INDIVIDUAL RESEARCH REPORTS 41. Leone, will be published as soon as possible in the journal Sierra Leone Studies. I wish to thank Professors J . H. Nketia and I . G. Wilks, and the Pro-Vice-Chance I lor, for allowing me to be absent during term; and to express my gratitude to Mr. M. Crowder and Mr. J. Hyde of the Institute of African Studies, Fourah Bay College, as the most instrumental of the many who made my work possible, successful, and extremely enjoyable. Paul Ozanne. COMPARATIVE STUDY OP THE VOLTA-COMOE LANGUAGES Both WESTERMANN AND BRYAN and GREENBERG recognise the genetic unity of a group of languages which they call Akan and which includes Akan (Twi-Fante), Anyi-Bawule, Nzema, Ahanta and the Guan languages. On the advice of my colleagues in Ghana I now call this group Volta-Comoe in order to avoid confusion with the Akan (Twi-Fante) language. I have for many years been interested in the comparison of Akan (Twi-Fante) with other languages of the Western Sudanic (Niger-Congo) family.' At an early stage I sought regular sound correspondences between the Akuapem Twi dialect of Akan, on which a vast amount of material was available in CHRISTALLER'S Dictionary of the Asante and Fante language called Twi, and the Common Bantu worked out by GUTHRIE from the languages of the Bantu group. The Bantu group was the only one within the Western Sudanic family for which ah overall system of sound correspon- dences and a list of starred forms had been established, and I had the privilege of access to GUTHRIE'S 'Index of Common Bantu starred forms' 1. See J . H. GREENBERG, Languages of Africa (Mouton, The Hague 1963). 42. INDIVIDUAL RESEARCH REPORTS (cyclosryled), an early version of part of his forthcoming Comparative Bantu, which summarised the results of the most recent and most reliable work in this field. I was unable to establish an overall system of sound correspon- dences, but a number of isolated correspondence sets did emerge, two of which, (1) Twi h?/hy9 • Common Bantu *p and (ii) Twi s • Common Bantu *t, were particularly well attested: ham hyeg henS hSra hye e-hS nig how homS huw a-huru ©-hum o-sa saw sa &-sa sa sag se ss n-stJ ar-sS so 30a soe sopa Common Bantu *pam *pep *pede *pemb *pi *pi *pimbo *pu *puuin, puumud *pud *pydo, *pudu, pudu *puti *ta *tap *tad *tatu *tat *tatud *ti *ton, *ton^, *tont *to *tu *tu *tuad *tuud *tuk . .. •' argue blow itoh (n) blow nose burn where? stiok (v) dry up rest (v) blow foam gun(l) war draw (water) cut open three tie untie say drip ash ear carry carry put down (load) abuse but Generally the transcription used in this report is CHRISTALLER'S, b in the case of Common Bantu GUTHRIE'S transcription is retained i in original form. hy before front oral vowels, otherwise h. INDIVIDUAL RESEARCH REPORTS 43. I visited the Ivory Coast in 19534, I9605 and 1962°, and Paris in 1955', to collect material on Anyi-Bawule, Abure and Betibe of the Volta- Comoe group as well as on a number of non-Volta-Comoe languages spoken in the lagoon area: Cama (Kyama, EbriS), Mbato (Gwa), Akie (Atttf), Krobu, Alaja (Alladian), Ojukru (Adioukrou), Abiji (Abidji, Ari) and Abe. From the material collected/ there emerged Anyi, Abure, Betibe, Cama and Mbato counterparts of the Twi - Common Bantu correspondence sets already noted, the resulting extended correspondence sets being as follows:- Cama Mbato Betibe Abure Anyi Twi Common Bantu P p f f 0 y/w(?)/ny/nw(?)? h/V *o > t t h y/w/ny/nw** s/zlO s *t These correspondences suggested that-the original sounds were p, t, and that these shifted to f, s in a common ancestor of the Volta-Comoe tongues represented, namely Betibe, Abure, Anyi and Twi; that later, the s shifted to h in a common ancestor of Betibe and Abure, and the f shifted to h in a 4. With the aid of a British Treasury 'Scarbrough' Postgraduate Studentship. 5. With the aid of a grant from the School of Oriental and African Studies, University of London. 6. With the aid of a grant from the Institute of African Studies, University of Ghana. 7. With the aid of a grant from the School of Oriental and African Studies, University of London. 8. Generally y/w before oral vowels and ny/nw before nasal vowels, and 9. y/ny before non-back vowels and w/nw before back vowels. In general the distribution is similar to that of Abure y/w/ny/nw (see footnote 8), though the situation before back vowels remains obscure. 10. z after n, otherwise s. £•* •'-5- 44. INDIVIDUAL RESEARCH REPORTS - IN common ancestor of Anyi and Twi; and that later still, the h shifted to y/w/ny/nw (i) in ancestor of Abu re, where it represented the original t, and (ii) in an ancestor of Anyi, where it represented the original p. The genetic interrelationships of the four Volra-Comoe tongues thus appeared to be as follows: Betibe There was further evidence for the validity of subgroup C in the form of sound correspondences suggesting two further soundshifts in a common ancestor of Anyi and Twi: (a) p (representing an original "checked" p, as distinct from the original "unchecked" p which shifted via f to h in a common ancestor of Anyi and Twi) apparently shifted to f, thereby filling a gap left by the evidently earlier shift of f (representing the original "unchecked" p) to h. (b) ts apparently shifted to s, thereby merging with an existing s. A subsequent glottochronological investigation carried out in collaboration with Professor MORRIS SWADESH during his term as Visiting Professor to the Institute of African Studies in 1965 confirmed these genetic interrelation- ships except that it showed Abu re as closer to the languages of subgroup C than to Betibe. The apparently anomalous figures for Abu re I put down to the fact that most if not all Abu res speak Anyi as a second language and • •»• M* INDIVIDUAL RESEARCH REPORTS 45. make abundant use of Anyi lexical items even when speaking their own language. From the material immediately available on the, remaining Volta- Comoe tongues it was clear that Bawule, Chakosi, Nzema and Atlanta ail belonged to subgroup C. It also seemed likely that the remaining Volta- Comoe languages, namely the Guan languages, belonged to group A but not to either of the subgroups B or C, for the following reasons:- (a) They appeared to belong to group A as they had evidently undergone the characteristic shift from p, t to f, s. (b) They did not belong to group B as they had not undergone the characteristic shift from s to h. (c) They did not belong to group C as they had not undergone the characteristic shift from f to h. Group A thus appeared to coincide with the Volta-Comoe group. The Guan languages were provisionally supposed to constitute a third subgroup D co-ordinate with B and C. It had emerged only shortly before this, from comparative wordlists collected by Mr. GILBERT ANSRE of the Institute of African Studies, that Guan was not just one language as was generally supposed, but a group: (or at least a number) of quite widely divergent languages. Guided by the evidence in Mr. ANSRE'S material as to the subdivisions within Guan, I decided to collect extensive comparative wordlists of rhe forms of Guan spoken in the Awutu, Larteh, Nkonya and Krachi areas with a view to a study of the sound correspondences. To this end I prepared a questionnaire containing about 1400 items, based mainly on the items represented in GUTHRE'S list of Common Bantu starred forms and in WESTERMANN'S list of Proto- West Sudanic starred forms but supplemented from a variety of other sources. I then made a hasty collection of the material, generally omitting items which presented difficulties; my object at this stage was to collect as large a number of root morphemes as possible in as short a time as possible. presented a selection of the material in Awutu, Larteh, Nkonya and Krachi, N o .l of a new Comparative African Wordlists series based on the question- I 46. INDIVIDUAL RESEARCH REPORTS naire referred to and published by the Institute of African Studies. The material suggested that in a common ancestor of the Guan languages more recent than Proto-Volta-Comoe, kw may have shifted to kp, thereby merging with an existing kp; this is admittedly very tentative as it is based on material on only four of the Guan languages, but it does tend to confirm the validity of the Guan group. I hope eventually to be able to establish an overall system of sound correspondences and a list of starred forms for the whole of the Volta- Comoe group. This has not so far been done for any language group in West Africa, and the Volta-Comoe group with its relatively large number of relatively closely related languages offers particularly favourable conditions for such a project. sound correspondences in 'Three Akan-Bantu sound correspondences', a paper which I read at the Fifth West African Languages Congress, Legon, 1965. I have already presented some Volta-Comoe ' Taking into consideration recent work by MR. COLIN PAINTER of the Department of Linguistics of this University, in collaboration with Professor SWADESH, on the Guan languages, by RON AND LYN STAN- FORD of the Institute of Linguistics (West Africa), which is affiliated to the Institute of African Studies, on Chakosi, and by MR. ALBERT KO0WO- MENSAH, a postgraduate student of the Institute of African Studies, on Ahanta, in addition to my own work, I now classify the Volta-Como© languages as follows:- 1. Ono Abu re (Aboure) Betibe (Vitre, Metyibo) 2. Tano (a) Bia (i) Any-i-Bawule-Chakosi Anyi-Bawule (Agni-Baoule) Chakosi (Anufo) INDIVIDUAL RESEARCH REPORTS 47. (ii) Nzema-Ahanta Nzema Ahanta (b) Akan (Twi-Fante) 3. Guan (Guang) Awutu (Afutu) Lete-Chiripon-Anum (Larfeh-Cherepong-Anum) Nkonya Krachi Anyanga Achode Nchummuru Nawuri Yeji Gonja (Ngbanyito) (and possibly others). This classification first appeared with different nomenclature in 'Three Akan-Bantu sound correspondences'. The nomenclature was extensively revised in consultation with the Seminar on the Akans, Legon, 1965. A fascinating aspect of the comparative study of languages is the light it can shed on the history of the peoples. I have recently paid considerable attention to this aspect of the study of the Volta-Comoe languages, and have reported some conclusions in 'Akan history: some linguistic evidence', to appear in a special volume of Ghana Notes and Queries devoted to the Akans. J . M. Stewart.