TRANSCRIPTION OF GEOGRAPHICAL NAMES IN CENTRAL ASIA, 659
this word, one could feel greater certainty that the u really was a u, and consequently a low vowel; but the Swedish u is a high vowel, which when it is long approximates to ü, and when it is short resembles an ö. If now the u in Ugendarja is to be read as a high vowel the g would be the right sign for the sound of the letter that immediately succeeds it, otherwise gh. Clearly the simplest way would have been not to discriminate at all between gh and g; in that way the interpretation of the sounds would have been clearer, though at the same time with a resulting loss in scientific precision.
But it was more difficult to distinguish between the closed e and the open a. Dr. Hedin had, it is true, made use of both these signs, but being a native of Stockholm he does not in his pronounciation of Swedish make any difference between these two sounds, though the natives of most other parts of Sweden keep them scrupulously distinct. Hence I could not be perfectly sure that he had in every case recorded. the right sound, that is if the dialect does possess both e and ä. And upon this point I was doubtful. Vâmbéry seems, it is true, (p. 13) to assume two sounds; but in Kimos I find only e and in Radloff only ä. I endeavoured therefore to follow the latter as being the greater authority, and consequently put ä in a great number of words, in fact in as many as was possible; but at length the excessive number of these ä's caused me to be doubtful, and on the whole I now think it would have been better to have left Dr. Hedin's orthography as it stood. It was then however too late.
Whilst the work of revising the earliest maps was in progress, the dialect appeared, as far as could be seen, to be in essential agreement with the East Turkish literary language, which conclusion I had arrived at when studying the list of names in Pelerm. Milleil., Ergänzhft 28. On Sheets 9, 10, and II of the Atlas there occurred however a few names, which pointed to the existence of quite a different dialect, varying in essential particulars from the one I was dealing with — on Sheet 9 Muhamed Kulluning-uji, to the Kullu of which the corresponding literary form is Kuli = »his servant», a form which occurs also in Shaw; Sheet to Ail-öllögön, the final syllable in which ought to be -gen in the literary language; Sheet I I Monauni-ollogho, the last member of which would be written olaghi in the literary language. This feature, the so-called »labial affinity in vocalic harmony», again made its appearance at a later stage in a vast number of names from the Lop country, and pointed to the so-called Eastern dialects (Radloff) where corresponding phenomena exist in the Teleut dialect and in the Altai dialect. In these cases it would be particularly inappropriate to normalize to the East Turkish literary language, because by so doing one would obliterate traces of national elements which have no immediate connection with the Kaschgar Turks, but on the contrary are possibly derived from the ancient Uigurs. For this reason I allowed these peculiar forms to stand unaltered.
Having reached this stage of the work, I was unfortunately prostrated with illness and for a considerable time totally incapable of work. In consequence of this Dr. Hedin had to continue unaided the revision of the maps and text and the application of the principles I had laid down. It was only when vol. II began to be printed that I was able to resume work. In consequence of this unfortunate break it is almost inevitable
IIIFカラー高解像度
