<p> This dissertation uses multilevel models to test the veracity of two competing theories regarding the effect of Aboriginal language use on socioeconomic well-being. The cohesion hypothesis suggests that Aboriginal language use will contribute to a sense of ethnic identity and, in turn, to socioeconomic prosperity. The ghettoization hypothesis suggests that Aboriginal language use will reduce well-being by contributing to social and economic isolation. </p> <p> Descriptive statistics from the 2001 Census of Canada support the ghettoization hypothesis. Compared to Aboriginal people who do not use an Aboriginal language, Aboriginal language users have lower levels of educational attainment, income, labour force participation and employment. Multilevel models however, demonstrate that neither hypothesis merits unqualified support. Aboriginal language users are predicted to have lower well-being than non-speakers under some circumstances- most notably in non-Aboriginal communities. Under other circumstances, however, the opposite is true. Tests of the mechanisms by which Aboriginal language use is supposed to affect well-being also have inconsistent implications. Additional research is proposed that might clarify the apparently complex relationship between Aboriginal language use and well-being. </p> <p> Aboriginal language use in Canada is declining very rapidly. Of the dozens of Aboriginal languages used in Canada today, only a few are expected to survive into the next century. This dissertation may provide guidance to Aboriginal leaders tasked with allocating resources, as well as to politicians and policy-makers faced with increasingly urgent demands to support Aboriginal language maintenance. </p> / Thesis / Doctor of Philosophy (PhD)