The P* reading group will be meeting againthis Monday, 22 January at 15:30 in room 002 (and on Zoom, at the following link). This week, Xuanda will be presenting original research in a talk titled Laryngeal Specification on Fricatives: Evidence from Perceptual Bias. Here is an abstract:
Compared with conventional phonological models that use a single [±voice] feature to represent all laryngeal contrasts, the theory of laryngeal realism advocates for two monovalent features—[voice] and [spread glottis]—to better capture laryngeal contrasts, depending on language-specific evidence from feature activeness. For example, English/Mandarin are considered [spread glottis] languages (sonorant devoicing), while Japanese is considered as a [voice] language (Rendaku). While a substantial portion of empirical evidence pertains to stop consonants, fricatives, particularly sibilants, receive comparatively less attention. Sibilants are often produced with strong air flow, high amplitude and pitch. These salient acoustic cues present challenges to tap into their underlying representation. To mitigate the impact of acoustic saliency, we introduce additional cognitive load (CL) by incorporating an extra task when listeners perform a phoneme (/s/-/z/) identification task. The pilot findings reveal a tendency among English and Mandarin listeners to misidentify /z/, while Japanese listeners show the opposite pattern. These asymmetrical identification behaviors are argued to support realist representations of fricatives, thereby supporting the theory of laryngeal realism.