Nelson Goering posted on Facebook as follows (I’ve added links and italics):
Roan eats a lot of banana these days, and as is inevitable in such circumstances, we got to talking about the word “banana”. English Wiktionary claims it goes back either to Wolof banaana, or to a similar word in a related language, but goes no further. Le Trésor de la langue française informatisé says it’s probably from “le bantou de Guinée” (the Bantu of Guinea): since Guinea doesn’t, as far as I know, have any native Bantu languages, I guess this either means Equatorial Guinea, or is using “bantou” in a very broad sense to refer to the larger Atlantic-Congo family.
I’m curious if anyone I know here has any further light to shed, both whether the Wolof/Atlantic-Congo origin is regarded as likely, and if so, if there’s anything to say about the history of the word *within* Atlantic-Congo.
I responded “There are a number of Africanists at LH, so I’ll post this and see what they have to say,” and I am now so posting it. Thoughts? (Forget the OED — its entry is from 1885.)
The Ghanaian Oti-Volta languages all borrow the Twi kwadu, not surprisingly, as does Moba, and even the relatively distant Mbelime.
Mooré, on the other hand, has bànándè, probably a back-formation (sound familiar?) from the plural banana; Gulimancema has (probably) borrowed the first element of its báanáantìbu “banana tree” from Mooré (the SIL dictionary does not give a word for “banana” as such.) Nawdm has àyàbà, which is obviously from Hausa in the first instance; I think Hausa got it from Yoruba. (I’m away from my books at present.)
Niggli says “emprunt français” for Mooré bànándè, but he’s often wrong about the source of loans. Dyula has bàràndá; a lot of Wanderwörter got to Mooré via Dyula, but I must say that in this case French does look more likely.
Assuming that the Wolof word exists and is not itself from Portuguese, Wolof looks as likely a source as any for “banana” to me.
To describe “banana” as “from an Atlantic-Congo language” is like saying “weekend” is “from an Indo-European language.” (Or “from Nostratic”, even.) It means nothing (even in the event that it might be technically true.)
References to “the Bantu of Guinea” are helpful inasmuch as they show that the writer has no idea what they’re talking about and are just bloviating.
Didn’t we discuss this before? (Generally a safe bet at LH.)
Another possibility is that French “Guinee” much like English “Guinea” previously described a much larger if vaguely defined region of coastal Africa adjoining the shores of le https://fr.wikipedia.org/wiki/Golfe_de_Guin%C3%A9e. There’s some Bantu toward the southern end of that coast. Although Wolof is at the other end …
https://en.wikipedia.org/wiki/Canaan_Banana
—
Banana was a controversial figure, especially after his criminal conviction. As president, he did not always command respect (a law was passed in 1982, banning Zimbabweans from joking about his surname).
—
Apologies if this has come up before.
Didn’t we discuss this before? (Generally a safe bet at LH.)
[2025] Banana is supposed to have come to Portuguese on loan from a West African language …
[2020] The word banana is said to have come to Spanish and Portuguese from Wolof. How did it get there?
[2012] Actually, Trond, there is a more northerly dialect, generally known as Banana-nana Fofana.
Assuming that Wolof banaana is not itself a loan, tracing its origin would be hard. Contrary to the overoptimistic statements of many of those who have attempted comparative work on the “Atlantic” languages, Wolof is not particularly close to other supposedly “Senegambian” languages. One of the few who have applied genuine rigorous comparative methods to languages in this grouping is John Merrill; in his 2018 thesis, “The Historical Origin of Consonant Mutation in the Atlantic Languages” he says
(I’m fairly sure that I’ve quoted this before, too, in service of casting shade on the as-yet-still-mainstream idea that Atlantic-Congo is a demonstrated real thing.)
Didn’t we discuss this before? (Generally a safe bet at LH.)
I had the same feeling, but couldn’t find the discussions; a tip of the Languagehat hat to Stu for doing so!
@JWB:
“Guinea” did indeed originally mean the part of West Africa south of the savanna zone (“West and Central Sudan”), and certainly, if “banana” really is originally from a West African language, it must be from one spoken in this region, as bananas are no more native to the savanna than they are to Europe. (And echt Bantu languages do indeed manage to creep in to the Guinea Zone at its Eastern margin.)
But, unhelpfully for the hypothesis, most of the Guinea-zone languages have words for “banana” which are nothing like “banana” at all (or even much like each other, mostly.) Wolof seems to be about the only exception (which looks a bit suspicious in itself.)
I see that Bété has batɛ, to add to the parade of the unbananalike Twi kwadu, Ewe akɔɖu*, Yoruba ọ̀gẹ̀dẹ̀ etc etc.
I must admit that there is a bit of a danger of circular argument here, as any real bananalike word that actually turns up could be falsely accused of being a mere loan from French or English or Portuguese. But ah hae ma doots; ah hae ma doots.
* A much more likely source of the Nawdm and Mbelime words than Twi, on reflection.
Rather belately, it’s occurred to me that bananas are not actually native to Africa anyway: which surely increases the likelihood that this Wolof banaana is actually a loan. Though that also raises the question of where all the nonbanana West African words came from. Repurposed words for “plantain”, presumably …
https://en.wikipedia.org/wiki/Musa_(genus)
But I’ve no idea how long banana-bananas have been in West Africa. Calling ROGER BLENCH (again …) Unfortunately, the paper I previously linked to by him seems to have disappeared along with his website.
His very long academia.edu page is still up.
We did the banana and the plantains, their cousin the enset, Wolof, Blench and all, back in 2020/21 (Direct link to the first banana in the thread linked by Stu above).
As my wife remembers vividly, the itinerant banana-sellers who used to come to our compound in Ghana would call out
Da’ kɔdʋ! Da’ kɔdʋ!
I was never able to parse this. Da’ is “buy” in Kusaal, and kɔdʋ is “banana”; but if it means “Buy banana(s)!”, you’d expect the imperative: Da’am kɔdʋ!
Unfortunately, it never occurred to me at the time to interrogate a banana-seller on this point.
On reflection, they were probably just being polite, and saying
M sɔsi ya ye ya da’ kɔdʋ.
I ask you-pl that you-pl buy banana
“I entreat you to buy bananas.”
with ellipsis of the main clause and the non-salient bits of the subordinate clause.
After all, it would be rude to command potential customers to buy your bananas.
Obviously there is further research to be done on the language of Kusaal street cries.
Wiki claims Magellan’s expedition encountered bananas in the Philippines (or maybe Marianas or maybe both) but the chronicler didn’t have a word for them so he described them in Portuguese as weirdly long figs. But the Portuguese were traveling to all sorts of places at similar times and presumably veterans of the different expeditions sometimes compared notes and culinary anecdotes and newly-acquired lexemes back in Lisbon …
The Malay banana-word “pisang” was brought to Europe in Dutch (where it now allegedly sounds “dated” but maybe “piesang” is still current in Afrikaans) and also supposedly found its way into Hobson-Jobson and various Scandinavian languages. The word in many Filipino languages is “saging” or something that looks similar to that, but I’m not sure if that lexeme got exported. Japanese has its own name for the so-called Japanese banana (Musa basjoo, fruit is inedible), which has long been cultivated in the Ryukyus, but seems to just use katakanified “banana” for sort you eat.
Oh and here’s an interesting banana-related cryptolect from Taiwan: https://en.wikipedia.org/wiki/Banana_colloquial_speech
G. Banane as defined in J. C. A. Heyse’s Allgemeines verdeutschendes und erklärendes Fremdwörterbuch, 1807, quoted in the 9th edition of Hermann Paul’s Deutsches Wörterbuch (1992):
— so it was still compared to a fig in Germany at the beginning of the 19th century.
The first attestation of banana everyone still cites is from Garcia de Orta (1563) Colóquios dos Simples e Drogas e Coisas Medicinais da Índia, available here (Colóquio 22; middle of the page in image file no. 210 at that site). Garcia de Orta describes bananas as a type of fig:
Garcia de Orta was resident in Goa for 30 years and had made the voyage down the Atlantic coast of Africa in 1534. He seems like an upright, serious kind of guy who wouldn’t just make stuff up or fob off to his reader any old story he considered dubious. For Garcia de Orta’s quelli, cf. Konkani केळें keḷẽ, Marathi केळे keḷe, Gujarati કેળું keḷũ, Hindustani کیلا केला kelā, etc.; for palã, Malayalam പഴം paḻaṁ; for piçã, Malay ڤيسڠ pisang; for musa, Arabic موزة mauzah (singulative of موز mauz). The question remains, what did de Orta mean by Guiné ?
For decades I have been charmed by the idea that banana is the singulative بنانة banānah of بنان banān ‘fingertips, fingers’ (cf. the expression a hand of bananas), picked up by the Portuguese somewhere and spread around the world. But where? The Arabic of northwestern Africa during Portugal’s various incursions into Morocco? The Arabic of the Indian Ocean traders on the eastern coast of Africa? But I could never find any evidence that would back any of this up. Garcia de Orta says that he found the bananas at Sofala in Mozambique to be especially delicious: e de muito bom sabor os de Çoffala ja os prouei, sã muyto gabados, eu os achei de bõ sabor (that would have been in 1534).
The other early attestation (1591) of banana that everyone cites is here, page 41 of the text (page 59 of the pdf), from Filippo Pigafetta, relaying information from a Portuguese trader, Duarte Lopes (a somewhat obscure figure), who had been in kingdom of Kongo for 12 years. (Searching on the word banana in the pdf will find it.)
@DE: I’ll just put this bit of Danish banana rock here.
https://www.youtube.com/watch?v=_cNRu80wCck
Banana cobs.
That is a lovely idea, Xerîb, and it would parallel the history of musa. When and how did the latter enter European languages?
The name Pigafetta always makes me think of “Mrs. Pigafetta Swims Well.”
“To describe “banana” as “from an Atlantic-Congo language” is like saying “weekend” is “from an Indo-European language.” (Or “from Nostratic”, even.) It means nothing (even in the event that it might be technically true.)”
Those are rather different cases! To say a loan is from an Indo-European language (let’s leave aside weekend, which is a late compound) can certainly be meaningful, and even helpful. If you’re a bit more cautious about attributing everything in Eurasia to Tocharian, it’s perfectly reasonable to suggest that Finnish mesi is a loan from an Indo-European language, and quite useful too for pursuing the further etymology of the word.
From your later comments, it sounds like we’re closer to Nostratic territory with Atlantic-Congo, which is interesting to learn. Though disappointing, since it sounds like modern Wolof might indeed be the end of the road. I’d naively assumed that one could at least back-project a modern Wolof word into a hypothetical earlier form, just as one could do with, say end, which if inherited, would phonologically have to go back to something like *andī̆- (comparative data specifies the exact stem class more precisely, of course).
Wolof really is related to at least some other “Atlantic” languages, and TBH I know much less about this than I do about Volta-Congo. But what I have read on attempted reconstruction at the level of the supposed top-level branches of Atlantic is notable for pretty gruesome methodological problems. Lower-level stuff (as you might expect) is often a lot better, and I find Merrill particularly good. The Atlanticists are learning that you need to learn to walk before you can run.
Though, as with “Khoisan”, actual specialists in the languages involved have all along tended to be a lot more sceptical about the high-level relationships than Greenbergian uberlumpers are. Way back in the eighties, W A A Wilson reckoned (using the voodoo science of lexicostatistics, but even so) that the branches of “Atlantic” were as remote from one another and as isolated within “Niger-Congo” as Ijoid (nowadays often regarded as an isolate.) The fact that the grouping-previously-known-as-Atlantic is, at the very least, not actually a single branch within “Atlantic-Congo” now seems to be generally accepted.
My guess is that at least parts of Atlantic really are related to Volta-Congo, but at a level where convincing demonstration by proper comparative methods is never going to be possible. But I think the tide is turning on all this lumpery: people seem often to express scepticism about Kru being part of “Niger-Congo” now, for example, and frankly there are better reasons for including Kru than any part of Atlantic.
But Greenberg’s zombie groupings are hard to kill off. “Ubangian”, on the other side of the supposed phylum, is another grouping which is long past its sell-by date. Far from being a genuine subgroup of “Niger-Congo”, it’s a Frankenstein’s Monster of some Volta-Congo languages along with several quite unrelated groups.
A lot of traditional subgrouping even within pukka Volta-Congo is really just based on vibes, too. There is no “Gur” …(at least, not as Manessy understood it, and as Greenberg did.)
Of course, all that is part of what makes comparative work on “Niger-Congo” so interesting. There is still everything left to play for (and vastly better and more copious data to play with than in Greenberg’s day.)
And once you shake off the illusion that Greenberg’s classification was in any way definitive, then you can get properly started.
Thanks for all that — it’s always interesting to learn where things are with languages/families/non-families I know (much) less about!
Pisang vel sim. was indeed a thing in Scandinavia, but it was before living memory. SAOB has the latest attestation in 1815 (apart from a dictionary entry in 1911), ODS in 1874. (And as a synonym in a lexicon of colonial goods in 1929).
We say banan now.
people seem often to express scepticism about Kru being part of “Niger-Congo” now, for example, and frankly there are better reasons for including Kru than any part of Atlantic.
Via a post yesterday by Mair at LLog: in a story at Sachan Journal on How Minnesotans are preserving six rare languages (five immigrant languages, one indigenous), one of them is a Kru speaker, who now teaches the language via Zoom to other adults and children. The journalist says Kru “is part of the Niger-Congo language family”, but I don’t mind that — at least she looked something up. I mind a bit that she calls it “the Kru dialect”. Also that the OED’s entry for Kru, revised December 2025, defines it as “A Niger–Congo linguistic group” without mentioning any dispute.
(The story has video clips of each of the six interviewees speaking their language with English subtitles.)
I wouldn’t have expected the OED to have caught up yet (if it ever does …) They don’t seem to currently have the advice of anybody with much of an interest in African languages, even, let alone specialised knowledge of the state of the art in comparative work.
While I’m being all splittery: though I long since realised that Greenberg’s “Ubangian” is Not A Thing, I only recently realised that his “Adamawa” is at least as bad, if not worse. Parts of it are undoubtedly Volta-Congo (like Tula-Waja and Leko); others show really no sign of any such relationship at all. (I started looking into this in more detail after reading the late Stefan Elders’ nice grammar of Mundang, one of the really-not-Volta-Congo kind. If it’s related at all to Volta-Congo, it’s at least as remotely as any part of “Atlantic.” There’s no question of it being part of a single real branch of Volta-Congo along with Tula-Waja or Leko.)
Greenberg’s whole Adamawa-Ubangi thing is a real unholy mess. A standing indictment of the flakiness of his language-classification methods. Right up there with his “Amerind.”
I only recently realised that his “Adamawa” is at least as bad, if not worse.
I think you might find some things to dislike in Pozdniakov and Vydrin’s recent Mande and Bantu in the Niger-Congo context. Besides that, what do you think of their central thesis, linking Mande and Bantu?
I think you might find some things to dislike in Pozdniakov and Vydrin’s recent Mande and Bantu in the Niger-Congo context.
Good grief, yes. An Awful Warning of the perils of trying to do comparative work without any grasp of the basic principles at all.
Frankly, comment is superfluous. (Though I will say that Bangime, which even Roger Blench thinks is an isolate, has nɛ̀ɛⁿ “four.”)
Pozdniakov has form: he’s published extensively on comparative Atlantic using similarly valueless methodology. For an antidote, I’d recommend John Merrill, who, unlike Pozdniakov, actually understands the comparative method and has done proper comparative work in Atlantic.
There’s no evidence worth a damn that Mande is related to Bantu.
Bantu is, demonstrably, related to Oti-Volta*. Any non-crank comparison of Mande with Bantu should instead work via comparison of proto-Mande with the common ancestor of Bantu and Oti-Volta. This is elementary stuff. Pozdniakov, of course, ought to be working to compare Mande with the common ancestor of “Atlantic” and Volta-Congo.
I was, incidentally, just looking at Dame Ndao’s nice account of Pepel, a Bak language of Guinea-Bissau. As usual with “Atlantic”, some striking typological similarities with Volta-Congo but precious little else to suggest an actual genetic relationship. But I noticed, alongside dɛ “eat”, dʊm “bite”, which is one of Pozdniakov’s pet words for “proving” relationships. It’s certainly pan-Volta-Congo, but I was reflecting just the other day that like the Ubiquitous Turtle, it’s odd that it should be so highly conserved when e.g. “child” or “woman” aren’t. Phonaesthetic?
* I discovered quite early that it’s fatally easy to get lookalikes between Bantu and Oti-Volta by cherry-picking from the vast array of proto-Bantu variants. Unlike P & V, I regard this as a bug rather than a feature.
This paper in a nutshell: Greenberg in, Greenberg out.
Forms like Kusaal (a)naasi “four” are indeed extremely widespread in West Africa and in Bantu. But when you turn up similarities like this in languages which show no other evidence of close (or, sometimes, any) genetic relatedness, the proper response is not “it’s statistically impossible that this is chance: this proves that the languages are genetically related after all!”
Statistically impossible it may be (though many such statements turn out to be mathematically mistaken.) But even if it is not coincidence, the correct response should be “why does this same form keep turning up in languages which show so little other evidence of being related? Should I be reconsidering my assumptions about the borrowability of words for small numbers? Has this been shown to occur elsewhere?”
I learned of it from Lameen’s Bluesky feed, and from Pystynen’s Txitter feed (@JLingPystynen, perusable through Nitter: pick a server from here, and substitute it for that other one). Lameen is cautiously encouraging, Pystynen more heartily so. Y’all play nice, now.
There’s so much wrong with this …
It’s a kind of Gish gallop …
https://en.wikipedia.org/wiki/Gish_gallop
Just picking one of so many: “bone.” They reconstruct this as PB *kʊdɪ (“regional”) and equate it with PM *xɔlɛ.
Now, as it happens, “bone” is reconstructable to proto-Volta-Congo. The actual PB form is *-kúpà (cf Swahili mfupa.) The POV form was *kṍb-; forms of the shape *kup turn up in several languages in between OV and Bantu. There’s an issue with the vowel correspondences, perhaps related to the fact the the vowel was nasalised (something lost in PB), but there’s little doubt that this is the form that should be compared with PM *xɔlɛ, not this cherry-picked “regional” form.
The lengthy list at the end is pure Mass Comparison stuff.
I wonder why Miyobe is called Miyove?
On crankery: how comapring PM with PB is different from comparing OV with PB? Otherwise, I don’t understand what people are arguing about. Either the results are sufficient to work on them or they aren’t.
It ain’t what you compare, it’s how you compare it.
In principle, there is no sin in comparing Oti-Volta and Athabaskan. But if anyone claims that their comparison has resulted in a demonstration of genetic affinity, a certain scepticism about the soundness of their methodology would be called for.
I am by no means hostile a priori to the idea that Mande might be genetically related to Volta-Congo. That would be way cool. (I hope that it really is, though I’m pretty sure that the question does not admit of a convincing answer either way through rigorous comparative methods.)
Nor do I think there’s anything at all wrong about investigating the question. What I object to is bad methodology, especially from eminent and learned academics. They should know better.
[In my day job, bad science kills and maims. The stakes in comparative linguistics are a bit less. Perhaps I should lighten up a bit …]
That frankly sounds like Proto-Bantu hasn’t actually been reconstructed at all, just a blur of forms that are old within or around Bantu, but no attempt to figure out how old exactly.
The IEists have noticed (slide 22–31) that PIE is actually in the same situation, and they still haven’t done a lot about it, though of course they have it easier than the Bantuists.
As the biologists found out around, very roughly, the same time: “nothing makes sense in evolution without a phylogeny”.
I should perhaps mention that proto-Mande is generally thought (including by Vydrin) to be of a time depth at least as great as PIE, and possibly a good bit more; both POV and PB are probably quite a bit younger than PIE, but their common ancestor must surely be at least as old as PIE. That’s before you start trying to factor in “Atlantic”, which is as internally diverse as it is, overall, different from Volta-Congo. So this Mande-Bantu thing is Altaic level stuff, or at least Nostratic level. By common agreement, there is zero morphological support for it. Zip. It’s all based on supposed lexical correspondences (along with a steadfast refusal to consider that a region currently notable for widespread extensive borrowing across multiple languages might perhaps have been notable for extensive borrowing in the past too.)
The default response to this kind of proposal should be O RLY?
Of course, P & V may wish to imply that West African languages change much more slowly than languages elsewhere …
Or they may wish to maintain that Mande and Bantu form a subgroup within Niger-Congo, thus accounting for their remarkable resemblance …
Feh.
@DM:
Very interesting slideshow (as you’d expect from the author.) Thanks.
The Bantuists are pretty clued up on issues like these, at least on a theoretical level; Guthrie was already careful to distinguish between the way comparisons of existing languages are notated, and the putative spoken last common ancestor of all the Bantu languages itself.
Current issues (among many others) are whether the weirdo Northwest Bantu languages can be made to spring from the same protolanguage at all as the “classic Bantu” languages that Guthrie and Meinhof etc mainly relied on; and the related question of just where the boundary between Bantu and not-Bantu even is, up there in Cameroon. Or if there is one …
I think what’s important is whether there are unusually many forms in PM similar to PB forms.
The interpretation (are they borrowings?) is an expression of [methodological?] beliefs which is not based on a “methodology”, unless accompanied by a respective study, and if it were accompanied by it, it is the study which would be valuable and interesting.
The single biggest problem is egregious cherry-picking, which unfortunately the Bantu material readily lends itself to.
Take “ashes”, for example, where they pick *bú in order to align it with a PM *bugu.
“Ashes”, like “bone”, can in fact be reconstructed to proto-Volta-Congo: POV *tó- PB *-tó (again, with cognates in other VC branches.) It is this etymon, therefore, that ought to be compared with PM, not the regional Bantu form *-bú.
This sort of thing is exactly the reason why you shouldn’t compare language groups piecemeal in this way. It just leads to nonsense results. It’s the same catastrophic fundamental error as was made by John Stewart in comparing Potou-Akanic and Fulfulde directly. It just doesn’t matter if you feel you’ve gone through all the right motions of setting up regular sound correspondences etc when you’ve undermined your whole basis at the outset like this.
It’s also not so impressive that you’re “comparing protolanguages” (hurrah!) when the protolanguages are (a) subject to the internal stratification and diversity problems that DM was mentioning, and (b) themselves only represent subbranches of subbranches (like proto-Bantu.)
The “unusually many forms in PM similar to PB forms” (leaving aside what constitutes “unusually”, exactly) that you collect by such methods are lookalikes, not cognates. It’s just Greenberg Mass Comparison in a nicer suit.
@DE, I meant the part where you say they should compare it not to proto-Bantu, but to proto-proto-proto (or are cranks). Can’t same be said about comparing Oti-Volta to Bantu?
“Atlantic” – but does it have anything to do with Mande? We don’t even know if it is NK.
“morphology” – you forgot to tell why morphological support is impossible.
They also seem to have distorted some PB forms to make them look more like PM. For example, PB “drink” should certainly be *ɲú or *ɲó, not *mó (cf also POV *ɲû-.) Quite why they think it matches PM *ɓɔ̃ɗɪ̃, I cannot tell.
If this is rigour, I’m a Chomskyite.
lookalikes – absolutely. But it can be valuable, and there is no need, as I said, to interpret it at this point.
Can’t same be said about comparing Oti-Volta to Bantu?
That’s a reasonable point, and in fact it’s why I’ve been cryptically alluding to “other languages” beside OV and Bantu above. I was just too lazy to cite them, but I didn’t make them up.
As a general point, too: Yes. I put a specific disclaimer in my POV tractate about this:
In other words, “what follows is for the lulz, and should be taken with a pinch of salt.” However, nobody (not even me) actually disputes than OV is related to Bantu: I’m not trying to prove a genetic relationship there, just looking at probable cognates in the light of an already-accepted and uncontroversial relationship.
One of the possible cognates there is cherry-picked, and should be excised. (“Cook.”) I’ll get round to it …
“Atlantic” – but does it have anything to do with Mande? We don’t even know if it is NK.
P & V assume it is. (They’re not alone, by any means: my view that any relationship is, as yet, unproven is a minority one.)
The point here is that even the maximalist lumpers do not dispute that “Gur” is closer to Bantu than “Atlantic” is; hence the point that “Atlantic” forms should be compared with forms common to Gur and Bantu, rather than with either alone. All the more so with Mande, which even megalumpers think is more remote from Bantu than Atlantic is: that is why, to be consistent with their own previously stated position on “Niger-Congo”, P & V should be comparing PM with some approximation to proto-Atlantic-Congo, not to a subbranch of a subbranch of a single branch of the Volta-Congo branch.
Unless, indeed, what they are claiming is that Mande is not an outlying first branch from Niger-Congo, but belongs to the core. That would not find a lot of support even among uberlumpers, to put it mildly.
“morphology” – you forgot to tell why morphological support is impossible
Sorry: I tend too readily to assume that things I’m familiar with myself are common knowledge.
The morphological support I was talking about is the famous noun-class system, on the one hand, and the system of verbal derivational suffixes (called “extensions” in the trade, after Bantuist usage.)
Within Volta-Congo, you can find a great deal of morphological evidence of this kind, at least when it comes to noun-class affixes. Enthusiasts for including “Atlantic” point to the fact that (some of) these languages have typologically very similar systems (though in my own view, they fail to demonstrate convincing correlations between form and meaning.) Similarly with (parts of) Kordofanian.
Mande has absolutely no trace of any of this; nor has anything like it been reconstructed for the umpteen-thousands-of-years old PM.
If I can be a bit devil’s-advocate-y about this, I think the absence of class affixes is less of a clincher than is sometimes thought; it’s clear from comparing e.g. Gur and Bantu that the position of these “affixes” was not fixed in the protolanguage: they must have been clitics, not affixes, and in relatively closely related languages they sometimes precede, sometimes follow the noun stem. So I don’t think it’s all that improbable that PM might just have dropped them.
Be that as it may, there is thus no morphological support for a Mande-Bantu connection, unlike the case within Volta-Congo.
Syntactically, Mande is weird, in a way remarkably well conserved within the group, and the only Volta-Congo languages like them in this are the Senoufo languages, where the explanation is obviously prolonged contact. But I’m not a Johanna Nichols style believer in the great persistence of syntactic features in general, and I doubt whether that really means very much.
DE, I think somoene here mentioned that Mande languages shed off their morphology. I don’t know if it is a hypothesis or consensus opinion.
But can we say that PM morphology is largely “unknown” rather than “different’?
No: there’s no more mystery about PM morphology than about the morphology of comparable protolanguages.
It definitely isn’t the consensus that Mande “shed” its morphology (specifically, noun-class affixes.) The usual view is that it never had that morphology to begin with.
There aren’t any plausible relics, even. Contrast, for example, Samba Leko: this is an “Adamawa” language that (unlike some) is certainly Volta-Congo, and has been very well described by Gwenaëlle Fabre. It has lost almost all traces of VC morphology (most words are monosyllabic); nouns usually do not inflect for number at all. Even so, there are a few brave holdouts like nɛ́ “person”, plural nɛ́b; compare e.g. Mooré nédà “person”, plural nébà.
Nothing like that in Mande.
Looking at P & V’s “cow” entry, which shows the problem that they have with not even considering borrowing.
There definitely is a very widespread West African root something like *nak “cow.” It turns up all over: POV *nâag-wá, Kassem nāa, Miyobe ì-náà Samba Leko nà, Kulango naà, proto-Eastern Grassfields *nàkˊ; Fulfulde nagge, Jamsay Dogon nàŋá; and nàa in the isolate Bangime.
Characteristically, P & V list this along with some reflexes of a completely different word, “meat”, POV *nêm-, Kassem nwǎm, Pana nàmá, proto-Bantu *-nyàmà. Because *nak, *ɲam, “cow”, “meat”, it’s all the same, right?
The point about this *nak is that it is shared by languages which are at most only extremely distantly related (Kusaal, Fulfulde, Dogon), or not related at all (Kusaal, Bangime.)
These languages do not share words like “dog”, “woman”, “eat”, “drink” … but they do share “cow”! Clearly this proves that they are genetically related!
P & V’s PM form here is *ɗĩɠã, and the actual Mande reflexes virtually all have the stem vowel /i/, not /a/; this is not addressed, much less explained. [The relatively few exceptions with /a/ seem to me most likely to be loans.]
The *nak form is altogether absent from Bantu, so P & V have to cheat rather with this one. Presumably the speakers of PB just forgot what cows were called. It happens … (though they didn’t forget what dogs or goats were called. Or bloody tortoises …)
Their conclusion: “It [what?] can undoubtedly be reconstructed for Proto-Niger-Congo.”
https://www.sciencedirect.com/science/article/abs/pii/S0305440311003232
“The earliest evidence for cattle in central West Africa south of the Niger River is at the site of Winde Koroji Ouest, ca. 2200–950 BCE in the Malian sahel.”
I wonder if it is just conceivable that (a) the *nak form is a very old Wanderwort that has spread along with its referent, and (b) PM *ɗĩɠã is quite unconnected with it …
Even, I, my brethren, have fallen into temptation and committed the foul sin of cherry-picking …
The “cook” equation I mentioned is POV *dʊ̂g-, proto-Bantu *dúg-. Looks great, eh?
There is cherry-picking on the Bantu side there: PB has a pleasing menu of “cook” words to choose from, including e.g. *tedɪk-, *tek-, *jambɪk- and *jipɪk-, none of which looked quite as appetising as *dúg-.
Furthermore, I hereby confess to cherry picking on the Oti-Volta side too: the evidence for POV *dʊ̂g- is
Kusaal dʋg, Mooré dʋge (and other WOV), Buli digi, Konni digi, Nawdm duug-.
The trouble with this is that WOV, Buli/Konni and Yom/Nawdm form one genetic branch of Oti-Volta. There are no forms like this in Gurma, Eastern Oti-Volta or Waama.*
So it is not legit to set it up for proto-Oti-Volta, and I have no business comparing it with Bantu.
I repent in dust and POV *tó- PB *-tó, but not in PB -bú, PM *bugu.
* There’s also some troubling evidence that the OV -g might be some sort of derivational suffix. There’s a Kusaal set expression dʋgʋb dʋt “cooking pots.” Here dʋgub is the gerund of dʋg, and is No Problem, but the plural of dʋk “pot” is normally dʋgʋd not dʋt. Probably all some sort of analogical resegmentation, but in the dark watches of the night, I am troubled, my brethren …
“Cow” is weird:/
Can we expect them to call domesticated cows (once introduced) same as, say, wild buffalo cows?
Presumably the speakers of PB just forgot what cows were called. It happens …
Well, the Slavs also did, innovating *korwa:, and still we reconstruct PIE *g_wow- based on the other branches.
(Ok, Slavic still preserves the root in *gowęd- “beef”, but IEanists would reconstruct *g_wow- anyway.)
At the present day, cattle-raising tends to be strongly associated with particular ethnic groups; in the Kusaasi area, with Mossi and Fulɓe. Kusaasi are typically cultivators rather than livestock-raisers.
When I first (years ago) noticed Fulfulde nagge “cow”, which surely must be connected with all these *nak words somehow, I wondered if the Fulɓe had actually introduced the word along with cattle-raising across the whole area (given that Fulfulde is clearly not at all closely related to e.g. Kusaal.) But I don’t think the timescales can be made to work; the relevant forms seem to go back to too many protolanguages. (Cf the impressively irregular plural niigi of Kusaal naaf “cow.”) On the other hand, this might be like the case of proto-Algonquian “whisky.” In Oti-Volta, “cow” typically inflects exactly like “snake”: e.g. Kusaal waaf, plural wiigi, so the irregularity may have simply arisen by analogy, rather than being evidence that the word was already present in proto-Oti-Volta.
But even if the Fulɓe were not the vector of *nak, its distribution screams “Wanderwort” to me. And proto-Oti-Volta (for example) was probably spoken as recently as ca. 3000 YBP, which is quite recent enough for the word to have been an old loan rather than inherited from proto-Volta-Congo.
Well, the Slavs also did
I was indulging in a bit of childish mockery.
The proto-Bantu “cow” word has long been tagged as an ancient loanword. (I forget whether it’s supposed to be from Afroasiatic or “Nilo-Saharan.”)
Proto-Bantu also lost the Volta-Congo word for “child”, but there is genetic evidence that the speakers nevertheless did actually have children.
…except for those who reconstruct *gʷéh₃-u-s ~ *gʷh₃-éw-s as explained here.
*h₃ is particularly likely to lurk somewhere undetected.
DE, I can’t speak about forms and its semantics is strange. But what are other reasons to think of wanderwords (unless we – unlike them – think Dogon, Mande, Atlantic can’t be in Niger-Congo)?
Bangimi is variously bad example. It is full of Dogon loans. And they call it “likely” Niger-Congo, not Niger-Congo. Their desire to list it is understandable (their unwillingness to list Kordofanian is less so).
Assuming (for the sake of argument) that Dogon, Mande, Atlantic and even Bangime are indeed all related to Volta-Congo, and that these *nak/*niŋa words are all inherited from some proto-Niger-Congo original (the exact form of which it is apparently unnecessary to specify): isn’t it a bit strange that the word for “cow” should be so highly conserved when the words for e.g. “man”, “woman”, “child”, “eye”, “mouth”, “goat”, “dog”, “tree”, “eat”, “drink” etc etc aren’t?
Do they actually call Bangime “likely Niger-Congo”? If so, that in itself tells you a great deal about their presuppositions. It’s “likely Niger-Congo” in the same sense that Tibetan is “likely Dene-Caucasian.”
unwillingness to list Kordofanian
There, I wholly agree with them.
Apparently even Greenberg hisownself said upfront that his reason for putting “Kordofanian” in “Niger-Congo” was entirely due to typological similarities. There’s pretty much zero lexically (and zero in terms of correlating the form and function of morphological affixes: all attempts to demonstrate the contrary that I’ve ever seen are exercises in egregious question-begging.)
G seems to have believed that the multiple-gender-paired-affix noun-class system was so typologically unusual that if languages shared it, the only possible explanation was common genetic inheritance.
“Probably bolong to Niger-Congo”, not “likely”. P. 27.
*belong:)
…except for those who reconstruct *gʷéh₃-u-s ~ *gʷh₃-éw-s as explained here.
I was not familiar with that proposal. Doesn’t look bad at first glance.
The entry itself though is contradictory, stating on one hand that the e-grade is not attested and then reconstructing a genitive singular *g_wew-s. (By the way, how do you create superscript characters here? I tried “sup” between “> <"*), but it didn't work.)
*) In the obverse order, but whenever I try to type it that way, it doesn't show.
@drasvi:
In a way, I kinda agree with them. I think that Blench’s conclusion that Bangime is an isolate came about because his judgment in such matters is quite sensible whenever he’s not hobbled by his apparent belief that Greenberg’s African classification is definitive and final. In fact, he keeps discovering isolates. So long as Greenberg never mentioned them, he seems fine with calling them that.
The sense in which I agree with P & V is that I think there is no less reason to suppose that Bangime is related to Volta-Congo than to believe that Dogon is related to Volta-Congo. Why strain at a gnat when you’ve swallowed a camel?
(My own gut feeling is actually that Dogon probably is related to VC, but so distantly as to be beyond demonstration by proper comparative methods. Happily, this feeling is probably unfalsifiable. That’s gut feelings for you.)
That’s Wikipedia/Wiktionary for you – most people edit it by adding or removing a sentence or two; hardly anyone goes through the whole article for consistency.
The reconstruction in the table is simply what the inflection of an acrostatic noun is expected to look like.
In the comment that “there is no trace of the e-grade (*gʷéw-) (except perhaps in the oblique Indo-Iranian stem gav-, through the absence of Brugmann’s law)”, the part in parentheses doesn’t mean much because Brugmann’s law is so hopelessly morphologized – if you’re going to find an analogical *e-grade instead of *o anywhere (in Indo-Iranian terms: full grade as opposed to lengthened grade; while I’m at it, in pukka Sanskrit: guṇa vs. vṛddhi), it’s going to be Indo-Iranian.
…and now I’m going to take a look at lore and order, by Laura “Norder” Grestenberger.
You don’t; <sup> and <sub> don’t work in this version of WordPress. Instead, you have to copy & paste the ready-made super- and subscript characters from elsewhere; I just open the Windows character map. (Zeichentabelle. Comes with Windows, but is hidden in Start > All Programs > Accessories [Zubehör] > System Programs, because that’s how Microsoft rolls.)
because it’s interpreted as an HTML tag that doesn’t exist, so it’s deleted. The trick is to spell out the HTML entity: > for >, < for <. (…and I typed those by using &; and so on.)
And I still use Babelmap.
By the way, how do you create superscript characters here?
I copy and paste from here.
Thanks for the technical help. As I am normally posting here on my phone, the Zeichentabelle (which I regularly use on my computer) doesn’t help, and I need super-/subscript characters rarely enough that I’ll probably continue to make do without them if I forget about the links kindly provided by ulr and LH 🙂
And now I’m off to watch Germany 🙂
I do it по-LHски.
That is I go to WP and search for superscipts.
DE, for them Kordofanian is NC.
If they look for their forms in all NC branches and even in Bangime, why not in Kordofanian?
Figures. If you have already decided (somehow) that Indo-European and Pama-Nyungan form a single genetic group, it becomes a lot easier to demonstrate that Uto-Aztecan also belongs, because you’ve greatly increased the number and variety of potential lookalikes. If you have enough lookalikes, you can easily find some “regular phonetic correspondences” with a bit of judicious selection, too.
In other words, premature assumptions about genetic relatedness lead to a sort of vicious circle, where your defective methodology will make it relatively easier to absorb yet more unrelated languages into your Blob, by transitivity of mistaken ascription of cognacy.
The method is different though: 1. find lookalikes in PB and PM 2. see what other branches also have them. So these other branches work as a filter. Besides, it is interesting to see where in NC those lookalikes can be found.
What depresses me about this is that what I’ve been saying about methodology in historical-comparative linguistics is not remotely original or controversial. Any decent introduction to the subject explains these kinds of pitfalls in detail: cherry-picking, exclusive reliance on lexicon … all this.
It really is hardly surprising that linguists from outside African studies are often disagreeably startled by the state of things in African comparative work. R M W Dixon is supposed to have said at the end of a seminar on Afroasiatic: “so, not a demonstrated group, then.” His views on “Niger-Kordofanian” and “Nilo-Saharan” are probably unsuitable for a family blog like this one, especially after all this time he’s spent in Australia.
There actually is quite a lot of good, entirely rigorous, bottom-up comparative work going on with many African language families. It’s this supposedly “high-level” stuff, at the level of Greenberg’s Big Four, which is, over and over, frankly bollocks. It won’t be possible to address questions at that level at all adequately until much more work has been done on the subgroups. Omitting that vital step leads, not to progress, but to a mere illusion of progress.
Proto-Bantu, for example, has a tiny consonant system. Why? Have there been a lot of mergers? Or have all the other VC groups developed a lot of splits, under conditions we don’t understand? The way to begin is to look at how Bantu fits into its Bantoid subgroup of the Benue-Congo branch of Volta-Congo. Meantime, you just cannot assume that PB *t corresponds to precisely one consonant in proto-VC (let alone “proto-Niger-Kordofanian”) and make pretty equations with proto-Mande as if you already knew the answer.
R M W Dixon is supposed to have said at the end of a seminar on Afroasiatic: “so, not a demonstrated group, then.”
Reminds me of Ben Bradlee in All the President’s Men: “Where’s the goddamn story?”
Oh, so 𐞴 for epiglottalization already exists, I can start claiming that [æ] is really [ɛ𐞴] – I just can’t see it in my obsolescent browser.
SCHLAAAAAAAAAAAAAAAAAAND
(I’m asportual, I just figured somebody had to say it.)
Looked like Ivory Coast was going to win, but Germany brought in Deniz Undav (a Kurdish Yazidi!) who tied it in the second half and then drove in the winning goal with just a few minutes left. Good game, like so many of these group-stage matches.
Of the fonts I have, the only ones which include 𐞴 (superscript ʢ) are the ones from SIL (Charis, Doulos, Gentium), even though the character has been included in Unicode since version 14.0, nealy five years ago.
Well, this is what Ringelnatz had to say, 126 years ago:
The whole poem (Fußball. (nebst Abart und Ausartung)) can be easily found on the web.
Are there any English poems where a line stops mid-word, just for the sake of rhyme?
There was a time when every German boy had read Karl May’s Durchs wilde Kurdistan, showing (in Ottoman times) a lot of sympathy for (completely fictional) Kurds and Yazidi.
@ulr
Re English poems with broken rhyme,
“Gerard Manley Hopkins’ poem The Windhover, for example, divides the word “kingdom” at the end of the first line to rhyme with the word “wing” ending the fourth line. Hopkins is rare in using the device in serious poems.”
https://en.wikipedia.org/wiki/Broken_rhyme
Are there any English poems where a line stops mid-word, just for the sake of rhyme?
Oh yes.
For the metre rather than the rhyme (but an excuse to link a great poem):
https://www.poeticous.com/w-h-auden/heavy-date
For the rhyme (unfortunately given here with incorrect line breaks):
https://genius.com/Tom-lehrer-we-will-all-go-together-when-we-go-lyrics
[Ninja’d by PP. No matter …]
@DE, where do those protoforms, from which both Proto-Oti-Volta and Proto-Bantu can be obtained, come from?
Your own reconstructions? [Numerous] articles? Some book?
I’m trying to understand to what extent they’re accessible to me – and to Pozdniakov and Vydrin.
@de, ulr
De’s examples are good. Another poet who is extremely playful is Ogden Nash, but I did not know of or find any examples of broken rhyme (he does a lot of odd rhymes and extremely long, short or varied line lengths).
Elizabeth Bishop, Pink Dog:
You could argue that the second “a” is also part of a broken rhyme, splitting a phonological word, not a written one.
(A wonderful poem, all-around.)
where do those protoforms, from which both Proto-Oti-Volta and Proto-Bantu can be obtained, come from?
Did I cite any? If so, I apologise and promise not to do it again.
In my POV treatise, I just juxtapose similar POV and PB forms and let the intrepid reader make of it what they may, as with POV *kṍb- and PB *-kúpà “bone.” This is, admittedly, lookalike stuff, but I think one can go a little further (using this word as an illustration.)
For example, there are POV-internal reasons to think that the *b in that form goes back to earlier *p.
The absence of a final vowel in the POV is to be expected, because there aren’t any POV *CVCV roots at all, so if you assume that the PB final *a is a real part of the root (as it surely is, CVCV roots being common in PB, with the final vowel not predictable) you have to conclude that somewhere in the prehistory of POV *CVCV roots just lost their final vowel contrasts. The tone structure of POV supports this idea: POV roots have a four-way tone contrast, pretty certainly going back to an earlier HH/HL/LH/LL.
The vowels don’t match. However, the POV form definitely had a nasal vowel, and PB has no vowel nasalisation. There are a few cases where POV and another branch of VC agree in having a nasal vowel, for example in “three”, POV *-tãɦ, Twi ɛsā, PB *-tátʊ̀. There are few examples, because, contrary to the impression given by works like P & V, there are really not all that many good comparative series to work with (I’ve managed to find only a couple of dozen or so between POV and PB, despite the fact that I haven’t been idle in searching, and that POV is very much more closely related to Bantu than Mande is – or might be.) Nevertheless, I think there’s enough to show that PB has lost a feature that was present in proto-VC here.
Accordingly, one might guess that the unexpected open vowel in POV *kṍb- may be the result of the nasalisation. POV did have nasal close vowels, but nasal open vowels were a lot commoner, so that’s fairly plausible.
So maybe something like proto-VC *kũpa (first tone maybe high, but there aren’t enough good comparanda to really make much of a guess at tones.)
But all this is based on a chain of shaky speculation, and (worse yet) is not rooted in any theory of how many distinct vowels and consonants proto-VC had, even. It’s all ad hoc. So this *kũpa is really just a sort of guess, not a reconstructed proto-VC form.
There are (alas) plenty of sources purporting to give proto-Niger-Congo protoforms. They are all at least as flaky as this; most of them don’t even attempt as much detailed justification as I just attempted.
For “Gur” protoforms, people cite Gabriel Manessy’s work of getting on for half a century ago. I have considerable respect for Manessy’s pioneering work, but “pioneering” is the word, and his reconstructed roots are frankly of little value. He was working with very much more limited material than is now available, and explicitly subscribed to an unfortunate doctrine that all CVC roots were actually CV plus a derivational suffix. In practice, this allowed him to ignore non-initial consonants altogether, and equate forms based entirely on their initial CV-; as he did not recognise that nasalisation was contrastive, and didn’t have the data to try to reconstruct tone, the result is … chaotic. But people still treat his reconstructions as state of the art.
My own POV reconstructions are better, but still miles away from where I would like. I only recently came to a conclusion (for example) that I should have seen long ago: POV had contrastive vowel length. You’d have thought this was obvious from series like
Kusaal saan, Mooré sã́anà, Buli (ní)cháanoā,
Konni cháánʊ̀, Yom cāānà, Nawdm sááná,
Gulimancema càanō, Mbelime sàànɔ̀, Waama caaŋo, all meaning “stranger”;
what kept me from seeing this was a doctrine that all Oti-Volta long vowels are of secondary origin (cf Miyobe ù-sánɛ̀ “stranger.”)
I still think this, in fact; but what I was not doing was thinking properly about when these vowel-lengthening processes operated exactly.
Some processes operate in individual branches or even in individual languages (e.g. in Mbelime nouns, root vowels are always long in open syllables unless previously followed by a consonant cluster, so the length in sàànɔ̀ actually tells you nothing at all about the length in POV.)
But obviously some of these lengthening processes had already operated in the last common spoken ancestor of the Oti-Volta languages. So POV *cã̂an-, not *cã̂n-. Duh …
Among other consequences, this means that vowel length could have been a conditioning factor on the development of postvocalic consonants, and at the moment I’m considering whether my current distinction between non-initial POV *k and *g might be invalid. That would have knock-on consequences elsewhere in the reconstruction.
My point in this, is that a protolanguage reconstruction needs to be of a whole system, not just isolated segments.
(I should perhaps, in self-exculpation, say that my account of POV/Bantu resemblances is only a very minor part of my POV treatise, and was something of an afterthought, that I just added because I thought it was interesting. I make no great claims for it – though it turns out that quite a bit of published stuff is even worse. My comparison with Miyobe, though … that is the bomb. Easily the best comparison of Miyobe and Oti-Volta EVAH.)
That should be “Twi ɛsã.” I’m sure you all spotted that.
More of Lehrer’s broken rhymes: https://genius.com/Tom-lehrer-smut-lyrics
My immediate reponse to things like P & V is actually to look through to see if they’ve got a comparative series that I’ve missed that looks hopeful.
I had a little flutter of excitement at their “egg” series, in fact; and their PM *ɣeli caught my eye as looking quite uncannily like Kusaal gɛl, Mooré géllé “egg” (POV *gêl-.) Sadly, I’m pretty sure that this is the Awesome Power of Sheer Coincidence: although PB *-gɪdɪ is cited there, several of the series I do have suggest that would correspond instead to a Kusaal gɛr, plural gɛya, Mooré gérè, plural gɛya, and the usual PB form actually seems to be *-gɪ́ anyway. Oh well.
I would think PB *-gɪ́ is more probably cognate with English “egg” …
To accentuate the positive: their PM *yɪ̃ “tooth” does look a whole lot like Kusaal nyin “tooth” and its kindred. And I mean, what are the odds on two CV syllables agreeing by pure chance? Waddya mean, “quite high”?
what are the odds on two CV syllables agreeing by pure chance?
Pretty much none, never, as long as they are my syllables.
But you’ve only got one syllable altogether!
Why is that?
I can claim any syllables as dependents for tax purposes.
Are there any English poems where a line stops mid-word, just for the sake of rhyme?
Whene’er with haggard eyes I view
This dungeon that I’m rotting in,
I think of those companions true
Who studied with me at the U—
—niversity of Gottingen,—
—niversity of Gottingen.
[…]
There first for thee my passion grew,
Sweet, sweet Matilda Pottingen!
Thou wast the daughter of my Tu—
—tor, Law Professor at the U—
—niversity of Gottingen—
—niversity of Gottingen.
Sun, moon, and thou vain world, adieu,
That kings and priests are plotting in:
Here doom’d to starve on water gru—
—el, never shall I see the U—
—niversity of Gottingen—
—niversity of Gottingen.
George Canning, The Rovers
Note that you can’t process a binary tree bottom-up.
If you do, enough to find ONE language without the word you need and you can’t reconstruct it to the root of the tree (the protolanguage).
Say, you see a word (with predictable sound changes) in proto-Indo-Arian, proto-Slavic and proto-Baltic… but not in proto-Irainian. Then you can’t reconstruct it in proto-Indo-Iranian. Then you can’t reconstruct it in proto-II-BS. And so on, you won’t be able to reconstruct it in Proto-IE as well.
proto-Indo-Arian
A major problem with the reconstruction in that case is that the branches are only of a similar substance, not the same substance.
Actually, that situation – assuming the word is missing in the rest of IE – means it’s equally likely for it to have been present in Proto-Indo-Slavic and lost in Iranian or to have been gained in Balto-Slavic and Indic separately (by chance, or by geographically improbable loaning from each other or from a common third source). It’s two steps either way: one gain and one loss, or two gains.
A major problem with the reconstruction in that case is that the branches are only of a similar substance, not the same substance.
I smell heresy!
Note that you can’t process a binary tree bottom-up
The problem with my *dug- “cook” example is that it’s only found in one subbranch of Oti-Volta and one subbranch of Bantu. If a plausible cognate could be found in some other branch of Volta-Congo, then it would probably be justifiable to suppose that all the subbranches of Oti-Volta and Bantu which don’t have such a form have simply lost it, but otherwise it’s likely that this is just a chance resemblance. Again, if *dug- were reconstructable to either POV or PB as the root for “cook”, it would be reasonable to assume that subbranches without it have lost it. But in fact that’s not the case in either, unless you adduce this very resemblance between the Oti-Volta and Bantu forms as evidence, which begs the question.
Both Oti-Volta and Bantu offer quite a few other words for “cook”, thus making one or another chance resemblance quite likely to happen.
To give you an idea of the potential scope for cherry-picking on the Oti-Volta side: Gulimancema has bìàlí “cook (porridge)”, kùdi “cook (food in general)” bógì “cook”, ŋààní “cook, bake”, nùgidí “cook (tubers)” …
“Cook” is problematic anyway in a lot of languages, because there tend to be quite a lot of unrelated words for the concept, depending on what you’re cooking and just how you’re cooking it. If you blithely ignore the nuances, you award yourself a lot of free choices for comparisons …
If you then say, “well, it’s statistically impossible that so great a resemblance in sound and meaning could be due to chance”, you’re committing a basic statistical error which is highly popular in medicine: seriously underestimating the degrees of freedom in the design of your experiment. This makes your statistics worthless.
Back to broken rhyme, here is a poem some might enjoy:
Scintillate (R. McGough)
I have outlived
my youthfulness
so a quiet life for me
where once
I used to
scintillate
now I sin
till ten
past three.
An early example is from the third “Stand” of Ben Jonson’s Pindaric ode “To the Immortal Memory and Friendship of That Noble Pair, Sir Lucius Cary and Sir Henry Morison”:
Apparently “twi-” is a combining form that in this quotation has the OED’s meaning “1.c. With noun in sense ‘twofold, double’.”
I might mention the related “linked rhyme”, where the rhyming sound is completed by the beginning of the next line. The usual example in English is from Hopkins’s “Wreck of the Deutschland”:
“She drove in the dark to leeward,
She struck—not a reef or a rock
But the combs of a smother of sand: night drew her
Dead to the Kentish Knock;”
(Probably most Hatters know that “leeward” has /u/ in the first syllable.)
I believe Hopkins found this poetic device in an obscure language of southwestern Britain.
Tom Lehrer’s “We Will All Go Together When We Go” (supra) has a linked rhyme in the first two lines as it’s printed at his site:
“When you attend a funeral,
It is sad to think that sooner o’
Later those you love will do the same for you.”
I believe Hopkins found this poetic device in an obscure language of southwestern Britain
Welsh traditional verse doesn’t actually do this, though GMH was indeed often trying to emulate
https://en.wikipedia.org/wiki/Cynghanedd
in English. English doesn’t really lend itself to this very well, though, and you pretty much have to be Hopkins to pull it off.
The things I learn…
The way he actually sang it was with a very long [l] that filled the pause between the lines and plausibly absorbed the [ɹ]. It works beautifully. (I can’t look for it on YouTube right now.)
The pronunciation dictionaries restrict this to nautical usage; but is leeward ever used non-nautically?
Sure — the Windward and Leeward islands in the Caribbean and in Tahiti, or the sides of any island in the tradewind zone or such.
They’re listed as a separate headword.
Even for sailing, how common is the “leward” pronunciation?
Leeward
https://m.youtube.com/shorts/QI4oMrH9tYM
Joe Bennett’s home sailing club is English and he trained in Southampton (don’t know where he is from).
Leward
https://m.youtube.com/shorts/O2EszrqjD0U
Jon Emmett is from Portland (or Weymouth) in the UK.
What do you say, AntC?
Hungarian, in keeping with its lack of maritime history, doesn’t seem to have a simple, native word for “lee”; it has szélárnyék ‘wind shadow’ (“possibly a calque of German Windschatten”, says Wiktionary), but that would never have survived centuries of being shouted by sailors without being worn down. The Leeward Islands are called Szélcsendes ‘wind-peaceful’.
in keeping with its lack of maritime history
Well, pre-WWII, it had a coastline, and it was once led by an Admiral…
I should have said that “leeward” is sometimes pronounced with /u/. The OED says
The only time I remember hearing it pronounced, it was /ˈluːərd/, by an American who didn’t have /lju-/, as I recall. The context was poetry—not “The Wreck of the Deutschland”, and not, I think, “The Waste Land”.
I confess I’m never sure how to pronounce that word.
I’m surprised they’re still showing /lju/ rather than /lu/; I bet that will be changed when this entry gets its full revision. They removed the /lju-/ variant from lucid and similar words not long ago. (That page should **show the date!** It was September 2024, but we shouldn’t have to hunt that down.) Meanwhile, the Routledge Dictionary of Pronunciation (2017), whose authors are also responsible for the OED3’s British and American pronunciation models, has “leeward² technical, shipping BR ˈluːəd; AM ˈluərd”.
…but the sailors were speaking Croatian, Venetian and maybe Istriot or so.
Likewise, German has words for “windward” & “leeward”, but they’re specialized nautical terms, not used otherwise*, and they’re blatantly obvious Low German loans: Luv, Lee. (…They seem to be nouns like the cardinal directions. They don’t take articles, though.)
* Even on maps, the islands are Inseln im Wind vs. Inseln unter dem Wind.
The Routledge dictionary was originally the Oxford Dictionary of Pronunciation. I remember it suddenly became unaffordable once it was taken over by Routledge. Unlike CEPD and LPD this dictionary only rarely lists variant pronunciations.
CEPT and LPD list both /luː/ and /ljuː/ for GB pronunciation.
(From ktschwartz’ link)
“For an increasing number of English speakers, words like mince and mints are homophonous.”
This caught me out in a crossword not long ago: for the setter (but not for me) “prince” and “prints” were evidently exact homophones.
I say “lucid” with /lju-/, which is therefore the only Truly Correct form.
I didn’t know about “leeward.” A “boatswain” word, evidently. Omit Needless Consonants!
I remember it suddenly became unaffordable once it was taken over by Routledge.
*utters ritual curse*
Ritual curses in hexameter.
The best kind of ritual curses.
What do you say, AntC?
I say /ˈluː(w)əd/ LOO-(w)uhd as a nautical term — because I’d be talking to other sailors.
The Islands I tend to pronounce (not that I often need to) as /ˈliːwəd/ LEE-wuhd — because I’d more likely be talking to landlubbers.
I don’t say it as /ˈljuːəd/ — that sounds like a landlubber trying too hard.
OK, I hereby adopt AntC’s pronunciations.
@DE: Welsh traditional verse doesn’t actually do this,
Thanks for the correction. I tried to figure out where I’d gotten the mistaken idea, but maybe I just made it up.
Well, Welsh traditional verse does do pretty much everything else, between multiple alliterations and internal rhymes, so it’s understandable.
Something like it can actually happen within a line:
Fel eco drwy waelod arch
“Like an echo through the bottom of a coffin”
There’s an internal rhyme od here.
(Example nicked from Mererid Hopwood’s Singing in Chains; the poem it’s from is Alan Llwyd’s Yn Eisteddfod Cwm Rhymni.)
Maybe that’s the kind of thing you had in mind.
Hopwood’s title is not, as you might think, an allusion to the joys of BDSM (though we Welsh yield to none in our embrace of Forbidden Pleasures), but refers to this poem:
https://poets.org/poem/fern-hill
[I’m not a great fan of Dylan Thomas, myself. I have tried.]
I quoted the “held me green and dying” bit at the end of an exam bluebook when I was 18 years old on account of having run out of time to try to coherently say what I was trying to say. The professor (https://en.wikipedia.org/wiki/Ramsay_MacMullen) commented positively on the poetic quality of the lines, but the grade for what I had managed to write was what it was.
Watering the Eddyshaw:
Abbie Hantgan’s study of loanwords in Bangime, The Small Bang: A pilot study investigating the origins of a language and population isolate through loan words. What do you think? At a glance, it looks careful and interesting.
I’ll have a look. She’s a fine scholar: I’d expect it to be good stuff.
Impressively aporetic …
I was a bit troubled by some of the assumptions about semantic drift, but it’s interesting stuff.
Of course, I liked the point that, not only is Bangime itself an isolate, but none of the neighbouring languages is uncontroversially a member of a larger family (like “Niger-Congo” or “Nilo-Saharan.”)
Naturally I was on the lookout for Oti-Volta-like forms, but unsurprisingly there wasn’t much. Surprising there were any at all, in fact.
The only really convincing one was various Dogon forms along the lines of màànàdʒí “okra”, which must surely have some connection with POV *mâ(ɦ)n-ʎɪ́ “okra.” (e.g. Kusaal ma’an.) I see that even Manessy had “proto-Central-Gur” *man.
Interesting on Bostoen’s Bantu *-bèdé “millet” being “an early loan into Nilotic, where it means SORGHUM today.” This actually looks rather like POV *bɪ́-fʊ́ “seed” (and various verbs meaning “ripen”), but these very short forms could easily resemble one another by pure chance, and the “seed > millet” shift seems a bit too handy, for all that it’s clearly possible. (The Gurma languages have all shifted POV *dɪ-fʊ “food” to “millet”, including as the name for the growing plant itself.)
The Dogon “fonio” words look very like the Eastern Oti-Volta, e.g. Ditammari pl īpɔ̄kà, Mbelime pl pūòhḭ̄, but it’s not reconstructable to POV and I think this is all just coincidence.
Bambara tìɡá “groundnut” looks a bit like POV *tím-ʎɪ́, especially given the fact that *m seems irritatingly often to turn up as a meaningless derivational suffix in Oti-Volta. But again, such short forms can very easily resemble each other by pure coincidence.
Still, the Dogon and Oti-Volta “okra” words certainly look connected, which is interesting. Wanderwort!
“For an increasing number of English speakers, words like mince and mints are homophonous.”
Example in the wild from the Financial Times crossword today:
Orders report on getting off. (5)
No offense beyond what is warranted, but… your review of their work here is so snide, uncharitable and distorted that I no longer think you have “any grasp of the basic principles at all” of actually demonstrating a relationship, instead of merely taking one as already given.
Either stop lying or actually read the work you are pretending to criticize. P & V make good, serious effort to minimize this bug by a priori limiting their comparison to the 700-ish most strongly established Proto-Bantu reconstructions.
I must highlight this as your most crucial failure, undermining everything else you say. You cry “it’s all cherrypicking” falsely and without substantiating. Cherrypicking is a lexeme-specific problem. Let me agree for the sake of the argument that they have, say, erred in cherrypicking a a regional Bantu term for ‘bone’ over the more widespread one. What precisely does this imply? 200+ other cognates will still remain, easily enough to constitute proof of relationship. Maybe you’ll go on to show dozens more of them to be cherrypicked! Again, so what: the remaining 100 would still stand as a proof of relationship. This form of evidence is inherently parallel, not serial, and is for that reason so fundamentally strong. It cannot be dismissed as a “Gish gallop”.
You have far more work to do in claiming their work to be “crankery” than, indeed, cherrypicking yourself a handful of comparisons that you see problems in — and using them as an excuse to dismiss everything else. This critique is lazy and dishonest.
—
No, his work e.g. in the recent Oxford Handbook of Atlantic is methodologically sound (and cited also for Glottolog), and also produces results rather similar to what you also complain should be recognized. Pozdniakov is no reflexive lumper: he excises basically the entirety of “South Atlantic” as not sufficiently demonstrated as related (including to each other), and limits the family only to what has been “North Atlantic” in earlier classifications. (He seems to still hold on to a working hypothesis that Mel, Limba etc. are all unassigned Niger-Congo; I would be more skeptical on this, but it is at least a possible hypothesis.)
False. This is entrenched denialism from someone who wants every excuse possible to deny a relationship.
The Uralic nature of Samoyedic, for an example I know extremely well, was established to general satisfaction by direct lexical comparison of Nenets(ic) and modern Finnish, with only minor support from anything else in Uralic and nothing else in Samoyedic. The error rate in this procedure was harsh; ca. 50% in a wordlist of 120 items; but the result stands and does not turn into “crankery” just because it was not routed through every assumable intermediate. And this would have been itself wrong — one of those intermediates would have ended up being “Proto-Finno-Ugric”, a pseudo-protolanguage we today think does not exist. After about two centuries of careful study, Finnic appears to be genealogically no closer to Ugric than it is to Samoyedic, even if Samoyedic has been lexically the most innovative of the three.
Quite analogously, why are you even so sure there would have to exist a clearly more recent ancestor of Bantu and Oti-Volta, than of Bantu and Mande, if no one has a solid reconstruction or clear subclassification of Volta-Congo? The appearence of P & V’s work has shifted my thinking more positive to the idea of Mande losing Proto-NC morphology by partial or full creolization (and not as per Greenberg etc. as an early outgroup that would leave before class morphology develops). If this is true, it could be in principle nested somewhere within conventional Volta-Congo — which, if we separate the divergent Adamawa and Ubangian pieces, looks itself very, very diverse already.
Lexical correspondences in basic vocabulary that display regular sound correspondences are valid evidence for a relationship. This is sine qua non for comparative linguistics. Every established language relationship can be shown by this evidence, and this evidence alone. Morphology is not at all a necessary criterion, given in particular the existence of isolating languages (though whenever present, morphology can certainly help greatly). Your ignorant blathering is equal to complaining that surely modern English cannot be Indo-European, it does not even possess a case system! (Or a reflex of *tu!)
If the point is not clear for someone, I recommend for one antidote going back to the Automated Similarity Judgement Program world-tree-of-languages results and checking just how many valid groupings end up being indeed recognized in it — with far more coarse (prescientific, even) methodology that has only finite understanding of phonetics and none of regular sound correspondence. Even most decently old/large families like Athabaskan, Central Sudanic, Dravidian, Nilotic, Northwest Caucasian, Turkic, Tungusic, Uralic turn up generally fine, and usually with their basic subgroups also well resolved. All this from nothing but a list of 40 basic words!
Correct, and given enough of them with enough regularity, this constitutes a demonstration of relationship. We could debate what is “enough” in either sense (I would think that 30–50 good examples is probably sufficient, and will insist that 100+ is beyond all reasonable doubt), but there is simply no discussion to be had if you cannot or pretend not to understand this elementary point at all.
One cannot find arbitrary amounts of good-quality lookalikes between arbitrary languages. The way to “generate” lookalikes is to relax judicious selection to freestyling semantics, and individual-language or shallow-subgroup cherrypicking. But what do we have in this study? “Found in the majority of Bantu subgroups” is not cherrypicking, it is a strong sieve. “Shares exact semantics and is found on the Swadesh list”, as in the majority of P & V’s comparisons, is the farthest possible thing from semantic latitude.
This would be the important principle that “reconstruction depth helps”. Proto-Mande and Proto-Bantu are old; therefore if related, they will have had less time to lose cognates, unlike narrow late groups like Leko or Oti-Volta. P & V indeed also show this remarkable resemblance does not exist if you instead take modern Mande and Bantu languages. That is, the reconstruction of Mande and the reconstruction of Bantu independently converge towards a single proto-language. This is exactly what should be expected! Latin and Sanskrit look more closely related than Spanish and Hindi do, and they even might look more closely related than Hindi and Persian do. Not only in their grammar but also in their basic lexicon.
“‘Moon’ can be reconstructed for Proto-Italo-Celto-Germanic: PC *mīns PI *mēns PG *mēnōþs. Therefore, it is this that should be compared with Proto-Slavic, not the regional Italic form luna.”
Conflicting lexical isoglosses happen! They may have even been actual proto-synonyms; or there may have been later drift towards similar directions, as we believe is the case in this IE example. This is not itself a reason to reject a comparison that works otherwise. This should be textbook understanding.
Even more damningly, if you actually read P & V’s work instead of just presuming from your high horse that surely it cannot work, you will also see (p. 153) that their comparison is not restricted to Bantu and Mande. *bú turns up likely cognate candidates widely in Benue-Congo, in Ekoid, Limbum, Kom, Ring, Upper Cross, Jukun, Edo; and parts of “Adamawa”, “Kru”, “Ubangian”, “Atlantic”. Best of all: a majority of these mean ‘dust’, which is also a sense found in Bantu and Mande! Therefore there is not even any real need to think this would be a competing Proto-NC term for ‘ashes’. More likely it is the Proto-NC term for ‘dust’, which in several languages (both some Mande and some Bantu) has shifted to ‘ashes’. Precisely analogous to the case of a Latin–Slavic comparison for ‘moon’.
There may be a formal source of error in listing this comparison in various tables as if were the Swadesh list item ‘ashes’, but to seize this as an excuse to dismiss the comparison entirely is again lazy and dishonest.
This is a methodological effect that follows from the fact that contact of specifically Bantu and Mande is implausible. Much of their comparative corpus consists also of terms for which borrowing is not particularly likely; in this light the wiser approach is to treat the full corpus as inherited, unless actual reason is found to reanalyse particular items as borrowings from this or that northern relative of Bantu.
Maybe ‘cow’ is indeed a borrowing. But are we expected to just believe for no reason that so are ‘nose’, ‘eye’, ‘tongue’, ‘tooth’, ‘mouth’, ‘blood’…
There is an entrenched ideological position (I have thought of calling it “Campbellism”) that pretends there exists such a thing as “widespread extensive borrowing” extending into core vocabulary, and uses this to deny more or less every language relationship proposal deeper than ca. two millennia that has not been textbook knowledge for at least a century. But no such thing exists. Even the most utterly loan-filled and relexified crypto-varieties like Helsinki slang (often noted e.g. for losing vesi in favor of a Russian loan voda for ‘water’) can be still typically classified lexicostatistically in their correct families (ASJP does just that).
Sigh. Either stop lying or actually read the work you are pretending to criticize. P & V do not equate these two roots, they recognize them as distinct: the former at pp. 320–321; the latter, since it does not possess a Mande cognate they’d have found, only in their outlines of the core vocabulary of Bantu (pp. 31, 127) and surveys of PB cognates elsewhere in Niger-Congo (chapters 5–6). Possibly some reflexes have ended up in the wrong etymon for now in their surveys, but this does not equal “all the same”.
Damned if you do, damned if you don’t. Comparing Mande and Bantu binarily seems to be, in your opinion, skipping too many steps. Yet comparing not only binarily would then be “absorbing yet more unrelated languages into the Blob”?
I am happy to see e.g. Atlantic has not been dragged into the fray, but the same principles that lead to its exclusion (lack of a Proto-Atlantic-Congo reconstruction that might be employed) will also lead to the exclusion of your own pet branch (lack of a Proto-Volta-Congo reconstruction that might be employed).
P & V make no such “assumption”. They merely do not find a regular correspondence to substantiate anything more; their Mande–Bantu reconstruction is not claimed to be a comprehensive model of Proto-NC. This is how we actually work out proto-language inventories (“answers”), you know? by establishing regular correspondences. If they’re regular then they’re regular, period.
This can and will allow one to, say, find regular correspondences that imply a loose approximation of Proto-Malayo-Polynesian from the comparison of Malay, Javanese and Tagalog (as was actually done in the foundational work on comparative Austronesian). Such foundational evidence does not somehow turn invalid just because e.g. Kaibobo or Tboli-Blaan was not considered in it. Maybe the evidence of Kaibobo still changes some of the specifics, but you’d have to actually show this, not just assume that it might.
—
Obviously I am not in unconditional agreement with everything P & V claim. The section on “FOUR and TONGUE” in particular is, yes, a provocation for little reason and would have been better omitted from the study. And pigheaded skeptics are, of course, also going to latch on these couple of paragraphs while discounting their other work. They are still correct in that claim, but it is only so post hoc, after already knowing these lexemes end up unexpectedly strong in Niger-Congo, and thus any principle like this would not be usable in working out the family’s boundaries beforehand. Some of it may be from the fact that they are built on particularly stable root consonants *m, *n, *l, *j, but this cannot explain the full effect; e.g. the relative stability of *j is strongly subject to macro-areal phenomena.
Any serious knowledge of multiple language families regardless shows that every one has phenomena like this, lexemes that are conserved beyond the average expectations and trivial to notice in all but the most phonologically outlandish daughter languages. *lima for ‘hand/five’ is a marker of this sort in Austronesian; *kala ‘fish’ is one in Uralic; in Indo-European one might pick *bʰer- ‘to carry’; in Semitic, *ʕayn- ‘eye’. They have not accidentally risen for “phonaesthetic” reasons, they are due to common descent.
—
I had to sleep on this comment in general. My assorted critique-of-critique above is strong, but I must go further. I have in the past put some stock on mr. Eddyshaw’s various critiques on the validity of Niger-Congo comparative work, but this I can no longer do. I do not say this lightly, and do so with my own seriousness as a linguist riding behind it: the man is on this topic a dilettante and luddite, a clueless buffoon hopelessly stuck in his own miniscule corner of the family. Not understanding a milestone in African comparative linguistics when it stands in his face could be forgivable. But he even opts to outright insult much more remarkable and accomplished linguists than himself as “cranks”, and dishonestly distort their work, because he is, yes, pre-disposed to dislike (and in denial about this) the results of their thorough and careful work. P & V’s work is exemplary and will be taught in textbooks for centuries to come. Meanwhile I advice all Languagehat readers to assign no credence or expertise to anything D. E. may claim on this topic.
Your long comment is so nasty and over-the-top that I will henceforth assign no credence or expertise to anything you say. You might want to reconsider your communicative style; once you start accusing other people of lying you might as well walk around with a sign saying “I am a crank.” And don’t start a comment with “no offense” when you are clearly dying to cause offense.
He is either lying or not even trying. I agree the latter is the more charitable interpretation and am refraining from locking down the former accusation (though for now I cannot be actually certain if I am correct in this). And clearly I do intend the due amount of offense at this drivel from D. E., the precise extent of which may remain to be specified.
If my sounding of the dishonest-reporting alarm is unwelcome: very well, I have liked your blog and its community in the past, but I have no inherent need to continue reading or commenting here. Take care people; those of you who care about truth more than cozying up to friendly buffoons know where to find me.
@JP
As one who has enjoyed your past contributions on this site (specifically those on or related to Uralic), I would be sad if you decided not to post here anymore.
You clearly felt you needed to better inform readers unacquainted with the results of the P-V paper. Those parts of your post were interesting and thought-provoking.
But you seem to have (or write in such a way you suggest you have) additional objectives that you might do better to unpack:
1. You seem to care very much about uninformed criticism/deliberate distortion of methodology used in longer-range comparative work like the P-V paper. This is “ein weites Feld”, and you may yet find you mostly agree with DE and could find, on reflection, the degree of passion you employ in your writing to be embarassing.
2. You seem to want people to be punished or humiliated for expressing their opinions, however ill-informed. I would suggest that, if you wish to be a teacher, mentor or collaborator, rather than a “lone wolf”, you might want to “dial this down” and prioritise communication of your knowledge and insights.
I don’t understand the accusation of lying either – why would anyone lie about such things? – but the actual substance of the actual linguistics in the comment looks pretty sound to me and shouldn’t be dismissed for the tone of the comment it’s in.
You do notice you’re doing the exact opposite here…? You’re responding to one wild accusation (of lying) with another (of wanting punishment) and add some patronizing on top.
but the actual substance of the actual linguistics in the comment looks pretty sound to me and shouldn’t be dismissed for the tone of the comment it’s in
I agree (up to point, anyhow.) I was, in fact, hoping that Pystynen would comment to redress the balance of my own negativity. It actually is true that I have a splitterist bias (which I am hardly unforthcoming about) which may well cloud my assessment of things like this paper; it is also undoubtedly true that I am a “dilettante”, whereas he is actually paid to do this sort of thing. (I have in fact commented myself elsewhere on the difference that this makes: even a very committed amateur is greatly handicapped by lack of an adequate academic training and support network, and I have specifically lamented that there is so little professional interest in proto-Oti-Volta. It is really unfortunate that no academic linguist knows more about that than I do.)
I see no point in trying to engage with his ad hominem stuff.
I will say that if P & V were commenters here, I would have been less rude about them; I think the tone of some of my own comments probably played a role in bringing down this rant on my head.
@dm
I was referring to the following paragraph in bold type in JP’s post:
—
I advice [sic] all Languagehat readers to assign no credence or expertise to anything D. E. may claim on this topic.
—
I read this as “wanting punishment”. I am sorry if I have misunderstood or misinterpreted this paragraph (it is possibly just some kind of summing-up of his previous arguments, but for me it seems to go beyond that).
Re patronising, I genuinely wanted to be helpful and highlight what was for me unnecessarily emotive language, which made it more difficult for me to engage with the content.
Well do I know how frivolity can suddenly be thrown back in your face. Much like DE, I am an incorrigible tease. However, he knows much more than I do, as compensation. And he has better manners, which cannot be said of Pystynen. Nobody gains from P going all Nicolas Cage on DE (Ghost Rider: Spirit of Vengeance).
A salutary reminder to myself to not fly off the handle over a bunch of ideas. There is a topic in IT which gnaws on my liver. I was so close to exploding in rage. Not gonna happen now.
Didn’t I say “Y’all play nice, now”? I did!
+1
(not sure if this convention is Russian or international)
I think some of JP’s criticism is deserved but not by DE.
A fierce attack on the Hattery wihout calling a specific person a “buffon” would make more sense.
It would make sense, would it? So everyone here is a liar except DE? Or what?
@LH, DE’s comments were writen according to our local norms and expectations.
Also I think JP knows Vydrin in person or if he does not, he respects him and respects Pozdniakov. You, in turn, respect DE. Of course you both are displeased.
While I’m not ready to criticise aforementioned “norms and expectations”, I see why JP’s anger is predictable.
I took a look at P&V. To start with, I have no skin in the game. I know nothing about any of these languages, and have no opinions about their classification except what I read here and there. I should also say that both Eddyshaw and Pystynen have done careful and methodical historical reconstruction on their language families and adjacent ones over many years, and my initial inclination is to trust both of them.
That said, P&V’s work here really does look to me Greenbergian in its quality. That is to say, completely unconvincing. To begin, P&V grade their cognate candidates, reasonably in principle, by criteria of phonological similarity, semantic similarity, and universality of distribution within each group. With the aim of demonstrating an above-chance similarity, they collect their 10 phonologically best-matched words out of the basic vocabulary list they use (table 7, p. 36). I use these as an objective sample of their work.
For each set I give the forms as given by them: gloss, Proto-Bantu, and Mande. Each pair is followed by the total number of comparanda in their lexicon containing that initial consonant correspondence, e.g. *d- : *d- occurs in four examples. I checked most of the Bantu reconstructions in their source for them, BLR3.
Gloss — Proto-Bantu — Proto-Mande
1. bark (of tree) — *koba — *koba (4)
Proto-Bantu *kóbá ‘skin, strap, girdle’. Mande *koba ‘bark’, based on Mogofin kóbólé ‘skin’, S. Bobo kōbā ‘bark, scales, shell’, Jalonke kóbólé ‘bark’, Susu kóbólè ‘bark’.
It seems that the Mande is best reconstructed as ‘skin’. The shift of ‘skin’ to ‘bark’ is much more likely than vice versa. ‘Skin’ seems a better gloss for both sides. The Proto-Bantu term seems like a specialized term. There are about a dozen other terms for human skin, animal skin, or fruit skin, and this one was cherry-picked for its resemblance.
2. bite/eat — *dʊ́m — *doɓũ (4)
Proto-Bantu ‘bite’. Mande ‘chew, eat’ (most reflexes mean ‘eat’)
3. long — *dàì — *dãa (4)
PB basic meaning ‘long’; derived meanings ‘far, long ago’. PM ‘far, distant’
4. lie (down) — *dáad — *ɗa (2)
PB ‘lie down, sleep; spend night; be fallow (field)’. In Mande more than half the glosses are transitive ‘lay down’, ‘put’, etc.
5. cold — *dìdì — *ɗɛ̃ɗɛ̃ (2)
PB ‘become cold, shiver’. Mande ‘cold’.
6. egg *gɩ́, *gɩ̀dí — *ɣeli (1)
PB gí, gé, géjí, gìdí, gìjé, gìjí, jájí, jágí, jákí
7. road — *gɩ̀dà — *ɠila (2)
8. go — *gɩ̀ — *gɛ (1)
9. say — *tɩ̀ — *tĩ (4)
10. walk — *támb — *taŋbi (4)
Mande ‘pass, surpass, go across’. Limited distribution in both families.
Looking at these, there is not a single term that looks like a good match phonologically and semantically at the same time. Some look similar enough, but are not: (5) on both sides one reduplicated syllable consisting of a consonant, a vowel, and a tone. All three in dì and ɗɛ̃ are as different from each other as the phonemes in English bit and peed. As a group, they do even worse. There is no regularity to be discerned for anything except *a : *a.
Other comparanda which got a high ‘reliability index’ (table 5, pp. 32–33) are phonologially worse: ‘mouth’, *nʊ̀à ~ *ɲʊ̀à : *ɗɔɣa; ‘dog’, *bʊ́à : *gbɛra; etc.
A final note about the Cow. DE was right to pick on that one, even if it isn’t representative. If there were no cows in Africa at PB-PM time, there was no word for them in the protolanguage. That’s basic. If I saw a reconstruction of Proto-Germanic containing a word for ‘telephone’, that would immediately color my impression of the quality of the entire work, and would make me suspect every other statement it contained.
@Y, I agree about cows.
However, there are wild cows in West Africa. Technically I can imagine similar semantic shift in many languages (in the context of similar culture) and it even could be a shift from what I call a ‘cow’ (wild) to what I call a ‘cow’ (domestic). I find that strange but I don’t know why exactly.
ASJP (“Automated Similarity Judgment Program”) got popular for a little while. Because computers, I guess? The idea was to take small vocabularies of thousands of languages, put them through a fairly simplistic word matching algorithm, and see if language families fall out. A number of versions of the global family tree are here (at the bottom of the page).
It succeeds in discerning young language families, though the internal subgrouping is often scrambled. At a greater time depth it succumbs to noise, and comes up with howlers, like Australian languages in Africa or American ones in PNG. It tends to easily combine unrelated areal groupings: I suspect that that is partly because of borrowings, but especially because at the noise level, languages with similar phonotactics will have more chance resemblances. I suspect that for that reason it recognizes some deeper legit families as well, not because of an actual lexical signal.
They included some artificial languages, which make for a nice control. In version 5, Sindarin and Quenya cluster with Maiduan of California, and Solresol with the Bolivian isolates Movima and Itonama.
Despite all that, the authors of that program can’t resist seeing newly discovered groupings in the far depths of the tree.
If anyone wants to try their own hand at comparison with proto-Mande, Vydrine’s comparative lexical database is, commendably, online: https://cormand.huma-num.fr/mandecomp/interface.php.
Personally, having read through the whole paper but never engaged much with proto-Bantu, I found it far superior to Greenberg – I don’t think that kind of sarcasm was warranted – but not as convincing as Pystynen does, insofar as the proposed regularities seem to be confined to the initial consonants. (And, yes, “cow” is a bit of a problem, though not necessarily an insuperable one – at first glance, “camel” seems to reconstruct back to Proto-Semitic…)
I found it far superior to Greenberg – I don’t think that kind of sarcasm was warranted
Yes, that was unfair – apart from anything else, they are actually looking for genuine phonological regularities, unlike Greenberg. It was intended as a bit of hyperbole, though. (As I said, if V & P were commenters here, I’d have been less sarcastic, or indeed not sarcastic at all. The amount of sheer work that’s gone into this deserves respect, too.)
The particular series I commented on were by no means the only ones I thought were seriously flawed.
I picked “cow” partly because it’s one that I have previously wondered about, and is a case where borrowing has quite certainly occurred in Bantu, and because of the mismatch between the quality of the evidence and V & P’s conclusion that it can “undoubtedly be reconstructed for Proto-Niger-Congo.”
Still markedly not Chadic.
And definitely not Indo-European …
“Cow” has been reconstructed for proto-Chadic:
https://en.wiktionary.org/wiki/Appendix:Proto-Chadic_reconstructions
As *ɬ-: vowels are for the weak.
By morphic resonance, a clue from the Independent crossword today:
Be visibly upset about that woman’s photo, man being biased? (6-7)
@Y: Thanks for the interesting comparisons. If those are the best phonological matches, I agree that they’re unconvincing (Edit: to someone like me who has never taken a linguistics class and may not be entitled to an opinion). But JP brought up the evidence that’s considered convincing for other families. It makes me wonder what the ten best matches between two Indo-European groups believed to be of similar age would look like. Or Uralic or other well-established relatives.
Edit: Or the complementary question: What would the ten best matches between two unrelated groups look like? Then, how much more do you learn by going to 30 or 50 or 100?
But these questions are not assignments for anyone.
A fragmentary map of cows: https://clics.clld.org/parameters/1007
Nice map!
Presumably the colour codes are families/phyla (though Kru has been banished from Atlantic-Congo, which seems harsh.) They evidently aren’t labels for presumed cognates.
Nice to see a couple of Oti-Volta ones, though I’m sad to see the wrong tone on the ethnonym for Mossi. (I wonder what the source was? The orthography of the forms given for Mooré, Gurenne/Farefare and Kasem is non-standard. Probably Manessy.)
Gulimancema (o)nùa (plural nè) actually is from the POV *naag- root, despite the vowel: the loss of the velar is regular in Gurma, and the vowel resulted from *naao > *nao > *noo > nua, where -o is the singular suffix for the “animal” class-pair/”gender.” The plural is similarly from *naai by monophthongisation: Gurma doesn’t have the actual passing-over-a-consonant umlaut of WOV (Kusaal naaf “cow”, plural niigi.)
In Proto-Bigot we have the alternating or reduplicated stem *niig-naag
It was international when Google+ still existed. I don’t know how widely it’s still known.
I was wondering about rind… but it’s probably a pretty recent derivative from a verb whose former causative is rend, so it probably started out with a much more abstract meaning and specialized later.
Wow, that looks harder than the IE one!
Some such work has been done; I can’t remember where to find it quickly.
What would the ten best matches between two unrelated groups look like?
Oti-Volta/Germanic (interpreting “best” as “most-regular-looking for initial consonant correspondences”):
Most initial consonants simply correspond exactly:
Kusaal bɛ’ɛd “bad”, English “bad.”
Kusaal bʋd “sow”, English “bud.”
Kusaal bin “excrement”, English “bin” (“reject, throw away.”)
Buli būuk “goat”, English “buck.”
Buli bàūk “swamp”, English “bog.”
Kusaal bul “spring up” (of water), English “billow”
Kusaal daar “day”, Old Norse dagr “day.”
— (In both languages, the -r is not part of the stem but a singular suffix: so a morphological match too!)
Farefare dàrgɛ̀ “drag”, English “drag.”
Mooré gẽde “mix up”, English “guddle.”
Ditammari kēētá “cut”, English “cut.”
Kusaal sɛɛd “plant” (verb), English “seed.”
Mooré sẽ “sew”, English “sew.”
Ditammari sipí “sheep” (pl), English “sheep” (also pl)
Mbelime hèndè “horn”, English “horn.”
Byali lōú “log”, English “log.”
Kusaal m “I”, English “me.”
Kusaal mu’ar “lake”, English “mere.”
Nateni níkú “neck”, English “neck.”
Dagbani yi “you” (pl), English “ye.”
Kusaal ya’am “yawn”, English “yawn”
Moba yènd̀ “one”, Scots “yin.”
However, in Oti-Volta, proto-Oti-Oder *f *θ fall together with *p *t:
Kusaal pɛ’ɛl “fill”, English “fill.”
Mooré pɛɛla “white”, English “pale.”
Kusaal tɛk “pull”, English “take.”
Byali tēēgə̄ “leg”, English “thigh.”
Buli tāin “stone”, Scots “stane.”
Kusaal tiig “tree”, English “tree.”
Mooré tã “three”, English “three.”
Further work will undoubtedly add to the evidence for this relationship.
I see that P&V do address something like my questions about top-ten lists.
Thanks for all the cognates, David E. I appreciate it that you didn’t cheat by using the well-known Scandi-Congo family (except once).
Lameen, P&V’s work is uncontroversially better than Greenberg in that they know the languages, and their data is accurate. Otherwise I don’t see how it is any better, and section §5.3 (p. 116), Greenberg’s hypothesis 80 years later, explicitly defends him with only minor criticisms of a few supposed etymons. They say,
How is calling them “Greenbergian” even an insult, from their point of view?
And what are these initial sound correspondences worth, anyhow, if the rest of the etymons don’t match at all? They seem like window dressing. You could add similar tables to any of Greenberg’s or Ruhlen’s proposals, and they would look similar: quite a bit of identical initial consonant matches, a few multiple correspondences with no explanation, and the rest of the word pairs not matching at all.
Even the toughest distant language relationships — AA, Sino-Tibetan, Armenian-French, what have you — started off with a few secure anchors, which hinted at the next steps. I don’t see a single such anchor here.
@JF:
Yeah: my most egregious sin in making that list was to cherry-pick comparanda to fit the hypothesis that the initials would correspond “regularly.” The way to avoid that (at the stage where it is still not known if the languages involved are truly genetically related) is to decide beforehand which words you’re going to compare (e.g. with a Swadesh list-style thing) and to stick to it pretty closely (avoiding semantic latitude as much as possible.)
P & V do in fact try to do this. Chwarae teg iddyn nhw.
I didn’t (much) exploit the fact that there are some thirty-odd Oti-Volta languages to pick lookalikes from, though I did abuse the vast number of English synonyms and near-synonyms. Proto-Bantu is uncomfortably like English there. This all introduces a great degree of freedom to pick lookalikes that will fit the phonological hypothesis.
And I only went for excessive semantic latitude a leetle, really. And no more than I’ve seen in some real cognates.
I mostly didn’t silently incorporate flexions into the stem I was allegedly comparing, though I do feel proud of Ditammari sipi “sheep”, where si is a plural noun-class prefix, and the stem is actually cognate with that of Mooré pésgò “sheep” (singular.) That one is true Mass Comparison level, though I say so myself.
I have compared forms shamelessly from individual languages instead of proto-Oti-Volta in cases where actual reconstruction within OV shows that the comparison cannot possibly be valid: for example, Dagbani yi “ye” derives from *ɲi, not *ji.
I could have come up with a lot more: that list took me about thirty minutes.
How is calling them “Greenbergian” even an insult, from their point of view?
I can’t vouch for their point of view, but in my own view Greenberg got considerably laxer in his standards over the years, even before he moved on from Africa to other continents. 1950s Greenberg was perfectly happy to admit an Africa with dozens of isolates or small families, alongside a couple of big families including Niger-Congo. I don’t think his methods were really sound at any period, but I think it would be a coherent position to defend early Greenberg while rejecting his later excesses.
Did Greenberg ever do proper historical linguistics, or did his early reputation rest mostly on his editing out the dross of earlier classifications, like Hamitic?
The shift of ‘skin’ to ‘bark’ is much more likely than vice versa
“Bark”, “shell” and “(fish) scale” go together in Oti-Volta: POV *pàk- or *pèk-. I don’t understand the morphological relationship between these roots (assuming there is one); there’s also POV *wák- “bark” or “scale.” But (human or animal) “skin” is always different: POV *gbãn-.
This seems to hold true for Grũsi, too, though the “skin” etymon, at least, can hardly be related to the Oti-Volta one: Kasem pʋgʋ “bark”, “scab” (on a wound), tɔnɔ “skin.”
Koromfe has pogəlle “bark, shell”, tɔ̃nɔ̃ŋ “skin.”
So “Central Gur” seems to show no propensity to conflate these.
On the other hand, Dyula has fàrá “scale, skin of certain animals, skin of fruit, peel, bark.” (Doesn’t say which animals, unfortunately.) But the dictionary gives gòló for “skin of human or animal.”
What I’m driving at here is that it would be good to try to get some sort of idea as to how likely a particular semantic shift or conflation is in a particular region.
For all I know, it just happens to be the case that “skin” and “bark” aren’t expressed by the same word in any West African language I’ve looked at (a minuscule fraction of all those that there are, and regionally limited too.) Perhaps it is perfectly usual for one word to have both senses in a lot of Mande languages. But it would be good to see the evidence.
Our subjective feelings about how likely such things are can’t be relied on. Some unlikely conflations (from a SAE standpoint) do turn up regularly in Oti-Volta, for example, e.g. sky/rain, stone/mountain, tendon/root. They all kinda make sense in retrospect, but they’d all strike me as quite a stretch if there weren’t actual evidence for the conflation, often within a single language. It wouldn’t do just to assume they were cromulent in comparative work without such evidence.
*dum “bite” turns up all over in Volta-Congo (and even in Bak, it seems.) It seems always to mean “bite”, and not “eat.” If it is only citeable as “eat” in Mande, I think that’s a problem with the equation that shouldn’t just be waved away. Sure bite -> eat is possible; still awkward if the meanings consistently differ between your group that you are trying to show is not an isolate and all the languages currently accepted as uncontroversially related.
For some context, I should perhaps say that *dɪ “eat” is about as plausible a candidate for a proto-Volta-Congo root as anything that there is. For all my strictures about premature setting up of protoforms, I am happy to bet anyone a thousand pounds on that one turning up in proto-VC, if we ever do get to the stage of actually properly reconstructing it. (By then, of course, a thousand pounds will probably just about buy you a beer. We’ve no idea how many distinct consonants proto-VC had, even. More than proto-Bantu, less than Ébrié … maybe …)
Gérard Dumestre’s 2011 Bambara–French dictionary has fàra as both ‘peau, épiderme’ and ‘écorce des arbres’, p. 294.
I remind that CLICS is a database of colexifications.
Perhaps it is perfectly usual for one word to have both senses in a lot of Mande languages.
In CLICS it’s a number of “Atlantic-Congo” languages and Mandinka: 1 out of 4 “Mande” languages in the list* of “bark” words (Mandinka, Soninke, Bambara – 2 words for bark, neither is fàra, Jenaama).
To my surprise Maltese ġilda:/ Wiktionary does not confirm that, Maltese “bark” is qoxra /ʔɔʃra/ (Ar. qišra) there, but I don’t have a Maltese dictionary and don’t even know if j-l-d can be applied to trees in Maghreb.
*Click “concepts” in the upper menu, in the field “name” type, say, bark and you will see a link to their “bark” page.
My perspective (I’m repeating myself in slightly different words here):
There is a certain threshold such that once similarities exceed it, we MUST interpret it somehow.
When those are found in basic vocabulary (rather than the local vocabularies for telecommunication:)) interpretation will require search for correspondences, exploration of local wanderwords, tests like that practiced by G. Starostin (more lookalikes among 25 most stable words than among words 26-50 will likely mean both are Niger-Congo, more lookalikes among words 26-50 than among the first 25 will likely mean contact).
But the question Q: “has this threshold been reached?” or else “how exactly similar these two lnaguages?” is arithmetical.
You can instead formulate it differently:
There is a thereshold within such that once exceeded within basic vocabulary, both langauges must be NC.
P and V prefer this formulation. I don’t understand why I should prefer it. Let’s answer Q first.
As for interpretation, we will learn more in the course of it.
drasvi: I don’t think anyone has ever defined that threshold well, and I’m skeptical that any single statistic would do it.
It does not reflect how science progresses in general. It’s not like you go from no to yes in one big leap. The question would be more like, is there a hint of something worth investigating (which might be a false lead)? Here, as I said, I haven’t seen any such hint.
Y, of course it’s a function: amount of attention we must give to it as a function of (similarity) (me)
And, accordingly, likelyhood of a given interpretation of similarity (say, that Mande is NC) as a function of (similarity) (P and V)
I only formulate my function in terms of “thresholds” because others prefer to think in thresholds. However, I still think that my formulation is better. It is a statistical or arithmetical question, it CAN be answered.
To me the question I ask myself isn’t “Are A and B related” (previously not shown to be related) but “Is there something suggestive of a relationship between A and B, drawing me to investigate further?”
@Y, I agree.
However, “too many” lookalikes in the list can also mean contact, and information about contacts is interesting too. Yet, first we need to know if there are too many lookalikes.
Also this reminds me Semitic-Berber-Chadic-Egyptian-Cushitic. There are meaningful results of comparisons between Semitc and Berber. Or Berber and Chadic. Some of them seem to point at old contacts (Lameen wrote about those), some are interpreted differently.
There is even consensus that such comparisions are not “crank” (I note that DE applied this word to the comparison, not to P or V personally, he didn’t call them “cranks”). But.
“I agree”
Moreover, all langauges can belong to one tree whose root is Proto-World. And? If God tells you that about Russian and Khoi-San or Mande and Kusaal or whatever pair – what this information will mean to you?
Nothing at all. I’d say “huh”. I find the details of historical linguistics fascinating. A family tree by itself, not very much.
Some unlikely conflations (from a SAE standpoint) do turn up regularly in Oti-Volta, for example, e.g. sky/rain, stone/mountain
Of these, stone / mountain doesn’t strike this speaker of SAE as unlikely. Look at what Americans call rock (from something you can hold in your hand to the one at Gibraltar); German Stein only means “stone” as a normal noun, but turns up in hundreds of names of mountains as second element.
“crank” (I note that DE applied this word to the comparison, not to P or V personally, he didn’t call them “cranks”)
You note correctly: I didn’t call them cranks, and I don’t think that they are.
Perfectly capable scholars have published comparative work that doesn’t stand up to proper analysis: a grisly example in African scholarship is Neil Skinner’s comparative dictionary of Hausa. Skinner was an eminent Hausaist; the dictionary is all but worthless. (This is not some “bigoted” view unique to myself: you can find a tactful review of the work by Paul Newman online.)
Skinner was an eminent Hausaist; the dictionary is all but worthless
Can confirm. It’s not that unusual, in fact; if I wanted to be undiplomatic, I could name notable Arabist linguists in the West with major achievements to their credit who nevertheless get the absolute basics of historical linguistics completely and publicly wrong, despite ample opportunity to learn. (In the Arab world, professors who insist on being vocally wrong about historical linguistics usually have obvious ideological and/or economic motivations for doing so, making it a rather different story.)
Incidentally, my feeling that Pozdniakov’s comparative work on Atlantic is methodologically unreliable is shared by at least one professional linguist who has also done comparative stuff in Atlantic (a fairly select group of people.) I’m away from my books just now, but could cite chapter and verse later if anyone cares.
“historical linguistics” – I only leafed through their work, but what I see looks more like lexicostatistics, which is “historical linguistics” too, but not the “let’s date the first Slavic palatalisation based on loans from Greek and known date of delabialisation of Greek y” variety which I usually mean by “historical linguistics”.
If their math is bad, it is bad. If their math is good, it is a result – which can be interpreted* by others.
Meanwhile the result is based on “historical linguistics” and comparative work in the usual for me sense, namely on Proto-Mande reconstructions. If those are good, we can’t question their knowldge of basics of historical linguistics.
* Say, the argument about borrowability (DE: Africa is notable for “widespread extensive borrowing”, JP: it does not extend into the most stable part of vocabulary). If it can be resolved slowly (in a study) or quickly (by a reference to a study), it still belongs to “interpretation” part. Same for the argument about morphology (I think many will disagree with such a requirement which means that Mande should not be compared to anything, period).
Re sky/rain: “sky” itself is of course a borrowing of a Norse word for “cloud.”
Indeed. I’m sure that tendon/root has parallels elsewhere too (that one struck me personally as the least weird of the three.)
Russian penis, horseradish,
English vagina, cat…
The “cat” one turns up in French, too. No idea if it’s calqued on English, or vice versâ, or if it’s some weird areal thing.
(If it’s the first, one might have hoped that the Académie Française might have stepped in to safeguard the purity of the language. Do not the heirs of Rabelais have the linguistic resources to craft genuine Gallic obscenities?)
As I said, https://clics.clld.org “Concepts” then type rain in “name”.
Not much for Africa. (But: Fulfulde kammu, Bangime ʒoɔ̃)
DE, Breton C’hoari koukou (c’hoari “play”) is a rather cute name for sex and must be a calque from French but can French coucou “peekaboo” mean it?
clics
Disappointing for Oti-Volta and “rain.” The only one I can see is tara for Gulimancema, which is wrong; I think it must be a Francophone attempt to spell taaga by ear.
I wonder where they sourced their materials? (It probably tells you somewhere, but I’m currently just on a phone, with a very flaky signal, which makes it hard to explore.)
Is there a way to get it to show colexifications specifically on the map? (Hausa has rain/water, a natural enough one in the savanna; the clics map for “rain” gives a disambiguated form “sky water”, but “it’s raining” in Hausa is actually just ana ruwa “it’s water-ing.”)
In those beautiful days when cam girls took money for undressing in private (today it is neither private nor undressing and to earn anything meaningful they repeat it thousands times a day) two freinds of mine wanted to create an account named “three pussies” (meanign two female humans and one cat). I was reminded about it when I saw Zwei Gazellen here.
C’hoari koukou
All the meanings of coucou I know are thoroughly innocent, but (a) I am not French and (b) I have always* led a life of strict Calvinist rectitude, confining my amusements to blameless pursuits like burning heretics.
You must ask a French person; ideally, a debauched one. There is no substitute for a capable informant when it comes to linguistic fieldwork.
Something to do with cocu rather than coucou, maybe?
* Always!
Segerer’s REFLEX database is better, they seem to have tried to digitise everything they could find, but one will have to process their lists manually. Also some glosses are in French.
@DE click “varieties“. Too see where lists for a specific language come from enter its name.
I noticed a concept “gazelle”, only in Chadic and Tanzania, from Kraft’s Chadic wordlists and Tanzanian wordlists.
Several Arabic women’s names (not to say “words”) can be translated to English as “gazelle” (say, Rim, or Maha unless you prefer “rhim” for Rim and “antelope”, “oryx” or whatever for Maha) so I looked what they have for Arabic and it is 5 wordlists, 3 of them are Koelle.
It seems they prefer complete wordlists to processing dictionaries and vocabularies from grammars.
___
“GIVE”
adḗdak – Koelle’s (Arabic) Wádai,
atáituka – Koelle’s (Arabic) Ā́dirar,
ātáitak – Koelle’s (Arabic) Bḗṛān
aʕtˤaː – literary
Messy. Not Messi.
@de
For gazelle names you could include zabi/dhabi (PS *ṯ̣aby(-at)- ). The Hebrew name Tzvi seems to be for men.
It also turns up, regionally, in German – not with the normal word for “cat”, but with a nickname that got attached to cats. My aunt had that nickname when she was very little.
I didn’t… and Ctrl+F confirms that the string gaz does not occur on the page. 🙁
It also turns up, regionally, in German
I assume you mean Muschi? Due to use in abysmally bad 80s movies à la Supernasen and similar “comedy”, that word is understood all across Germany.
[Muschi] is understood all across Germany.
English “snake pussy” would be rendered as Schlangenmuschi or simply Sushi. But “hog” is currently in fashion among the young, at any rate those of the Canadian persuasion.
You could be called “gazelle” in Kusaal (pretty much any noun is potentially a personal name), but it would mean something along the lines of your mother having seen a gazelle when you were born, rather than being a reference to/hope for your personal beauty and grace.
Complete absence of “gazelle” words over most of Africa in the clics thing. Bizarre. I see that it is indeed based on wordlists they chanced to have available (presumably in digital form) rather than dictionaries; the African ones at least, in turn, must themselves be based on old and obsolete wordlists, judging by misspellings like tara for taaga. You’d need to go back to many-decades-old collections by non-linguists to get a form like that.
Mind you, doing it properly would be a huge task.
I was recently reading that R M W Dixon (who is admittedly prone to such remarks) says that it would have been better for linguistics if WALS had never been created …
Understood, yes, but I haven’t heard or seen it in use in Austria. On the very rare occasions I’ve encountered mentions of the referent in the appropriate register, the word has been the old one.
Koelle? Obsolete?!
Another “gazelle” name is Greek “Dorcas”. Wiktionary defines δορκάς as “roe, gazelle, antelope”.
For anyone who’s interested, Wikt says of the etymology:
Koelle? Obsolete?
Absolutely. An astonishing achievement of lasting value, and sometimes the only record of lost languages. But, as a source for modern languages, where later information is available (often abundant information), obsolete indeed.
A Koelle-era (or even Prost-era) account of Kusaal might say that kʋk represented a “colexification” of “ghost”, “chair” and “mahogany”, which are identical segmentally but belong to three different tone classes.
More seriously, I don’t see how a study based on worldlists (even reliable ones) can distinguish between colexification and homophones (of the sort English abounds in.) Thus Kusaal siig “life force” is completely homophonous with siig “Anogeissus leiocarpa tree”, as a result of historical sound changes (Buli has, respectively, chiik and siik.)
It’s even more complicated when accidental homophony has led the speakers to regard the relationship as colexification, as with Kusaal yam/ya’am “gall bladder; common sense, mind.”
I don’t think there can be any substitute for asking the speakers, case by case, “are these the same word?” (often a difficult question.) A good dictionary might help, with how it groups entries, but even these can go wrong; for example, Naden’s Mampruli dictionary puts sàkki “agree” and sákki “suffice” under one entry.
I think compendia like this (and WALS) are like Mass Comparison: potentially good for suggesting hypotheses which can then be tested rigorously. But they are not a sound basis for drawing conclusions from directly at all; and they have, in practice, been abused in just that way.
According to Beekes, the δ-forms are “perhaps folk-etymological, after δέρκομαι”; there are variants with ζ- and ι-; according to Beekes the latter may be Celtic (Galatic) loans.
“Cry, weep” and “ferment” (as in beer) is another Oti-Volta one; it works even when the actual roots involved aren’t cognate, so definitely colexification.
It’s another one where you go, “yeah, I can see that” retrospectively. I wouldn’t be surprised if it turns up elsewhere too.
I break a stick, I flog a child, I do not play
I do not dance, I pray to God, I call a slave…
(Koelle)
@DE, it was irony:) (You said “decades” but https://en.wikipedia.org/wiki/Polyglotta_Africana is 130 years old)
However it is valuable in many ways. E.g. you can find there some infromation about Abd ul Kā́dir (who gave forms from Ā́dirar and Bḗṛān Arabic), p. 21.
“Messy” – I didn’t mean to criticise CLICS, I don’t know what they can do.
However it is valuable in many ways
Absolutely. For quite a lot of reasons, including his beginning of African language classification, a remarkable amount of which has stood the test of time. As I said, an astonishing achievement.
Koelle wrote a Kanuri grammar, too. TBH, it’s not very good; but again, it’s fairly amazing that he managed to do it at all. And writing a good Kanuri grammar is beyond the range of most of us lesser mortals, anyway (it’s that kind of language …)
I break a stick, I flog a child, I do not play
I do not dance, I pray to God, I call a slave…
We’ve all had days like that. Métro, boulot, dodo …
Is Koelle just an older spelling of Kölle?
Kölle Alaaf.
@PP (gazelle names), it was me, not DE:) Thanks, I didn’t remember Zabi!
@DM (Zwei Gazellen), ti’s in the image. Two images of
a young lady and a pet gazelle* a pet gazelle and a young lady a pet lady with a young gazellea young gazelle with a pet lady in the first two rows. The first of them is a book page with this inscription. Perhaps subscribed by the editor rather than painter and I don’t know if it is a painting or drawing or what.*pet gazelles
Is Koelle just an older spelling of Kölle?
The man himself seems to have written it “Koelle”:
https://en.wikipedia.org/wiki/Sigismund_Koelle
His (Brit) descendants pronounce it “Kelly.” And why not?
Hebrew Tsvi is a calque of German/Yiddish Hirsch. Why that became a European surname or given name is another question.
The man himself seems to have written it “Koelle”
Yes, I presumed as much; I was wondering about the name itself. It looked like “oe” = ö, but having been burned by Lueger (which is three syllables) I didn’t want to take anything for granted.
The “Kelly” version seems to support an original “Kölle.”
“Sigismund Kelly” sounds like a character out of Ulysses.
@JF, I think Maha (where h stands for [h] and not [ħ] ḥ or [x], [χ] ḫ) too is named so because of Arabian Oryx’s beautiful eyes. See https://en.wikipedia.org/wiki/Maha_(name) and for the eyes in question “Arabian Oryx” in Google images.
We don’t have anything like “doe-eyed” in Russian.
@Y
Re Tsvi, etc.
“Certain animals are traditionally associated with common Hebrew first names. (In part, these associations are based on Jacob’s blessings for his sons, Bereshit 49.) The German words for these animals were used as secular first names (Hebrew “kinnui”) and often became family names, e.g.
Judah – Loew, Loeb, etc; Spanish Leon ‘lion’
Issachar – Baer, Beer, Berl, Perl, etc. ‘bear’
Naphtali – Hirsch, Herz(l), etc.; Slavic Jellin(ek) ‘deer’
Asher – Lamm, etc. ‘lamb’
Ephraim – Fisch(el), etc. ‘fish’
Joseph – Stier; Ochs ‘bull; ox’
Benjamin – Wolf, Wulf, etc.; Spanish Lopez ‘wolf’
Joshua – Falk, Falik, etc. ‘falcon’
Jona – Taube, Teuber, etc. ‘dove’ ”
https://www.jewishgen.org/infofiles/namfaq0.htm
This seems to have been a naming practice particular to Jews. However I do not think this is an only-Ashkenazic practice. From a list of Sephardic names:
Abastado –see Joseph
Abas/Abasa/Abaz/Abbas–see Judah
Lavi/Leao–see Judah
Cason–see Ephraim
Aguila–this is like Ger. Adler, Russ. Orel
Albas/Albasa/Albaz–see Joshua
Allouch/Allouche/Allush–see Asher
Arias/Aries/Arie/Arieh/Arye–see Judah
Aza(e)ncot/Azancoth/Azankot–see Naphtali
Cervo/Corzo–see Naphtali
Columbus–see Jona
Corde(i)ro–see Asher
Crief/Krief–see Asher
Djam(o)us–see Joseph
Douieh–see Benjamin
Elofer–see Naphtali
Falc(a)o/Falcom/Falcon–see Joshua
Gazal/Gazale/Sevy–see Naphtali
Ghozal/Pombo–see Jona
Izerzel/Izerzer–see Naphtali
Lobata(m/n)/Lobato/Lobo–see Benjamin
Peix/Peixoto/Peixotto/Pexe–see Ephraim
Saba–see Judah
Tourinho/Touro–see Joseph
Tsevi/Tsvei/Tzevi–see Naphtali
Ts(e)via–feminine of above!
Uchoa–see Benjamin
Yamus–see Joseph
There are others, esp. goat/kid, for which I do not know the Biblical personage (although both have special significance there) or Ashkenazi equivalent. Note I did not find Sephardi “bear” names (except maybe Dibbo), however they do not seem to have been averse to pigs!
https://genealogia.org.mx/sephardic-names-translated-into-english/
An interesting paper pointing to problems with “Eskimo-Aleut” (though not to the extent of casting doubt on its actual existence):
https://www.researchgate.net/profile/Jason-Rogers-12/publication/392077129_A_Model_of_the_Origins_and_Development_of_Aleut/links/684748add1054b0207fadea7/A-Model-of-the-Origins-and-Development-of-Aleut.pdf
Interesting in itself, I thought; also a lot of familiar methodological questions (though I know hardly anything about the actual languages themselves.)