Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Individual differences in the effects of offline consolidation and variable practice on extraction of linguistic regularities
Poster Session A, Wednesday, September 30, 11:00 am - 1:00 pm, Wangari Maathai
Tali Bitan1,2, Rabab Fadul1, Eva Kimel1,3; 1University of Haifa, Israel, 2University of Toronto, Canada, 3University of York, UK
Language learning involves both arbitrary, item-specific knowledge, and systematic regularities that can be generalized. These two types of knowledge, suggested to rely on distinct neurocognitive systems, may evolve in different temporal trajectories. Memory consolidation processes, by which new memories are transformed into qualitatively different representations, occur offline, after the end of training, during both sleep or wake. During consolidation, new memories can be integrated into existing networks, and shared aspects of memories may be extracted, leading to better generalization. While there is ample evidence for the effect of consolidation time, including sleep, on learning arbitrary, item-specific, knowledge, the evidence for its effects on the extraction of grammatical regularities is mixed. We investigated the effect of time for consolidation- time difference between exposure and test, and the benefit of training variability, on extraction of linguistic regularities. Variable training enhances generalization across domains, including in grammar learning. Since both, training variability and time for consolidation, improve generalization by extracting shared aspects of memories, we hypothesized that increasing the time interval between training and test would enhance the effect of stimulus variability. We also tested whether individual differences in phonological abilities and working memory modulate these effects in learning items and regularities. To test these questions, we used an artificial language, in which participants learned new vocabulary words and practiced making plural inflections, by adding one of three suffixes to the stem, determined by the final syllable on the singular form. These regularities were not taught explicitly. Words were aurally presented together with pictures of familiar objects. Participants were then tested on their vocabulary knowledge, and their ability to generalize inflection regularities to untrained words. Our participants were 122 native Arabic speaking adults, randomly assigned into four groups. We manipulated time for consolidation between-subjects, by conducting testing either immediately or 24 hours post-training. Stimulus variability was manipulated by including a large (42) vs. small (21) number of words, while keeping a constant total number of practice trials and rules. We also tested participants’ working memory and phonological abilities in their L1 to test their effect on regularity extraction, and their interaction with variability and time for consolidation. Results showed that phonological abilities modulated effects on extraction of regularities: individuals with high phonological abilities benefited from high variability for generalization and performed better immediately after training. In contrast, individuals with low phonological abilities did not benefit from variable exemplars and required extra consolidation time to achieve the same level of generalization. On the contrary, vocabulary learning was negatively affected by high exemplar variability and was associated with working memory, indicating separable mechanisms for item-specific and regularity extraction in language learning. In addition to their practical implications for teaching these results suggest that individual differences may explain the mixed evidence for the role of offline consolidation in extraction of linguistic regularities.
Topic Areas: Multilingualism, Morphology