SANDAWE, less correctly Sandawi, a small tribe of about 15,000 persons living in the Kondoa-Irangi district of Tangan yika Territory, Africa, between the rivers Bubu and Mponde.
Attention was drawn to their language by the late Dr. Oskar Bau mann, who visited their country in 1892, and was able to collect some specimens, being especially struck by the remarkable fact of its containing clicks (q.v.), which had up to that time only been observed in the Hottentot and Bushman languages of South Africa, and in those adjacent Bantu languages which had bor rowed them. It was investigated at a later date by E. Nigmann (1909) and, more scientifically, by Dr. Otto Dempwolff (1916).
The only English writer who has treated the subject in any detail is F. T. Bagshawe, Political Officer at Mbulu, who was in close contact with these people from 1917 to 1923.
It has been assumed, though without sufficient reason, that all languages containing clicks must be related to each other, and that there can be no possible relationship between a click-lan guage and one where clicks are absent. That this position, how ever, is by no means generally accepted, appears from the disputes which have arisen over the question whether the Hottentot and the Bushman groups (neither consisting of a single language) have a common origin, or are entirely distinct. (See AFRICA, Ethnology.) The three clicks found in Sandawe seem to be identical with three of the Nama clicks—those which have been adopted into Zulu and are represented in the current orthography by c (dental), q (cerebral or "retroflex") and x (lateral). The last-named is not produced in quite the same way as by the Nama, but, as observed by Dr. Dempwolff, the sound is exactly the same. Both Nigmann and Bagshawe have noticed the fact that the Sandawe clicks form a distinct syllable, at any rate in some cases (it is not clear from the latter's account whether the rule is invariable), whereas in Zulu they form an integral part of a syllable, the pronunciation of which constitutes one of the greatest difficulties to a European learner.
The word-stress appears to be extremely variable and is often the only distinction between words of similar form, but different meaning (e.g., non "elephant" and nod "claw"). Intonation ("tone") is also, as in the Hottentot and Bushman languages, a very important feature : tsa with a high tone means "pot," with a low tone "tear." Both these points await fuller investi
gation.
A large proportion of the Sandawe vocabulary can be corre lated with Nama words, and a less amount with other Hamitic languages, such as Tatoga, Iraku and even Somali. But this is an unsatisfactory basis for deducing relationship, otherwise San dawe might be claimed as a Bantu language, since Dr. Dempwolff gives a longer list of words which can be identified as Bantu (obviously borrowed from Swahili, Gogo, Hehe, Irangi, etc.), than of those which he has traced to Nama. A far more trust worthy criterion is to be found in the grammatical framework of the language and here there is little room for doubt. Sandawe possesses grammatical gender, suffixes to denote the plural, case endings for nouns (an "instrumental case" is almost identical, both in form and meaning with that found in Nama), and other features in common with Nama and also, to a less extent, with other Hamitic languages. One peculiarity, the use of a different form for a verb in singular and plural (e.g., the "he runs," giribe "they run"), is neither Hamitic, Bantu nor Sudanic. The order of words in a simple sentence (subject, object, verb : "he cattle herds," instead of "he herds cattle") and the position of the genitive (possessor preceding the thing possessed : Humbu tse, "ox's head," not, as in Bantu, "head of ox") are the same as in Nama, but differ from the normal Hamitic type. In both cases there would appear to be sufficient reasons for the difference.
Professor Alfredo Trombetti is inclined to think that the Sandawe are, of all existing peoples, those most nearly related to the Hottentots.