Graduate Student, Linguistics
Teaching Associate, Research Assistant
About
Children are nearly universally successful at acquiring their native language. Adults, on the other hand, display great variance in their success at learning other languages. What accounts for this finding?
The classical view of language acquisition rests on notions of innate symbolic rules, which cannot be learned but must, instead, be hardwired in the brain. However, more recent research has repeatedly demonstrated the power of memory-based learning mechanisms in language learning (e.g., associative learning, statistical learning, implicit/explicit learning).
On this view, children are predicted to be more successful than adults at language acquisition because of general properties underlying all human learning mechanisms: memory, conscious awareness, attention, etc. My research is aimed at exploring this thought-provoking view.
In particular, I am most interested two problems:
(1) Clarifying the role of conscious awareness in adult second language learning (when is consciousness necessary for learning, if at all?). This question addresses SLA theories relating to the implicit/explicit interface (Ellis, 2005), the noticing hypothesis (Schmidt, 1990), and theories from cognitive science, such as the self-organizing consciousness (Perruchet & Vinter, 2002), higher-order thought theory (Rosenthal, 1997), and the radical plasticity thesis (Cleeremans, 2008).
(2) Clarifying the nature of the learning mechanisms that support language learning (is learning driven by principles of associative memory and/or probabilistic learning?). This question addresses SLA theories based on associative and probabilistic learning mechanisms (Ellis, 2005) and declarative and procedural memory systems (Ullman, 2004).
To test these ideas, I conduct behavioral learning experiments that investigate different aspects of L2 development. These experiments are based on well-established methods from implicit and statistical learning research in cognitive science. I then use computational models to simulate my behavioral experiments (PARSER and simple recurrent neural networks). I simulate human behavior with these models, often getting a good fit to the data. This, in turn, sheds light on the nature of the cognitive mechanisms of adult SLA.
I take the view that the better we understand the role of these mechanisms in SLA, the more we can design L2 teaching materials to exploit adult learners' biases and maximize their learning outcomes and make language teaching more efficient.
Contact Information
| Homepage: |
