Prosodic focus marking in Canadian English Jiseung Kim, Anja Arnhold University of Alberta Jiseung1@ualberta.ca, arnhold@ualberta.ca Abstract The current study investigated how broad focus, narrow focus and given information are produced in Canadian English. Given previous findings that showed different varieties of English signal information structure differently, we hypothesized that the effects of focus on acoustic correlates involving duration, f0, and intensity would manifest differently in Canadian English than in Mainstream American English (MAE). Thirty- eight native speakers of Canadian English produced 24 short transitive sentences in different focus conditions: broad focus and narrow focus in different locations (Subject, Verb, Object). A total of 2,736 words were analyzed. While some acoustic correlates such as duration and maximum intensity replicated the same patterns as previous findings in MAE, mean intensity and f0 measures showed different patterns. These results suggest that speakers of Canadian English may employ a different set of acoustic correlates from speakers of MAE. The study sheds light on the role of dialect in the production of focus and givenness and expands our knowledge about the fine- grained details of the phonetic realization of prosodic focus marking in English. Index Terms: information structure, pitch accent, Canadian English 1. Introduction While it has been established that focus is prosodically marked in English through changes in f0, intensity, and duration ([12, 14, 16, 20, 26]), compared to our knowledge of focus realization in other languages there are relatively few studies that comprehensively examine different acoustic correlates associated with focus, using modern statistical analyses to characterize the phonetic details ([8]). In the most comprehensive study to date, [8] investigated whether differences between focus location (S, V, O), breadth (broad and narrow), and type (contrastive and noncontrastive) are systematically distinguished, and which acoustic features are associated with these distinctions in American English. They showed that narrow focus was marked by longer duration, higher f0 mean and maximum, and higher maximum intensity compared to given information. In addition, focus breadth distinction (broad vs. narrow object focus) was marked by longer duration, higher f0 mean and maximum, and higher intensity, whereas contrastiveness was marked by longer duration, higher intensity and lower f0 mean and maximum. There are additional inconsistent findings regarding the fine-grained details of f0 modulation in focus marking. Studies did not agree whether focused words are characterized by higher f0 than unfocused words ([3, 22]) or post-focal f0 drop ([12, 13]). Moreover, most previous studies examined a single focused word in the same location in an utterance without considering the different locations of the word in the utterance or the prosodic effects of the surrounding words ([19]). Some of the earlier studies have often built on findings from different languages, despite the fact that different languages mark focus differently ([9]). Recognizing these aspects, the current study is motivated to provide further insight into the mechanisms of prosodic focus marking in English. In addition, it is still unclear whether and how the phonetic details of prosodic focus marking systematically differ between Canadian English and MAE. There is substantial evidence showing that prosodic focus marking is realized differently in regional dialects ([15, 21]). [10] examined four pitch accent types of Southern Standard British English (SSBE) and showed that the relation between pitch accent types and information status is not aligned with that of MAE. [1] reported two perception experiments. In Experiment 1, the eye- movements of 42 native listeners of Canadian English were tracked while listening to the auditory stimuli in SSBE. Unlike the British English listeners who associated rising pitch accents with givenness in [10], the Canadian English listeners did not make the same association. The results of Experiment 2 further showed how focus marking in Canadian English and SSBE differs: Canadian English listeners did not accept rising accents in Canadian English speech regardless of whether the target word signaled newness or givenness, but they rated rising accent on given words highly when listening to SSBE speech, suggesting that prosodic focus marking in Canadian English differs from that in SSBE. Moreover, [18] found that duration and intensity are two primary markers of focus while f0 played a minor role in British and Irish English, contrary to the findings reported in [8]. On the other hand, [27] reported both intensity and f0 were consistently used to mark focus across varieties of South African English, while duration effects were found in some varieties. [2] (for Minnesotan and Southern Californian) and [11] (for Southern and Midland) provide cases where there are phonological or phonetic differences in focus marking between regional varieties of American English. A close examination of the phonetic realization of focus in Canadian English is necessary to shed light on the widespread assumption that Canadian English resembles American English in many aspects. The current study investigated how various acoustic correlates are systematically modulated to distinguish constituents in broad focus, narrow focus, and given constituents in Canadian English. The main goal is to characterize the phonetic details of prosodic focus-marking and to shed light on the role of regional dialect as a potential source of phonological and/or phonetic differences between prosodic focus-marking in Canadian English and MAE. Speech Prosody 2022 23-26 May 2022, Lisbon, Portugal 97 10.21437/SpeechProsody.2022-20