Young participants (8-16-year-old) and adults (18 years +) from three different language backgrounds (Yorùbá speakers in Nigeria, Yorùbá speakers in Canada, and English speakers in Canada), participated in four perception experiments (one AX Discrimination, two Identification, and one Lexical), and one production (imitation) experiment. These experiments investigate the effects of age and tone language context on participants' tone perception and production accuracy. The experiments further

more »

... stigate participants' perception and production of the three Yorùbá basic tones and the two complex tones in addition to the effects of attrition on the perception and production accuracy of the young Yorùbá speakers in Canada. The results reveal that globally, adults are more accurate than the young participants in tone perception and production. The results also reveal that while the young Yoruba speakers in Nigeria outperform the other young groups in all experiments, there is no significant difference between the Yorùbáspeaking and the English-speaking adults in tone discrimination and in the identification of the three basic tones. In the identification of the contour tones and in tone production, the Yorùbá speakers in Nigeria are more accurate than the Yorùbá speakers and the English speakers in Canada. The Yorùbá speakers in Canada are more accurate than the English speakers in Canada only in tone production. In the lexical recognition experiment, which involves only the Yorùbáspeaking groups, the Yorùbá speakers in Nigeria are more accurate than the Yorùbá speakers in Canada. The perception of the three basic tones reveals that while there is no significant difference in the identification accuracy between the High and the Low tones, these two tones are identified more accurately than the Mid tone. The results further reveal that the adult English speakers in Canada tend to identify the High tones with significantly more accuracy. Further analysis reveals High-Mid confusion in addition to the effects of the mid-range position of the iii Mid tone; participants often misperceived both the High and the Low tones as Mid tone. The production results highlight the importance of the salient acoustic cues of the Low-tone as the tone is produced more accurately than either the High or the Mid tone. While also revealing High-Mid tone confusion, the production results further reveal that both the High and the Low tones are often misrepresented as Mid tones while Low tones rarely attract misrepresentation of the other tones. In the identification of the contour tones, the contour High tones (in LH pattern) are identified more accurately than the contour Low (in HL pattern) with further observation that the English-speaking Canadian adults tend to identify the contour-High tones significantly more accurately than the contour Low. Similarly, in Production, participants imitate the contour-High tones more accurately than the contour-Low tones. The performance of the English-speaking Canadian adults concerning the basic High tone and the contour High tones lends some support to the PAM-S model. The basic High tone is possibly assimilated to the English primary stress, while the contour High tone is comparable to the English iambic metre. Finally, the demographic factors reveal that 'longer stay in Canada' by the young Yorùbá speakers in Canada significantly lowers the group's accurate performance in tone perception and production. Older age at testing contributes to improved tone perception but not production while being born outside Nigeria does not significantly affect participants' tone perception and production skills. iv Preface This thesis is an original work by Saliu Ayowumi Shittu. The research project, of which this thesis is a part, received research ethics approval from the assisted in suggesting and applying the appropriate statistical models used in analyzing the results for all the five experiments, discussion of the results of the experiments, chapter reviews, and manuscript editing. Dr. Benjamin V. Tucker assisted in the initial experimental design, and in suggesting and reviewing the statistical models for the experiments, statistical analyses of the results, revising the chapters, and editing the manuscripts. Dr. Evangelia Daskalaki, a member of my supervisory committee, assisted with the review of the thesis especially the sections related to bilingualism and heritage language.