A reference to “Akan speakers (generally called Minas or Coromantees in the Americas)” led me via Google Books to Kwasi Konadu’s The Akan Diaspora in the Americas, which, after establishing to the author’s satisfaction that “Mina” (a term used primarily in non-English-language sources) does refer specifically to Akan speakers, continues with the following extremely interesting discussion (pp. 13 ff.):

The Mina experience thus raises some critical questions about African social formation and cultural transformation in the Americas. Did Africans use language, “religious” affiliation (e.g., as adherents to African and African-based spiritualities in the Americas, Islam, or Christianity), or the structures of African polities as remembered from Africa, or did they use all three in varying degrees and as principles by which to organize themselves? If the mechanism of organization was primarily language, did cultural groups identify themselves and others by the principal and perhaps mutually intelligible languages they spoke? What might have been the decision-making process of bi- or multilingual speakers from contiguous areas and those accessible by land and water? Africans may have identified with localized or broader polities in West Africa as a source of security and thus would have given their loyalty to those bases of social unity, and this would have been true for centralized Akan polities. However, religious affiliations via Islam or Christianity would have been meaningless for most Akan, who were non-Christian and non-Islamic and had been that way for centuries. …

The topography and commercial networks of West Africa, as well as the roads and waterways that facilitated those networks, also assisted the convergences of cultural traditions through either the volition of the groups involved or political domination. As a result of either process, a kind of multilingualism and multiculturalism emerged in a region where a high degree of diversity or ethnicity has been greatly exaggerated. It has been easy to imagine African diasporic groups as “ethnies” or “nations” as broadened expressions of “ethnicity,” but these approaches have not fully grappled with the issues of African language competency among scholars who do diasporic and “Atlantic” history and the locating of “Africa” and African lives in European scribal sources. These sources still confound the process of locating African identities, and they continue to shape the perspective of scholars who remain true to their perceptions of “nations.” The “nation” (nación, nação; Spanish and Portuguese, respectively) and “country” (terre, French) of Africans inventoried in neo-European documents were based less on Europeans’ perceptions of African polities and more on language (or at least what they believed were a plethora of indiscriminate languages). Thus, these “nations” and “countries” were early inventions that anticipated the concept of “ethnicity,” a colonial creation based largely upon the European idea of a nation and built upon the early process of determining African language or “ethnolinguistic” groups as found in the ethnographies and vocabulary lists compiled by seventeenth-to-nineteenth-century European missionaries. Though commerce was the common denominator between Africa and Europe of the late fifteenth and nineteenth centuries, very few European merchants were and even fewer scholars of “diaspora” are fluent in African history and languages. Scholars have thus grounded [...] their works in the linguistic knowledge gleaned from European missionary ethnographies and lexicons. If 537 million (or 82.6 percent) of 650 million Africans of non-Arab and non-European origin used no more than ten to twelve root African languages—as bi- and multilingual speakers, based upon a mutual intelligibility of at least 85 percent—as of 2004, we cannot start and end with Joseph Greenb[e]rg’s classification or a missionary’s catalog of African languages in order to clarify the linguistic or “ethnic” map of the African past. Rather than debate the extent of African ethnicities in the Americas—a discourse that privileges the “nation” and “ethnic” notions such as “Mina” and “Coromantee”—perhaps we should start with the Africans’ core understandings and identities in Africa and then use both as a compass to chart their shapes in the Americas.

It’s a shame that so much scholarship on non-European parts of the world, particularly Africa, has been done primarily via European languages, and it’s great that scholars like Konadu are remedying the situation and providing fresh perspectives.
(Can I just say, by the way, how much I hate the term “ethnie,” borrowed from French by Anthony D. Smith in his The Ethnic Origins of Nations? What’s wrong with good old ethnos, if “ethnic group” is too long for you?)


  1. Garrigus Carraig says:

    I will be sure to tell sometime commenter komfo,amonan about this. Thank you!

  2. David Eddyshaw says:

    “If 537 million (or 82.6 percent) of 650 million Africans of non-Arab and non-European origin used no more than ten to twelve root African languages”
    Seems pretty dubious, though I suppose it could be made forcibly trueish by interpreting “root” language as broadly as you like. (Chadic? Bantu?)
    The south of Ghana is unusually homogeneous linguistically, which may give speakers of Twi/Fante/Akwapem (one of those languages without a traditional name for the whole dialect complex, for which the name “Akan” has been made up) a rather odd idea of Africa’s linguistic variety.
    Where I lived in Bawku, in the far north of Ghana, on the other hand, I crossed a linguistic frontier from Kusaal to Hausa whenever I walked into town, and many of my colleagues could speak half a dozen local languages fluently, some of which were no more “mutually intelligible” in themselves than Turkish and English.
    What certainly is true is that there are a lot of African lingua francas (necessarily so, in view of all this language diversity.)

  3. Yeah, I definitely got the feeling that his analysis was Akan-centric, but that’s perfectly natural. Bring on the Mande-centric, Fula-centric, and other analyses! Let a hundred language-family flowers blossom!

  4. J.W. Brewer says:

    To some extent scholars may be prisoners of the existing documentary record. Once you get past the last century or two, which languages existed in written form where documents have survived? Wikipedia indicates that some coastal West African languages relevant to the cross-Atlantic diaspora (e.g. Wolof and Yoruba) had a pre-colonial tradition of literacy in scripts derived from the Arabic alphabet. But how old are the oldest surviving texts in those languages and how numerous are they? Further down the coast, there are apparently surviving (Latin-alphabet) texts in Kikongo going back at least to the 17th century, but it sounds like those are mostly Christian catechisms and the like.

  5. One thing my research has indicated in extremely multilingual Papua New Guinea, where linguistic exogamy is almost mandatory to prevent inbreeding, is that language serves as much as a marker of difference as a marker of common identity–despite the implication of the Tok Pisin term wantok (< ‘one talk’), someone who speaks the same language. Multilingual kinship networks have been much more important than people who happen to speak a common language, and so has the ability to switch language affiliations when necessary for survival.
    Of course, PNG is one of the most linguistically diverse refuges on earth, like the Amazon, the Caucasus, parts of India, Indonesia, and the Mandara mountains of Nigeria and Cameroon. On the last, see Scott MacEachern’s chapter on “Setting the Boundaries: Linguistics, Ethnicity, Colonialism, and Archaeology South of Lake Chad,” in Archaeology, Language, and History: Essays on Culture and Ethnicity, ed. by John Edward Terrell (Bergin & Garvey, 2001). I have a chapter in the same volume on “Language, Culture, and Community Boundaries around the Huon Gulf of New Guinea” (the three have rarely coincided).

  6. Sounds like a great book, but I knew before I looked it up that I wasn’t going to be able to afford it. Price: $119.95; 7 used from $36.00. Sigh. That’s what libraries are for, I guess…

  7. marie-lucie says:

    At a conference I recently attended, one of the major academic publishers was advertising a book for 500 euros. It was big, but still …

  8. “(Can I just say, by the way, how much I hate the term “ethnie,” borrowed from French by Anthony D. Smith in his The Ethnic Origins of Nations? What’s wrong with good old ethnos, if “ethnic group” is too long for you?)”
    Can I just say that ‘thede’ is better than either? This is the same argument my mother and I had over “arugula”.

  9. If it had been used in the last 500 years, maybe.

  10. marie-lucie says:

    At a conference I recently attended, one of the major academic publishers was advertising a book for 500 euros. It was big, but still …

  11. David Eddyshaw says:

    @J W Brewer
    I think there are Hausa texts in the Ajami (Arabic-derived) writing going back to the seventeenth century.
    I was astonished when I first went to West Africa to find out just how much actual *history* there is for the area, from long before the Europeans put their noses in along the coast, because of Muslims literate in Arabic.
    Wikipedia says there are Swahili texts in Arabic letters from the early eighteenth century.

  12. “from long before the Europeans put their noses in along the coast… Swahili texts in Arabic letters from the early eighteenth century.” The Portugese were on the coast in the 15th century.

  13. David Eddyshaw says:

    Yup. And the history goes back before the Portuguese. Muslims literate in Arabic.

  14. David Eddyshaw says:

    (My remark about history was supposed to be parenthetic, rather than to do with the Hausa and Swahili texts.)

  15. J.W. Brewer says:

    My question I suppose should have been more focused on surviving early texts in languages such as the “Akan” grouping, i.e. those spoken in the subsets of Africa relevant to the “diaspora” in the New World, which has to do with the dynamics of the trans-Atlantic slave trade, which of course varied somewhat over time and also over place (e.g. primary source regions for those who involuntarily ended up in Brazil not identical to those who involuntarily ended up in the British West Indies). Swahili was on the wrong side of the continent (although it was the lingua franca of the separate slave trade there for the Arab market); I think of Hausa as being spoken pretty far inland, but I don’t know how far away from the coast those unfortunate enough to be shipped off may have been taken from. Thus I don’t have a sense of whether any surviving early texts in Hausa might provide useful insight into the original ethnic/linguistic/cultural groupings/identities of the involuntary West African immigrants to the New World, which is what I take Konadu to be interested in. If the only or primary texts we have access to are those generated by Europeans in European languages, we can accept the point about how those may be distorted/biased/unreliable in various ways, but if we don’t have a different POV available from surviving texts of indigenous origin from the same time period (or e.g. texts in Arabic which if not of indigenous origin at least have a different POV, although wikipedia indicates Ibn Battuta didn’t get anywhere near the coast on the western side of the continent), then the question becomes what’s the historian to use instead? Or is this the sort of move where one clears the ground by pointing out possible inaccuracies/limitations in the only documentary record there is, and thus justifies a much more speculative/conjectural approach in finding/interpreting evidence for ones rival account of what really happened?

  16. David Eddyshaw says:

    As far as I know there isn’t any documentation of Akan before the Europeans came.
    The homeland of Hausa certainly is inland, in what’s now Northern Nigeria and the neighbouring bits of Niger. It’s spread far beyond its original homeland, partly as the first language of a large ethnic Hausa diaspora who are found in commercial quarters of cities and towns all over the place, and partly as a major West African lingua franca. In Nigeria it’s very widely used by Northerners of all sorts of ethnic groups, notably by the Fulani, the dominant group in the old Sokoto Caliphate, who mostly are first language Hausa speakers now. In Ghana it was once the major lingua franca of the north but is tending to be replaced by Akan and English nowadays, though it’s still widely used. It’s quite strongly associated with Islam; if you see somebody dressed as a Muslim in Ghana it’s reasonable to address him in Hausa as a first guess.
    About the languages of transported slaves: the Ashanti were great slave traders themselves, and certainly sent along a number of war captives from their battles with their northern neighbours. I remember a northern Ghanaian firend saying that he though black Americans tended to look like northerners rather than southerners to him. But the unequivocal traces of African languages that made it to the new world seem as far as I know to be mostly southern languages like Yoruba.
    Mind you, there is a lot of shared culture anyway across West African languages. The trickster rabbit who gets into trouble all the time due to his own overelaborate schemes turns up to my knowledge in Kusaal and Hausa. Hausa actually has the Tar Baby story (in a somewhat raunchier version.) I’ve seen this motif in black American folklore attributed to American Indian sources, but that hardly squares with Brer Rabbit’s vigorous existence in West Africa. What’s interesting in this context is that Rabbit seems to be the hero of Savanna-zone stories, rather than coastal (the Ashanti have a trickster spider instead …)

  17. From what I understand, the Spider (Anansi) stories were, er, respeciesized with Rabbit once they got to North America, because that was the local trickster animal. In the islands, this didn’t happen, and so the trickster remained a spider. I never knew that there were already rabbit versions of the stories further inland. Still, it’s not surprising that rabbits should independently be tricksters in different parts of the world. Consider Peter Rabbit.

  18. If it had been used in the last 500 years, maybe.
    Didn’t stop Tolkien from reviving mathom (1275), dwimmerlaik (1400), attercop (1658), or frith (1584).

  19. J.W. Brewer says:

    David Eddyshaw obviously knows much more about West Africa than I do; as noted above there were some cryptic suggestions on wikipedia that Arabic-derived-script literacy existed fairly early for Wolof and Yoruba (which are languages unquestionably relevant to the cross-Atlantic diaspora), but no indication as to the extent to which such texts have survived and if so from how many centures back and whether their content is useful for this sort of historical inquiry (i.e. if they were primarily vernacular Muslim catechism-type works, they wouldn’t provide much more insight than the Kikongo catechisms produced by Portuguese missionaries other than on purely linguistic questions).

  20. David Eddyshaw says:

    It’s just occurred to me that the fact that traces of Savanna-zone languages haven’t been very forthcoming in studies of the language of transported slaves could be a straightforward consequence of the fact that these languages are often (Hausa excepted) more poorly documented to this day than the forest-zone and coastal languages; I doubt whether there are all that many people in a position to recognise such traces anyhow.
    Moreover (generalising unwarrantably from Ghana, the country I know most about) there may be more language diversity in the Savanna zone than further south, so that captives from these areas would have been mixed together with speakers of different languages even before leaving Africa, which might result in less transmission of their language to the Americas.
    People must have done genetic studies on the African origins of black Americans. Don’t know if any of it is sufficiently soundly based or detailed enough to shed any light on questions like these. There certainly are notable physical differences between typical northern and southern Ghanaians (which even I could pick up on, and which are very evident to local people) so I imagine the thing is possible in principle.

  21. J.W. Brewer says:

    DE: there are now a number of entrepeneurs in the U.S. marketing home-DNA-test kits to African-Americans curious about their ancestry (which often lacks a documentary record all the way back to arrival in the New World, much less before that), promising to do a genome analysis telling them where in Africa (by “tribe” and/or modern nation-state) their ancestors are statistically likely to have originated. is an example of such an operation. I don’t know whether the accuracy of the results lives up to the marketing. Presumably the geographical-origin database they’re testing against is based on recently-collected DNA of current inhabitants of Africa, and who knows if they’ve adjusted adequately for possible substantial intra-African population movements subsequent to the 17th/18th century time of involuntary Atlantic crossing (you’d hope they’d at least be smart enough to exclude people from Liberia/Sierra Leone whose ancestors came back across the Atlantic rather than having been in Africa all along . . .). Whether even within West Africa today a genetic analysis will match you up with a reasonably high likelihood of being a speaker of a specific language or bearer of a particular ethnocultural identity (as opposed to being in a particular geographical range with a bunch of different languages / ethnocultural groups of differing degrees of relationship to each other) I don’t know.

  22. David Eddyshaw says:

    @John Cowan:
    I didn’t know there were trickster spider stories in the West Indies. Thanks.
    I’ve just discovered from Wikipedia that Anansi too has his Tar Baby story.
    What an interesting world this is.
    Peter Rabbit is probably the ultimate source …
    On the ultimate interrelatedness of all things:
    I came across in a book of Kusaal folk stories a tale which was plainly the same as the Pardoner’s Tale from Canterbury Tales. I believe it’s ultimately a Jataka story that made its way through Pahlavi to Syriac to Arabic (and thence to West African) and to Hebrew and then Latin and hence to Western Europe ..

  23. What an interesting world this is.

  24. Denizens of the Hattery (Hatterites?) with a literary bent may be interested in the House of Anansi.

  25. David Eddyshaw: I know a thing or two about creole languages, and it is quite clear that languages of the Savanna zone played little if any role in the genesis of creole languages of the Americas: African-origin vocabulary in these languages derives almost exclusively from coastal languages, for instance.
    Some more basic contributions to the linguistic make-up of creoles do give us a glimpse as to what languages were prestigious/dominant at the time these languages emerged.
    For example, nearly all Atlantic English-based creoles have an Igbo (South-eastern Nigeria)-origin second person plural pronoun, realized as UNU or UNA: the most excentric English-based creole of the Americas, Saramaccan (spoken in Inland Suriname) has basic interrogatives (“who, “what”) deriving from Fon.
    Other examples: Palenquero (Spanish-based creole, spoken in Colombia) has a number of elements of Kikongo origin, notably a pluralizing prefix MA-. Papiamentu, a Spanish (originally Portuguese) creole, has a third-person plural pronoun /nan/ that seems to derive from Wolof. Finally, Berbice Dutch Creole, a now-extinct language once spoken in inland Guyana, had a huge borrowed element (lexicon and affixes alike) from a coastal Nigerian language, Eastern Ijo.
    The same is true of phonology: some creole languages have African-origin phonemes in some subset of their vocabularies: Haitian Creole, for instance, has nasal /i/ and /u/ as phonemes in a number of words (all of Fon origin).
    As you can see, all of the African languages listed above have a common denominator: all are coastal.

  26. And don’t forget Yoruba (which extends to the coast, though it’s mainly inland).

  27. David Eddyshaw says:

    Thanks, Etienne – interesting. Though I still wonder how likely it would be that words of (say) Dagbani origin would be recognised in American creoles even if there were any, given the paucity of descriptions of the language compared with Igbo and Yoruba and Wolof etc.
    However, I must admit, a priori, it would hardly be surprising if seaborne European slavers ended up largely with hauls of coastal victims.
    I’ve idly wondered in the past where Nigerian Pidgin 2pl “una” came from. Never occurred to me to think of Igbo (not least because I don’t know any Igbo, I guess.)
    I’ve encountered the theory that Nigerian Pidgin and its relatives in West Africa and beyond originated from relexification of an existing pre-European creole which drew its vocabulary from African languages. Is that a viable idea? Nigerian Pidgin always struck me as being a pretty typical West African language in everything except the eerie sense of deja vu inspired by the lexicon.

  28. “However, I must admit, a priori, it would hardly be surprising if seaborne European slavers ended up largely with hauls of coastal victims.”
    Actually it would be surprising, since Europeans bought slaves from a trade that spanned the trade routes, and since they were buying them from those very coastal groups.

  29. David Eddyshaw: granted, Savanna languages are underdescribed compared to coastal languages: but since most African words in creole languages of the Americas have an identified etymology, there can’t be that many words deriving from Savanna (or little-known coastal) languages. I thus doubt any new discoveries could change the picture significantly. It is telling that Hausa and Fulani are both well-described languages, both lingua francas of the West African Savanna, and both equally marginal when it comes to their (lexcial or grammatical) contribution to Creoles of the Americas.
    The origin of Nigerian pidgin and related English-based pidgins and creoles of West Africa is contentious. The theory I find likeliest is that sketched by John McWhorter (see his “Towards a new model of creole genesis” and “The missing Spanish Creoles” for more details): all English-based creoles of the Atlantic area have a common (pidgin) ancestor originally spoken at the Cormantin trading fort, located on the coast of what is today Ghana. This pidgin was transplanted to the Americas with the first generation of African slaves in the West Indies. It then died out in West Africa, but was subsequently re-introduced (at first in Sierra Leone) from the Americas.
    It indeed used to be believed that ALL Atlantic creoles (not just the English-based ones) had a common pidgin Portuguese ancestor, but that theory has been pretty much given up today: it has a kernel of truth to it, though, inasmuch as there does exist a Portuguese element common to all Atlantic English creoles. Which is unsurprising, since Portuguese was the first European language to be brought to coastal West Africa and was widely known and used by the English sailors who followed them.
    David, Jim: what I am talking about is *language*. Even if large numbers of non-coastal individuals ended up enslaved, coastal languages would end up in a dominant position: within a group of slaves of mixed origin who were not shipped to the Americas immediately the language of their coastal captors would quickly acquire a dominant position as the LINGUA FRANCA of the group, so that Akan and Fon and Wolof and other coastal languages would indeed be the dominant languages among the first generation of slaves arriving in the Americas .
    This being said, though…
    I believe Joseph Greenberg published a study on a first-generation African slave in the colonial Southern U.S. who kept his master’s records…in Arabic! I’m not sure we know what the slave’s ethnicity was. I also believe that some early slave revolts in Brazil involved leaders who wrote one another in Arabic (to maintain secrecy), and finally I believe there are reports from colonial Saint-Domingue (AKA Haiti) about slaves using Hausa as a secret language to conspire against their masters. So Islam and Savanna languages did play some role in the slave societies of the Americas.
    (I *might* be able to dig up references on any or all of the above, assuming anyone is interested of course).

  30. I have an Arabic autobiography written by a slave in the U.S., “The Life of Omar Ibn Said, Written by Himself” (1831). Ala Alryyes, who introduces it, says “Omar’s knowledge of Arabic, as revealed in the narrative, sets him apart from other early African-American writers of slave narratives in that Omar was literate before being captured and wrote in a language that most of his enslavers could not read.” Omar was from Futa Toro.

  31. David Eddyshaw says:

    Thanks. I see your point about there not being many words in American creoles that aren’t already spoken for etymologically.
    The idea that English-based creoles might have died out in West Africa and been reintroduced from the Americas is entirely new to me. Very interesting, though very counterintuitive. I was only exposed to Nigerian Pidgin after I’d learnt some Kusaal and Hausa, with the result that I was much more struck by its resemblance to other West African languages than by its resemblance to English. But I suppose that is easy enough to explain even if the language was a sort of Liberian returnee. Mind you, John McWhorter seems to have some pretty farfetched ideas about Celtic influence on English if I remember right ….
    I for one would be very interested in any reports of Hausa as a secret language among Haitian slaves, if you have time to track them down.

  32. David Eddyshaw says:

    Quite right, of course, too, on the question of language as distinct from origin of the speakers. Even if my Ghanaian friend’s intuition was correct, that black Americans are descended more from savanna people than forest or coast people, they still could very well have brought coastal languages with them to the Americas rather than their original languages for the reasons you point out.
    I suppose this is actually one case in point of what Konadu is talking about.

  33. David Eddyshaw:
    West African pidgin Englishes are more “Sierra Leone” returnees than Liberian ones: indeed the varieties of English spoken in Liberia are quite unlike the others that are found in West Africa. And what was originally brought to the Americas from Cormantin was almost certainly a pidgin which never creolized in Africa, so its subsequently dying out in Africa is not very surprising.
    As for the reference you requested, I located it again after slaving TWO MINUTES over a hot keyboard (long live the Internet!): here it is:
    A. M. D’Ans. 1996. “Essai de sociolinguistique historique, à partir d’un témoignage inédit sur l’emploi des langues, notamment africaines, en Haïti, au cours de la guerre de libération et des premières années de l’indépendance”. Published in the journal ÉTUDES CRÉOLES of that year.
    (I don’t have the page number for the article, but the reference to Hausa [the only explicitly identified language in the article] is on page 115. Hope you find this useful).

  34. David Eddyshaw says:


  35. “David, Jim: what I am talking about is *language*. Even if large numbers of non-coastal individuals ended up enslaved, coastal languages would end up in a dominant position: within a group of slaves of mixed origin who were not shipped to the Americas immediately the language of their coastal captors would quickly acquire a dominant position as the LINGUA FRANCA of the group”
    Etienne, I wonder how much the economics of the slave trade would have conduced towards this, or much contact at all.

  36. Jim: one aspect of the slave trade which must be remembered is that slaves were often held captive in the vicinity of the coastal trading posts for months before being shipped to the Americas: under such circumstances a common language would be required, as the newly-captured slaves needed to communicate with one another as well as with their captors: there is some evidence that (in some cases probably pidginized forms of) various coastal African languages were in use around these trading forts, and pidginized forms of European languages as well.

  37. John Cowan says:

    I assume the “debate about arugula” mentioned above has to do with the use of the “foreign” word rather than the “native” word rocket. But while arugula does look less English than rocket, both are ultimately diminutives < Latin eruca ‘colewort’ (which is the Linnaean name of the genus), one < an unspecified Italo-Romance language (not Standard Italian, in which it is rucola), the other < French roquette < older Italian rochetta.

Speak Your Mind