Historical Phonetics of Amharic.

• Author: Leslau, Wolf

The national language of Ethiopia, Amharic, as well as the other Ethiopian Semitic (also known as Ethio-Semitic) languages, are of particular interest to the Semitist, as well as the general linguist. In addition to their obvious Semitic features, the Ethiopian Semitic languages, and particularly the still-spoken languages, underwent influence from the Cushitic substratum in phonology, morphology, and syntax. Since the book under review deals with historical phonetics, special mention should be made of the glottalized or ejective pronunciation of some consonants, of the rounded or labialized consonants, and of the considerable number of palatalized consonants, features that are not known in the other Semitic languages. While the first study of Amharic goes back to the seventeenth century, the study of the historical phonetics of the language is still in its infancy. It is therefore gratifying to have a treatise dealing with this subject - a treatise that is challenging and informative, and displays original ideas concerning the historical phonetics of the language. The observations that follow reflect my interest in the subject: I have devoted the last five years to the compilation of a "Reference Grammar of Amharic."

P. 13. It should be stressed that t in the sequence t-d of Amharic tammada 'harness' lost its glottalization in Chaha damada through a process of assimilation, t-d becoming d-d. As for d in daqa 'laugh' in Western and Northern Gurage, it developed through dissimilation from saqa > Gafat saqa by assimilation > *taqa (s becoming t in South Ethiopic) > daqa by dissimilation.

P. 14. I agree with the author that the labiovelars are phonemic and did not necessarily develop "from plain velar consonants in the presence of rounded vowels o, u." Cushitic influence should not be excluded. Incidentally, the labiovelars are not generalized in Ethiopian Semitic. They do not occur in Tigre or in Harari.

We have no proof from Geez itself that the vowels a, i, u were long. Chances are that Geez had vocalic length, to judge from Tigre, Harari, and some Gurage dialects such as Ennemor and Selti.

P. 15. I do not think that Geez forms such as -ewwo (from *gabaru-o > gabgruwwo > gabarewwo) or -eyyo (from *gabarki-o > gabarkiyyo > gabarkeyyo) are to be considered diphthongs. The -w(w) of gabarewwo or -y(y) of gabarkeyyo are simply semivowels, w, y, to avoid the meeting of u-o. Note that in Amharic neither -w nor -y is geminated so that in a form such as *naggaru-at > naggaru-wat and in a form such as *tenagri-at > tenagri-yat, -wat, -yat begin a new syllable, and they are not a part of a diphthong.

P. 22. Note that p may also be replaced by f: thus, polis, bolis, and folis 'police'.

P. 25. The statement, "the only exception is the Geez word her 'good' which became car in Amharic," gives the impression that h became c. What seems to be the case, however, is that Ar. hayr became *her > *ker in Amharic > car with palatalization of k.

The pronunciation of the glottal stop ?? in the position V??C is a debatable question. I personally detect the sound of the glottal stop in ma??zan, me??rab, as against mazan, maezan that the author hears.

P. 27. There is no proof for the statement...