Poster C5, Friday, August 17, 10:30 am – 12:15 pm, Room 2000AB

Does alphabetic orthography influence sound variation and change? Evidence from Hong Kong Cantonese

Yubin Zhang1, Kisa Sze Wai Chan1, Caicai Zhang1,2;1The Hong Kong Polytechnic University, 2Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences

Sound change is a complex process driven by many factors. In Hong Kong Cantonese (HKC), several sound pairs, e.g., l-/n-, g-/gw-, k-/kw-, ng-/0- (zero initial), ng/m, -n/-ng, -t/-k, high-rising/low-rising tone, mid-level/low-level tone, low-falling/low-level tone, are in the process of variation and merging. In this study, we aim to examine the influence of learning an alphabetic script on sound variation and change in HKC. The official writing system in Hong Kong is Chinese, a logographic script, which maps signs (i.e., Chinese characters) onto spoken words at the syllabic level, without revealing fine-grained phonemic information. Nonetheless, some Cantonese speakers have mastered Jyutping, a Romanization system of Cantonese, in adulthood. We attempt to test the hypothesis that Jyutping expertise is associated with more accurate production and perception of potential sound mergers, possibly via improved phonemic awareness and representations. Twelve Cantonese speakers with Jyutping proficiency and 12 matched controls were recruited. Their production and perception of potential mergers were tested by an elicited production task and a discrimination task. In the production task, they were required to read aloud monosyllabic Chinese characters, representing minimal pairs that contrast the merging sounds, within a carrier sentence. The accuracy and goodness of their production were assessed by two independent phoneticians. In the discrimination task, they had to judge whether the two words in a minimal pair were the same or not. Their reaction time and sensitivity index (d’) were analyzed. We found that the Jyutping group achieved significantly higher accuracy in their production of g-, ng-, and n- than the control group, whereas they made more errors for 0-, presumably due to their overcorrection of 0- to ng-. The Jyutping participants obtained higher rating scores in their production of g-, gw-, l-, n-, ng, and -ng than the controls, while their production of 0- again received lower scores. As for the discrimination, there was a non-significant trend that the Jyutping group obtained higher d’ than the control group for all the potential mergers. The results indicated that learning an alphabetic orthography could influence sound variation and change to some extent, consistent with previous findings that alphabetic orthography reorganizes and enhances spoken language processing. Interestingly, Jyutping expertise appeared to exert more effects on distinctions involving the presence or absence of a certain segment or a static feature, e.g., g-/gw-, 0-/ng-, and l-/n-. In contrast, its influence on sounds that utilize dynamic acoustic information, like formant transitions in -n/-ng and -t/-k, and suprasegmentals like tones, seemed to be limited. We argue that the alphabetic orthography may influence sound variation and change by providing general orthographic labels of speech categories, but it may benefit less phonological distinctions that rely on the richness of fine-grained acoustic details. These findings imply that the acoustic-orthographic-phonological interaction might need to be incorporated into models that deal with cognitive components of sound change (e.g. Ohala, 1990).

Topic Area: Phonology and Phonological Working Memory

Back