Another in my occasional series featuring dubious statements about language by scholars in other fields: Nelson Goering posts on FB about a book on Vietnamese history he’s reading in which the author claims the Vietnamese word for water, nác/nước, derives from “one of the archaic Vietnamese terms for woman (nàng).” (This is in service of a point about “a possible Earth Mother cult,” which makes me sigh an entirely separate sigh.) Now, I know next to nothing about Vietnamese etymology, but I know enough to be skeptical about that, and sure enough, a quick visit to Wiktionary informed me that nước is “From Proto-Vietic *ɗaːk (‘water’), from Proto-Mon-Khmer *ɗaak (‘water’).” (If you’re curious about that archaic Vietnamese term for woman, nàng, it’s from Proto-Tai *naːŋᴬ ‘lady.’) I realize the Vietnamese word starts with n- and the protoforms with d-; I am assuming this is a regular or at least explicable development or it wouldn’t be stated with such assurance, but as I say, I am ignorant about Vietnamese etymology, so if there is reason to think Wiktionary is wrong (which, of course, it not infrequently is) and the historian is right, by all means let me know.
Seems to be yet another instance of the sadly familiar phenomenon of an otherwise well-educated person being totally unaware that scientific historical linguistics even exists.
Even so, it’s an odd one. I wonder if it’s actually a traditional Vietnamese folk etymology?
I’ve read of implosives “commonly” becoming nasals before. I find it hard to imagine, but I have very little experience with implosives…
(The context was Indo-Uralic comparison, and no examples were given for the assertion that the sound change is known elsewhere. Anyway, I find both the evidence used to support the idea that the PIE plain-voiced plosives were implosives and the evidence that they correspond to PU nasals quite a bit easier to explain in different ways – even within an IU framework.)
This combines two of my biggest gripes:
1. Non-linguists wading into a linguistic problem with all the humility and thoughtfulness of Ned Kelly entering a bank, and
2. Anthropologists or historians assuming that people in the past had no material or non-material culture that was not directly indicative of their deepest spiritual beliefs.
Double dipping pet peeves. Couldn’t be happier than I am right now.
[one of my biggest gripes] 2. Anthropologists or historians assuming that people in the past had no material or non-material culture that was not directly indicative of their deepest spiritual beliefs.
Not quite sure what you mean here.
Why anthropologists and historians in particular, why “people in the past” especially? Is there any human attribute or behavior that has not been subjected to merciless interpretation? As the outward sign of some deep spiritual thing or other ? By self-appointed interpreters not only of the indicated two kinds ?
This did not start with cigars. With less cognitive corrosiveness in the air, baseball might pass as an example of empty-headed enjoyment.
In the case of “water”, Vietnamese seems to be the sole language in the entire family to show an initial n reflex of proto-Mon-Khmer *ɗ:
https://en.wiktionary.org/wiki/Reconstruction:Proto-Vietic/%C9%97a%CB%90k
However, “five” seems to be another example of the same outcome:
https://en.wiktionary.org/wiki/Reconstruction:Proto-Vietic/%C9%97am
so maybe it is indeed regular, weird though it be.
Eastern Oti-Volta has quite a few cases of initial n for proto-Oti-Volta *d (and of m for *b), but in such cases the following vowel was always originally nasalised in POV, e.g. Nateni nɔ́ɔ̄kú “Parkia biglobosa tree”, beside e.g. Kusaal duan [dũ̀ã].*
(The same kind of thing seemed to be going on with Pozdniakov and Vydrin’s proto-Mande forms in that paper, which IIRC always reconstructed initial *ɓ *ɗ + nasalised vowel rather than initial *m *n.)
But as far as I can see, contrastive vowel nasalisation isn’t involved in these proto-Mon-Khmer reconstructions. Perhaps the path in Vietnamese was something like ɗ > l > n.
*The intermediate stage in the shift *d > n before nasal vowels in Eastern Oti-Volta was probably [l]: none of these languages has a contrast of /l/ vs /d/, and the written d in their standard orthographies seems to actually be realised [l] often, or even perhaps always. There’s a regional aversion to voiced plosives.
Ben Kiernan (author of the history book) didn’t make this “water < woman” thing up himself, he footnotes the whole paragraph to… let’s see if Google Books is going to be generous today… yes, it is! The Golden Serpent: How Humans Learned to Speak and Invent Culture by Huynh Sanh Thong, best known for translating Vietnamese poetry into English. Self-published — and considering that Huynh had taught at Yale, been director of Yale’s Southeast Asian Refugee Project, and published his translations with the Yale University Press, if he couldn’t get any publisher to take it, it must have been way out there.
Looks like the book is rare and not online as a whole, but somebody posted a few scanned pages on a blog: it opens with the discovery that woman, womb, and worm sound very similar, “Instead of arbitrary randomness, I suspected some structural principle at work here”, and goes on from there for 800 pages about ancient matriarchy and snake worship. Also, worship of Mother Nature, therefore French mer is derived from mère, just as English sea is derived from she, and German See from sie.
Ah! A mystery solved!
Western Oti-Volta lacks the root *ɲɪ- “water”, found not only in all the rest of Oti-Volta but commonly elsewhere in Volta-Congo to boot.
Instead, WOV has a quite unrelated root, seen in e.g. Mooré kòóm, stem kò-.
Hitherto, the origin of this root has been unknown, but I now realise, in the light of Prof Thông’s wisdom, that it is cognate to the root ka- of Kasem kaanɩ “woman.” It is clear that the speakers of proto-WOV worshipped a mother goddess associated with water, in contrast with their neighbours, who associated water with rain and the male sky-god.
It seems likely that the favour of this earth goddess was a material factor in the military conquests of the WOV-speaking Mossi-Dagomba states. (Compare Ishtar/Inanna, goddess of sex and fertility, and also of war.)
It is also surely not a coincidence that the Western Oti-Volta words for “snake”, e.g. Mooré wáafó, begin with wa-. (With the -fo element, compare fu “woman”, fu “womb” and fu “viper” in Chinese.)
And so I learn about Ned Kelly. Humility and thoughtfulness indeed.
I’d love to see his reaction to the fact that See “sea” is indeed a she, but See “lake” is a he.
Ferlus, “Histoire abrégée de l’évolution des consonnes initiales du vietnamien et du sino-vietnamien”, explains (p. 117): “Les préglottalisées ɓ ɗ ˀj sont spécifiques au PV et au PVM. Elles sont aujourd’hui représentées par les nasales m n nh avec des tons de la série haute. Le processus de transformation a dû être ɓ ɗ ˀj > ˀm ˀn ˀŋ > m n ŋ et les notations par les phonogrammes montrent clairement qu’aux XV-XVIe elles étaient déjà des nasales, glottalisées ou non.”
It does not reassure one regarding Ben Kiernan’s scholarly judgment that he is apparently unable to tell that this particular work of Thông’s is tripe. No actual knowledge of linguistics should be needed to see this, I would have thought, though of course it is immediately obvious to anybody with even the slightest acquaintance with either linguistics or ethnography.
Concerns about Kiernan’s judgment, especially regarding this particular book of his, seem to have been raised on other grounds by experts.
https://en.wikipedia.org/wiki/Ben_Kiernan
it opens with the discovery that woman, womb, and worm sound very similar, “Instead of arbitrary randomness, I suspected some structural principle at work here”, and goes on from there for 800 pages about ancient matriarchy and snake worship. Also, worship of Mother Nature, therefore French mer is derived from mère, just as English sea is derived from she, and German See from sie.
O…. K. We see now the level of scholarship on display, and I regret having toned down my original wording (e.g., the phrase began as “ludicrous statements about language”) in an attempt at humility. “Be more Ned Kelly!” I should remind myself.
oy. i’m with MK, with an added pet peeve about the primeval-goddess-cult stuff, which always comes back to that nazi gimbutas.
But what you can do if you’re a monotheist or atheist and still like the idea of such a cult?
When I was a child I giggled at Soviet museums who subscribed numerous objects “a cultic object” when (as I thought) they didn’t know what it is.
“2. Anthropologists or historians assuming that people in the past had no material or non-material culture that was not directly indicative of their deepest spiritual beliefs.”
This is not the only example, and he’s prone to finding Significance of various kinds in language wherever he can. A little later, on p. 55, he says this:
“To this new home [in modern Vietnam] they brought their Cham language, which today still preserves its ancient word for “sunset,” meaning “sun plunge into sea,” a vivid memory of their former home on Borneo’s west coast.” (Cited from Ian W. Mabbett, “Buddhism in Champa”)
I haven’t looked into this, but my first thought is that someone could use expressions like “the sun was sinking” (an old expression in Germanic, cf. Gothic “þan gasaggq sauil”) to claim that English preserves an “ancient memory” of being on the west coast of the Black Sea or something (or the Baltic maybe, depending on what image you’re trying to conjure). Maybe the Champa phrase really is more specific in preserving a fossilized lexical item for “sea” in sunset (that actually would be interesting), but I’m more than a little skeptical, and figure the odds of getting any reward for following this one up are minimal.
His general working method, at least for linguistics and archaeology, seems to be to more or less repeat any interesting quote (he inserts many quotes into his running text) or factoid he’s come across. At least on the linguistic side, this leads to a bizarre (and I think sometimes outright contradictory, though I’m reading this book in a “snatch a couple of paragraphs while the toddler is playing with his toys” way, so I’m not always paying that much attention) mishmashes of what seem to be normal kinds of historical linguistics and rather unlikely claims (some of which I’d chalked up to Kiernan misunderstanding his sources, but now I’m not inclined to be even that generous). His footnote placement also makes it a bit unclear what’s coming from where sometimes, at least for the hurried reader. I’d actually thought this particular claim, about water and women, was uncited, but I guess I missed the attribution to Thong.
When I first read this, I’d vaguely wondered if there might be some kind of morphological process in early Vietic/Austroasiatic that might make the connections claimed by Kiernan seem at least superficially plausible (though of course not necessarily correct), or at least, with DE, if this might be a traditional folk etymology he’d picked up somewhere. I didn’t realize he was making use of such blatantly pseudoscientific sources.
When actual cognates include wheel/chakra and futile/God, it’s telling that the words these cranks claim to be related are never really that different. Where are the crackpots trying to say that “danger” derives from “Republican” or that “money” is an offshoot of “sadness?”
“woman, womb, and worm”.
Well… Thong is a translator of poetry and thus poet. Analogies and poets work with associations, and analogy could have played some role in phonological development of woman. The problem with analogy is that you can’t be sure it’s a factor.
“womb, and worm”
These remind me the odd Russian pair chrevo and cherv’ (“womb, belly” and ‘worm”). Wiktionary says PBS *ker-m- and *kirm-, the second is a well known IE root, but I don’t about the first.
“analogy could have played some role in phonological development of woman.”
Do you mean analogy from womb? That’s highly implausible, given that we have spellings like wummon already in Ancrene Wisse, at a time and place where womb would have been roughly wǭmbə. I think we can, actually, be sure such an analogy didn’t take place.
If you mean the possible analogical influence of the pseudo-umlaut in the plural in many dialects, that’s plausible — but of course that’s just normal analogy in the linguistic sense, which isn’t terribly similar to the analogies of poets (hence etymology not working at all like Thong seems to think).
Let’s say, when poets speculate about language, I don’t call it “pseudoscience”. What else they should speculate about? Sex? There is no shortage of specialists who study this, they too will object.
@NG, I didn’t mean any word in particular. I think this specific combination of contraction and rounding in “woman” is not predictable and expected. It well could be still wyfman or wyvman or anything else in modern English. How can one know for sure if other wo- or wu- words made it more attractive or convenient to speakers or if semantics of those played any role?
But why not womb. Co-lexification womb-mother is not uncommon in Europe, sometimes even womb-woman (even though I don’t associate words for women with words for wombs).
Don’t know about Gothic, but in German, sinken does not entail “into water”; dann versank die Sonne is probably out there somewhere… oh, als die Sonne versank is actually a Google suggestion.
(Die Sonne sank does not entail “all the way down”.)
When a speculation is presented as science and even the existing tests are ignored, I do call it pseudoscience; whether it comes from a poet or someone else doesn’t change that.
Examples?
I mean, “womb” in German is Gebärmutter, sure, but that’s a metaphor, not a colexification.
“contraction” – or assimilation and then contraction?
“Co-lexification” – or rather derivation as in matrix. But I think in some languages it’s co-lexification.
@DM, Thong’s formulation cited by ktschwarz is careful enough.
It should not be CITED as a scientific fact.
“Examples?” – I meant words similar to matrix, or Greek metra, thus not co-lexification.
But mátka (literary Russian “womb” literary Belarusian and Polish “mother”) could mean “mother” in some Russian varieties (and “womb” in Middle Polish and dialects says Wiktionary).
In many languages such words also can refer to dams, queen-bees, other-than-human mothers etc.
Breton mamm is 1. mother 2. female 3. womb. (Wiktionary in etymology of “womb” also mentions gwamm “woman, wife” but it is not the main word for these in textbook Breton)
I’m sure there are others.
The Kusaasi sun “falls” rather than “sinks.” Curiously, it doesn’t seem to “rise” at dawn; it just “appears.” However, by midday it has “germinated.” No doubt all this has profound anthropological and ethnohistorical significance. (It is clear, for example, that the ancestors of the Kusaasi must have had a liking for mixed metaphors.)
None of the Western Oti-Volta languages seems to derive “womb” from “mother” or “woman” (or vice versâ); for “womb” they use derivatives of the verb which appears in Kusaal as du’a “bear/beget.”
Pu’alim “femininity” (from pu’a “woman”) does appear in Kusaal for “female genitalia” (and similarly in Mooré), but that’s just a parallel formation to daalim “masculinity/male genitalia.” “I kneed him in the manhood.”
Hmph. GPC reckons that Welsh bru “womb” may be a loan from Irish. Well, I suppose, if we needed to borrow our word for “child” from Latin, why not? (Croth is the more usual word for “womb.” But that seems to be etymologically pretty opaque too.)
In a number of Semitic languages (Arabic, Hebrew, Akkadian) mercy and womb are same or similar (which does make some say that as men don’t have one, they also don’t have the other).
@de
Matasovič has:
—
*bruson- ‘abdomen, womb’ [Noun] GOID: OIr. bru [n f\ W: MW bru [m] ‘womb, belly’, bron [f] ‘breast’ BRET: OBret. bronn ‘breast’, MBret. and MoBret. bronn ‘breast, teat’ [f] CO: Co. bron ‘breast’ PIE: *b h rews- ‘belly’ (?)(IEW: 170f.)
*druto- ‘foolish, lewd’ [Adj] GOID: Mir. druth [o] ‘professional jester, fool’ W: MW drut (GPC drud) ‘dear, valiant, foolish, foolhardy’ GAUL: Truti-knos ‘Druti filius’ (Todi), Gallo-Latin *drudo- ‘dear’ > Fr. dru ETYM: W drud has an unexpected u, rather than i < PCelt. *u. Maybe it was influenced by Gallo-Latin *drudo- and its Romance reflexes, as Pokomy thinks.
—
Maybe instead of MW bru being a loan from OIr, there was a Gallo-Latin *brusu "belly, womb". There is a late Latin brutis "Latin speaking wife of a German soldier". However there is also an Old French bru/brut "daughter-in-law". This probably does not work either phonetically or semantically.
brusu
No, I don’t think that could work semantically at all, and I’m dubious about the phonology too.
French bru seems to be from Frankish, and cognate with “bride.”
Matasovič’ easy conflation of wombs and breasts strikes me as … daring.
Bron is “breast” in Welsh, not “belly” or “womb.” GPC attributes it to a Celtic *brusnā, alongside Old Irish bruinne from *brusni̯o-. Wiktionary actually links the latter with Welsh bryn “hill” too, and GPC concurs. (Hadn’t occurred to me before, but it makes sense once it’s been pointed out, I suppose.)
“Womb” itself is of course from Old English wamb “belly.” Kusaal has the same sort of shift: unless you’re being particularly Anatomically Correct or pedantically medical, you just say pʋʋg “belly” for a woman’s uterus. (In Kusaal, it’s used metonymically for “pregnancy” too.)
“I don’t think that could work semantically at all” – “At all” can be an exaggeration: “womb” is used in the Bible in the sense “woman”. The example in wiktionary is about [slave]girls qua spoils of war, and I’m not in the mood to check other contexts. The meaning marked as “Biblical Hebrew” and “poetic”.
Moro optimistically, cf. Romanian word for “man”:)
Huỳnh Sanh Thông’s Wikipedia page says “In accordance with Vietnamese custom, this person should be referred to by the given name, Thông”, and consistently does that (as seen here in 2005 at VIETNAMESE NAMES). On the other hand, Yale’s obituary says “Huynh dedicated much of his life to bringing the achievements of Vietnamese literature and culture to Western audiences” and so on, and Kiernan’s footnote is to “Huynh, Golden Serpent, 9–25”, so I tentatively went with that.
Let’s see: Chicago (2024, probably same in previous editions) says: “Since Vietnamese persons are usually referred to by the last part of their given names (Premier Diem, General Giap), they are best indexed under that form.” The Oxford Guide to Style (2002) says the same, also with the example of General Giap.
However … they were Vietnamese, but Huỳnh Sanh Thông was Vietnamese-American. He became an American citizen and wrote in English. The Library of Congress catalog has him as “Huỳnh, Sanh Thông”; I think (correct me if I’m wrong) that would have come from the publishers, and I think (again, correct me if I’m wrong) publishers wouldn’t do something ridiculous to their author’s name. So I’d come down on the side of referring to him as “Huỳnh” as the author of a book written in English, and likewise in the bibliography.
(Why Kiernan’s bibliography spells the author’s name with diacritics on first reference, but “Huynh” without diacritics on second and subsequent references, is a question for Kiernan and his OUP copyeditors. The diacritics are there on the book cover.)
If I were to follow the Way of Thông (or Huỳnh), I would of course link Kusaal pʋ̄ʋg “uterus” (stem pʋ̀-) with pū’a “woman.” (In some old compounds, the stem form of “woman”, usually pù’a-, actually has got worn down to pʋ̀-, e.g. in pʋ̀kɔ̄ɔnr “widow”; compare dàkɔ̄ɔnr “bachelor, widower, unmarried man”, and dāu “man.”)
[Pay no attention to the fact that pʋ̄ʋg does not actually have the core meaning “uterus”, or the Goddess will be angry.]
Proto-Oti-Volta *pó(g)-wá “woman” actually is a bit of a mystery. It doesn’t seem to have any cognates outside the group; Kasem kaanɩ (stem ka-) probably does represent an older, much more widespread, root for “woman.”
It’s not from “womb”, though.
“800 pages about ancient matriarchy and snake worship”
1. Eve
2. Medusa Gorgon
3. “Every woman must have a snake!” – a song by БГ, a musician of piercing spiritual vision.
(У каждой женщины должна быть змея)
I was going to say “bride”, but the t of German Braut comes from the High German Consonant Shift (late 6th century), and the original d is preserved in bride as expected. So that doesn’t work.
…though that’s “bearded”, not just “beard”.
Pudgy by way of Scandi-Congo.
I’m not sure we have enough data to fully sketch out the semantic range of Gothic sigqan*, but I don’t think that would bother the kinds of people who’d want to make hay out of sinking suns.
@dm
Dariusz R. Piwowarczyk, “A Note on the Etymology of Brūtes/Brūtis ‘a (Latin-Speaking?) Bride’”
Classica Cracoviensia
vol. XXVII (2024), pp. 377–387
https://journals.akademicka.pl/cc/article/download/6182/5745
Well, technically Blackbeard, Barbe-Bleue, Barbarossas and Langobards are synecdoches and various Abu Lihyas are not. Different derivation, but is the meaning different from “The Red-bearded” etc?
@DM:
To be fair, there’s not a lot in the way of potential cognates for POV *dà-wá “man” either, except for Miyobe ù-lā “husband” (I’m pretty sure that Miyobe is Oti-Volta’s closest relative, and the *d/l correspondence is apparently regular.)
[The Kasem word for “man” is baarʋ, which is obviously the source of the Romanian word. The Kasem dictionary illustrates the word with what seems to be a picture of Nelson Mandela, looking suitably Manly.]
On the other hand, POV nîd-wá “person” seems to go way back. Maybe even North Atlantic cognates (cf Fulfulde neɗɗo “person”, though one problem with this is that the *d of the POV form may be derivational or epenthetic rather than part of the actual root.)
Thanks, have downloaded, will read!
…speaking of which… the sun rising & setting in German is aufgehen & untergehen, and the latter is what’s used for sinking ships and not much else (the world does it when it ends, empires do it when they fall…).
And there is Romanian femeie “woman” (from family, not from femina or femella), analogous to “beard” rather than “bearded”… (I wonder why such shifts for both words)
Oh, fascinating.
untergehen …. the world does it when it ends, empires do it when they fall…
as in Spengler, “Der Untergang des Abendlandes”
In Swahili (at least among the actual Muslim Swahili themselves) you don’t ask a man “how is your wife?”, even if you know that she’s been ill; instead, you’d ask “how is the family?”
No idea if a similar thing obtained in traditional Romanian society, though I wouldn’t be surprised if it did.
No such tabus among the Kusaasi; the only way you could really offend somebody with such questions would be by not going on long enough: “How are you? How did you sleep? How’s the work? How are the wives? How are the children? How are your siblings? How are the in-laws? How are the grandparents? How are the grandchildren? How are your aunts? How are your uncles? How are your cousins? How are your nephews and nieces? How is the hamster? …”
(The only linguistic trap is that every answer has to be “fine!” Only once you’ve got the greetings finished can you say that your hamster [or aunt, as the case may be] has, unfortunately, just been eaten by a hippopotamus [or by Freddie Starr, as the case may be.])
https://www.inpublishing.co.uk/articles/e28098freddie-starr-ate-my-hamster-crowned-the-suns-most-iconic-cover-14939
My friend’s* girlfriend’s** brother’s wife***’s uncle has been eaten by a crocodile.
* not sure if ‘s can work in chains**** or I need “of” here.
** perhaps ex. And in Russian it is девушка, “[his] girl, young woman” so repetition of words is avoided.
***Ethiopian
****if it can, then my teacher can say “my student’s freind’s girlfriend’s…” and so on.
Chain ’s’s as much as you like. Anthropologists and linguists use these chained esses all the time to describe complex family relationships (“the word for ‘mother’s father’s mother’ is the same as that for ‘father’s father’s mother’,” like that.)
Sorry about the uncle.
futile/God? I looked in AHD, it says they’re from two distinct PIE roots; the OED says the origin of “God” beyond Germanic is “very uncertain”. So what’s going on? The claim is probably coming from one of Simon Roper’s youtube videos. He traces them both back to PIE *ǵʰew- ‘pour’: Latin fūtilis was supposedly ‘easily pours out, leaky, unreliable’; God is what libations are poured to. However, both of those lines of descent are disputed.
Pokorny traces “God” to a different root *ǵʰewH- ‘call, invoke’, and this is also in AHD and Mallory/Adams. The OED’s revision (2014) says it’s disputed between the ‘pour’ and ‘invoke’ options; the latter “has often been favoured, but it poses phonological problems.” (Their original edition from 1900 already said it was uncertain between these two options.) Wiktionary at Proto-Germanic *gudą adds a couple of other proposals, one from Kroonen’s Proto-Germanic dictionary, the other from Beekes saying it’s non-Indo-European.
The derivation of Latin fūtilis from “easily pours” has also been challenged by De Vaan’s etymological dictionary, which has it from *fūtō ‘strike’ instead. “Easily pours out” is a definition from Lewis and Short; the Oxford Latin Dictionary doesn’t have this definition, but instead “1 (of vessels) Brittle, fragile. … 2 (of things) Vain, unprofitable, pointless, futile …”, and further similar definitions that don’t involve pouring or leaking.
Simon Roper is generally a good guy who tries to use good sources, but (as he clearly says) he’s not a linguist and sometimes makes mistakes. The problem here is that he presents futile/God as absolutely certain cognates. A real linguist would find out that these particular words are *not* certain, and would know how to pick out a pair that is (like wheel/chakra).
I like the pair Rus[sia] and Nga.
(Of course, it’s a borrowing, but the thread is Vietnamese…)
I like the one from Kroonen. With “pour” we run into morphological difficulties (the word looks like it should refer to sacrificed liquids), with “invoke” the extra *-H would probably have interfered with Verner’s law.
Beekes saying something is non-IE is rarely newsworthy.
I liked it too. Partly for different reasons: it allows interesting comparisons with the Slavic word and others. In a different sense of “like”.
“how is the family?”
perhaps compare yiddish פֿרעג־נישט | freg-nisht [don’t ask] for a cohabiting lover* and פּלונית | ploynes [whatsername] for a wife.
That’s an important reason – finding separate explanations for all these words is less parsimonious.
Have read. Something seems to be missing in its argumentation. It effortlessly reconstructs Proto-Northwest-Germanic *brūðiz for the bride word, then quietly assumes that that’s a Verner variant and came with a Grimm variant *brūþiz which would have survived in East Germanic, where it’s attested in the Gothic acc. sg. bruþ “daughter-in-law” and nowhere else. This *brūþiz would have been borrowed as brutis, from which brutes and bruta (also attested) would be expected developments within Vulgar Latin. And then the paper is suddenly over. Trouble is, Gothic had final devoicing of fricatives, so it isn’t evidence for *brūþiz specifically. Any other Germanic-internal reasons for supposing a Grimm variant once existed are obscured by the fact that, according to the paper, the IE etymology of the word (if any!) is itself quite unclear.
Also interesting, as the paper points out, is how old the word is: the oldest inscription that contains it (in the expected oldest form, brutis) is from between the years 131 and 250. That’s really early for anything Germanic in Pannonia, and I’m wondering if we should actually look at Temematic/Crotonian – if we could.
I googled this line from Thong “all share what the richest man on earth does not possess” (about the vagina).
The context is his derivation of Bantu class 1, 3 and 6 prefixes from some [primordial?] mother word. I picked it at random. The response of Google AI:
Contextual advertising or not, but soooo subtle….
* the [link] leads to the blog post already linked above by ktschwarz. That’s where I took the line, though it speaks about vagina, not womb and says nothing about “famous Vietnamese philosophical riddles”. Perhaps the riddle is AI’s own invention.
“his derivation of Bantu class 1, 3 and 6 prefixes” – very, very, very silly. I mean, I still hold to what I said about poets, but they could try at least:(
Pointless to make merely factual points when dealing with a type like Thông, but the *m- in the Bantu class 1, 3 and 6 prefixes (*mʊ *mʊ and *ma respectively in proto-Bantu) is a Bantu innovation, confined to nouns, that doesn’t appear in the corresponding verb agreement prefixes (*ʊ *gʊ *ga) even in Bantu*.
There is a melancholy aspect to seeing a mind with sufficient curiosity about Language to find out about Bantu morphology debase itself with this kind of intellectual self-abuse. Reading it would be like rubbernecking at a car crash.
He must be a bad poet. Real poets have more respect for facts.
* Actually, Bantu class 6 is a Frankenstein’s monster; its morphology reflects the plural of class 5, and did not have initial *m in proto-Volta-Congo; the prefix pair is cognate to the class suffixes seen in e.g. Mooré kúgrì “stone”, plural kúgà. But the accreted initial *m in Bantu nouns led to this being merged with an originally quite different noun-class affix *ma-, which had *m all along in proto-Volta-Congo. It’s the marker of a class for liquids: its cognate turns up as a suffix in e.g. Mooré kòóm “water.” Even North Atlantic has a ma “liquids” noun class.
I would be astonished if Thông doesn’t have a little squeal of delight over *ma and “water” …
Real poets have more respect for facts.
How do you feel about Robert Graves?
For the original distinction between the plural of Class 5 and the “liquids” class, in another branch of Volta-Congo which, like Bantu, prefixes its noun-class affixes, cf Lelemi (a “Ghana-Togo Mountain language”): lɛ̀nimì “eye”, plural ànimì; lɛ̀nyi “tooth”, plural ànyi; but ntu “water”, nnɔ “oil” (not *àtu *ànɔ.)
There may actually have been two distinct “liquids/substances/abstracts” *mV classes in proto-Volta-Congo: proto-Oti-Volta had a *ma “water” class distinct from a *mʊ “fire” class, and Miyobe does too. Unfortunately, most Volta-Congo languages have preserved a lot less of the class affix system than Oti-Volta and Bantu, so it’s not really possible to confirm this. But the Lelemi “liquid” class prefix n- looks more like a reflex of *mʊ than *ma, for example.
How do you feel about Robert Graves?
His poetry respects facts. It’s his “nonfiction” prose that’s unpoetical … *
https://www.poeticous.com/robert-graves/ogres-and-pygmies
* OK, bit of an epicycle there … I concede the point.
True, and well played!
DE, well, it was me who thought it is “class 6”. Thong lists Kaonde words like mateshi “mud”.
Perhaps such theories don’t look so idiotic when you know literally nothing about historical linguistics. But how someone can spend several decades and 800 pages in a book and at the same time be so uncurious about it is above me.
Well it just is Class 6, from a Bantuist point of view. The conflation of Class 5 plural with (one of?) the older “liquid” classes was already there in proto-Bantu.
The Bantuists tend to downplay the whole singular-plural pairing thing in their analyses, which makes more sense in Bantu that in e.g. Oti-Volta or Grũsi, where the pairings are more consistent.
The 5/6 pairing, though, in my own opinion, is one of the better pieces of evidence for Volta-Congo membership: *dɪ/*ga (or *rɪ́/*ɣá for POV) is just so damn arbitrary, and, unlike the “human” *ʊ/*ba (POV *wá/*bá) pairing, it doesn’t even have any real semantic correlation.
At a pinch, you might try to maintain that the extremely strong semantic associations of the “human” pairing (and of the “liquid” *mV class, come to that) might have caused those clitics (as they must have been originally) to spread by borrowing (such things happen, as witness English “they”) but *li/*ŋa (or whatever) looks so left-field that it’s hard to believe anything but common inheritance could explain it. YMMV …
I agree wholeheartedly about the incuriosity. I think that, at that level, it counts as wilful ignorance: not so much a misfortune as an actual moral failing.
Konni, incidentally, one of the two Oti-Volta languages most closely related to Western Oti-Volta, has done some weird thing with its singular noun-class suffixes which looks almost like the exact opposite of the Bantu Class 6 conflation.
Both Konni and its closest relation Buli have distinct indefinite and definite suffixes for each number in each noun class. In Buli, this looks like a relatively straightforward case of the indefinite forms continuing the old inherited stem-plus-class-suffix forms, while the definite forms have added an “article” with the form of the corresponding third-person pronoun, which usually looks just like the class suffix but without so much wear and tear. So “log” in Buli goes
sg indef dùòk def dùòkū
pl indef dààtà def dààtàŋā
(where “Class 6” ŋā has ousted the original ta of this class)
But in Konni, except in the “human” class, every indefinite singular is formed by removing the proper class suffix and adding ŋ instead; “log” again:
sg indef dàáŋ def dààkʊ́
pl indef dààtɪ̀ def dààtɪ̀tɪ́
(Cf Mooré ràoogó “log”, plural ràadó.)
I think what’s going on with the Konni non-human indefinite singulars is that the ŋ is the reflex of the “liquids/substances/abstractions” class suffix *mʊ, used originally in an abstract derivational sense, so dàáŋ was (as it were) “a logness.” I’ve no better ideas, anyway.*
This would be a real (albeit peculiar) semantic thing though, not the consequence of an accidental falling together of originally distinct forms, like the Bantu Class 6 conflation.
* Another possibility, albeit not a very plausible one, might be that this is linked somehow with the almost equally bizarre Waama habit of making definite forms by infixing -m between the stem and noun-class suffix.
tīībú “tree”, tīīm̄bú “the tree”
búkā “dog”, bún̄ká “the dog”
But, quite apart from the meanings being the exact opposite, the only evidence that this is not just some weird Waama-internal development is that there seem to be an awful lot of nouns across all of Oti-Volta with an apparently meaningless derivational suffix *m, alongside synonymous stems without that suffix.
Further research is needed …
DE, thx!
I called it Class 6 with some hesitation and wondered about the liquid class too. But I didn’t know the history.
It is! This is unexpectedly fascinating.
A lot turns on where Waama actually fits in to the subclassification of Oti-Volta languages. It’s synchronically pretty aberrant in a lot of ways, and it might be the Anatolian of Oti-Volta (so the infixed definite -m- could then be an archaic feature, preserved in Waama but lost elsewhere), but firm evidence is frustratingly lacking.
Waama is traditionally classified as Eastern Oti-Volta, but that is definitely wrong; that idea arose from its geographical position and from the fact that it shares a good number of areal phenomena with EOV, especially in phonology – most of which are also shared by the unequivocally Western Oti-Volta language Nõotre, which seems to have wandered into the area a few centuries ago.
There is some reason to think Real EOV and Gurma may together form a genetic branch of Oti-Volta, but it’s not conclusive, and even if it were, it wouldn’t solve the question of just how Waama relates to “Outer Oti-Volta” (WOV-Buli/Konni-Yom/Nawdm, which is a demonstrable genetic unity.)
Looking to Oti-Volta’s own external relationships: unfortunately, Miyobe, though quite close to Oti-Volta lexically, is itself pretty aberrant phonologically: it shares the EOV Sprachbund phonological mergers (and then some), and seems on top of that to have deleted all final syllables in polysyllabic words, so it’s not a great control on a reconstruction of POV. (It’s also SOV, but there is internal evidence that pre-POV may have been SOV too.)
The next-closest relationship of Oti-Volta is with Grũsi, and there you’re talking about something probably getting on for as remote as Celtic-Germanic. If not more so.
[I despair sometimes of conveying just how distant the relationships are even within – certainly real – entities like “Central Gur”, that Greenbergly classifications talk about as if they were something like Italic. They’re not.]
Sub-sub-subbranches do seem small when you’re looking from the top. But I’m not sure classifications themselves talk about degrees of similarity or time-depths of those.
Also terminology. If IE was African, Rusyn would be
1. a language in the Rus-Belorus branch of Dnieper- Northern Dvina branch of Danube-Dnieper (not to be confused with Danube-Rhine) branch of Danube-Neman branch of Neman-Indus branch of Indus-Rhone.
2. a language in the [same for Pannonian Rusyn]
instead of “some Slavic language”.
From Piwowarczyk’s “note” linked above.
I wonder what it means in Medieval Greek (and how common it is in texts).
Green 1998 is Language and History in the Early Germanic World and Lehmann 1986 is A Gothic Etymological Dictionary – I’ll look them up but those are books about Germanic.
Sub-sub-subbranches do seem small when you’re looking from the top
I suspect that the internal structure of Volta-Congo may eventually turn out to be quite a bit flatter than the Greenberg picture suggests, with more top-level subgroups and less nesting.
The trend does seem to be that way, as more proper comparative work gets done; traditional Greenbergish groups like “Kwa” and “Gur” are getting broken up into components which are probably not really any closer to one another than to other groups within Volta-Congo.
But then, doing subgrouping on the basis of shared innovations is hardly feasible when we are so far from having any worked-out notion of proto-Volta-Congo that we don’t even know how many consonant phonemes it had (even approximately.)
But Pannonia was a border province, and north of the border/the Danube was the territory of the Suebic Quadi (allies of the Marcomanni).
…Yes. I should have thought of that, thanks.
FWTW, there’s no reason to think these people spoke any kind of East Germanic…
@DM, (as I understand), you’re skeptical about the reconstructed PWG form. What other options you have in mind?
No, I find the PWG form perfectly convincing, indeed obvious – the trouble is that I can’t get it to account for the t in all the attested Latin forms. Instead, d would be expected there.
The linked paper tried to solve that problem by assuming that the *ð form came with a *þ alternant. Given the lack of an etymology (as the paper emphasizes), all I can say is “why not”; check out this article, especially the short paragraph “In other parts of speech”.
However, the only other evidence for *þ is in fact equally compatible with *ð; it’s the Gothic outcome, attested only as acc. sg. bruþ, which is ambiguous because Gothic had word-final devoicing of fricatives and consistently spelled it out.
Until the mid-20th century French was the one modern foreign language every educated person in Europe was supposed to know. Above all, it was the language of diplomacy. Germany as a state didn’t even exist in a meaningful way until 1871. People who could read Goethe in the original (like Carlyle) were a rarity.
That post ended up in the wrong thread…
Sorry ulr, I “moved” mine
Indeed, “it is said” is not very precise or helpful. From a very cursory search on the online TLG, the earliest occurrence seems to be in the 7th-century Chronicon Paschale, then in the Life of St. Spyridon (“7th-11th century”).
But a by-form βροῦτα occurs a bit earlier in two different works of John Lydus (6th century).
Both forms are recorded in the LBG, under the headword βροῦτα.
Notably, Lydus derives the noun from the name Brutus:
De Magistratibus 1.33: τελευτήσαντος δὲ τὸν βίον τοῦ Βρούτου, δημοσίῳ πένθει ἐτίμησαν τὸν νεκρὸν καὶ βρούτας τὰς σφῶν γυναῖκας ὠνόμασαν ἐξ αὐτοῦ διὰ τὴν σωφροσύνην
“when Brutus had died, they honored the dead man with a public mourning and named their own wives brutae after him on account of their discretion”
De mensibus 4.29: ἐτιμήθη οὖν […] παρὰ ταῖς Ῥωμαίων γυναιξὶν ὁ Βροῦτος μετὰ θάνατον δημοσίῳ πένθει, οἷα ἔκδικος τῆς σωφροσύνης· καὶ Βρούτας ἑαυτὰς ἠξίουν ὀνομάζεσθαι πρὸς τιμὴν Βρούτου.
“Among the Roman women, Brutus was honored with public mourning after his death, as if he were the avenger of prudence; and they took it upon themselves to be called Brutae in honor of Brutus.”
It would be interesting to know whether this folk etymology already happened in Latin or was Lydus’ invention. But note that an a-stem bruta is also attested in the second Erfurt Glossary (9th century, but possibly relying on older sources, according to this very recent paper, which discusses our words on pp. 300–302; the author does not exclude a Verner variant).
The Chronicon Paschale (p. 201 Dindorf) explicitly states that the βρούτιδες were young women who married soldiers (ὁ Ῥῶμος […] ἐξεφώνησε νόμον ὥστε λαμβάνειν τοὺς στρατιώτας πρὸς γάμον παρθένους, ἃς ἐκάλεσαν βρούτιδας); it does not trace the word back to Brutus, nor could it, since it’s describing an even earlier moment in Roman history, the time of Romulus and the rape of the Sabine women.
Confusingly enough, the Suda (β 560) at first glance seems to gloss Βρούτιδες as ‘Sibyls and prophetesses’. But I suspect that the entry’s actual point might be that, just like ‘Sibyl’ was the archetypal name for a prophetess, so the common noun βροῦτις stemmed from some individual’s proper name.
Roberto Batisti, thanks!
However, “the Roman women … took it upon themselves to be called Brutae” is said about the Latin word, not a Greek one. Same for the Chronicon.
Logically, wives of soldiers once were a distinct social stratum. Normal household depended on both men’s and women’s labour, wives of soldiers, whether they travelled with the army or not are a whole different story.
Russian soldátka is formally not too different from simple “soldier’s wife” (lit. “a soldieress” – except the feminine suffix ka instead of sha as in generál’sha) but for me is more like a term on its own for members of said stratum in imperial Russia. Perhaps because it fell out of use.
From Augustus to Septimius Severus, serving Roman legionaries were not actually legally permitted to marry, though the actual practical working out of this ban seem to have been complicated (as you might expect.)
An even more recent paper by Rübekeil mentions hints at Grimm/Verner variants of other nouns, most prominently the “elk” word.
Best Academia abstract ever: “Permission is granted by the publishers to post this file on a closed server which is accessible only to members (students and faculty) of the author’s institute. It is not permitted to post this PDF on the internet, or to share it on sites such as Mendeley, ResearchGate, Academia.edu.”
Oh, so the word uxor was banned. That helps!
That reminds me of looking for the earliest attestation of Storebælt, which turned out to be a royal letter of privilege affording free transportation on the ferries to the monks of a specific monastery in Odense — and their wives! (In connection with the idea that the “bælt” names in Denmark are connected with the Baltic word for sea — which would be balt vel sim. — and not with the article of clothing).
Interesting. I know the two names as Belt in German, so I was thinking of the English word the whole time.
Lithuanian for “sea”, however, is jūra, which is also the name of a river somehow (…and yes, the Jūra flows directly into the sea). Baltas just means “white”, though it does seem to be cognate with the Slavic “swamp” word…
But that’s now. The Wends had all of Falster at the time, there’s actually Fehmern Bælt south of Falster as well and I was told that the Balts called themselves that because they lived by the (Baltic) Sea. (Every other piece of water in Denmark that could be a bælt is a sund).
No -h- in Femern Bælt (previously Femer Bælt).
Fehmarn, in Danish Femern, is Germany’s 3rd-largest island, and nowadays also the 2nd-largest town in Schleswig-Holstein. It often gets final stress because of the a.
Huh:
Dieter Berger, DUDEN Geographische Namen in Deutschland (1993), disagrees:
Huh indeed. I didn’t scroll that far… and the source is lost at sea.* Seems like an early [v] was interpreted as /f/ instead of /w/; the island’s Germanic language history seems too complex to unravel any details, though, or for that matter to clarify whether *a > o had happened on the Slavic side (settlement by Saxons, Frisians and Danes in addition to the Slavs; evidently Christianized from Denmark). Anyway, the final stress seems original then, and the eh is a slightly desperate interpretation of ъ…
* “Fehler 404
Ahoi! Diese Seite ist wohl auf hoher See verloren gegangen.”
Edit:
That makes sense if the r can be explained.
The oldest forms of the name have no a: Fembre, Vemere, Ymbria (Danish)
FWIW, non-initial stress in older words and toponyms is alien to Danish, and so it’s ‘Femern. Being unstressed, the vowel in the second syllable is more or less neutralized so it could reflect e or a and we wouldn’t be able to tell. We are all supposed to know that it was Valdemar I (the Great) who vanquished the Wends and burned their effigy of Svantevit, but that was on Rügen (in 1159).
eher liegt ein germanisches *Fimber zugrunde, das zu altsächs. fimba ‘Haufe’ und aisl. fimbul ‘groß’ gehören kann.
So Fimbulwinter is ‘a winter spent on Fehmarn.’
Curiouser and curiouser.
July 25, 2026 at 10:06 am
whether *a > o had happened
The oldest forms of the name have no a: Fembre, Vemere, Ymbria (Danish)
Wherever the “a” in Fehmarn is from, there are other cases of /a/ where common Slavic has /o/ in the region, e.g., Stargard (here additionally the metathesis of liquids doesn’t seem to have happened yet.)
All the way to Stargard Szczecziński.
“which is also the name of a river”
Persian daryâ “ocean, sea, river” and Amu Darya.
…So Darius is etymologically Genghis Khan?
Encyclopedia Iranica has “Old Persian Dāraya-vauš, which is composed of the present stem dāraya- “hold” and the adjective vau- “good,” must be translated as “holding firm the good” (cf. analogous expressions in Vedic texts) or the like. ”
Would dm’s alternative work (i.e, with the long a in daryā?).
https://www.iranicaonline.org/articles/darius-i/
Ah, I forgot the lengths.
Wiktionary (دریا) gives two transliterations: daryā and daryâ (plural daryā-hā and daryâ-hâ. That’s cute (but where is daryaa?).
Arabic
نهر nahr, nahar “river”
بحر baḥr “sea”
too form a variously (e.g. graphically: بــ نــ h ḥ) rhyming pair for me.
(I think they rhyme to speakers as well to an extent, but I don’t know if that affected anything. Wiktianary gives PS *nahar for one and PWS *baḥr for the other: “Replaced the original word *tihām– in West Semitic, but was in turn replaced in Northwest Semitic by *yamm-.”)