EEG and MEG evidence of a predominant number code in bilinguals and its significance for developmental dyscalculia

Abstract

When talking about bilinguals it is evident that they manage with two codes for the language. However, bilinguals also manage two codes for the representation of numerical facts and, even balanced bilinguals, have a preference for one of these codes for arithmetic fact retrieval or accessing to core magnitude representations. This preferred code is usually the language in which early acquisition of math has taken place (LLmath). The current thesis is aimed to investigate these preferences in the codes for math in the most basic representations. Although balanced bilinguals switch between languages back and forth without any effort we hypothesize that they will behave as unbalanced bilinguals when switching between the codes for math having an asymmetric switch being more difficult to switch into the more dominant code (LLmath) but following the same neural mechanisms as language switches. This preference for the LLmath will be reflected in bilingual children with developmental dyscalculia showing only distance effects in the LLmath and no effects in the other language. Four experiments are conducted, Experiments 1 to 3 investigate the dominance patterns in the codes for math in balanced bilinguals and their neural networks using ERP, MEG and source analyses using overt and masked priming switch tasks. And Experiment 4 investigates the preferences in these codes in bilingual children and in a normal developing control group with dyscalculia using ERP and source analyses in a magnitude comparison task. Results show that indeed, there is an unbalance in the codes for math and that the switch mechanisms for the codes for math are similar to those of general language-switch mechanisms. Additionally, bilingual dyscalculic children show magnitude effects in the LLmath but not in the other language (as opposed to the controls who show effects in both languages).