PLENARY SPEAKERS

ROWENA
KASPROWICZ

(University of Reading, UK)

Young learners are thought to have a greater propensity for implicit learning than older learners do. However, opportunities for drawing on implicit mechanisms for learning L2 morphosyntax are limited in the minimal exposure context of primary school classrooms in England. Although the effectiveness of explicit instruction has been established for adolescent / adult learners, the extent to which explicit information and practice is useful for young L2 learners is poorly understood. Further, despite grammar being a core component of England’s primary foreign language (FL) curriculum, research-based resources to support this are severely lacking and grammar teaching is perceived to be a substantial challenge (Language Trends survey 2014/2015).

To address this need, the current study reports on the design and evaluation of a digital grammar learning game. The game utilised the input-based ‘Form-Meaning Mapping’ approach (i.e. task-essential attention to form-meaning connections) (VanPatten, 2015) to teach verb agreement for tense and number in L2 French. Previous research has established the effectiveness of ‘Form-Meaning Mapping’ for young FL learners for comprehension and production (Kasprowicz & Marsden 2017; Marsden & Chen 2011).

The web-based game was deployed in eight UK primary school classrooms with 150 beginner learners of L2 French (aged 8 to 11). Learning behaviour was tracked over three weeks (180 minutes of gameplay). 34,467 data-points were recorded and provided detailed information about global gameplay behaviours (trends in accuracy/response times) and individual learning trajectories (critical learning points). Global accuracy in the game was high, indicating learning of the grammatical patterns introduced. Nevertheless, more fine-grained analysis revealed variation in the speed and rate of learning between individual learners. Further, learners’ L1 metalinguistic knowledge and their inductive and deductive language analysis abilities had predictive power over performance.

Taken together, the findings demonstrate: the potential of digital games, informed by SLA theory, for facilitating explicit grammar learning among young learners; a role for individual differences in language analytic abilities in mediating learning; and, more generally, the potential of online game data in SLA research.