Prediction error, processing elements, and the development of early linguistic skills: An integrated EDL-PRIMs model
Yang Ji · 2022
Statistical learning and cognitive processing perspectives independently acknowledge the crucial role of learning in early linguistic acquisition. In the current paper, simultaneous acquirement of lexicons and procedures is implemented by uniformly integrating prediction error-driven learning (EDL) within the primitive information processing element architecture (PRIMs), assuming a statistical learning perspective in learning both forms. The first proof-of-concept simulation study extends the statistical learning perspective in syntactic acquisition based on EDL-PRIMs procedural learning. The second simulation study demonstrates procedural and lexical learning from syntactic, lexical, and hybrid patterns and simulates corresponding skill development by modeling age-related efficiency enhancement. The simulation studies are related to empirical focusing preferences based on operation latencies through a resource availability perspective. The paper discusses the EDL-PRIMs approach and simulated results linking to empirical evidence and computational theories of early linguistic development.