On 17 June 2026, at the invitation of the School of Foreign Languages and Cultures at Nanjing Normal University, Prof. Robert J. Hartsuiker of Ghent University, Belgium, delivered a lecture entitled “Towards a Revised Developmental Account of Second Language Syntactic Representations.” The talk was chaired by Prof. Xu Xiaodong and drew an engaged audience of faculty and students.
Prof. Hartsuiker received his PhD from Radboud University Nijmegen in 1996 and now is the senior full professor of cognitive psychology at Ghent University. He serves as the research director of the Faculty of Psychology and Education at Ghent University, Belgium and is the former president of the European Society for Cognitive Psychology (ESCoP). His work spans bilingualism, L2 processing and learning, syntactic planning, speech errors and monitoring, and disfluency. He has authored more than 150 peer-reviewed articles, with close to 15,000 citations and an H-index of 60, making him one of the most influential scholars in psycholinguistics today.
The lecture comprised two parts. In Part 1, Prof. Hartsuiker proposed a developmental account of L2 syntactic representation. He introduced structural priming as a powerful tool for tapping into learners' syntactic representations. Revisiting Bock's (1986) classic studies, he explained the basic priming effect and the lexical boost effect. He then turned to cross-linguistic priming and the shared syntax hypothesis for bilinguals.
Addressing the central question of whether syntax is shared from the very start or becomes shared only with growing proficiency, he reviewed evidence showing that proficiency modulates priming: between L1 and L2, higher proficiency yields stronger priming, whereas within the L2, higher proficiency yields weaker priming. This motivates a developmental model of bilingual shared syntax that runs from transfer/copying, through easy-structure-first and item-specific stages, to within- and finally cross-language abstraction. To test the model, his team combined artificial-language learning for experimental control with longitudinal studies of real classroom learners for ecological validity. The findings show that syntactic sharing arises remarkably fast but depends on structural complexity; that a similar, simpler structure can partly “block” the learning of a dissimilar, harder one; and that brief interventions with related primes can substantially boost acquisition. These results partly support the account while yielding surprises, pointing to the need for an updated model reconcilable with error-driven learning theories.
In Part 2, Prof. Hartsuiker presented a lexicalist view of syntactic representation through three recent studies. Study 1 showed that the lexical boost depends on repetition of the syntactic head (the verb) rather than on adjacency, supporting head-centered models. Study 2, on the German nominative-dative alternation, argued that priming can operate at both syntactic and conceptual levels, possibly unified within a hierarchy of argument structures. Study 3, on Dutch passives, he proposed that similar constructions can “contaminate” one another through horizontal links, reflecting a construction network perspective.
A lively discussion followed on experimental design and suggestions for future research. Broad in scope and rigorous in argument, the lecture showcased frontier research on L2 syntactic processing as well as a methodology that integrates artificial and natural language, production experiments and computational modeling—offering valuable inspiration for psycholinguistic and SLA research within our School.
(Written by Chi Zhang)