2 There are already some excellent treatises on the subject; see especially Pike, K. L., Tone languages, Ann Arbor, University of Michigan Press, 1948, and see also Welmers, W. E., ‘Tonemics, morphotonemics and tonal morphemes’, General Linguistics, IV, 1, 1959, 1–9, and the introductory section in Wängler, H.-H., Zur Tonologie des Hausa, Berlin, Akademie-Verlag, 1963.
3 Non-tonal languages, such as Swahili, are outside the scope of this treatise.
4 ‘The science of tonetics and its application to Bantu languages’, Bantu Studies, II, 2, 1924, 75–106.
5 Following Laman's nine tone-points for Kongo, as set out in The musical accent or intonation in the Kongo language, Stockholm, 1922.
6 In this, and in his subsequent Xhosa work, he used rather than - for the sign of the High tone.
7 ‘Venda: tonal structure and intonation (part i)’, African Studies, XXI, 2, 1962, 51.
8 ZDMG, LV, 1901, 607–82.
9 ‘Die Verba des Ṭṣivẹṋḓa1’, MSOS, VII, 1904, 12–31. This was followed up much later by their ‘Wörterverzeichnis der Venda-Sprache’, Jahrbuch dtr Hamburgischer Wiss. Anstalten, XXXVI, 1918, and DrSchwellnus's, Paul E.Luvenda grammar ya u talukanya Tshivenda, Pretoria, [c. 1935], and his Kima le kxaló le mešitó ya dirêtó ‘Weight, tone, and poetic diction’, Pretoria, 1942—this time in Pedi.
10 Thus possibly setting the pattern for Laman and Doke (see above).
11 Herein after referred to as Suto-Chuana.
12 First used by Jones in his Tones of Sechuana nouns (International Institute of African Languages and Cultures, Memorandum Vi), London, 1928.
13 ‘Sotho-Nguni orthography and tone-marking’, BSOAS, XIII, 1, 1949, 200–24.
14 ‘Step down’ was regarded as the normal feature of junction here, and therefore not indicated. Under ‘step up’ was to be understood a junction where the tone fails to step down; it may have the same level as the preceding High tone, though the general tendency is for the pitch to be slightly higher. (In subsequent works I reversed the words to ‘down step’ and ‘up step’ to conform with general usage.)
15 The book, which was published posthumously, unfortunately contains certain errors in printing, which render some parts of the tonal discussion unintelligible, and which the author, had he lived, would undoubtedly have corrected.
16 Exceptions are the Tshiveṋda-English dictionary, by van Warmelo, N. J., Pretoria, 1937, and the Zulu-English dictionary, by Doke, and Vilakazi, , Johannesburg, 1948, in which Doke's numbers are again placed above the words concerned; this practice, however, is largely dropped in the abridged English-Zulu dictionary of Doke, , Malcolm, , and Sikakana, , Johannesburg, 1958, apparently on grounds of economy.
17 Exemplified by the following works: Westphal, E. O. J., Kwangari: an index of lexical types, London, SOAS, 1958; idem, The sentence in Venda, Ph.D. thesis, University of London, 1955; idem, ‘Venda: tonal structure and intonation’, African Studies, XXI, 1962, 50–69, 123–73; Köhler, O., ‘Das Tonsystem des Verbum im Südsotho’, Mitt, des Inst. fur Orientforschung, IV, 3, 1956, 435–74; Beuchat, P.-D., ‘Tonomorphology of the Tsonga noun’, African Studies, XVIII, 3, 1959, 133–45; idem, ‘Additional notes on the tonomorphology of the Tsonga noun’, African Studies, XXI, 3–4, 1962, 105–22 (these authors use the type of tone-marking described later in the present article); Sandilands, A., Introduction to Tswana, Tigerkloof, London Missionary Society, 1953(here the tone levels are indicated in some of the exercises by lines superimposed on and partly obscuring the text).
18 Westermann's linguistic activities cover a large period of time (1907 to 1961) and a large area of Africa, ranging from Ewe to Shilluk. There is no need to enumerate his publications here, but it is worth recording a note of appreciation on the policy of the German and Swiss Churches of last century, of sending eminent language scholars to their mission fields, who produced excellent studies of the local languages. Christaller is one example, Westermann another; in East Africa names such as Krapf and Meinhof are to this day held in respect.
19 e.g. e, o, the only exceptions being vowel length, shown by a macron, and nasalization, shown by a tilde. Christaller also used the Lepsius alphabet.
20 e.g. ε Ɔ, used in posthumous editions of Christaller's works.
21 Later known as the International African Institute.
22 Practical orthography of African languages, revised ed., London, OUP, 1930.
23 The phonetic and tonal structure of Efik, Cambridge, Heffer, 1933; An introduction to the Ibo language, Cambridge, Heffer. 1936; The pronunciation of Twi, Cambridge, Heffer, 1939. See also Crosby, K. H., An introduction to the study of Mende, Cambridge, Heffer, 1944.
24 Dictionary of the Asante and Fante language, second ed., Basel, 1933.
25 Taken from Schachter, P., ‘Phonetic similarity in tonemic analysis’, Language, XXXVII, 2, 1961, 231–8. Note that the final lowered High tone is lower in pitch than the initial Low tone.
26 Stevick, E. W. and Machiwana, Kingston, Manyika step by step, Cleveland, Transvaal, 1960.
27 Redden, J. E. and Owusu, N., Twi basic course, Washington, D.C., 1963.
28 I shall return later to the principles underlying the unmarking of syllables—or ‘zero tone-marking’ as it was later called.
29 Abraham's Falling tone mark here, and his Rising tone mark in Tiv, (⋏represent ‘an arrow pointing downwards’ and ‘an arrow pointing upward’ respectively, and are intended to suggest automatically to the reader the direction of tone movement.
30 Cantonese primer, Cambridge, Mass., 1947.
31 The new official Chinese Latin script: Gwoyeu romatzyh, London, Probsthain, 1942.
32 Chinese sentence series, London, Probsthain, 1942–1944 (with Simon).
33 ‘A transcription for Cantonese’, BSOAS, XIII, 3, 1950, 725–45.
34 Structure drill in Cantonese, London, Lund Humphries, 1954.
35 Only a few of the many meanings of the words cited are given here.
36 Outstanding names here (in chronological order) are: Whitehead, Lekens, Hulstaert, Burssens, Carrington, Stappers, Willems, Meeussen, de Boeck, van Avermaet, Coupez, Larochette, Mamet, Maes, Ntahokaja, Rodegem, Bapfutwabo, Jacobs, Kamanzi, van den Eynde, Kadima, Matsene, de Witte, Hagendorens, Kagame, Rood, Vansina; and new names are for ever appearing in the prodigious output from this field.
37 In Notes on the tonal system of Northern Rhodesian Plateau Tonga, London, HMSO, 1962. Note that, whereas most writers place the down step symbol ! before the syllable concerned, Mrs. Carter places it before the vowel letter, because in the language she studied the pitch of the consonants b and m is the same as that of the preceding vowel.
38 In ‘Morphotonology of the Tonga verb’, Journal of African Languages, II, 1, 1963, 72–92.
39 See particularly The Eastern Sudanic languages, I, London, OUP, 1940.
40 op. cit., ch. iii and xviii. For the benefit of those unversed in this method of recording, the following are the values in the chromatic scale I used: d de r re m f fe S si 1 ta t d'.
41 See Ashton, , Mulira, , Ndawula, , and Tucker, , A Luganda grammar, London, Longmans, Green, 1954; also Kitching, and Blackledge, , A Luganda-English and English-Luganda dictionary (revised by Mulira and Ndawula), London, SPCK, 1952. D. T. Cole, in ‘Some features of Ganda linguistic structure’ (cyclostyled 1964), marks both High and Low tone.
42 See Taylor, C., A simplified Runyanlcore-Rukiga-English and English-Bunyankore-Rukiga dictionary, Kampala, Eagle Press, 1959.
43 With this distinction that the normal down step symbol! could not be used in Luganda because of its possible confusion with the apostrophe demanded by the standard orthography; the stroke / was used instead.
44 Which, however, would be wrong, as Low toneme may also occur in this position.
46 Dictionnaire lomóngo-franÇais, Tervuren, 1957.
47 The High toneme syllable 'zi, having a voiced initial consonant, has a pitch pattern shortened to _, whereas the Low toneme syllable hla, having an unvoiced initial consonant, has descending pitch (hard to distinguish from Falling tone) in this context.
48 See Meeussen, A. E., Essai de grammaire rundi, Tervuren, 1959.
49 See Coupez, A., Grammaire rwanda simplifiée, Usumbura, 1961.
50 Example supplied by Meeussen, who uses double vowels to indicate syllable length; in those systems where double vowels are not used, different conventions are found with Rodegem, Ntahokaja, and Kagame to indicate the various tones on long as opposed to short syllables.
51 See also his The role of tone in the structure of Sukùma, London, SOAS, 1959.
52 Tones of Sechuana nouns, 14.
53 Suto-Chuana,108; ‘Sotho-Nguni’, 216. An antepenultimate High tone causes lowering of a final High tone when the word stands in isolation, but not when it stands as Subject.