Emma Marsden

Filling a significant gap in SLA research in highly instructed settings, this book explores language proficiency in learners of French and Spanish in English schools. This detailed study investigates aspects of linguistic knowledge and proficiency, and the relationships between them, to offer insights into language learning, teaching and testing.

Tests of Aptitude for Language Learning (TALL) is a newly developed battery for L1 Chinese speakers that operationalizes language aptitude based on the early stages of Skehan’s (2016) Stages Approach through parallel aural and written suites. Each suite comprises subtests measuring associative memory, phonetic coding ability, language analytic ability, phonological short-term memory, and executive function of working memory. Using openly available data, this study examined TALL’s predictive validity through multiple regression analyses. Results from 165 L1-Chinese L2-English learners showed that the composite scores on the aural and written suites accounted for 12.5% and 13.1% of the variance in L2 proficiency, respectively. Models based on individual subtest scores accounted for 17.0% of the variance in L2 proficiency in the aural suite and 19.7% in the written suite. Dominance analyses identified language analytic and phonetic coding abilities as the strongest predictors, whereas other components contributed minimally. These findings provide initial evidence that TALL scores predict L2 proficiency. The study also offers a replicable procedure to facilitate further validation of this new battery. (Received 4 November 2025; revised 7 June 2026; accepted 6 August 2026) Keywords: aptitude battery; language aptitude; predictive validity

This study investigates whether classroom training on prediction can facilitate development of predictive processing using case information in second language (L2) German and increased accuracy in case production. We compared a group of intermediate-level German learners that received training ( N = 28) with another that did not ( N = 13). The training comprised explicit instruction on German subject and object case followed by structured input activities with subject–verb–object (SVO) and object–verb–subject (OVS) sentences, in which case-marking on the sentence-initial definite article ( der : nominative; den : accusative) provided a cue to its function (subject or object), and hence to the upcoming word order and sentence meaning. The effects of training were measured through a visual-world eye-tracking experiment, an aural comprehension task, and an oral production task. In the eye-tracking experiment, participants listened to SVO and OVS sentences while viewing two pictures differentiated by the reversal of the subject and object (e.g., a cook calling a waiter and a waiter calling a cook). If the case-marking on the first article is processed, this can facilitate looks towards the target picture even before hearing the second noun. Pretest results showed that the two groups were similar, with no evidence of case-driven eye movements, and low accuracy in comprehension and production of case in OVS sentences. In posttests, the training group made gains in all measures, which were maintained at a delayed posttest. The posttraining eye gaze data showed increased looks to the target picture during audio presentation of the second noun, though not before this point: thus, case-driven prediction was not established, but earlier processing of case-marking was. The comparison group made relatively smaller gains in comprehension and production, and showed no evidence of real-time processing of case. These findings suggest that targeted grammar comprehension training instigated a change in real-time processing and production behavior.

[This chapter explores the contribution made by the Universal Grammar approach (the generative linguistics approach) to our understanding of second language