## ABSTRACT

Mathematics education research has considered language issues for bilingual mathematics learners (for a review see Moschkovich, 2010). Interest in codeswitching while doing mathematics generated many questions: Why do learners codeswitch while doing mathematics? How does codeswitching reflect mathematical reasoning? How does codeswitching, especially the different connotations of a word, impact mathematical thinking? These questions reflect an intuitive model of language as an individual activity and a simple relationship between language and mathematical thinking. Some of the assumptions underlying such questions include: (a) speakers have a reason (conscious or unconscious) for switching languages, (b) these reasons are purely cognitive, (c) bilingual speakers, two languages are separate systems so that word connotations arise only if a mathematics word is actually spoken, and (d) the relationship between language and thought is simple, mechanistic, and unidirectional so that a spoken word impacts or reflects thinking in a way that is easily accessible to a speaker or an analyst.