Shatou near Shumchun. Mr. Tang can speak SC, but his pronunciation is recognizable as non-native. In this, he differs from his father, who in his lifetime knew only the local dialect, and from his son, who is more fluent in SC than in KHW; even more so from his grandson, a child of 3 who, although he is being raised in KHW and understands the dialect, is acquiring SC.
The data consist primarily of the reading pronunciations of about 430 Chinese characters, for the most part those listed in the first pages of the Fangyan Diaocha Zibiao, a questionnaire commonly used in surveys on Chinese dialects. The characters in these lists are chosen so as to allow a quick characterization of the main phonological features of a Chinese dialect. Additional reading pronunciations were collected, as well as a few lexical items and samples of spoken dialect. The phonetic transcriptions were checked out for consistency with the informant, especially with respect to finals and tones.
It might be argued that reading pronunciations serve the purpose of reading aloud texts in classical Chinese, and do not necessarily coincide with the colloquial. However, the spoken forms within the data, and a short exposure to the dialect outside working sessions convinced me that the same phonological system underlies the reading pronunciations and the colloquial. In a few cases, a difference was found between reading and spoken forms. Such cases seem restricted to a phonologically well-defined category of words which show similar double readings in SC. There are undoubtedly a number of theoretical readings within the data, as a reading was obtained for each character, however unlikely to occur in spoken KHW. Still, no effort was made to find out which readings were theoretical, as phonology was the only concern of this survey.
In the following pages, the Meyer-Wempe (MW) system of romanisation will be used for transcribing SC sounds, and a system close to the International Phonetic Alphabet will be used for transcribing KHW sounds. An asterisk precedes reconstructed forms.
Page 165
Page 166