How do young children develop structured mental representations for the vast number of words in any given language? How do bilinguals deal with the even more daunting task of organizing two lexicons? Our research attempts to provide insights into the understanding of these questions from computational and neurocognitive perspectives. In particular I will present DevLex, a computational model of the development of lexicon (Li et al., 2004, 2007). Of particular relevance to the issue of dynamic self-organization is the model’s ability to demonstrate how early learning impacts later development, that is, how early learning leads to dedicated cognitive and neural structures that affect the shape and outcome of later development (positively or adversely). In the case of monolingual acquisition, rapid vocabulary growth observed in early childhood (the so-called “vocabulary spurt”) is predated and prepared by the system’s building of a structural representation that sets up the basic organization of the lexicon. In the case of bilingual acquisition, the structural consolidation of the first-language lexicon adversely impacts the representation and retrieval of the second-language lexicon, resulting parasitic L2 structure due to reduced plasticity in the organization and restructuring process (Hernandez & Li, 2007). These findings point to the developmental dynamics in which mechanisms of learning interact with the timing and history of learning to determine developmental trajectories.
|Publication status||Submitted - 2008|
|Event||XVI International Conference on Infant Studies - Vancouver, Canada|
Duration: 27 Mar 2008 → 30 Mar 2008
|Conference||XVI International Conference on Infant Studies|
|Abbreviated title||XVI ICIS 2008|
|Period||27/03/08 → 30/03/08|