¹ Northcote Thomas, W., ‘Who were the Manes?’, Journal of the African Society, xix (1919), 176–88; xx (1920), 33–42.Google Scholar

² Hair, P. E. H., ‘An ethnolinguistic inventory of the lower Guinea Coast before 1700: Part I’, African Language Review, vii (1968), 47–73, 52, 58–9Google Scholar. The Mani vocabulary also included Manding words, as well as some which have not been identified but seem to be vaguely ‘Mande’.

³ Earlier versions of this article have appeared in my Ph.D. thesis, ‘A history of the Galinhas country, Leone, Sierra, c. 1650–1890’ (University of Birmingham, 1979)Google Scholar and in a paper given at the School of Oriental and African Studies on 16 January 1980. I should like to thank David Dalby, Paul Hair, Matthew Hill, David Sapir, Elizabeth Tonkin and William Welmers for their assistance. Since writing the first version of this article, I have obtained a copy of Svend Holsoe's paper, ‘The “first” Vai migration’, presented at the Sixth Annual Liberian Studies Research Conference (Madison, 1974).Google Scholar Some of the material used by Holsoe is the same as mine; but I believe my conclusions are sufficiently different to warrant a second attempt to deal with the same question.

⁴ Dapper, Olfert. Naukeurige Beschrijvinge der Afrikaensche Gewesten… (Meurs, 1668)Google Scholar. A little of Dapper's information appears in two earlier works: Janssonius, J., Nouvel Atlas (Amsterdam, 1641)Google Scholar; Vienna, Österreichische Nationalbibliothek, ‘Atlas Blaeuw’.

⁵ Hair, ‘Inventory, I’; idem, ‘An early seventeenth-century vocabulary of Vai’, African Studies, xxiii, iii–iv (1964), 129–39.

⁶ The Vai spoken in the seventeenth century was not, of course, identical to the Vai spoken today. It was probably closer to Manding than it is now.

⁷ Alternatively, Vai may originally have been the language of the Karou, who passed it on to the Puy, Vey and Kquoja. The weakness of this hypothesis is that the people up-country who spoke a language related to Kquoja (i.e. Vai) were called ‘up-country Kquoja’, not ‘up-country Karou’. For further discussion of the identity of these groups see Jones, ‘Galinhas’, 32–3.

⁸ Dapper, , Beschrijvinge, 383.Google Scholar

⁹ Ibid. 382–3.

¹⁰ Bird, Charles S., ‘The development of Mandekan (Manding): a study of the role of extra-linguistic factors in linguistic change’, in Dalby, David, ed., Language and History in Africa (London, 1970), 146–8, 153–8Google Scholar. Bird uses the term ‘Mandekan’ for what I prefer to call Manding (the group of dialects which includes Mandinka, Maninka, Dyula and Bamana).

¹¹ According to Welmers, ‘it is beyond question that the Kono(–Vai) divergence from Manding took place before Maninka was recognizable as a dialect distinct from Bambara and many others’ (personal communication, 12 March 1980). He also argues, contradicting Bird, that ‘Vai and Maninka are hardly more closely related to each other than are English and German’.

¹² Bird, , ‘Mandekan’, 152–3.Google Scholar

¹³ Dalby, David, ‘The Mel languages: a reclassification of Southern “West Atlantic”’, African Language Studies, VI (1965), 12.Google Scholar

¹⁴ Personal communication, 12 March 1980.

¹⁵ Turay, A. K., ‘Language contact: Mende and Temne – a case study’, Africana Marburgensia, xi, i (1978), 55–73Google Scholar; idem, ‘Loanwords in Temne: a study of the sources and processes of lexical borrowing in a Sierra Leonean language’, Ph.D. dissertation, University of London, 1971.

¹⁶ Sigismund Koelle, W., Polyglotta Africana (London, 1854Google Scholar; reprint ed., with introduction by P. E. H. Hair and David Dalby, Graz, 1963).

¹⁷ Pereira, Duarte Pacheco, Esmeraldo de Situ Orbis (Côte Occidentale d'Afrique de Sud Marocain an Gabon), ed. Mauny, Raymond (Bissau, 1956), 94–8Google Scholar; Fernandes, Valentim, Description de la Côte Occidentale d'Afrique (Sénégal au Cap de Monte, Archipels ), ed. Monod, T., da Mota, A. Teixeira and Mauny, R. (Bissau, 1951), 80.Google Scholar

¹⁸ The town Bop, mentioned by Fernandes, may have been Gbap, on the Bum Kittam. Pacheco Pereira mentions a town called Quynamo which, judging by his description, was probably on the navigable part of the Sewa or Waanje. Both places apparently belonged to the ‘Bolões’. See Jones, , ‘Galinhas’, 45–6.Google Scholar

¹⁹ Pereira, Pacheco, Esmeraldo, 98.Google Scholar

²⁰ Holsoe, Svend E., ‘The cassava-leaf people: an ethnohistorical study of the Vai people, with particular emphasis on the Tewo chiefdom’, Ph.D. dissertation, Boston University (1967), pp. 60–8.Google Scholar

²¹ Hair, P. E. H., ‘Some minor sources for Guinea, 1519–1559: Enciso and Alfonce/ Fonteneau’, History in Africa, iii (1976), 30–1.Google Scholar

²² Idem, ‘Inventory I’, 49.

²³ Swindell, Ken, ‘Mineral deposits’, in Clarke, John I., ed., Sierra Leone in Maps (London, 1966), 90–1Google Scholar; Jones, A. E. Nyema, ‘Mineral resources’, in Gnielinski, Stefan von, ed., Liberia in Maps (London, 1972), 86–7.Google Scholar

²⁴ Pacheco Pereira seems to imply that gold from Coya reached the coast not only via the Ryo dos Monos but also via the Ryo das Palmas (probably the Bum Kittam): Esmeraldo, 96. Certainly by the 1520s the Portuguese were obtaining gold mainly from the latter: Magalhães-Godinho, Vitorino, L'économie de /'empire portugais aux XV^e et XVI^e siècles (Paris, 1969), 209.Google Scholar

²⁵ A further complication is the existence of another Koya chiefdom, on the lower part of the Rokel River. This seems too far north to fit Pacheco Pereira's description; but the name may have been brought up the coast by the Kquoja/Mani invaders in the sixteenth century. For an early reference to this Koya see Schlenker, C. F., A Collection of Temne Traditions, Fables and Proverbs… iv (London, 1861Google Scholar; reprint, Nendeln, 1970), 2, 4.

²⁶ Person, Yves, ‘Les Kissi et leurs statuettes de pierre dans le cadre de l'histoire ouest-africaine’, Bulletin de l'I.F.A.N., xxiiiB (1961), 19.Google Scholar

²⁷ Idem, ‘ Ethnic movements and acculturation in Upper Guinea since the fifteenth century’, African Historical Studies, iv (1971), 681.

²⁸ Pereira, Pacheco, Esmeraldo, 184Google Scholar (note by the editor, R. Mauny ).

²⁹ Birmingham, Church Missionary Society Archives, C/A1/0 135, Koelle, S., ‘Account of a journey into the Vei-country’ (1850–1851)Google Scholar; Creswick, H. C., ‘Life amongst the Veys’, Transactions of the Ethnological Society, (n.s.) vi (1868), 354–7CrossRefGoogle Scholar; George, Ellis, W., Negro Culture in West Africa (New York, 1914), 27–8.Google Scholar

³⁰ Brun, Samuel, Schiffarten in Etliche Newe Länder und Insulen (Basel, 1624), 51Google Scholar; Donelha, Andre, Descrição da Serra Leoa e dos Rios de Guiné do Cabo Verde, ed. da Mota, Avelino Teixeira, with notes and English translation by Hair, P. E. H. (Lisbon, 1977), 1041Google Scholar; de Laet, Joannes, Iaerlyck Verhael van de Verrichtinghen der Geoctroyeerde West-Indische Compagnie…, 4 vols. (The Hague, 1931–1937), 1, 104Google Scholar; Dapper, , Beschrijvinge, 379, 384Google Scholar; Villault, N., Relation des Costes d'Afrique (Paris, 1669), 103.Google Scholar

³¹ PRO, CO 267/467/36, King-Harman 23 December 1903 to Secretary of State; Massaquoi, Momolu, ‘The Vai people and their writings’, Journal of the African Society, X (1911), 459–66Google Scholar; Klingenheben, August, ‘Vai-Texte’, Zeitschrift für Eingeborenen Sprachen, xvi (1926), 111–31Google Scholar; Massaquoi-Fahnbulleh, Fatima, ‘Autobiography’ (typescript, 1940)Google Scholar; Jangaba, S., Johnson, M., Traditional History and Folklore of the Vai Tribe (Monrovia, n.d., ?1954), 1–11Google Scholar; Morrison, C. R., ‘The Vai-speaking people of the Southern Province’, Sierra Leone Studies, xxii (1939), 104–7Google Scholar; Holsoe, , ‘Cassava-leaf people’, 62Google Scholar; interview with Braima Vawa Kemokai (Malema), 23 December 1977.

³² ‘Jomani’ may signify ‘Dyomande’, an alternative name for the Kamara clan.

³³ It is worth quoting Professor Welmers's comments on the word said to have been uttered by Kamara 11fai: ‘I had heard it in oral citation from two speakers of Vai. It is an “ideophone”, with a characteristically exaggeratedly long vowel…with an initial v and with low tone: vààài. The present name of the language, on the other hand, has a short a and a high tone: váí… Early records up to and including Koelle suggest that the tribe and language name was originally v⋯í, with a different vowel. My Vai informant, Father C. K. Kandakai, who must be about 60 years old now, recognized v⋯í as a form he had heard in earlier years. Apparently the vowel a in the name is the result of an Americo-Liberian distortion; and it surely has nothing to do with the ideophone said to mean “bravely”. The citation of the ideophone as an etymology appears to date from a time when the proposer could handle written materials; he knew no way of distinguishing, in writing, what I have distinguished above’ (personal communication, 12 March 1980). In this connexion it should also be noted that Dapper gives the name of one of the Vai-speaking groups in the seventeenth century as Vey: Dapper, , Beschrijvinge, 384.Google Scholar

³⁴ The symbol of the hunter-founder in Mende histories is the subject of a forthcoming article by Matthew Hill.

³⁵ Kromah, K. Senesee, A Brief Historic Account of the Kromah and Massaquoi Houses of the Vai Tribe in Grand Cape Mount County (Bobalor, 1961).Google Scholar

³⁶ Binger, L. G., Du Niger au Golfe de Guinée, 2 vols. (Paris, 1892), 11, 151.Google Scholar

³⁷ Delafosse, Maurice, ‘Les Vai, leur langue et leur système d'écriture’, l'Anthropologie, x (1899), 132Google Scholar; idem, ‘Un état negre: la république de Liberia‘, Bulletin du Comité de l'Afrique Française, 1900, Supplément, 191. Another author, relying apparently on traditions in the Bondoukou area, states that the Huela (closely related to the Ligbi) ‘pushed on’ from Bondoukou to the coast, ‘where they became the Vai’. In other words, the Vai reached the sea from the North East, not the North West (Tauxier, L., Le Noir de Bondoukou (Paris, 1921), 65–6)Google Scholar. There are still supposed to be traditions in Wenchi and Demisa (east of Bondoukou) supporting the theory that the Vai came from there (Goody, Jack, ‘The Mande and the Akan heartland’, in Vansina, Jan, Mauny, Raymond and Thomas, L. V., The Historian in Tropical Africa (London, 1964), 195).Google Scholar

³⁸ Yves Person, ‘En quête d'une chronologie ivoirienne’, in Vansina, , Mauny, and Thomas, , Historian, 322–36, pp. 328, 336Google Scholar; idem, ‘Ethnic movements and acculturation in Upper Guinea since the fifteenth century’, African Historical Studies, iv (1971), 676.

³⁹ Dalby, David, ‘Distribution and nomenclature of the Manding people and their language’, in Hodge, Carleton E., ed., Papers on the Manding (Bloomington, 1971), 6.Google Scholar

⁴⁰ According to Welmers, Manding appears to have undergone a change from k to s before high front vowels, while Kono, Vai, Ligbi and Huela/Numu have not shared in this change (personal communication, 12 March 1980).

⁴¹ Person, Yves, ‘Tradition orale et chronologie’, Cahiers d'études africaines, vii (1962), 470–1Google Scholar; Ibid. ‘En quetê’, 326; Ibid. ‘Ethnic movements’, 676. In fact, ‘Galinhas’, was not used as an ethnonym until the twentieth century.

⁴² Ibid. ‘Tradition orale’, 469–74.

⁴³ Ibid. 463.

⁴⁴ Perhaps it is unfair to compare Person's approach with Alex Haley's search for ‘Roots’ in Jufure; but the general idea is the same.

⁴⁵ Person, , ‘Ethnic movements’, 674Google Scholar. Cf. Hair, P. E. H., ‘Ethnolinguistic continuity on the Guinea coast’, J. Afr Hist, viii, ii (1967), 266.Google Scholar

⁴⁶ Person, , ‘Kisi’, 16–17.Google Scholar

⁴⁷ Dalby, David, ‘Mel languages in the Polyglotta Africana. Part II: Bullom, Kissi and Gola’, Sierra Leone Language Review, v (1966), 151.Google ScholarMa- is used for fluids in a large number of ‘ Niger-Congo’ languages (many of them in Central Africa) and in a few others, such as Hausa. Rivers with the Ma- prefix are found also in northwestern Liberia and southeastern Guinee; but since these countries do not have maps on the scale 1:50,000, they have not been included in Fig. 4.

⁴⁸ Hill, Matthew H., ‘Speculations on linguistic and cultural history in Sierra Leone’, paper presented at the Conference on Manding Studies, S.O.A.S. (London, 1972), 1–2.Google Scholar Dr Hill has informed me that his recent examination of pottery collected by Creighton Gabel in the Vai area yields no support for his proposed identification of Kono-Vai with ‘Sefadu-Tankoro tradition’ pottery (personal communication, 6 March 1979), There is, of course, no need to assume that the ancestors of the Kono and Vai used a single type of pottery by which they could easily be identified.

⁴⁹ Dalby, David, ‘The extinct language of Dama’, Sierra Leone Language Review, 11 (1963) 50–4.Google Scholar

⁵⁰ Hill, Matthew H., ‘Ceramic seriation of archaeological sites in Sierra Leone, West Africa’, Ph.D. dissertation, Southern Illinois University (1970), 69–71Google Scholar; Hair, , ‘ Inventory I’ S3–7IGoogle ScholarDapper, , Beschrijvinge, 386.Google Scholar

⁵¹ The ‘Hondoish’ peoples called monkey bridges Jenge-jenge, because they wobbled (Dapper, , Beschrijvinge, 476Google Scholar). This resembles the Mende jεηjεη, meaning ‘with rhythmical up and down movement of the body’, though it might conceivably be related to the Manding dyingi, ‘to slant’: Innes, Gordon, A Mende-English Dictionary (Cambridge, 1969), 35Google Scholar; Delafosse, Maurice, La langue Mandingue et ses dialectes, 2 vols. (Paris, 1955), II, 164Google Scholar.‘ Hondo’ may have been a misreading of‘ Kondo’, the name of a nineteenth-century confederation of tribes around Bopolu (Holsoe, Svend E., ‘The Condo confederation in Western Liberia’, Liberian Historical Review, iii, i (1966), 5Google Scholar). But it need not follow that Hondo occupied the same territory as Kondo did in the nineteenth century, nor that it consisted of the same tribal mixture. What matters here is that Hondo included territory north of Cape Mount and its language was not closely related to Vai.

⁵² Hill, , ‘Ceramic seriation’, 72Google Scholar. Cf. Ibid. ‘Towards a cultural sequence for southern Sierra Leone’, Africana Research Bulletin, 1, ii (1971), 9.

⁵³ Dapper, , Beschrijvinge, 386, 406.Google Scholar

⁵⁴ Hill, , ‘Speculations’, 4Google Scholar; Dapper, , Beschrijvinge, 476Google Scholar. The Konde-kquoja too were Hondoish and yet spoke a language similar to Kquoja.

⁵⁵ Hill, , ‘Ceramic seriation’, 59–63.Google Scholar

⁵⁶ Dyen, Isidore, ‘Language distribution and migration theory’, Language, xxxii (1956), 611–26.CrossRefGoogle Scholar The first of these principles may be compared with the botanical notion of an ‘agricultural cradle’; cf. Portères, Roland, ‘Berceaux agricoles primaires sur le continent africain’, J. Afr. Hist., Ill, ii (1962), 195–210, p. 197.CrossRefGoogle Scholar

⁵⁷ Hill, , ‘Ceramic seriation’, 60.Google Scholar

⁵⁸ Sapir, , ‘West Atlantic’, 48–9.Google Scholar Links between the Krim, Gola and Kisi are stressed in some traditions: interview with Braima Sowa (Bandajuma), 1975; Joseph J. Tucker, ‘History narrated to me by my grandpa, father and some elders’, manuscripts in the possession of S. S. Minah (Boma Sakrim).

⁵⁹ Welmers, , ‘Mande languages’, 18, 24Google Scholar (with the caveat: ‘Whatever the absolute validity of such figures may be, I believe they clothe our Mande skeleton with flesh that is more than illusory’). Dalby (‘Mel’, 6) suggests that the proto-Mel ‘may well be of similar antiquity to the proto-Bantu’, but refrains from giving any date: Hill's deduction that this means 2,500 years is unwarranted.

⁶⁰ Bird, , ‘Mandekan’, 150.Google Scholar

⁶¹ Rodney, Walter, A History of the Upper Guinea Coast, 1545 to 1800 (Oxford, 1970), 64Google Scholar; Hair, P. E. H., ‘From language to culture: some problems in the systematic analysis of the ethnohistorical records of the Sierra Leone region’, in Moss, Rowland P. and Rathbone, R. J. A. R., The Population Factor in African Studies (London, 1974), 79.Google Scholar

⁶² Atherton, John H., ‘Excavations at Kamabai and Yagala rock shelters, Sierra Leone’, West African Journal of Archaeology, 11 (1972), 67.Google Scholar For early European information, see BJones, , ‘Galinhas’, 79–81.Google Scholar

⁶³ BHill, , ‘Speculations’, 8–11.Google Scholar

⁶⁴ See Jones, , ‘Galinhas’, 464–70Google Scholar and the articles by Holsoe and MacCormack in Suzanne Miers and Igor Kopytoff, , eds., Slavery in Africa: Historical and Anthropological Perspectives (Madison, 1977).Google Scholar

⁶⁵ Hakluyt, Richard, The Principall Navigations, Voiages and Discoveries of the English Nation (London1589; repr. 1965), 526–8Google Scholar André Donelha, , Descrição da Serra Leoa e dos Rios de Guiné do Cabo Verde, ed. Avelino Teixeira da Mota, with notes and English translation by Hair, P. E. H. (Lisbon, 1977), 12Google Scholar; Alvares, Manuel, ‘Descrição geográ;phica da Provincia da Serra Leoa’ (Lisbon:Biblioteca da Sociedade de Geografia)Google Scholar; d'Almada, André Alvares, Tradado breve dos Rios de Guiné do Cabo Verde (1594)Google Scholar, reprinted in Brasio, Antonio, Monumenta Missionária Africana: Africa Occidental, series 2, vol. 3 (Lisbon, 1964), 229–378, pp. 358–75Google Scholar. I am grateful to Paul Hair for allowing me to consult his translations of the last two sources.

⁶⁶ Rodney, , History, 43Google Scholar; Miller, Joseph P., ‘Requiem for the “Jaga”’, Cahiers d'études africaines, xiii, 1 (49) (1973), 121–49, pp. 122–5.CrossRefGoogle Scholar

⁶⁷ Hair, , ‘Inventory II’, 226 and notes 13–15.Google Scholar

⁶⁸ For evidence of the use of Vai as a trade language in the nineteenth century, see Anderson, Benjamin, Narrative of a Journey to Musardu, the Capital of the Western Mandingoes, together with Narrative of the Expedition Despatched to Musahdu by the Liberian Government, 2nd ed., with introduction by Fisher, Humphrey (London, 1971), 39Google Scholar; Hartert, Heinrich, ‘Die Veys’, Globus, liii (1888), 236–7.Google Scholar

⁶⁹ An interesting study could be made of the diseases prevalent among the Vai in the early seventeenth century, as described in Dapper, , Beschrijvinge, 400–1.Google Scholar

⁷⁰ The growth of European trade in the fifteenth and sixteenth centuries may have helped to reorientate the Vai; but one might equally expect it to have had the opposite effect of consolidating long-distance trade routes to the savannah.

⁷¹ The only published research, as far as I am aware, is Gabel, Creighton, Borden, Robert and White, Susan, ‘Preliminary report on an archaeological survey of Liberia’, Liberian Studies Journal, v, ii (1972–1974), 95–6.Google Scholar