Authors:
Bordag, Denisa, denisav@uni-leipzig.de, Universität Leipzig
Opitz, Andreas, andreas.opitz@uni-leipzig.de, Universität Leipzig
Berulava, Hans-Georg, hans-georg.berulava@uni-leipzig.de
Keywords: memory, reading, L1 L2 differences, content, surface linguistic form
Abstract:
In our eye-tracking study, we assessed cognitive differences in reading between L1 (German) and advanced L2 (with L1 Czech) readers by comparing which type of information they predominantly retain in memory. Previous research indicated that L2 readers maintain more surface linguistic information (SLI) in their mental text model compared to L1 readers. Our aim was to explore how this relates to the retention of content information (CI).
Participants (L1: 64, L2: 64) first read a short text containing critical sentences designed to test the retention of SLI and CI. Then, they viewed pictures depicting the critical CI sentences, which were either congruent with the text or incongruent (involving a factual deviation). Finally, participants read single sentences that either appeared exactly as in the text or were manipulated with respect to SLI: active sentences were changed to passive and vice versa, and attributes either preceded a noun or followed it as a clause and vice versa. Additionally, critical CI sentences were presented again—half after a congruent picture and half after an incongruent one. The whole design was repeated four times (with four different texts, with 802 to 930 words each). Thus, there were two main sentence manipulations (CI and SLI) and two main conditions: ‘identical’ (a sentence that was preceded by the identical attribute type or grammatical voice in the text) and ‘changed’ (a sentence that was preceded by a different voice or attribute position in the corresponding critical sentence in the text). The rationale was that retention effects—measured as increased fixations in the incongruent/changed condition compared to the congruent/identical condition—could occur only if participants retained the respective information from the text in their memory.
The results revealed that L1 readers showed a significant retention effect only for the CI manipulation, both during picture viewing and sentence reading. In contrast, L2 readers showed the retention effect only for the SLI manipulation. These findings highlight different cognitive processes employed by L1 and L2 readers in reading comprehension and memory retention, and underscore the need for tailored approaches in instruction and assessment.