十六世紀における西康省チベツト語天全方言について : 漢語・チベット語単語集いわゆる丙種本『西番館譯語』の研究

I. Tibetan written and spoken language. II. Sifan A and Sifan B--the written Tibetan of Amdo in the fifteenth century and the T'ien-ch'üan dialect of Hsi-k'ang in the sixteenth century. III. Phonemic system of Sifan B. IV. Grammatic form of Sifan B. V. Text of the Chinese-Tibetan vocabulary. 1. The Tibetan written language has two types. One is the written language which developed along sanskrit lines after the Devanāgari script was introduced in the seventh century, and which was largely composed of a sanskritized translated vocabulary and style. In 826 this written language was reformed by King Khri-sde srong-btsan, and became what is known as Classical Tibetan, as found in the works of the Buddhist Tripitaka. The second type of the written language had no direct connection with sanskrit, and represented the spoken form of the Tibetan language as it existed in the ninth century. The present writer refers to this language, which has been established from Turfan, Turkestan, and Tun-huang manuscripts, as Ancient Tibetan. After the ninth century this Ancient Tibetan came under the influence of Classical Tibetan and a new written language evolved. I refer to the language as found in the non-canonical literature as Written Tibetan. While the written language was undergoing these changes, the spoken language was also following its own development. Source materials on the old forms of spoken Tibetan are extremely scarce, and it is possible that among the numerous Tibetan literary works surviving there is not a single source for the spoken form of Tibetan later than the tenth century. There is, however, one excellent source which records the Tibetan spoken language of one particular area. This is the Chinese Tibetan vocabulary known as Hsi-fan-kuan i-yü. Although this work does not use Tibetan characters and transcribes spoken Tibetan in Chinese characters, it serves as an excellent text for the spoken Tibetan of the sixteenth century, despite the limitations imposed by the structural differences in the Tibetan and Chinese phonemic systems. I conclude that here is represented the Tibetan language of the T'ien-ch'üan 天全 dialect of Hsi-k'ang 西康 province, for the reason that in its "Place Names Part" we find the name T'ien-ch'üan Liu-fan Chao tao ssŭ 天全六番招討司 listed directly after the names of Pei-ching 北京 and Nan-ching 南京. It is quite conceivable that the spoken Tibetan recorded here is still in use today. The object of the present study is an examination of this text, the Hsi-fan-kuan i-yü. 2. In Hsi-fan-kuan i-yü series, there are also several Tibetan-Chinese vocabularies. In a previous study I indicated (taking the texts in the Paris Asia Association and the Toyo bunko as representative) that the written Tibetan of the Amdo area was recorded there. Provisionally this has been referred to as Sifan language A and the above-mentioned dialect of the T'ien-ch'üan area as Sifan language B. Sifan language A is recorded in a vocabulary used as reference in translating letters sent from Tibet by a member of the Ssŭ-i-kuan 四夷館. The work which records Sifan language B is a pocket Tibetan glossay used for reference by an employee of the Hui-T'ung kuan 會同館, acting as interpreter for a Tibetan envoy. Thus the latter is an excellent source for information on the spoken form of the Tibetan of the time. For example, while "sea" is transcribed 児甲木錯 rgya-mtsho in Sifan A, in Sifan B it is written as ȡen-tsho ; while "border" is 薩木塔 sa-mtha in Sifan A, in Sifan B it is 三塔 santha. In the former, the writer is aware that "sea" is formed by a combination of rgya and mtsho and "border" by a combination of sa and mtha. The latter work, however, differs in that represents a record of the language as actually heard by the interpreter. Not a few examples similar to the above can be detected. 3. In order to infer exactly what sort of phonetic form the Chinese characters which transcribed Sifan B were intended to represent, I have found that it is essential to refer to the languages which are most closely related to Sifan B, such as Written Tibetan and forms of modern Tibetan dialects. "Rabbit", for example, is transcribed as "里公", and judging from the Chinese phonetic form, would be inferred to be li-kung. In actuality, however, the form ri-gong can be inferred because of the presence of ri-gong in Written Tibetan, of ri-kong in the Lhasa dialect, of ɻə3-goŋ3 in Chamdo, and of ri-γoŋ in Amdo. In each instance, both in Written Tibetan and in modern dialects, the initial is r- and not 1- ; the vowel is -oŋ and not -uŋ. In the Chinese of the such a distinction could not be made. Thus the word form meaning "torrent" in Sifan B and the word meaning "ditch", also in Sifan B, are both transcribed in Chinese by "瀧" luŋ In actuality, the former may be presumed to have had the form roŋ and the latter the form luŋ. I have transliterated the Chinese phonetic form used here on the basis of the system devised by Hsü Hsiao 除孝 of the Ming dynasty in his Ssŭ- ...