadce-18.2018.87
Wang, H; Ding, X. Xiandai Hanyu shuangbinju fenlei yanjiu zongshu [A review of the classification of Chinese double-object structures]. Liaodong Xueyuan Xuebao: Shehui Kexue Ban; 2014; 16,
Wang, M. Yuyan shuiping ji renwu leixing dui dier yuyan chanchu zhong jiegou qidong de yingxiang [The influence of language proficiency and task type on structural priming in L2 production]. Xiandai Waiyu; 2009; 3, pp. 276-286.
Xu, H. Shuangyu gongzuo jiyi he eryu shuiping dui kua yuyan jufa qidong xiaoying de yingxiang [The effect of bilingual working memory and L2 proficiency on cross-linguistic syntactic priming]. Waiyu Jiaoxue Yu Yanjiu; 2014; 46,
Xu, K; Zeng, T. Cross-linguistic syntactic priming as rational expectation for syntactic repetition in the bilingual environment. PLoS ONE; 2024; 19,
Yang, J; Zhang, YX. Juzi chansheng zhong de jufa qidong [Syntactic priming in sentence production]. Xinli Kexue Jinzhan (Advances in Psychological Science); 2007; 15,
Ye, T. (2024). Shared Syntax or Separate: Cross-linguistic Structural Priming with Non-canonical Sentences on Chinese-English Bilinguals. Master’s thesis, Georgetown University.
Zhang, JJ; Lu, SM. Han-Ying shuangyuzhe yan yu chansheng zhong de jufa qidong xiaoying [Syntactic priming in speech production of Chinese-English bilinguals]. Xi’an Waiyu Xueyuan Xuebao; 2012; 20,
Zhang, JQ; Wang, Y. Yinni liuxuesheng Hanyu juzi chansheng zhong de kua yuyan jufa qidong [Cross-linguistic syntactic priming in Chinese sentence production by Indonesian students]. Xinli Yu Xingwei Yanjiu; 2012; 10,
Zhou, L. Z. (2009). Jiyu HSK dongtai zuowen yuliao ku de liuxuesheng shuangbin jiegou pianwu yanjiu [A study of double-object structures errors by international students based on a dynamic HSK writing corpus] (Master’s thesis, Beijing Language and Culture University).
© The Author(s) 2025. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.