CAHG Culture and Lexicon.

Dmitry Pruss sent me the Nature Human Behaviour article “Deep history of cultural and linguistic evolution among Central African hunter-gatherers,” by Cecilia Padilla-Iglesias, Javier Blanco-Portillo, Bogdan Pricop, Alexander G. Ioannidis, Balthasar Bickel, Andrea Manica, Lucio Vinicius, and Andrea Bamberg Migliano (Open access), and the abstract certainly makes it sound of LH interest:

Human evolutionary history in Central Africa reflects a deep history of population connectivity. However, Central African hunter-gatherers (CAHGs) currently speak languages acquired from their neighbouring farmers. Hence it remains unclear which aspects of CAHG cultural diversity results from long-term evolution preceding agriculture and which reflect borrowing from farmers. On the basis of musical instruments, foraging tools, specialized vocabulary and genome-wide data from ten CAHG populations, we reveal evidence of large-scale cultural interconnectivity among CAHGs before and after the Bantu expansion. We also show that the distribution of hunter-gatherer musical instruments correlates with the oldest genomic segments in our sample predating farming. Music-related words are widely shared between western and eastern groups and likely precede the borrowing of Bantu languages. In contrast, subsistence tools are less frequently exchanged and may result from adaptation to local ecologies. We conclude that CAHG material culture and specialized lexicon reflect a long evolutionary history in Central Africa.

Dmitry quoted this bit:

For example, the word ngbídí (see Supplementary Table 13 for variants) denotes a musical bow with two strings that is played exclusively by female CAHGs in the exact same manner by both eastern (Efe) and western (Aka and Baka) groups living over 2,000 km apart and not found in any other population. Another shared word exclusive to CAHGs is ngombi, denoting the harp present in the Baka, Babongo, Aka and Batwa in the west (the latter being at least 400 km away from the other 3 groups). It is also relevant that Aka (Bantu C10 speakers) and Baka (Ubangian speakers) neighbours speak very distinct languages and yet share various unique words denoting exclusively shared musical instruments, such as haka (ankle rattles), bogongo (zither harp), pole.pole (seed whistle), mobio (flute) and mokinda (single-skinned drum). The presence of these instruments and words shared between different CAHG groups but not with their respective Bantu-speaking and Ubangian farming neighbours nor with other farming groups speaking related languages has previously been interpreted as evidence for their descent from a common ancient population.

What do Hatters think?


  1. (Apologies to anyone who saw this post before I remembered to provide a link for the article!)

  2. David Marjanović says

    Sounds good to me, but I won’t have time to read the paper soon.

  3. David Eddyshaw says

    I’m away from home and not in a position to look at this properly. However …

    Their evidence for ongoing cultural exchange among these groups (which seems quite solid – a very interesting thing in itself, incidentally) is at odds with their claim that this is all evidence for an original shared culture. The claim to have distinguished these possibilities by the wonders of statistics alone incites my usual scepticism.

    In this regard, the fact that the supposed conserved words are confined to specific cultural areas is a bug, not a feature. This screams “diffusion”, not “retention.”

    The actual word comparisons excite some suspicion. The data seem to consist of just marking an etymon as present or absent, and the actual words in these (very different)  languages are not given. I can’t find what actual language (if any) the names given to the actual etyma are in. This is like asserting that it is certain that the Kusaal word ligidi “money” is a cultural loan from (proto-)Songhay *nogru: it may be so, but simple assertion is just not good enough – you’ve omitted all the necessary argumentation if you do that. Reliably identifying loanwords is even harder than reliably identifying cognates: this is not something to be handwaved away as unproblematic or trivial. Anyone who thinks otherwise is not a competent judge of the basic materials of the study.

  4. David Eddyshaw says

    If, on rhe other hand, the various words are almost all practically identical (as they seem to imply), that is a more or less conclusive argument against them being a common inheritance from a lost CAHG protolanguage, unless this is to be dated very recently (compare the unequivocal proto-Oti-Volta cognates Kusaal kɔlig “river”, Moba kpenu, Mbelime huonu; Mooré sɛ̃ “sew”, Moba nyal: the Bantu expansion can hardly have been a significantly more recent development than the Oti-Volta expansion.)

    [Actually, I’ve maligned them: they do give more detailed linguistic data in the supplementary materials. Unfortunately I can’t look at it properly on a mobile phone. I don’t think that answers my more general points, though.]

  5. Dmitry Pruss says

    The authors actually mention recent or ongoing travel between ethnic groups, even far-flung,, to participate in ceremonies, as well as a strong connection between musical instruments and ceremonies. It made me rethink my initial wow, like, there must be an ancient substrate, in favor of “cool, this must be a very special recent superstrate!”

  6. Their evidence for ongoing cultural exchange among these groups (which seems quite solid – a very interesting thing in itself, incidentally) is at odds with their claim that this is all evidence for an original shared culture.

    That bothered me as well. But like Dmitry, I find it cool even if it’s not ancient!

  7. David Eddyshaw says

    I agree. In fact, that seems to be the actual point of the paper, a fact which is not altogether clear from the way they have presented it. My points about language change and loanwords would actually support their main thesis, if so.

    A relatively shallow timescale for cultural borrowing would also make sense, given that the various CAHG groups have been losing ground to farmers at an accelerating pace relatively recently. There may well have been a lot more contact between different groups of forest dwellers not too long ago.

  8. David Eddyshaw says

    I like CAHG as a euphemism for “pygmy”, btw. Has potential. I look forward to reading about cahg hippos likewise.

    To be fair, the word is certainly regarded as pejorative by at least some of the peoples in question. And it is in any case objectively unsatisfactory in making a presumption that the particular bodily habitus is necessarily associated with a particular culture, in other words, that all such groups are “the same” in some significant way. (The habitus isn’t associated with a particular language or language family, of course – as this paper rightly makes clear.)

    I hope that the authors of this paper have not simply carried over this presumption from “pygmies” to “CAHG.” Have to look at what groups they specifically name, I suppose. There certainly are forager groups who aren’t “pygmies.” They may well have taken that into account.

  9. Trond Engen says

    My first impression was that the names of the instruments are too disparate to represent one substrate, and I was starting to look forward to coming up with an alternative hypothesis, but it seems that everybody, including the authors of the study, beat me to it.

    I’ll instead suggest that evidence of recent regular intergroup travel for ceremonial purposes may suggest that there until recently also was a shared language, or at least a shared ceremonial language, and there might stil be vestiges of that to uncover.

    And now I’ll read the paper.

  10. Trond Engen says

    I have read the paper, slept on it, and now I’ve looked at it again. It’s a very interesting finding that words for musical instruments are shared between widely disparate CAHG groups but not with the Bantu farmers whose languages they’ve adopted. The genetics, statistics, etc., add little or nothing at all.

    A couple of paragraphs were especially cringeworthy:

    For example, the two groups with the most similar general lexicons were the Aka and the Eastern Batwa (PMI = 0.71), which share no musical instrument words.

    This just shows the uselessness of the lexicostatistical approach. The Aka and Batwa languages are genetically unrelated, and they are spoken on opposite sides of the Congo Basin. Thus, it’s the baseline for accidental similarity.

    This ability of maintaining large-scale social networks throughout the evolutionary history of our hunter-gatherer ancestors might represent a fundamental step in the divergence between hominins and our closest relatives.

    Non sequitur. Long distance networking may well be a human superpower, but hunter-gatherers surrounded by farmers do not represent the primordial state of humankind.

    But I liked this:

    In fact, it has been reported that rituals in which the Aka and Baka use similar musical instruments are also named the same in both populations, such as the ritual to invoke the success of a quest (zɔ̀bɔ̀kɔ̀, ndàmbò, è.sà and mbèlà), to prepare for the death of an animal (mò.nzòlì and kóbá) or to celebrate the first capture of an important animal by a young man (mò.póndí).

    The Aka and the Baka are neighbours in the Northwestern Congo Basin. The Aka are Ubangui speakers and the Baka are Bantu speakers*. I guess it’s essentially the same ethnonym too.

    Overall the distribution of music-related words across Central Africa points to an ancient linguistic network interconnecting western and eastern populations. We believe that extending our procedure may lead to the identification of other ancient shared linguistic terms in cultural domains known to precede the arrival of farming groups, such as vocabulary related to forest, forest life, rituals regarding the forest and kin relationships.


    * I think this is much more complicated, and that’s the case for other groups as well. I would have liked the paper to go deeper into the linguistic variation within the different groups, and perhaps the ambiguity of the ethnic designations. Without that, the whole idea of measuring cultural, genetic and linguistic distance between the groups is of little value.

Speak Your Mind