Asymmetry in the retention of content and surface linguistic information during reading in L1 and L2

This eye-tracking study investigates how native (L1) and non-native (L2) German speakers retain content and surface linguistic information during reading, drawing on the Construction-Integration Model of text comprehension. Participants read narrative texts, followed by picture and sentence reading tasks designed to assess memory for content and surface linguistic forms (e.g., grammatical voice, attribute position). Results reveal an asymmetric retention pattern: L1 readers demonstrated stronger retention of content information, indicated by longer fixation times on semantically incongruent pictures and sentences. In contrast, L2 readers showed enhanced retention of surface linguistic forms, evidenced by extended fixations on sentences with altered surface structures. These findings align with the Shallow Structure Hypothesis and the Declarative/Procedural Model, suggesting that L2 readers rely more heavily on declarative memory for surface forms due to less automatized syntactic processing. By directly comparing L1 and L2 retention patterns, this study provides novel insights into the mental representation of text in L2 readers, highlighting an increased retention of surface information that is accompanied by reduced content retention.