| { | |
| "paper_id": "O07-1018", | |
| "header": { | |
| "generated_with": "S2ORC 1.0.0", | |
| "date_generated": "2023-01-19T08:07:55.743382Z" | |
| }, | |
| "title": "The Role of Sound Change in the Speech Recognition System: A Phonetic Analysis of the Final Nasal Shift in Mandarin", | |
| "authors": [ | |
| { | |
| "first": "James", | |
| "middle": [ | |
| "H" | |
| ], | |
| "last": "\u694a\u5b5d\u6148", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "", | |
| "institution": "National Yunlin University of Science and Technology", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "", | |
| "middle": [], | |
| "last": "Yang \u570b\uf9f7\u96f2\uf9f4\u79d1\u6280\u5927\u5b78\u61c9\u7528\u5916\u8a9e\u7cfb", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "", | |
| "institution": "National Yunlin University of Science and Technology", | |
| "location": {} | |
| }, | |
| "email": "" | |
| } | |
| ], | |
| "year": "", | |
| "venue": null, | |
| "identifiers": {}, | |
| "abstract": "Over the past decade, computational linguists have been striving to design a speech recognition system that is able to identify standard speeches and to accommodate sound variables caused by different individual accents. Furthermore, some speech recognition programs have been able to learn and identify distinctive sound frequencies due to the user's age and gender. Nevertheless, regular sound alterations that occur in different varieties of a language have never been seriously considered in the design of the speech recognition system. Accordingly, this study proposes to incorporate the socio-phonological information about regular sound modifications to enhance the performance of Automatic Speech Recognition. To illustrate this point, this study investigates and analyzes the acoustic variation of the syllable-final nasal shift from the velar to the dental, which has been discovered to be one of the distinctive sound features that makes the variety of Mandarin spoken in Taiwan differ from that spoken in China. Following the phonetic analysis, this study discusses the effect of the nasal merger on the development of phonology-dependent speech technologies. It concludes by proposing a preliminary resolution to the identification of syllable-final nasals for the design of Automatic Speech Recognition.", | |
| "pdf_parse": { | |
| "paper_id": "O07-1018", | |
| "_pdf_hash": "", | |
| "abstract": [ | |
| { | |
| "text": "Over the past decade, computational linguists have been striving to design a speech recognition system that is able to identify standard speeches and to accommodate sound variables caused by different individual accents. Furthermore, some speech recognition programs have been able to learn and identify distinctive sound frequencies due to the user's age and gender. Nevertheless, regular sound alterations that occur in different varieties of a language have never been seriously considered in the design of the speech recognition system. Accordingly, this study proposes to incorporate the socio-phonological information about regular sound modifications to enhance the performance of Automatic Speech Recognition. To illustrate this point, this study investigates and analyzes the acoustic variation of the syllable-final nasal shift from the velar to the dental, which has been discovered to be one of the distinctive sound features that makes the variety of Mandarin spoken in Taiwan differ from that spoken in China. Following the phonetic analysis, this study discusses the effect of the nasal merger on the development of phonology-dependent speech technologies. It concludes by proposing a preliminary resolution to the identification of syllable-final nasals for the design of Automatic Speech Recognition.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Abstract", | |
| "sec_num": null | |
| } | |
| ], | |
| "body_text": [ | |
| { | |
| "text": "My motivation to explore the nasal merger of Mandarin spoken in Taiwan originates from an incident in my life. My brother's first baby was born on August 31 st , 1999. He gave his son a name called Geng-ren /k\u0259\u014b.\u0290\u0259n/ 1 ( \u8015\u4ec1, meaning \"to cultivate benevolence\").", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1." | |
| }, | |
| { | |
| "text": "Interestingly, I found, as a native speaker of Mandarin in Taiwan, that I would easily mispronounce his name as Gen-ren [k\u0259n.\u0290\u0259n] (\u8ddf\u4eba, meaning \"to follow people\"), rather than its standard pronunciation.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1." | |
| }, | |
| { | |
| "text": "Since the name was subject to mispronunciation and misunderstanding, I suggested to my brother that he should change the name; hence, he later selected another name Jia-he (\u5bb6\u548c, meaning \"harmony in the family\"). Because of this interesting incident, I came to realize that many native speakers of Mandarin in Taiwan seem to merge the syllable-final velar nasal /\u014b/ with the dental nasal /n/, hence neutralizing such minimal pairs as geng /k\u0259\u014b/ (\u8015, \"to cultivate\") and gen /k\u0259n/ (\u8ddf, \"to follow\"). To explore this possible sound change, I later conducted a speech production experiment, which is discussed in the subsequent section.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1." | |
| }, | |
| { | |
| "text": "To investigate the possible nasal merger observed above, I addressed three research questions:", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Speech Production Experiment", | |
| "sec_num": "2." | |
| }, | |
| { | |
| "text": "1) Is the syllable-final nasal modification a free variation or a conditioned alteration?", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Speech Production Experiment", | |
| "sec_num": "2." | |
| }, | |
| { | |
| "text": "2) Does it occur in Mandarin spoken in Taiwan, China, or both?", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Speech Production Experiment", | |
| "sec_num": "2." | |
| }, | |
| { | |
| "text": "3) Is it an ongoing or complete sound change?", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Speech Production Experiment", | |
| "sec_num": "2." | |
| }, | |
| { | |
| "text": "To address these questions, I invited 30 native Mandarin speakers to participate in the speech production experiment, including 11 males and 19 females, who were students of the University of Hawaii at Manoa. Fifteen of them were from Taiwan, and another fifteen were from China. They were all young adults with the average age of 27, the eldest subject being 36 years old and the youngest one being 21.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Speech Production Experiment", | |
| "sec_num": "2." | |
| }, | |
| { | |
| "text": "For this experiment, I designed a questionnaire to understand the subjects' basic sociolinguistic backgrounds. In addition, I also created sixty easy and interesting riddles to elicit spontaneous speech data. The answers to the riddles included the test words needed for this study--that is, words which end with three types of rhymes: -ing, -eng, and -ang. ", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Speech Production Experiment", | |
| "sec_num": "2." | |
| }, | |
| { | |
| "text": "Regarding the first research question as to whether the final nasal shift from the velar to the dental (/\u014b/>/n/) occurs without syllabic constraints or appears only in certain environments, the results show that the nasal merger is not a free variation, but a conditioned sound change, which can be formulated by the following phonological rule:", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "(1) Nasal Fronting:", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "/\u014b/ [n]/{i, \u0259}_____.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "For example, the word \u7d93 jing (/t\uf0fei\u014b/, \"pass\") is regularly pronounced by the Taiwanese respondents as \uf90a jin (/t\uf0fein/, \"gold\") according to Rule 1. Furthermore, this change in the rhyme also causes lexical neutralization. For instance, the word jing-yu (\u9be8\u9b5a, whale) is recurrently pronounced by the Taiwanese respondents as jin-yu ( \uf90a \u9b5a , goldfish).", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "Consequently, this nasal merger leads to lexical neutralization, creating homophones sharing the nasal endings with different meanings.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "Notably, the nasal merger in Mandarin spoken in Taiwan is not only conditioned by the preceding vowel (Rule 1) but is also blocked by the bilabial onset, which is regularized by the following rule:", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "(2) Vowel Labialization:", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "/\u0259\u014b/ [o\u014b]/[labial] _____", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "To my knowledge, such discovery has not been specifically analyzed in any previous studies.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "For instance, the sound meng (/m\u0259\u014b/) is pronounced as mong (/mo\u014b/), rather than men (/m\u0259n/) As Figure 1 displays, the syllable-final velar nasal /\u014b/ in MT merges nearly completely with the dental nasal /n/ when preceded by /i/ or /\u0259/. By comparison, MC in general does not undergo the nasal shift.", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 95, | |
| "end": 103, | |
| "text": "Figure 1", | |
| "ref_id": "FIGREF0" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "according", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "Taken together, all of the Taiwanese respondents underwent the nasal fronting (Rule 1).", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "More than 95 percent of the time they displayed the nasal shift when the final nasal was preceded by the vowel either /i/ or /\uf0ab/. By comparison, Tse's 1992 survey suggests that 73% of his Taiwanese informants could not distinguish the syllable-final nasal minimal pairs. Accordingly, the final velar nasal merger with the dental has evolved into a nearly complete status in MT.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "However, the final nasal modification described above only appears sporadically in MC.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "Specifically, Rule 1 occurs only 20 percent of the time in the Chinese responses.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "Comparatively speaking, Rule 1 occurs more than 95 percent of the time in the Taiwanese responses, making MT differ significantly from MC (P<0.05). While some of the Chinese informants from southern China also consistently undergo the nasal shift , the findings do not allow this study to conclude that the final nasal modification occurs regularly in southern China. First, although three of the speakers from southern China recurrently displayed the nasal merger, a couple of them were able to pronounce the test words according to the standard pronunciations without changing the final velar nasal into the dental one.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "Additionally, the number of the speakers from southern China was very few; only five informants participated in this study. Accordingly, the nasal shift from the velar to the dental seems to present an ongoing phonological process of confusion and interchange in southern China. To confirm whether the nasal merger is common in southern China, an empirical and quantitative study is needed in the future.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "In summary, the results demonstrate that the nasal shift (Rule 1) is nearly complete in MT, leaving very few lexical \"residues,\" which are usually high frequency words, such as sheng (\u751f, life) and qing (\u6e05, clear), instead of shen (\u8eab, body) and qin (\u89aa, kin), respectively.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "Theoretically speaking, this nasal shift is consistent with Zee's prediction of the nasal shift from -ng to -n in Chinese dialects (1985) and is opposed to M. Chen's theory of unidirectionality of the nasal shift from -n to -ng (1972, 1973, 1975) . However, the syllable-final velar nasal remains unchanged when preceded by the non-palatal low vowel /a/.", | |
| "cite_spans": [ | |
| { | |
| "start": 131, | |
| "end": 137, | |
| "text": "(1985)", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 228, | |
| "end": 234, | |
| "text": "(1972,", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 235, | |
| "end": 240, | |
| "text": "1973,", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 241, | |
| "end": 246, | |
| "text": "1975)", | |
| "ref_id": null | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Findings", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "The perceived transcriptions of the nasal shift in question are also supported by acoustic analyses. First, the spectrogram of the word jing-yu (\u9be8\u9b5a, whale) pronounced by the Taiwanese Mandarin speaker shows that the phonological change is a syllable-final nasal shift, rather than a nasal deletion with the preceding vowel nasalized, as shown below: The darker parts of the sound reflect the spectrograms of the vowels, while the whiter parts signal those of the consonants (Ladefoged, 2003) . In addition, the sound analysis via Pratt (a computer program for sound analyses) demonstrates that the syllable-final nasal still remains as shown in the middle whiter part of Figure 2 .", | |
| "cite_spans": [ | |
| { | |
| "start": 474, | |
| "end": 491, | |
| "text": "(Ladefoged, 2003)", | |
| "ref_id": "BIBREF5" | |
| } | |
| ], | |
| "ref_spans": [ | |
| { | |
| "start": 671, | |
| "end": 679, | |
| "text": "Figure 2", | |
| "ref_id": "FIGREF1" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Phonetic Analyses", | |
| "sec_num": "4." | |
| }, | |
| { | |
| "text": "The speculation that the sound alteration in question is a nasal deletion with the preceding vowel nasalized can also be cleared by comparing the spectrograms of the two sounds ji\uf029 -yu and ji-yu, as shown below. Figure 3 (the spectrogram of the sound ji\uf029 -yu ) does not display any whiter part between the vowels /i/ and /y/, but only parallel lines equally black present before the vowel /y/. By contrast, Figure 2 exhibits a whiter part between the two vowels /i/ and /y/. Therefore, the nasal does not disappear but remains.", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 212, | |
| "end": 220, | |
| "text": "Figure 3", | |
| "ref_id": "FIGREF2" | |
| }, | |
| { | |
| "start": 407, | |
| "end": 415, | |
| "text": "Figure 2", | |
| "ref_id": "FIGREF1" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Phonetic Analyses", | |
| "sec_num": "4." | |
| }, | |
| { | |
| "text": "Furthermore, Figure 4 (the spectrogram of the sound ji-yu) also looks distinct from Figure 2 . Because the spectrogram of the word jing-yu (\u9be8\u9b5a, whale) pronounced by the look similar to that of ji-yu, the sound modification in question should not be the nasal deletion.", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 13, | |
| "end": 21, | |
| "text": "Figure 4", | |
| "ref_id": "FIGREF3" | |
| }, | |
| { | |
| "start": 84, | |
| "end": 92, | |
| "text": "Figure 2", | |
| "ref_id": "FIGREF1" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Phonetic Analyses", | |
| "sec_num": "4." | |
| }, | |
| { | |
| "text": "Nonetheless, the spectrogram of the word jin-yu (\uf90a\u9b5a, goldfish) pronounced by the same Taiwanese informant looks similar to Figure 2 , although their durations and intensities are slightly different, as presented below: To summarize, the spectrograms presented above indicate that the syllable-final nasal changes from the velar to the dental when the preceding vowel is /i/ or /\uf0ab/, instead of vanishing with the preceding vowel nasalized. shen-gao (\u8eab\u9ad8). Consequently, when a Taiwanese Mandarin speaker says jing-yu (\u9be8\u9b5a, whale) as jin-yu (\uf90a\u9b5a, goldfish), the system might misidentify the intended word as jin-yu (\uf90a\u9b5a, goldfish) rather than jing-yu (\u9be8\u9b5a, whale), hence retrieving the wrong word. To resolve word-identification problems like this, a possible approach might be to train the user to produce the standard pronunciations of the minimal pairs differing only in nasal endings.", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 123, | |
| "end": 131, | |
| "text": "Figure 2", | |
| "ref_id": "FIGREF1" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Phonetic Analyses", | |
| "sec_num": "4." | |
| }, | |
| { | |
| "text": "Specifically, the user might be trained to articulate the minimal pairs as the standard pronunciations so that the recognition system is able to retrieve the right words.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Implications for Automatic Speech Recognition", | |
| "sec_num": "5." | |
| }, | |
| { | |
| "text": "Such sound training, however, is neither easy nor realistic. While the speech recognition program may provide the standard pronunciations of the syllable-final nasal minimal pairs for the user to imitate and articulate, there is evidence that most people tend to mishear new sounds and mispronounce them according to their habitual articulation (Ohala, 1992 (Ohala, , 2001 ). Furthermore, some recent research has demonstrated that Taiwanese Mandarin speakers tend to preserve their distinctive and unique sound features as their Taiwanese identities, instead of following the standard pronunciations associated with China (Hsu, 2005 ).", | |
| "cite_spans": [ | |
| { | |
| "start": 345, | |
| "end": 357, | |
| "text": "(Ohala, 1992", | |
| "ref_id": "BIBREF6" | |
| }, | |
| { | |
| "start": 358, | |
| "end": 372, | |
| "text": "(Ohala, , 2001", | |
| "ref_id": "BIBREF7" | |
| }, | |
| { | |
| "start": 623, | |
| "end": 633, | |
| "text": "(Hsu, 2005", | |
| "ref_id": null | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Implications for Automatic Speech Recognition", | |
| "sec_num": "5." | |
| }, | |
| { | |
| "text": "Accordingly, a viable resolution might program the automatic speech recognition to distinguish syllable-final nasal minimal pairs, as represented in Table 2. ", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 149, | |
| "end": 157, | |
| "text": "Table 2.", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Implications for Automatic Speech Recognition", | |
| "sec_num": "5." | |
| }, | |
| { | |
| "text": "The percentage of the nasal merger is obtained by excluding the pronunciations that follow Rule 2.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "", | |
| "sec_num": null | |
| } | |
| ], | |
| "back_matter": [ | |
| { | |
| "text": "To conclude, this study has analyzed the phonetic attributes of the syllable-final nasal shift from the velar to the dental that occurs nearly completely in Mandarin spoken in Taiwan.To accommodate this nasal shift, a computational linguist needs to include as many minimal pairs ending with the nasal rhymes as possible in the corpus of the speech recognition system.In addition, to improve the efficiency of the word identification, this study has also proposed to investigate the discourse where each of the minimal pairs appears. In short, future research into the syllable-final nasal minimal pairs and the context of their usages is needed to enhance the identification accuracy of the speech recognition system.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "annex", | |
| "sec_num": null | |
| } | |
| ], | |
| "bib_entries": { | |
| "BIBREF0": { | |
| "ref_id": "b0", | |
| "title": "Production and perception of syllable final [n] and [\u014b] in Mandarin Chinese: An experimental study", | |
| "authors": [ | |
| { | |
| "first": "J", | |
| "middle": [ | |
| "K" | |
| ], | |
| "last": "Tse", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1992, | |
| "venue": "Studies in English Literature", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "143--56", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "J. K. Tse, \"Production and perception of syllable final [n] and [\u014b] in Mandarin Chinese: An experimental study.,\" Studies in English Literature, pp. 143-56, 1992.", | |
| "links": null | |
| }, | |
| "BIBREF1": { | |
| "ref_id": "b1", | |
| "title": "Sound change in syllable-final nasal consonants in Chinese", | |
| "authors": [ | |
| { | |
| "first": "E", | |
| "middle": [], | |
| "last": "Zee", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1985, | |
| "venue": "Journal of Chinese Linguistics", | |
| "volume": "13", | |
| "issue": "", | |
| "pages": "291--330", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "E. Zee, \"Sound change in syllable-final nasal consonants in Chinese,\" Journal of Chinese Linguistics, vol. 13, pp. 291-330, 1985.", | |
| "links": null | |
| }, | |
| "BIBREF2": { | |
| "ref_id": "b2", | |
| "title": "Nasals and nasalization in Chinese: Exploration in phonological universals", | |
| "authors": [ | |
| { | |
| "first": "M", | |
| "middle": [], | |
| "last": "Chen", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1972, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "M. Chen, \"Nasals and nasalization in Chinese: Exploration in phonological universals,\" University of California at Berkeley, 1972.", | |
| "links": null | |
| }, | |
| "BIBREF3": { | |
| "ref_id": "b3", | |
| "title": "Cross-dialectal comparison: A case study and some theoretical considerations", | |
| "authors": [ | |
| { | |
| "first": "M", | |
| "middle": [], | |
| "last": "Chen", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1973, | |
| "venue": "Journal of Chinese Linguistics", | |
| "volume": "11", | |
| "issue": "", | |
| "pages": "38--63", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "M. Chen, \"Cross-dialectal comparison: A case study and some theoretical considerations,\" Journal of Chinese Linguistics, vol. 11, pp. 38-63, 1973.", | |
| "links": null | |
| }, | |
| "BIBREF4": { | |
| "ref_id": "b4", | |
| "title": "An areal study of nasalization in Chinese", | |
| "authors": [ | |
| { | |
| "first": "M", | |
| "middle": [], | |
| "last": "Chen", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1975, | |
| "venue": "Journal of Chinese Linguistics", | |
| "volume": "3", | |
| "issue": "", | |
| "pages": "16--59", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "M. Chen, \"An areal study of nasalization in Chinese,\" Journal of Chinese Linguistics, vol. 3, pp. 16-59, 1975.", | |
| "links": null | |
| }, | |
| "BIBREF5": { | |
| "ref_id": "b5", | |
| "title": "Phonetic data analysis: an introduction to instrumental phonetic fieldwork", | |
| "authors": [ | |
| { | |
| "first": "P", | |
| "middle": [], | |
| "last": "Ladefoged", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2003, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "P. Ladefoged, Phonetic data analysis: an introduction to instrumental phonetic fieldwork. Oxford: Blackwells, 2003.", | |
| "links": null | |
| }, | |
| "BIBREF6": { | |
| "ref_id": "b6", | |
| "title": "What's cognitive, what's not, in sound change", | |
| "authors": [ | |
| { | |
| "first": "J", | |
| "middle": [ | |
| "J" | |
| ], | |
| "last": "Ohala", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1992, | |
| "venue": "Lingua e Stile", | |
| "volume": "27", | |
| "issue": "", | |
| "pages": "321--362", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "J. J. Ohala, \"What's cognitive, what's not, in sound change,\" Lingua e Stile, vol. 27, pp. 321-362, 1992.", | |
| "links": null | |
| }, | |
| "BIBREF7": { | |
| "ref_id": "b7", | |
| "title": "An Account of Sound Change", | |
| "authors": [ | |
| { | |
| "first": "J", | |
| "middle": [ | |
| "J" | |
| ], | |
| "last": "Ohala", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2001, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "J. J. Ohala, \"An Account of Sound Change,\" University of California, Santa Barbara: The LSA Institute, 2001.", | |
| "links": null | |
| }, | |
| "BIBREF8": { | |
| "ref_id": "b8", | |
| "title": "The leveling of neutral tone in Taiwan Mandarin and its new identity", | |
| "authors": [ | |
| { | |
| "first": "H.-J", | |
| "middle": [], | |
| "last": "Hsu", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": null, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "H.-J. Hsu, \"The leveling of neutral tone in Taiwan Mandarin and its new identity,\" in", | |
| "links": null | |
| } | |
| }, | |
| "ref_entries": { | |
| "FIGREF0": { | |
| "uris": null, | |
| "type_str": "figure", | |
| "num": null, | |
| "text": "to Rule 2, the vowel labialization rule. Obviously, this sound modification displays articulatory assimilation because the vowel is labialized due to the influence of the initial bilabial consonant /m/. This sound change, however, might not merely occur because of sound assimilation, but it might also exist to constrain the creation of homophones; for example, if the word meng (/m\u0259\u014b/, \u5922, dream) is changed according to the nasal fronting rule, it would become neutralized with another word men (/m\u0259n/, \u9580, depressed), thus resulting in homophones. By contrast, if the word meng is changed according to the vowel labialization rule, it would be pronounced as mong (/mo\u014b/), which does not appear in Mandarin vocabulary. Therefore, the vowel labialization rule does not simply occur for ease of articulation but may also fill the vocabulary gap while avoiding creating homophones. In addition, the results demonstrate that the syllable-final nasal shift described above occurs mainly in Mandarin spoken in Taiwan, instead of China. Specifically, when the preceding vowel is /i/, the final nasal merger occurs 96 percent of the time in the native speakers of Mandarin from Taiwan (MT). By contrast, the final nasal alteration takes place only 38 percent of the time in the native speakers of Mandarin from China (MC). Moreover, when the preceding vowel is /\u0259/, the nasal shift occurs 95 percent of the time in MT, while only 3 percent of the time in MC. Nevertheless, the nasal merger never occurs when the preceding vowel is /a/. The occurrence percentage of the nasal merger in Mandarin is displayed as follows: Occurrence percentage of the syllable-final nasal merger in three environments 2" | |
| }, | |
| "FIGREF1": { | |
| "uris": null, | |
| "type_str": "figure", | |
| "num": null, | |
| "text": "Spectrogram of the test word jing-yu (\u9be8\u9b5a, whale) pronounced by the Taiwanese informant" | |
| }, | |
| "FIGREF2": { | |
| "uris": null, | |
| "type_str": "figure", | |
| "num": null, | |
| "text": "Spectrogram of the sound ji\uf029 -yu" | |
| }, | |
| "FIGREF3": { | |
| "uris": null, | |
| "type_str": "figure", | |
| "num": null, | |
| "text": "Spectrogram of the sound ji-yu At a glance, none of Figures 3 and 4 look similar to Figure 2." | |
| }, | |
| "FIGREF4": { | |
| "uris": null, | |
| "type_str": "figure", | |
| "num": null, | |
| "text": "Spectrogram of the test word jin-yu (\uf90a\u9b5a, goldfish) pronounced by the same Taiwanese informant The spectrogram similarity between Figures 2 and 5 indicates that the Taiwanese Mandarin speaker does not distinguish the minimal pairs jing-yu (\u9be8\u9b5a, whale) and jin-yu (\uf90a\u9b5a, goldfish), but pronounces them the same, changing the final velar nasal into the dental. Furthermore, this acoustic analysis is also supported by the discovery from my interview with the Taiwanese informant after the riddle game. By comparison, the spectrogram of the word jing-yu (\u9be8\u9b5a, whale) pronounced by the Chinese informant is distinct from the spectrogram of the same word pronounced by the Taiwanese informant, as shown below: Spectrogram of the test word jing-yu (\u9be8\u9b5a, whale) pronounced by the Chinese informant Apparently, Figure 2 does not look similar to Figure 6; accordingly, it is safe to infer that the final velar nasal does not retain in MT, but must shift into another, either the bilabial or the dental. Finally, the comparison between Figure 2 and the spectrogram of the bilabial nasal demonstrates that the nasal does not shift to the bilabial because the two spectrograms are distinct from each other, as shown in Figure Spectrogram of the sound jim-yu The spectrogram comparison between Figures 2 and 7 eventually assures us that the syllable-final velar nasal does not switch to the bilabial but the dental." | |
| }, | |
| "FIGREF5": { | |
| "uris": null, | |
| "type_str": "figure", | |
| "num": null, | |
| "text": "As phonetically presented above, the syllable-final velar nasal shift to the dental regularly occurs in the variety of Mandarin spoken in Taiwan. While this nasal merger in phonologically independent words might not influence computer intelligibility, it might interfere with the computer word-identification process of the syllable-final nasal minimal pairs. For instance, because of the nasal shift, Taiwanese Mandarin speakers tend to say the word gao-xing as gao-xin, changing the velar nasal into the dental one. This sound alteration might not mislead the speech recognition system to misidentify gao-xing (\u9ad8\u8208) as gao-xin (\u9ad8\u4fe1) because no such word as gao-xin (\u9ad8\u4fe1) exists in Mandarin. Nonetheless, this nasal shift has resulted in considerable homophones, neutralizing many minimal pairs, such as jing-yu (\u9be8\u9b5a, whale) and jin-yu (\uf90a\u9b5a, goldfish), and sheng-gao (\u5347\u9ad8) and" | |
| }, | |
| "TABREF1": { | |
| "html": null, | |
| "num": null, | |
| "content": "<table><tr><td>Riddle</td><td>Answer</td><td>Pinyin</td></tr><tr><td>\u665a\u4e0a\u6703\u767c\u5149\u7684\u6606\u87f2\uff0c\u662f\uf9fd\u9ebc\u87f2? (What kind of insect lights in the evening?)</td><td>\u87a2\u706b\u87f2 (Firefly)</td><td>Y\u00edng-hu\u014f-ch\u00f3ng</td></tr><tr><td>\u5728\u6d77\u908a\u6307\u5f15\u8239\u51fa\u5165\u6e2f\u7684\uff0c\u662f\uf9fd\u9ebc\u5854? (What kind of building stands on the coast and guides boats in and out of a harbor?)</td><td>\u71c8\u5854 (Lighthouse)</td><td>D\u0113ng-t\u0103</td></tr><tr><td>\u5168\uf94a\u6253\u662f\uf9fd\u9ebc\u7403\uf9d0\u904b\u52d5\u7684\u7528\u8a9e? ( \"Home run\" is a term of what sports?)</td><td>\u68d2\u7403 (Baseball)</td><td>B\u00e0ng-q\u00edu</td></tr><tr><td colspan=\"3\">The test words were randomly mixed with 10 irrelevant words in order to avoid the subject's</td></tr><tr><td colspan=\"3\">awareness of what words were being examined. I interviewed each participant at a time and</td></tr><tr><td colspan=\"3\">tape-recorded each interview. The subject was first asked to answer the questionnaire which</td></tr><tr><td colspan=\"3\">included questions concerning his or her basic sociolinguistic background. Next, the</td></tr><tr><td colspan=\"3\">interview proceeded with a riddle game, which was carried out in a relaxed atmosphere to</td></tr><tr><td colspan=\"3\">collect data from the subject's virtually spontaneous utterances. The informant was told to</td></tr><tr><td colspan=\"3\">give the answer to every riddle as soon as possible. If having no idea about the answer, the</td></tr><tr><td colspan=\"3\">informant would be given clues to say the test word. Furthermore, if the response was</td></tr><tr><td colspan=\"3\">perceived to be not loud enough for sound analyses, the informant would also be asked to say</td></tr><tr><td colspan=\"3\">the answer once more and aloud. The recorded data were later analyzed on the computer</td></tr><tr><td colspan=\"3\">using Praat, a sound-editing program. The findings are presented in the ensuing section.</td></tr></table>", | |
| "text": "some examples of the riddles used for speech data collection", | |
| "type_str": "table" | |
| } | |
| } | |
| } | |
| } |