Processing and retention of canonical (SO) vs. non-canonical (OS) sentences in L1 and L2 German

• Applicant: Professorin Dr. Denisa Bordag
• Subject Area: General and Comparative Linguistics, Experimental Linguistics, Typology, Non-European Languages
• Term: since 2026
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 581504405

Non-canonical sentence structures in German (e.g., Die Krankenschwester hat der Chefarzt gefragt.) have consistently been shown to pose processing difficulties and lead to systematic misinterpretation errors in both L1 and L2 speakers. Various linguistic and non-linguistic factors have been shown to influence the processing of these structures. While previous research has focused primarily on real-time comprehension, this project also examines how mental representations of canonical versus non-canonical sentences decay over time. Specifically, it investigates how sentence canonicity affects the processing, retention and decay of different types of information in L1 and two L2 groups (L1 Czech and Chinese/Vietnamese) that exhibit different degrees of overlap with German in word order flexibility and morphosyntactic templates. The two studies within this project will explore: (1) the origins and persistence of thematic role misassignments—whether they arise during processing or in post-interpretive processes and (2) the effects of morphosyntactic cues such as case marking. In addition to linguistic manipulations, we consider individual cognitive differences that may modulate processing and retention outcomes. By including measures of WM capacity, updating and interference control, we determine the extent to which challenges with non-canonical word orders, misinterpretation effects, and the apparent differences between L1 and L2 are influenced by differences in cognitive resources. Methodologically, the project builds on our previous research on the retention of content and linguistic surface information during reading in both L1 and L2. It integrates this line of research with results from prior studies on the processing of non-canonical word orders, particularly those in which the order of agent and patient is reversed in subject-object (SO) and object-subject (OS) sentences. By combining these perspectives, the project seeks to determine how different types of linguistic information are processed (and retained or lost) in canonical (SO) and non-canonical (OS) structures, and how these processes differ depending on L1 background and cognitive capacities. To address these questions, the project employs an innovative eye-tracking paradigm developed and established in our previous studies, enabling the implicit testing of information processing and retention through repeated sentence exposure in reading experiments. The project adopts a multifactorial interaction framework, explicitly considering the contribution of multiple factors to L1 and interlanguage sentence comprehension (canonicity, the relative role of semantic role assignment versus word order, the degree and type of inflectional marking and resulting ambiguity, cross-linguistic influence, and cognitive constraints), with the focus on how these factors jointly shape language processing and retention across different language acquisition contexts.

• DFG Programme: Research Grants