On the Relationship between Grammatical Units, Tonal Units and Physiological Constraints on the Respiratory System in GWA
暂无分享,去创建一个
In 1 9 6 5 , a mere eight years ago. when S T E W A R T proc\>< hie The typology of the Twi tone system (with comments by SCHACHTER ai.«l G E L M E R S ) [16]' the phenomenon of terraced tone languages was not widely recognised and was a point of discussion for only a handful of West African specialists. By now, however, the type is well known and often enough discussed not to warrant a full treatment, so I shall mention only a few salient points for the benefit of non Africanists. As S T E W A R T pointed out, CHRISTALLER had noted in his Grammar of the Asante and Fante language (1875, p. 15) [2] that "every syllable has, in comparison with neighbouring syllables, either high, low, or middle tone" , the latter defined as "high tones abating by one step or successive steps". In 1 9 5 9 , W E L M E R S in Tonemics, morphotonemics and tonal morphemes, [ 1 7 ] spoke specifically of "terraced level languages" as opposed to "discrete level languages" saying of the former "low pitches . . . represent the toneme low. But of the two nonlow tonemes, one must be described as 'same as preceding non-low', and the other as 'lowerthan preceding non-low'", which he calls "same" and "drop" . In 1 9 6 0 , W I N S T O N described essentially the same tonal system as Twi and set up a phoneme of downstep in his The 'Mid Tone' in Efik. [18] The following year, 1961, SCHACHTER in Phonetic similarity in tonemic analysis> with notes on the tone system of Akwapim Twi, [13] set up three tonal phonemes: high, low and high-change while S T E W A R T in his 1962 Ph. D. thesis on Fante [15] worked out the position which he was to adopt in his 1965 paper. [16] Positing high, low and falling ( = high -flow) tones, he says "Where a high tone is followed by one or more low tones which are followed in turn by a high tone, the second high is appreciably lower in pitch than the first, so that when the high tones of a phrase are interrupted by low tones at a number of points, the high tones descend in pitch by a series of steps from the beginning to the end of the phrase". The relationship between the Base Tones and the Intonational Terracing is well illustrated by Kofi hwèhwé Kwàdwó _1—] L H L H L H Kofi looks-for Kwadwo Kwàdwó hwèhwé Kofi [_ ] L H L H L H Kwadwo looks-for Kofi. Since this important article by S T E W A R T , SCHACHTER and F R O M K I N in A phonology of Akan [14] have given us a generative approach to the same subject, saying on pp. 105 sqq. "At the systematic-phonemic level, all Akan vowels are marked for one