Statistical learning (SL) is typically defined as a domain-general mechanism by which cognitive systems extract regularities from sensory input to discover its underlying structure. As such, SL is taken to underlie a range of cognitive faculties, with a particularly important role in language acquisition and use. In this talk, Dr. Noam Siegelman will take an individual-differences approach to examine SL as a theoretical construct and its role in language learning and processing. First, he will review studies examining the correlations in performance across different SL tasks, and between SL tasks and linguistic outcomes. Based on the observed pattern of correlations, showing surprising degree of modality- and material-specificity, Dr. Siegelman will present a framework for understanding SL as a multi-faceted theoretical construct, where computations across modalities and domains vary in terms of the available information present to the learner. He will then proceed to present an empirical investigation of this framework focusing on SL in the visual modality. Dr. Siegelman will conclude by discussing the implications of this approach for understanding the role of sensitivity to statistical regularities in language, focusing on reading behavior across writing systems that differ in the information they present to readers.