有一场主题是咧讲台语个「轻声」：To spread or not to spread: Neutral tone syllables in Taiwanese Southern Min 有兴趣个侬，准讲时间有拄好，会当去听看觅仔，罔参考。 --- Tsing Hua Workshop on Prosodic and Metrical Phonology 2020 A workshop series in celebration of the 35th years of linguistics at Tsing Hua Temporospatial: June 24 (Wed), 2020; HSS Building, B305, National Tsing Hua University ALL ARE WELCOME! 2:00 p.m. ~ 2:40 p.m. Just how metrical is the autosegmental-metrical model? Carlos Gussenhoven (Raboud Univeristy Nijmegen/National Chiao Tung University) 2:40 ~ 3:30 On the existence of word stress in Chinese Feng-fan Hsieh (National Tsing Hua University) 3:30 ~ 3:50 Break 3:50 ~ 4:30 To spread or not to spread: Neutral tone syllables in Taiwanese Southern Min Cheng-yen Roger Liu (National Tsing Hua University) 4:30 ~ 5:10 What is a boundary tone? Kristine Yu (University of Massachusetts, Amherst/National Chiao Tung University) Abstracts: Just how metrical is the autosegmental-metrical model? (Carlos Gussenhoven) Metrical representations of sentences, whether grids, trees, or s-w annotations on the constituents in the prosodic hierarchy, have been claimed to account for (a) Iambic Reversal in English (cf fífteen mén , but fiftéen) and (b) pitch accent locations on words more generally. I argue that the descriptive goal in (a) has not been achieved, while that in (b) has never been tried. Both problems concern the distribution of pitch accents. These can be accounted for with the help of an abstract location marker for the association of tone, as advocated by Goldsmith (1977). Parsimonious descriptions are shown to account for a dozen generalizations governing pitch accent locations in Nubi, Persian and English. Of these, only English Iambic Reversal has a phonological description, the others being informed by the morphology, the syntax, or focus. In conclusion, the Autosegmental-Metrical model does not distinguish itself from the general model of autosegmental phonology. On the existence of word stress in Chinese (Feng-fan Hsieh) In this talk, I argue that there exists word stress in Mandarin, even though it is well-established that Chinese stress is not perceptually salient, nor is it instrumentally confirmed at all. The direct evidence comes from the existence of word-level footing in (disyllabic) coordinative compounds (e.g., dà-xiǎo ‘big-small: size’), which is in many ways reminiscent of binominals in English (e.g., bread and butter, trick or treat, etc.), whereby the word order is not reversable (e.g. ??butter and bread; also, ??xiǎo-dà ‘small-big’). More precisely, from a corpus of 8,267 coordinative compounds, it is found that the distributions of the following phonetic properties are significantly biased towards the initial syllable of coordinative compounds, namely that (a) onset aspiration, (b) vowel lowness, (c) postnuclear glides, (d) nasal codas, and (e) Tone 1 (high level tone: 55), suggesting that the disyllabic coordinative compounds do form a metrical constituent, or, a trochaic foot exhibiting a strong-weak stress contour. In addition, the so-called neutral tone syllables constitute yet another evidence from the existence of foot in Mandarin Chinese. Precisely, vowel reduction, lenition and so on result from the lexically-specified extrametricality on neutral tone syllables. In conclusion, stress is overtly realized in Mandarin Chinese, albeit in a meager manner. To spread or not to spread: Neutral tone syllables in Taiwanese Southern Min ( Cheng-yen Roger Liu) It has long been noted in the literature that there are two broad types of neutral tone in Taiwanese Southern Min (TSM). Some neutral tone syllables bear a (default) low tone (dubbed “non-tone spreading” neutral tone, NTS), whereas the overall surface tone specification is contextually dependent in still others (dubbed “tone-spreading” neutral tone, TS). Neutral tone syllables are in general found in function words (postverbal particles, sentence final particles, etc.) and, more importantly, both types of neutral tone, i.e. TS and NTS, are found in particles with similar/identical grammatical functions, indicating that TS and NTS may not be grammatically modulated. One of our major findings is that, details aside, the choice between NTS and TS is predominantly determined by the manner of interlude consonants. Precisely, obstruent consonants “interrupt” TS, while NTS is not compatible with a sonorant interlude. Consequently, the distinction between TS and NTS may be reduced to a “blocking effect” of interlude consonants, which is reminiscent of depressor consonants in Southern Bantu languages or Wu Chinese. Nevertheless , it is equally remarkable that neutral tone syllables may occur when some specific grammatical features are being realized in TSM. We argue that grammatical tones of this sort are better analyzed as a floating low tone, leading to NTS. Finally, possible consequences for a proper treatment of the Southern Min tone sandhi systems will also be discussed. What is a boundary tone? (Kristine Yu) Prosodic boundary tones exhibit the phonological property of being conditioned on the appearance of prosodic constituent edges and the phonetic behavior of being aligned to the edge of a prosodic constituent. They thus must be temporally located at constituent edges in the phonological grammar. Why does this happen? Many phonological analyses have effectively stipulated this by marking boundary tones with a diacritic, e.g., H%. We show a way to have the edge behavior of boundary tones fall out of where they come into the grammar, following Pierrehumbert and Beckman (1988)'s idea that boundary tones are associated to higher-level prosodic nodes. What is a boundary tone? A boundary tone is a tone that enters the derivation of a prosodic tree when a prosodic constituent is built. --

※ 发信站: 批踢踢实业坊(ptt.cc), 来自: 111.255.65.185 (台湾) ※ 文章网址: https://webptt.com/cn.aspx?n=bbs/TW-language/M.1592212559.A.0D6.html ※ 编辑: CCY0927 (111.255.65.185 台湾), 06/15/2020 18:27:40