The primary object of this thesis is the analysis of responses to the phonological section of the Linguistic Survey of B.C.: Postal Questionnaire (PQ3) which was designed by James Poison and Dr. R. J. Gregg at the University of B.C. In analysing the responses to the questionnaire, major emphasis has been placed on the distribution of variants in the province as a whole, on the shifting distribution of variants from older to younger generation speakers and on the isolation of dialect areas within the province. With regard to the isolation of dialect areas, the overall area distribution is the focus of attention; discussion of regional age variations is of necessity brief since sample sizes in age/area categories are for the most part too small to allow a definitive analysis. In addition to the analysis of PQ3 data, correlations have been made with data from other B.C. dialect surveys at both provincial and regional levels. Responses to other surveys at the provincial level are for the most part in agreement with those of PQ3> however, for some items significant variations occur. Comparison of responses at the regional level is difficult due to small and/or disparate sample sizes. The analysis of PQ3 data and the comparison of the results of this survey with those of other B.C. surveys indicates the necessity of a more concentrated study of B.C. speech. To this end, some alternative approaches to the study of British Columbia dialect have been suggested. / Arts, Faculty of / Linguistics, Department of / Graduate

The idea that Canada consists of “two solitudes” (MacLennan, 1945), according to which the two dominant (English and French) linguistic groups live in separate worlds with little interaction or communication, has also received attention in sociolinguistic circles (e.g. Heller, 1999). This thesis examines this claim further, by comparing the content of English and French Canadian newspapers. More specifically, the thesis compares how English and French serve different purposes in three coexisting conceptualisations of national identity in Canada: Quebec national identity, English Canadian national identity, and pan-Canadian national identity. In each corresponding national identity discourse, the nation and its language(s) are imagined differently. With a corpus of 7.5 million words in English and 3.5 million words in French, the thesis employs corpus linguistics and discourse analysis tools to test the salience of these ideologies and discourses, as well as to compare and contrast findings across languages. Adopting the theoretical framework of language ideologies (e.g. Woolard, 1998; Milani and Johnson, 2008), it seeks to contextualise languages with regard to discourses of national identity. In other words, the thesis compares and contrasts language ideology findings within the three discourses examined. More specifically, three research questions are addressed: (1) How do the French and English Canadian media discursively represent languages and language issues in the news? (2) How do these representations differ? (3) How do the different representations relate to understandings of national identity in Canada? The findings indicate that French and English serve predominantly different purposes, thus helping to reinforce the image of a Canada comprising “two solitudes”.