Humans are remarkably sensitive to the statistical structure of language. However, different mechanisms have been proposed to account for such statistical sensitivities. The present study compared adult learning of syntax and the ability of two models of statistical learning to simulate human performance: Simple Recurrent Networks, which learn by predictive computation, and PARSER, which learns chunks as a byproduct of general principles of associative learning and memory. In the first stage, a semiartificial language paradigm was used to gather human data. In the second stage, a simulation paradigm was then used to compare the patterns of performance of the SRN and PARSER. After the human adults and the computational models were trained on sentences from the semiartificial language with probabilistic syntax, their learning outcomes were compared. Neither model was able to fully reproduce the human data, which may indicate less robust statistical learning effects in adults; however, PARSER was able to simulate more of the adult learning data than the SRN, suggesting a possible role for chunk formation in early phases of adult learning of second language syntax.

A growing body of evidence demonstrates that individual differences in declarative memory may be an important predictor of second language (L2) abilities. However, the evidence comes from studies using different declarative memory tasks that vary in their reliance on verbal abilities and task demands, which preclude estimating the size of the relationship between declarative memory and L2 learning. To address these concerns, we examined the relationship between verbal and nonverbal declarative memory abilities within the same task while controlling for task demands and stimulus modality, to estimate the upper bound of the relationship between verbal and nonverbal declarative memory. Results indicate that when task demands and stimulus modality are controlled, verbal and nonverbal declarative memory abilities shared a medium-to-large amount of underlying variance. However, future studies should exercise caution in appraising associations between declarative memory abilities and L2 learning until a more precise understanding of the underlying mechanisms is achieved.

Semantic verbal fluency (SVF) tests are sensitive to the declines in semantic processes found in persons with mild cognitive impairment (MCI) and Alzheimer’s disease (AD). Though most settings use the total number of words generated as the primary index of performance on SVF tests, there is reason to believe that SVF interval scores (e.g., the number of words generated at four, 15-second intervals) or performance slopes derived from these interval scores (i.e., the change in rate of word generation across the four 15-second intervals) may be even more sensitive to cognitive changes in the early stages of MCI. The goal of the current review is to provide preliminary evidence on the use of interval scores in persons with MCI due to AD relative to healthy controls (HCs). Following PRISMA guidelines, this systematic review synthesizes existing literature across PubMed, PsychINFO, and Web of Science electronic databases that utilize SVF interval data in individuals who are on average 60+ with a clinical diagnosis of MCI in the context of AD. Database searches identified 494 articles. After the screening process, eighteen eligible articles (n = 1,081 participants) comparing persons with MCI to HCs were retained. The most common interval methods used were two 30-second intervals, four 15-second intervals, and six 10-second intervals, and the “animals” prompt was most frequently used. Results revealed worsening word generation across intervals and initial performance was consistently worse in persons with MCI due to AD. Only one study did not show added utility of interval scores over traditional total scores. These initial findings encourage further investigation into the potential benefits of using interval scores and/or performance slopes from SVF testing for clinical and research applications.