Resolving Fundamental Debates in Statistical Learning: The Power of Process Dissociation

Statistical learning (SL) is a fundamental unsupervised mechanism enabling the brain to extract environmental regularities and build predictive models. Despite its central role in language acquisition and skill development, the field remains divided by persistent theoretical inconsistencies. This manuscript argues that these impasses stem from a failure to account for a fundamental principle of cognitive psychology championed by Larry Jacoby: the absence of "process-pure" tasks. We highlight how the lack of process purity fuels three major, unresolved debates that have hindered scientific progress. First, a long-standing tension persists over the relationship between executive functions and statistical learning, specifically whether they operate through cooperative or competitive neurocognitive networks. Second, the developmental trajectory of statistical learning remains contested, with conflicting evidence on whether children are superior learners, weaker learners, or whether performance is largely age-invariant relative to adults. Third, a profound disagreement exists concerning the impact of "offline" states such as mind wandering, which are variously described as either detrimental or facilitatory to learning. By applying a Jacoby-inspired framework, we propose that disentangling auxiliary cognitive processes from raw probability detection is necessary to reconcile these conflicts and ensure the validity of research in implicit memory and predictive processing.