1. Cortés, Vicenta, La esclavitud en Valencia durante el reinado de los Reyes Católicos (1479-1518) (Valencia, 1964).Google Scholar It must be noted that references in the indexes of African providence and anthroponyms do not always tally with references in the text, particularly in spelling; and it is not clear whether the lapses in proofreading are confined to the indexes or affect readings in the text. The same records, but for an earlier period, are analyzed in A. Teixeira da Mota, “A entrada de escravos negros em Valência de 1445 à 1482 e a mudança da via trans-Saariana para a via Atlantica,” forth-coming.

2. The entries (summarized in Spanish but originally in Catalan) relate either to the payment by merchants of a tax on slaves, as recorded by one official, the Mestre Racional; or to the related presentation of slaves to another official, the Bayle, in order to confirm the status of slavery, notionally recorded as deriving from Just War (ibid, 19-20). The latter act was recorded in Presentaciones y Confesiones de Cautivos and individual slaves were therein identified by racial category, provenance, sex, age, value, and name (African and Christian, either or both). When the slave was a non-minor and could speak an Iberian language or an African language for which an interpreter was available, some of this information was obtained on oath; and slaves were further questioned as to the legal condition of their parents, the circumstances in which they were reduced to slavery, and any masters they may have had in Iberia before reaching Valencia, though these latter details were unfortunately only occasionally recorded (ibid, 66-70). The list of Black African slaves analysed in the present paper begins in the 1480s and early 1490s with tax entries, but thereafter becomes one wholly derived from presentations. In theory there should have been matching entries for tax and presentation (cf. items 193, 194) but for various reasons, mainly the existence of gaps in each set of records, this hardly ever occurs, or is at least recognizable. However, since tax payments for large consignments of slaves do not record details of individuals, it may be suspected that there is more doubling of entries than can be recognised.

3. Alonso, Vicenta Cortés, “Procedencia de los esclavos negros en Valencia (1482-1516),” Revista española de antropología americana, 7(1972), 123–51.Google Scholar Provenances are tabulated in her Lista no. 1 but, partly because assumed variants have been conflated, the list is incomplete (for instance, Rio de los Esclavos and Visala are missing). Identifications are based solely on Curtin, P.D., The Atlantic Slave Trade: A Census, (Madison, 1969)Google Scholar, whose identifications are in turn based either on names employed during the later slave trade, especially names derived from Koelle's, S.Polyglotta Africana (listed 291–98)Google Scholar, or on names found among slaves in early Spanish America (listed 97-98). The source for the identification of names found in early Spanish sources is Beltrán, G. Aguirre, La poblacián negra de México, 1519-1810, (Mexico, 1946)Google Scholar, a work notable at the date of publication for its serious attempt to list early slave trade ethnonyms and relate these to the most up-to-date lists of modern ethnonyms then provided by Africanist studies. The influence of Aguirre Beltrán is shown by the failure of later writers to identify correctly names not found in his list. Thus, most significantly, a series of Americanist scholars -- e.g. Bowser, F.P., The African Slave in Colonial Peru, 1524-1650 (Stanford, 1974)Google Scholar tables 40-41 -- confuse the Kokoli of western Guinea with the Kotokoli of central Guinea, the latter being mentioned in Curtin's identification of Koelle's nineteeth-century list. Cortés makes the same mistake (131 and on the same page she misreads Curtin and relates the Ibo to the Temne). Yet the Kokoli appear regularly in primary Portuguese sources on eastern Guinea. Thus the weakness of all the identifications of early ethnonyms deriving from Aguirre Beltrán is their lack of acquaintance with the primary Portuguese sources. These sources were the basis of the identifications presented in Hair, P.E.H., “Ethnolinguistic Continuity on the Guinea Coast,” JAH, 8(1967), 247–68CrossRefGoogle Scholar; “An Ethnolinguistic Inventory of the Upper Guinea Coast Before 1700,” African Language Review, 6(1967), 32–70Google Scholar; “An Ethnolinguistic Inventory of the Lower Guinea Coast,” ibid, 7,(1968), 47-73; 8(1969), 225-256; “Problems in the Analysis of the Ethnohistorical Records of the Sierra Leone Region” in Moss, R.P. and Rathbone, R.J.A., eds., The Population Factor in African Studies (London, 1975), 82–83.Google Scholar

4. In totals for the whole period, the double-counting noted in footnote 2 is probably more than compensated for by gaps in the records for the certain years. For a discussion of the extent and irregular progress of the trade in Black Africans, see Cortés, , “Procedencia,” 137–38Google Scholar and the graphs, 150-51. She lists a total of only 5,132 “negros” in a table which contains arithmetical errors. Black slaves supplied by the Saharan trade were of course to be found in Iberia many centuries before the Portuguese discoveries; but the Valencian records first record a “natural de Gujneua” (native of Guinea) in 1457 (ibid, 127). For the trade in black slaves at Valencia between 1444 and 1482 and the evidence of a transition from the Saharan to the marine route see da Mota, A. Teixeira, “Alguns aspectos da colonizacão e do comércio maritimo dos Portugueses na Africa Ocidental nos séculos XV e XVI,” Anais do Clube Militar Naval, 106(1976), 677–710Google Scholar [reprinted as no. 98 of the série separatas of Centro de Estudos de Cartografia Antiga, Junta de Investigacões Cientificas do Ultramar, Lisbon, 1976] 680-81, 703 n. 12; and the article by the same scholar cited in note 1. The only African ethnonmy found in these earlier Valencian records is “Jaloff” (from 1470), but the female name “Bambada’ (1464) may represent the ethnonym Bambara.

5. On baptismal names see cortés, , “Procedencia,” 143–45.Google Scholar It is not clear what the irregular recording of baptismal names tells us about the regularity or baptism of slaves.

6. This reading is given in cortés, ibid. 128n25 and cited as deriving from item 610 of her Esclavitud, where in fact the text gives only “Bañul, Jalof.” (cf. 130n33). It is thus not clear whether the extended form is a gloss or the original wording. (Esclavitud is defective in that the entries are summarized without an explanation of the principles followed or the citation of a typical origingal entry.) The use of “Jalof” in a very general sense is found in a text on the “cantino” Map of 1502 where it is stated erroneously that Sierra Leone provides slaves, “sine Jelof, some Mandinga, some Cape.”

7. For the identification of localities in Iberia and North Africa (a well as for the identification of non-African personal names) the indexes in cortés' Esclavitud have been used.

8. If the names as recorded simply represent the Valencian clerk's attempt to write down African terms pronounced by an African - which musut have seemed gibberish to the clerk - then he must have spelled the names according to the contemporary orthographic forms for the corresponding sounds, exact or approximate, in Catalan. For guidance as to the sounds represented by late-medieval Catalan orthographic items I am indebted to Sr.J.L. Marfany, of the School of Hispanic Studies, University of Liverpool. However, there is evidence that the record was sometimes influenced by Portuguese orthography, presumably indicating that the clerk was copying a document accompanying the slave from Portugal or from a Portuguese vessel. Thus the term “Çape” (i.e. Sape), which was often carelessly copied without the cedilla in contemporary Portuguese sources, takes this form in the Valencian documents, making the pronounciation “Kape.” Doubts about whether the clerk heard the African term correctly are thus compounded by doubts about which orthography, Portuguese or Catalan, produced the form of the African name in the record.

9. Because of the rule in Spanish that the masculine plural is employed when items of both genders are aggregated, the sex of some slaves with African names cannot be deduced, e.g. “…presenta dos negros, Zamba y Gomba” (item 689). Are both slaves male, or only one, and if so which one? External evidence in this case indicates that the first is a negro, the second a negra. Further, since the editor may not have known from the names that even one of the slaves was male, the masculine plural must cover cases where the sex of both is unknown. However, in Cortés' later study, individuals whom I count as of uncertain sex are confidently identified (e.g. Stona, uncertain in item 749 becomes “el niño” on p. 143). Conceivably there is some indication of sex in the original entry which has been lost in the summary, but since the sex is wrongly stated in certain cases where there is no grammatical ambiguity (e.g. “un negro Sora,” item 1061, becomes “la muchachita Soria,” on p. 144) the sexing of those of uncertain sex may instead be only a slip. My insistence on the uncertainty regarding sex in some instances helps to explain why my totals differ from those of Cortés.

10. For “Zamba” in 1465 see Teixeira da Mota, “Entrada de escravos.” “Cumba,” as the name of a female slave apparently from the Gambia, is recorded in a Portuguese document of 1493 (Brásio, A., Monumenta missionaria africana: Africa ocidental, 1st ser. vol. 1, (Lisbon, 1952), 557–58Google Scholar). In Seville slave documents of the last quarter of the fifteenth century “Conba” occasionally appears as a female name, once attached to a slave from “Mandinga” (personal communication from Teixeira da Mota). “Cumba” and even “Conba” may represent “Kumba,” as “Gomba” almost certainly does, but a Valencia entry of 1470 gives the female name “Çomba” (and the cedilla was more often omitted than added wrongly), which must be a different anthroponym. Modern colloquial Portuguese has a term “cumba,” meaning “strong, determined, provocative, swaggering” and even “bewitching” - adjectives which might be applied to a slave girl. However, Teixeira da Mota tells me that he has never come across “Cumba” as a name or nickname in early Portuguese documents apart from the African instances, which makes it unlikely that the 1493 “Cumba” was other than the African anthroponym “Kumba.” For sources on “Zamba” and “Kumba” see Table II.

11. The Efik names are from Goldie, Hugh, Dictionary of the Efik Language, (Edinburgh, 1874, reprinted 1964)Google Scholar, end of Efik-English section.

12. Houis, M., Les noms individuels chez les mossi (Dakar, 1963) 75.Google Scholar

13. The detailed evidence for the conclusions that follow is in the notes to Table I.

14. Law, Robin, The Oyo empire c.1600-c.1836 (Oxford, 1977), 126.Google Scholar

15. Bowser, , African Slave in Colonial Peru, 279.Google Scholar