A Dialectal Squirrel.

I ran across a reference to Mandelstam’s 1922 essay Конец романа [The end of the novel] and wound up rereading it; it’s a short and fairly boring analysis of how the novel, which flourished in the 19th century with its focus on individual psychology, was now passé because individual psychology was no longer important. In the first place, that’s a silly way to look at literature; in the second place, most of the essay is written in a formulaic style barely recognizable as Mandelstam’s prose; and in the third place, the novel never died, so the whole idea is moot. (He says the swan song of the classic European novel was Romain Rolland’s Jean Christophe, which was wildly popular a century ago; anybody know if it’s still worth reading?) However, it comes alive at the very end:

Очевидно, силою вещей современный прозаик становится летописцем, и роман возвращается к своим истокам, к “Слову о полку Игореве”, к летописи, к агиографии, к “Четьи Минеи”. Снова мысль прозаика векшей растекается по древу истории, и не нам заманить эту векшу в ручную клетку.

Clearly, by the force of circumstance the contemporary prose writer has become a chronicler, and the novel is returning to its sources, to the Lay of Igor’s Campaign, to chronicles, to hagiography, to the Chet’i Minei. Once more the thought of the prose writer hastens like a squirrel over the tree of history, and it is not for us to lure this squirrel into a cage.

The last line refers to a very famous passage from the opening of the Lay of Igor’s Campaign:

Боянъ бо вѣщій, аще кому
хотяше пѣснь творити,
то растекашется мыслію по древу,
сѣрымъ волкомъ по земли,
шизымъ орломъ подъ облакы

For the vatic Boyan, when he wished to make a song for anyone, hastened in his thought over the trees, like a grey wolf over the earth, like a dusky eagle beneath the clouds.

We discussed this passage back in 2011, and there is much back-and-forth in the comment thread about a common but unfounded suggestion that мыслію ‘in thought’ is an error for мысью ‘like a squirrel,’ and Mandelstam is clearly working from that reading — except that he replaces мыс(л)ью with векшей, where векша [veksha] is a dialect word for ‘squirrel.’ He must have liked that word, because he also uses it in the last stanza of a 1937 poem:

И век бы падал векши легче,
И легче векши к мягкой речке —
Полнеба в валенках, в ногах…

The age could fall lighter than a squirrel,
lighter than a squirrel to the soft stream.
Half of the sky is wearing winter boots.
  (Translated by Richard and Elizabeth McKane.)

So my question for Russian-speakers is: how does the word векша sound to you (if it’s familiar at all)? Is it rustic, amusing, quaint, what? And my question for English-speakers is: are there any English dialect words for ‘squirrel’? You’d think there would be, but I haven’t turned up any. (We discussed the older word aquerne, from Old English acweorna, here.)


  1. (I assure you, it is pure coincidence that there are two squirrel posts in a row. I once had three Latin-titled posts in a row, equally unpremeditated.)

  2. Red squirrels are called “boomers” in the Appalachian mountains. I’m surprised there aren’t more words for such common little animals.

  3. 45 years ago I attempted to read Jean-Christophe in English translation several times, and abandoned it each time. I haven’t looked at it since. I assume that it belonged to my mother.

    In Le Guin’s essay “Science Fiction and Mrs. Brown” (1975), she restates this view thus:

    So then let me play my own enemy for a little, and try to argue the other side: the antinovel, or postnovel, point of view, which says that science fictioneers will never be novelists, and a good thing too.

    From this point of view, the novel, the novel of character, is dead — as dead as the heroic couplet, and for the same reason: the times have changed. Such writers as [Edmund] Wilson and [Margaret] Drabble are mere epigones, draining the last dregs of an emptied cask; such writers as Bhattacharya and Garcia Marquez flourish only because their countries are marginal to the place of origin of the novel, which was late in arriving at the periphery and correspondingly late in dying there. The novel is dead; and the task, the hope, of a new form such as science fiction is not to continue the novel, or to revitalize it, but to replace it.

    There is, really, no Mrs. Brown anymore. There are only classes, masses, statistics, body counts, subscription lists, insurance risks, consumers, randomly selected samples, andvictims. Or, if somewhere beyond all the quantification some hint of quality remains, some wisp of Mrs. Brown, she is not to be reached any longer with any of the traditional tools of fiction. No one can catch her. She has been too profoundly changed by our life, and too rapidly changed. Mrs. Brown herself has attained the speed of light, and become invisible to our finest telescopes. What is “human nature” now, who dares talk about it seriously, in 1975? Has it any recognizable relation to what was called “human nature” in the novel a century ago, which we now see as one tiny, limited fragment of the vast range of human variety and potentiality?

    The subject matter of the novel was the conscious, articulate portion of the minds of certain Europeans and North Americans, mostly white, mostly Christian, mostly middle class, mostly quite unaffected by science and, though affected by technology, totally uninterested in it; a handful of natives intensely interesting to the to the ethnologist because of their elaborate developments of manners, and their extraordinary absorption in interpersonal relationships. They thought their nature was human nature; but we don’t, we can’t. They thought themselves a norm; we have no norm. Through technology, which lets us travel and converse, and through such sciences as anthropology and psychology, we have learned too much about the complexity and variety of human behavior and the even vaster complexity of the human mind, conscious and unconscious; we have learned, that is, that we really know almost nothing at all. Nothing solid is left, nothing to take hold of.

    [Comparison of Dickens’s character Mrs. Gamp with a hard-to-delineate 1975 American analogue omitted.] She doesn’t amount to enough. She is a drifter, a pawn, a fragment, jagged bits of a person never annealed, never grown to a whole. Is there enough of her, indeed, to enter a novel as a real character, enough to paint a portrait of? Isn’t she, aren’t we all, too battered, too changed and changeable, too whirled about, future- shocked, relativized, and inconstant, ever to sit still for a painted portrait, ever to stay still long enough that the slow, clumsy art of the novelist can catch up with us?

    Click, the camera-eye — a moment, not a person, not a portrait, only a single moment implying nothing before or after, no continuity, click. And the whirr of the movie camera, catching the moment as it dissolves into the next, unrelated moment. These are our arts. The technological arts, dependent upon an incredible refinement of machinery and a vast expense of mechanical energy, expression of a technological age. There is poetry, still, but there is no more Mrs. Brown. There are snapshots of a woman at various moments. There are moving pictures of a woman in various places with various other persons. They do not add up to anything so solid, so fixed, so Victorian or medieval as a “character” or even a personality. They are moments; moods; the poetry of flux; fragments of the fragmented, of the changing of the changed.

    Do we not see this foreshadowed in the art of Virginia Woolf herself?

    And what is science fiction at its best but just such a “new tool” as Mrs. Woolf avowedly sought for fifty years ago, a crazy, protean, left-handed monkey wrench, which can be put to any use the craftsman has in mind — satire, extrapolation, prediction, absurdity, exactitude, exaggeration, warning — an infinitely expandable metaphor exactly suited to our expanding universe, a broken mirror, broken into numberless fragments, any one of which is capable of reflecting, for a moment, the left eye and nose of the reader, and also the farthest stars shining in the depths of the remotest galaxy?

    If science fiction is this, or is capable of being this, a true metaphor to our strange times, then surely it is rather stupid and reactionary to try to enclose it in the old limits of an old art — like trying to turn a nuclear reactor into a steam engine. Why should anyone try to patch up this marvelously smashed mirror so that it can reflect poor old Mrs. Brown — who may not even be among us anymore? Do we care, in fact, if she’s alive or dead?

    Well, of course Le Guin does.

  4. I think векша was the name of the smallest coin in ancient Russia. Didn’t know it actually meant a squirrel.

  5. I don’t know whether the New England Transcendentalists count as dialect, but Thoreau told us that the red squirrel was also called chickaree and Emerson’s “Fable” about the mountain and the squirrel gave the latter protagonist as bun (the non-diminutive of bunny, I suppose).

  6. January First-of-May says

    And my question for English-speakers is: are there any English dialect words for ‘squirrel’? You’d think there would be, but I haven’t turned up any.

    I’m not really an English speaker, but a Google search attempt turned up scug or skug (supposedly “Hants.” – whatever that means – according to an 1840 dictionary; shows up in many other places), con (said to be “Cumbrian” on one website, but I was unable to find more references – it didn’t help that the word is very common in other meanings), and chickaree (apparently a common term, of onomatopoeic origin, for some North American squirrel species; might not really count as “dialectal word for ‘squirrel'” as such because it appears to be limited to those two or three species).
    Oh, and apparently bun used to refer to squirrels as well as rabbits (as also noted by the commenter above).

    There are apparently also surprisingly many dialectal and otherwise local terms for assorted “ground squirrels” (Marmotini) – which might or might not be included in your “squirrel” category.


    I think векша was the name of the smallest coin in ancient Russia. Didn’t know it actually meant a squirrel.

    I wondered why that word seemed familiar! IIRC, веверица is another contemporary synonym (i.e. an even more archaic Russian word for “squirrel” that also referred to a small trade value).

    To clarify, both terms comes from the period where coins, as such, were not minted in Russia (or, at least, not in any significant quantities), and foreign coins and assorted other objects were used for trade; so far as I understand, there is some considerable debate whether the low-value trade objects denoted by векша and веверица were actual squirrel skins or not (IIRC, modern opinion is on the “not literally squirrels” side).

  7. Hants. is Hampshire.

  8. The OED historical thesaurus includes cat-squirrel (grey s.), picket-pin (ground s.), mountain boomer (American red s.) and the obsolete calaber (originally the fur, ultimately named after Calabria).

  9. January First-of-May says

    Hants. is Hampshire.

    That’s what I thought it was, but I wasn’t sure enough to say it definitely.

  10. Native Russian speaker here, Leningrad/SPb born (1969) and bred. To me, векша feels archaic or dialectal (Siberian?), something one may stumble upon in Dal’s dictionary for instance; definitely not a part of one’s active vocabulary. Then again, there exists a surname Векшин (if memory serves it was used in the hugely popular miniseries “Место встречи изменить нельзя”), so there.

  11. To me, векша feels respectably old-fashioned, archaic perhaps; if dialectal, then Northern, from the woods rather than from the steppes. It’s more or less obvious (although possibly wrong) that белка used to denote some other furry animal – squirrels are seldom if ever white.

    I suspect that I learned веверица from Nabokov, who hypothesized its affinity with “vair” (which I had only seen in a poem by Edith Sitwell).

  12. Trond Engen says

    The equation “small coin” = “squirrel hide” is very old. There’s an ancient Germanic borrowing in Finnish that shows this brilliantly

    (No references or further details. It’s Christmas Eve and I’m cooking.)

  13. To me, векша feels archaic or dialectal (Siberian?), something one may stumble upon in Dal’s dictionary for instance; definitely not a part of one’s active vocabulary.

    To me, векша feels respectably old-fashioned, archaic perhaps; if dialectal, then Northern, from the woods rather than from the steppes.

    Thanks, that gives me a good sense of it! And thanks to those who supplied English dialect words; the problem with them is that all of them are meaningless except to speakers of the dialect, whereas векша, though archaic/dialectal/old-fashioned, is apparently well enough known to Russian-speakers to be usable in literature. I was hoping for some way to translate it without using “squirrel,” but apparently not.

  14. It’s Christmas Eve and I’m cooking.

    What are you making?

  15. You also hear “tree-rat” sometimes. I don’t know if that helps.

  16. What are you making

    I’m afraid I must disappoint the Norwegian-American. We eat turkey on Christmas Eve. With rødkål, peas, prunes, and potatoes — and my wife’s waldorf salad. But we’ll have ribbe tomorrow.

    Merry Christmas to all!

  17. Thanks, and god jul to you all!

  18. I thought I’d look into the question of English dialect words for ‘squirrel’. Unfortunately, the EDD Online is now the EDD Offline, and the versions at the Internet Archive are hard to use. I found that Amazon is selling the whole work as an e-book for two bucks, but it’s not searchable. So no luck.

  19. A sad off-topic: it was announced today that Andrei Zaliznyak (Андрей Зализняк) died aged 82. He was a wonderful linguist (who proved inter alia that the abovementioned “Слово о полку Игореве” was a genuine 12th century text rather than a later forgery), and his yearly reports on newly found and deciphered birch bark manuscripts, fortunately still available on YouTube, are a delight to watch.

  20. The equation “small coin” = “squirrel hide” is very old.

    So I guess a hundred squirrel hides are worth a deer hide, or buck.

  21. E. M. Forster wrote an appreciation of the recently deceased Romain Rolland for The Listener. Forster aims to explain why Rolland had been forgotten since just before the first world-war. (What he has to say about Germany and heroes is no doubt colored by this being written in March ’45, in the midst of the second.)

  22. At one point, DARE queried its respondents about the word “squirrel,” but the results suggest that for the more part a squirrel is a squirrel is a squirrel. http://dare.wisc.edu/survey-results/1965-1970/fishing-hunting-wildlife/p27

  23. marie-lucie says

    January: chickaree (apparently a common term, of onomatopoeic origin, for some North American squirrel species

    “Onomatopoeic origin” is often blamed for the lack of possible cognates. I think it is more likely that the word is of American Indian origin. Onomatopoeia, an imitation of a sound or sound sequence is common for names of birds with distinctive cries, but not so much for mammals, even small ones.

  24. A sad off-topic: it was announced today that Andrei Zaliznyak (Андрей Зализняк) died aged 82.

    Damn, I’m sorry to hear that. Thanks for sharing the news.

    So I guess a hundred squirrel hides are worth a deer hide, or buck.


  25. The Comparative Siouan Dictionary shows various Siouan languages with ‘squirrel’ words such as Omaha-Ponca sį́ga, and warns, “it looks as if this is a diffused form more than a collection of cognates. Look-alikes in Tunica and Keresan would tend to confirm this latter analysis.”

  26. “Flickertail” is a name for the Richardson’s Ground Squirrel, and North Dakota is called the Flickertail State.

    And California Ground Squirrels are sometimes known as “beach squirrels,” at least in coastal San Diego, due to their dwelling on and in the coastal cliffs.

  27. Thanks, I can see using “flickertail” even if it’s not known to many readers as a word, because it’s transparent enough to give a general idea!

  28. January First-of-May says

    Damn, I’m sorry to hear that. Thanks for sharing the news.

    Me too. I just barely missed his annual lecture a few years back, and wanted to visit one eventually.
    I think I got to meet him in person at least once, but I’m not even sure of that…

    I think it is more likely that the word is of American Indian origin.

    This is effectively my default conjecture for weird words limited to North America, but all the sources I consulted said “onomatopoeic”, so I went with that.
    (Mind you, there are plenty of other sources for weird North American terms – both bison and grizzly have perfectly fine European etymologies…)

  29. The one thing about chickaree is that it’s attested as early as 1829 (Fauna boreali-americana, or, The zoology of the northern parts of British America: containing descriptions of the objects of natural history collected on the late northern land expeditions, under command of Captain Sir John Franklin, R.N.). That limits the languages from which the word could have come. That book considers the word an onomatopoeia, based on the squirrel’s call, and lists two native words, neither of which resembles chickaree (“Aroussen. Hurons” and “Annekcootchas. Cree Indians.”)

  30. January: Mind you, there are plenty of other sources for weird North American terms – both bison and grizzly have perfectly fine European etymologies…

    These two are not what I would call weird: le bison is a French word with nothing odd about it, of Germanic origin according to the TLFI; and grizzly is probably based on grizzled, referring to the silvery tips of some grizzlies’ fur.

    Y: the squirrel’s call
    I will have to bow to the opinion of persons who have heard a squirrel, I haven’t, although I have seen many.

    Chickaree is very similar to chickadee, the onomatopoetic name of a bird.

  31. Grizzly, at least, is perfectly transparent in English.

  32. David Marjanović says


    Blaukraut bleibt Blaukraut und Brautkleid bleibt Brautkleid.

    (Ten times fast.)

    Called -kohl in the north and -kraut in the south, and Rot- or Blau- in a bewildering patchwork depending on how much acid is involved in the traditional local way of preparing it. If you add apples, as they do in Vienna for example, it turns a redder shade of purple than otherwise.

  33. We call it czerwona kapusta ‘red cabbage’ in most varieties of Polish, but here in Wielkopolska (“Greater Poland”, including the urban dialect of Poznań) it’s modra [mʷɜdrɔ] kapusta ‘blue cabbage’.

  34. Ксёнѕ Фаўст says

    Also in (Upper) Silesia modro kapusta.

  35. Allan from Iowa says

    My boyfriend’s family (from the West Coast, with roots in Utah and Oklahoma) says “squinny” for the striped ground squireel.

  36. It seems they’ve got squinnies (Ictidomys tridecemlineatus) in Des Moines, too.

  37. A great word — I have to admit I’d be tempted to use it to translate векша!

  38. marie-lucie says

    Wikipedia has squinney for the singular.

  39. Russian Wikipedia gives векша as a synonym of белка обыкновенная (Sciurus vulgaris). Strangely, it indicates final stress in this word (векша́), which I’m pretty sure is a mistake, perhaps showing that whoever put it there was not really familiar with its pronunciation.

    In Poland we have four species of sciurid rodents: the alpine marmot or świstak (Marmota marmota), found only at high elevations in the Tatra Mountains, two rare and critically endangered ground squirrels (sousliks) of the genus Spermophilus (Pol. suseł, Russ. суслик), and the Eurasian red squirrel or wiewiórka (= белка обыкновенная). The last one, being common, eminently cute and synanthropic, is one of those wild animals that practically every child has seen. I don’t know of any dialectal synonyms (the Old Polish variant wiewierzyca originally referred to the squirrel itself but survived longest in the sense ‘squirrel skin’). The word has been covered here on LH.

  40. Strangely, it indicates final stress in this word (векша́), which I’m pretty sure is a mistake
    This dictionary says you’re right.

  41. David Marjanović says

    I’ve been bold and fixed it. (It still has to be approved.) I had to cite this page, though, because academic.ru is blacklisted as spam…!

  42. I had a vague feeling that I had seen something similar somewhere. As it turns out, it was an inscription saying “Towarzystwo Akcyjne Warszawskiej Fabryki wyrobów Metalowych – Marcin Weszicki” on an old mess tin. (Well, the original inscription is in Russian, which the software won’t let through).

    And now back to linguistics:

    Here are some of the lexicons people are making available to be viewed online or downloaded.
    Please let us know when you put your lexicon online or if you know of others that exist by emailing support1@lexiquepro.com.

  43. My grandma called вЕкша my little step sister when she behave bad.

  44. January First-of-May says

    and his yearly reports on newly found and deciphered birch bark manuscripts, fortunately still available on YouTube, are a delight to watch

    Finally got around to watching them this week, and yes, they’re a delight!

    (Though I only seriously considered posting about that on LH when I happened to notice him, in the 2017 report, giving “Slavomir” as an example of a pre-Christian name [31:31 on the video, phrase starts a few seconds earlier], and found it a sufficiently interesting coincidence to be worth posting here.)

  45. The equation “small coin” = “squirrel hide” is very old.

    History and Etymology for tiyn
    Kazakh tiɨn, tɨyɨn kopeck, literally, squirrel, squirrel skin (formerly used as currency)


    From Proto-Turkic *dEgiŋ (“squirrel”).

    IPA(key): /tijin/
    تىيىن • (tiyin)

    1. squirrel (mammal)
    2. kopeck (one-hundredth of a ruble)


  46. David Eddyshaw says

    The equation “small coin” = “squirrel hide” is very old

    Similarly with cowries in West Africa:


    In many languages “cowries” remains the ordinary word for “money” (e.g. Kusaal la’af “cowrie”, ligidi “cowries, money.”) Yoruba has a whole series of numbers for counting money based on owó “cowrie.”

    The conversion rate when cowries ceased to be legal tender was 20,000 cowries to five shillings. According to Teach Yourself Yoruba, older people (as of 1969) still used ẹgbàá “2000 cowries” for “sixpence”, and ọkẹ kan “one sack” for “five shillings.”

    Presumably by inflation, Kusaal yɔlʋg and Hausa jaka “sack” mean “£100” (or 100 Ghanaian cedis, which was very much less than £100 by the time I lived in Ghana.)

  47. Kusaal la’af “cowrie”, ligidi “cowries, money.”

    I’m almost afraid to ask, but are those related?

  48. David Eddyshaw says

    Oh, yes. I’m glad you asked! So glad!

    The whole noun class is moribund in Kusaal, but still pretty robust in some other Oti-Volta languages. The umlaut of /a/ to /i/ before pl -i is regular, at least insofar as it makes sense to call any member of such a small class regular; so also naaf “cow”, pl niigi, waaf “snake”, pl wiigi. In some words the pl vowel has been backported into the sg, eg miinf “okra seed”, pl miini beside ma’an “okra plant”, pl ma’ana.

    The sg la’af “cowrie” (which can also mean “very small coin”) is the regular outcome of *lagfʊ by rules which delete *g after low vowels, with laryngealisation of the vowel if it is short; these rules are still synchonically active.

    The -d- of the pl ligidi actually is exceptional, but not unparalleled (cf wief “horse”, pl widi; insertion of d turns up within a number of noun paradigms, sometimes as a repair strategy when a stem ending in a vowel adds a flexion beginning with a vowel.

    The prehistory of the noun class plural ending -i in Oti-Volta is unclear; like the much commoner pl ending -a, it has almost certainly lost an original initial consonant, but it’s not clear what that consonant originally was, or even what its position of articulation was. The various different Oti-Volta branches have adopted different strategies for the case where these endings are added to stems ending in vowels. There’s evidently been a whole lot of levelling and analogy.

  49. I love that stuff, so I’m glad I bit the bullet and asked!

  50. Trond Engen says

    There should be a map dividing the world into pre-monetary currency areas.

    How and when did Cowrie money spread to West Africa?

    Synchronity: Also, how do commodity money fint into Graeber’s accounr of the money/barter dichotomy?

  51. David Eddyshaw says

    This paper, linked form the one I linked above, is quite forthcoming:


  52. David Eddyshaw says

    It’s an interesting question what the various West African words for “cowrie” meant before there were cowries for them to refer to. (I never thought about this before.)

    The *lag- etymon which shows up in Kusaal must surely itself be old*. Its flexion is certainly of an ancient type, and AFAIK there isn’t any other obvious candidate with the same sg/pl patterning that might have served as a template for analogical remodelling of a loanword (a thing which is in general pretty widespread: elsewhere I’ve mentioned the highly irregular Kusaal lɔr “lorry”, plural lɔɔm; motor cars were probably not a feature of early Oti-Volta culture, nevertheless.)

    Moreover, the same etymon is extremely widespread within all of Oti-Volta, and the forms look as if they were all inherited from the protolanguage. Still, even before this particular question occurred to me, I noticed some oddities: for example, Waama, which is about as remote as can be from Kusaal within Oti-Volta, has díkítífā sg “cowrie.” This looks obviously remodelled from the plural: but Waama doesn’t have umlaut before the plural noun class suffix -i, so this suggests that the word is actually borrowed from Western Oti-Volta (cf Kusaal ligidi: the initial d for l and devoicing of the stops are areal features of the Atakora département of Benin where Waama is spoken.)

    Word for “money” are obviously good candidates for borrowing on first principles, after all.

    *It is homophonous (including tone) with the stem that turns up in Kusaal as lauk “item of goods”, pl la’ad “goods, merchandise”; as one of the two semantic fields associated with the noun class that la’af ~ ligidi belongs to is “small round things that usually come in large numbers”, I suppose that it’s conceivable that the primary meaning might have just been “seed-like things used in commerce.”

  53. Trond Engen says

    David E.: This paper, linked form the one I linked above, is quite forthcoming:

    Whoops, I missed your link, so thanks! Very interesting.

  54. David Eddyshaw says

    It’s a very interesting account both in economic and human terms. There are also numerous passages which appeal to my vicarious pride in the independent and enterprising spirit of the vigorous peoples of French West Africa, e.g.

    An additional influence can be discerned in the administrators’ reactions to the rejection of the franc. Reading the periodic reports of the province, one cannot avoid the impression that, for the administrators, the stubborn refusal of the local population to abandon the shells was an affront to their vision of the world, to their authority,and even to their male ego. For many second-generation administrators of colonial West Africa, the cowry was a symbol of a primitive life style. How could such an insignificant shell stand in the way of their desire to steer the “big children” under their care toward the course of better existence? The preference for the cowry was an unambiguous contradiction of their understanding of civilizing mission, which threatened the wished-for meaning of the colonial conquest. Finally French suspicion of gender insubordination may have also arisen. The fact that women vendors refused to accept in the marketplace the French coins brought by their subordinates to buy food may have wounded their pride as Frenchmen.


    The population found this piece and the similar two-centime piece too light. The centime pieces that had been distributed immediately returned, because the people of the province made all the payments to the administration in them, to rid themselves of the coins. In response, in 1908 the administration declared that it would not accept the centime coins back, except as a small proportion of payments that were made mostly with larger denomination coins. Thus, the centime piece turned from being special-purpose money to having almost no purpose at all. Realizing that this economic policy did not inspire confidence, in 1909 the administration reversed its decision and declared that it would take all the centime pieces without limit. This time it ended up quickly with a much larger number of centime pieces than it had issued.

    Interesting about how the five-franc piece, specifically, was eventually successful. To this day, people in Burkina Faso and Togo count francs in fives, so that (for example) in Mooré a hundred-franc piece is called a pisi “twenty.”

  55. Stu Clayton says

    one of the two semantic fields associated with the noun class that la’af ~ ligidi belongs to is “small round things that usually come in large numbers”, I suppose that it’s conceivable that the primary meaning might have just been “seed-like things used in commerce.”

    What’s the other semantic field ? “Large non-round things that usually come in small numbers” ? Or perhaps “any kind of serrated leaf” ?

    What would be an analogy in English ? Let’s take the class of nouns that have only one form for singular and plural (ie have neither a singular nor a plural). Does this class have more than one “semantic field” associated with it ? What advantages derive from thinking about word forms and meanings in this way ?

  56. January First-of-May says

    What would be an analogy in English ? Let’s take the class of nouns that have only one form for singular and plural (ie have neither a singular nor a plural). Does this class have more than one “semantic field” associated with it ?

    Well, if it’s the class I think it is, it does have at least one semantic field, as witnessed by the relatively recent (Algonquian) loanword moose being placed into that class precisely because it fell into an appropriate semantic field (as discussed previously on LH).

    That said, African noun classes are usually treated as if they were genders rather than declensions; I’m not sure to what extent this is an appropriate way of thinking about them in (what I understand to be) a typical West African situation where there’s about a dozen (or more) different ones and none of them really correspond all that well to what we think of as noun gender.

  57. John Emerson says

    On squirrelskins (above): Susan Sontag’s birth father, Jack Rosenblatt, dealt in furs in Inner Mongolia but died there in on 1938!when she was 5. Since I’ve known that I’ve thought that the world would be a better place if she had spent a couple of her teen years in Mongolia helping out, but fate would not have that.

  58. David Eddyshaw says

    What’s the other semantic field ?

    Large animals.
    There have been spirited attempts to bring this together with “small round things usually seen in large numbers”, which boil down to making the singular suffix “singulative” and proposing that the animals are of sorts normally seen in flocks/herds, but frankly that doesn’t really work (e.g. “cow”, sure, but not “horse” for example.)

    It seems much more likely that two originally distinct classes have simply fallen together formally, a process which has demonstrably happened elsewhere in Oti-Volta.

    As January says, Niger-Congo noun classes are basically genders, except that there are normally many more of them than three, and sex is the one thing that they don’t mark. Like grammatical genders, noun classes correlate with semantics but very imperfectly, to the degree that the meaning of a noun only predicts its class accurately in relatively marginal cases. It works a bit better the other way round with two of the classes in Oti-Volta, one of which has exclusively human membership, and another of which is nearly all liquids, substances, and abstractions. Otherwise, the bigger the class, the less correlation with meaning (though there is always some correlation.)

    Kusaal, like most of Western Oti-Volta, has given up grammatical agreement, so the classes no longer function like grammatical genders; but they’re still a thing for noun morphology.

  59. David Eddyshaw says

    The correlation of noun class with meaning is robust enough that it is exploited in noun derivation, as, for example, siinf “bee” beside siind “honey”, or Kʋsaas “Kusaasi people” beside Kʋsaal “Kusaal language.”

    So specialisation of the stem *lag- “item for trade” into “small round item for trade, usually seen in large numbers” by inflecting it with the fu/i class suffixes, is in accordance with how these languages work.

    It would imply that all the Oti-Volta languages other than the Western group have actually borrowed the word for “cowrie” from Western Oti-Volta (because of the umlaut thing), but that is historically quite plausible in view of the fact that the dominant precolonial kingdoms in that area used Western Oti-Volta languages.

  60. David Eddyshaw says

    The paper I linked above also points out that the cowrie performed much better in the end with respect to inflation than either the French franc or (even) gold. Anybody who changed all their francs into cowries and held their nerve up until the final demonetisation of the cowrie* would have made a huge financial killing. Bitcoin, nothing.

    *And beyond: they are so valuable now that it would be (as the author says) it would probably be worthwhile for some entrepreneur to reestablish the import trade in cowries to West Africa.

  61. Fascinating!

  62. John Cowan says

    Grizzly, at least, is perfectly transparent in English.

    So much so that I have seen it in edited print for grisly (a grizzly horror is not one caused by the sight of a bear) and in Dhalgren for gristle-ly, where it is used to describe the tactile quality of a human clitoris.

    at least insofar as it makes sense to call any member of such a small class regular

    Come, come: most of the surviving members of the n-declension in Modern English has perfectly regular plurals: ox:oxen, box:boxen ‘generic computer’, VAX:VAXen, Macintosh:Macintoshen. It’s true that child:children is a double plural, and fox:vixen, well, ….

  63. January First-of-May says

    I like to joke that “chicken” is actually the plural – the singular being “chick”.

    Bitcoin, nothing.

    The point with bitcoin, as I understand it, is that it had a microscopically low value for a while; anyone who bought bitcoins in 2010 and still has them would be ludicrously wealthy now, because at the time they were routinely traded in hundreds and thousands for nominal amounts in regular currency.

  64. David Eddyshaw says

    surviving members of the n-declension in Modern English

    A couple of centuries ago, when cowries first came on the scene in West Africa, the fu~i noun class and its umlauting plural were very likely fully productive in Western Oti-Volta languages. Even now, in the two Western Oti-Volta languages which still have the whole system of grammatical agreement by noun class, Farefare and Boulba, the class keeps its own separate 3rd person pronouns and so on. So, historically, it’s much less of a relic than the “ox(en)” type in English. Even in Kusaal, there are at least a couple of dozen words which always or sometimes show endings from this class (including malif “gun”, loaned ultimately from the Arabic midfaʕ, which has got assimilated to the pattern of mɔlif “gazelle” and makes its plural as mali, though sadly not *mili. The equivalent of “boxen”, I guess.)

  65. John Emerson says

    The noun classes of those African languages reminds me of the differences in classifiers/measures in Chinese. Sometimes the Chinese is similar to the English, as when you say “a pound of beans” or “ a quart of water”. And in general you use the empty classifier “ge” which just is grammatical :” yige dongxi “ = “one thing”. But many classes of nouns have designated classifiers, so you say “yizhi yu” which means “a length of fish”, like “a length of rope”, since most fish are sort of long. There are quite a number of such classes with designated classifiers, and usually they’re hard to make real sense of. Unfortunately, it’s been 35+ years and that’s all I remember; Classical Chinese, which I mostly study, doesn’t use classifiers.

    The jokes come easily, though I doubt that the Chinese find them very funny. I once showed my class a picture of a flat fish and asked if it would be called “yizhang yu”, with the measure for flat things. They liked me and were polite, and they would have been polite anyway because I was The Teacher, but I had the idea that they thought that it was a dumb joke, and that if they laughed they were just humoring me.

    The jok

  66. January First-of-May says

    The noun classes of those African languages reminds me of the differences in classifiers/measures in Chinese.

    I wanted to mention East Asian classifiers as a comparison, but didn’t recall the correct term for those and wasn’t sure if “count words” would be understood correctly.

  67. David Eddyshaw says

    It’s true that child:children is a double plural

    It’s just occurred to me that the mysterious -d- in ligidi “cowries” and its Western Oti-Volta cognates is also very likely to go back to a double plural (the analogy I drew with wief “horse” pl widi above doesn’t really work, because the -d- there is stem-final, and has been lost in the sg via a consonant cluster assimilation rule df -> ff.) The first pl suffix is the of the gʊ/dɪ class to which Kusaal lauk “piece of goods”, la’ad “goods” belongs, and the fʊ/i plural suffix has been added after that.

    This is actually not at all farfetched for Oti-Volta; there are a good many examples within the family of stem-plus-class-suffix combinations being reinterpreted as bare stems, with further suffixes being superadded; it’s a fairly natural consequence of the way that bare noun stems constantly occur as complete (though right-bound) words before adjectives and demonstratives. Furthermore, vowel-initial flexional suffixes get reinterpreted by taking stem-final consonants as part of the flexion, consonant-initial flexions get reinterpreted as vowel-initial by resegmenting the consonant as belonging to the stem, and hilarity ensues.

    @JE, JFoM:

    Greenberg himself hypothesised that Niger-Congo noun class systems originated from something like a classifier system, and there are indeed a lot of analogies even synchronically, though NC noun classes are much closer to grammatical gender in terms of apparent arbitrariness.
    There’s quite a literature about the whole grammatical-gender/noun-class/classifier thing; Alexandra Aikhenvald, in particular, has written whole books about it.

    From an Anglophone standpoint (though not so much from the Welsh) it’s remarkable just how very popular more or less arbitrary grammatical gender systems are among the languages of the world. Must fulfil some basic human need …

  68. PlasticPaddy says

    Clearly you did not absorb enough vitalist/animist principles whilst in Africa and remain an unrepentant Calvinist, hardened against the syllogism:

    All things have life.
    Living things are not eternal.
    Since there is life, life must be maintained by reproduction or spontaneous generation.
    There is evidence for reproduction but no clear evidence for spontaneous generation.
    Therefore most things we observe result from reproduction.
    Reproduction is most commonly sexual.
    Therefore most things have sexes.

  69. David Eddyshaw says

    And yet, in Ghana, nothing has a sex, including people (linguistically.) Even the Hausa have abandoned their beloved masculine/feminine grammatical gender, inherited from the legendary Proto-Afroasiatics, the People Without Vowels.

    These are deep waters.

    “God” is one of the very few Maasai nouns which is always feminine gender (along with “meat”, “milk” and “fire.”) They do these things differently in East Africa.


  70. David Marjanović says

    and fox:vixen, well, ….

    That’s just vowel reduction gone a little bit too far: it corresponds precisely to Fuchs : Füchsin. …OK, there’s also a Northwest Germanic u/o mess in there.

    “yizhi yu”

    No, zhī is the one used for land animals; the one for “elongate objects like ‘street’, ‘fish’, ‘pants'” is tiáo.

    Must fulfil some basic human need …

    Looks like it’s a surprisingly stable sink for various entropic processes.

    They do these things differently in East Africa.

    Does remind me of “women, fire, and dangerous things”.

  71. Does remind me of “women, fire, and dangerous things”.

    Exactly what I thought of.

  72. John Emerson says

    Thank you David. I had forgotten and was too lazy to look it up.

  73. Unfortunately, the EDD Online is now the EDD Offline

    It appears to be online again and searchable.
    The dialect words included for squirrel are:

    Cat-swirrel; Coggy; Con (also conn); Pug; Puggy; Scopperil (also scopperel, scopperill, scawprel, scop-a-diddle, scoperal, scoperel, scoperell, scoperil, scopper, scopperalt, scopperdil, scoppril, scoprel, scopril, skoperil, skopperdiddle, skopril); Scropel; Scrug; Scug; Squib; Squog; Swirrel; Vair

  74. OK, there’s also a Northwest Germanic u/o mess in there.

    Well, that gets us to OE fox:fyxen, but then when /y/ changed at the end of the OE period, the feminine form became /vɪksən/ in West Midlands ME, but /fʊksɘn > fʌksən/ in East Midlands.

  75. John Cowan says

    Perhaps the objection to fuxen was that it sounded too much like a ME infinitive.

    One of Larry Niven’s intelligent species are the fuxes, where the allusion is plainly deliberate. Their lifecycle is “virgin female with six legs” > mother with four legs > male with two legs”, with the two litters of eggs gestated in the dropped hindquarters.

  76. David Eddyshaw says

    Just happened to be looking at the Kusaal malif “gun” (ultimately from Arabic, as I said above) when it occurred to me to wonder why such a recent invention has such a nice short word for it in English; and discovered that it seems to be from the girl’s name “Gunnhild”:


    Who knew? (Well, not me, obviously.) Like “Big Bertha” avant la lettre.
    What is it with artillerymen and girls?

  77. David Marjanović says

    Well, a girl with an unusually fitting name, very much unlike Bertha (*Berhta).

  78. It is rude to tell a lady she’s a pleonasm.

  79. David Eddyshaw says

    True. The young lexicographers of today have no manners.
    At least it’s better than being called a nonce form.

  80. Trond Engen says

    You shouldn’t discuss a lady’s forms at all, let alone the compounding elements.

  81. ktschwarz says

    Since cowrie money was discussed above, possibly this is the right place to consider aggry beads as a medium of exchange as well. Their origin is much debated. Language Hat commented on another thread in 2006:

    I once had a very interesting exchange with the head of etymology at Merriam-Webster about the word aggry (a kind of millefiori glass bead); his explanation for the odd collection of proposed West African cognates in the big Websters can be summed up as “the guy who did African etymologies back in the ’20s had those words in a shoebox.”

    The shoebox etymology in Merriam-Webster’s Third Unabridged was:

    [of African origin; akin to Hausa gori snail shell used as an ornament, Twi agriratwefă a weight of gold, gyirapaw, a charm]

    … which was so unconvincing that the entry at m-w.com has been changed to “of unknown origin”. Merriam-Webster’s definition needs work, too: “a variegated glass bead found buried in the earth in Ghana and in England”. England??? Probably a misreading of the OED’s old definition from 1884: “coloured and variegated glass beads of ancient manufacture, found buried in the ground in Africa; they closely resemble the glain neidyr or adder stone of the Britons.” That resemblance was the fancy of Edward Bowdich (Mission to Ashantee, 1819), a colonial writer of the sort who liked to draw correspondences everywhere, as if towards some theory of Universal Folkloristics.

    OED revised aggry in 2012, still unable to find a source, but listing parallel words in other colonial languages:

    Origin unknown. Probably < a West African language. Compare Dutch agrie …, German Aggri (… perhaps < English).

    The substance is mentioned slightly earlier in English as Akori …, with which compare Portuguese †cori …, early modern German accarey …, early modern Dutch akori … It is unclear how closely these are related to English aggry ; the various forms may reflect borrowings of related words from two or more West African languages.

    There had been a claim in the meantime from M.D.W. Jeffreys, a British colonial administrator and anthropologist, that “My own researches show that when beads displaced the cowrie shell as currency, the local names for the cowrie shell — kori, akori, aggrey — were transferred to the displacing bead” (South African Archaeological Bulletin, 1959). Jeffreys included this in a seven-page list of “Words and Etymologies for the Oxford English Dictionary” collected from colonial sources, published in 1963; it appears that the OED did put this on file and use it, since several of his suggested quotations appeared in OED2, and a few more so far in OED3. So presumably they looked at his proposed aggry < cowrie but rejected it as not even worth mentioning.

    In this day and age, though, it’s not enough to throw out bad etymologies, since there’s always somebody eager to dig them out of the rubbish heap and insert them on Wiktionary, which now proclaims that aggry is from Hindi kauṛī, source of cowrie. You have to explain why they’re wrong. Is there enough known about the history of cowrie money to disprove this origin? David E?

    Above, DE asked:

    It’s an interesting question what the various West African words for “cowrie” meant before there were cowries for them to refer to.

    Also, the converse: why wasn’t the name cowrie imported along with the thing, when they were first introduced to West Africa?

  82. David Eddyshaw says

    The Hausa for “cowry” is in fact wuri, though guri does seem to occur as a variant. Mysteriously, the plural is kuɗi (usually = “money”); Western Hausa dialects have kurɗi, which is marginally less weird.

    The peculiar initial mutation looks like something from Fulfulde, but the k/w alternation is not a Fulfulde one, and anyway the Fulfulde word is seerre (plural ceede.) Dunno. It looks like a loan from somewhere.

    Gude means “trinket, jewel” in Twi. Given that aggrey beads seem to be associated with Ghana, Twi may be a better bet anyway. “Cowries” itself in Twi is ntrama or serewa.

  83. David Eddyshaw says

    Robinson’s dictionary speculates that Hausa wuri might be from Songhay, and says that ibn Battuta says that the Songhay used cowries as money in the fourteenth century.

    It doesn’t look much like the only word for “cowry” I can find in a Songhay language, though (Humburi Senni nò:rò.) Lameen will know …

  84. Could Arabic كودة kawda or kūda ‘cowries’, altered by passage through mediating languages, be the source of Hausa kuɗī, kur̃ɗī ? And could Arabic ودع wadaʿ, wadʿ, ‘seashells, cowries’ (colloquially also wudaʿ ?) similarly be the ultimate source of Hausa wurī̀ ?

    Note the remarks on the names of cowries in Egypt by the medieval Egyptian historian al-Maqrizi (1364–1442) in his work كتاب إغاثة الأمة بكشف الغمة Kitābu ʾiġāṯati l-ʾummati bi-kašfi l-ġummati (‘Book of the Relief of the Community through Uncovering of the Grief’). Yule translates these remarks in Hobson-Jobson here:

    c. 1420. — “A man on whom I could rely assured me that he saw the people of one of the chief towns of the Said [Upper Egypt] employ as currency, in the purchase of low-priced articles of provision, kaudas, which in Egypt [Cairo, Upper Egypt] are known as wada, just as people in Egypt use fals [a small copper coin]” — Makrizi, S. de Sacy, Chrest. Arabe, 2nd ed. i. 252.

    Text for the curious:

    فأخبرنى مَن لا أتّهم أنه شاهد فى بعض مدن إقليم الصعيد أهلها يتعاملون فى محقّرات المبيعات بالكودة وتُسمى بمصر الودع كما يتعامل أهل مصر الآن بالفلوس

    And de Sacy’s comments on the text are on page 253 here, including a mention of continued use of kūda in Egyptian Arabic in the 18th and 19th centuries. (Kazimirski, vol. ii, page 942, near the top of the left column, also gives the word as كوذة kawḏa.)

    Ibn Battuta however uses the word ودع wadaʿ for ‘cowries’ in his description of commerce in the Mali Empire here during his visit to West Africa (1352–1354). Evidently these cowries were brought to West Africa over land through Egypt from the Red Sea and Indian Ocean. Ibn Battuta uses the same word ودع wadaʿ in his description of the Maldives, apparently the main source of money cowries in later medieval times.

    Most sources agree that Egyptian Arabic كودة is a loanword from South Asia. Compare Hindi-Urdu कौड़ी کوڑی‎ kauṛī ‘cowrie’, noted in previous comments in this thread. This descends from Old Indic kaparda-, kapardikā, which in turn is possibly of Dravidian origin. Interestingly, although the Maldives were the main source of money cowries, the usual Dhivehi word used for cowries is apparently ބޮލި boli (actually, ‘mollusc shell’, in general?).

    Thomas Burrow proposed the ultimate Dravidian origin of Sanskrit kaparda- in ‘Dravidian Studies VII’ Bulletin of the School of Oriental and African Studies vol. 12, no. 2 (1948), pp. 365–396 (some abbreviations silently expanded):

    kaparda- m. a small shell or cowrie; braided and knotted hair, RV., etc.
    [Prakrit kavaḍḍa, kavaḍḍia; Hindi kauṛī, Panjabi kauḍ, Sindhi koḍ̱u]. Tamil kōṭu crookedness, flexure; shell, chank; coil of hair. Tamil kōṭu < *kavaṭu < *kapaṭu < *kapar-t/d- has already in the earliest form of the language altered as much from the original form of the word as Modern Indo-Aryan from the Sanskrit form. Sanskrit here has preserved a Dravidian word in an exceedingly ancient form.

    I am not sure if Burrow still upheld this in later years.

  85. As an aside, the Modern Hebrew word for ‘cowrie’ is פִּי הַכּוּשִׁי pi hakushi, lit. ‘negro’s mouth’. kushi is now derogatory, having at some point switched pragmatic positions with שָׁחוֹר shakhor ‘black’. As far as I can tell, pi hakushi is still the official zoological term, and it appears without embarrassed comment at the Hebrew Academy’s website.

  86. David Eddyshaw says


    Very interesting thoughts!

    According to Newman’s History of the Hausa Language, Chadic non-initial *r became Hausa y or i, which would mean that the “cowry/money” words would pretty much have to be loans from somewhere, but (a) there seem to be an awful lot of exceptions to this presumed law in inherited Chadic vocabulary and (b) the weird flexion looks like pretty good evidence of loanword status anyway.

    Given some of the transmogrifications that originally Arabic words have sometimes gone through on their way to Hausa via Berber, Kanuri and whatever, both your proposed origins look within the bounds of possibility to me.

    My favourite (possible) example of a strangely transmuted Arabic word is the Arabic حَتَّى, which has been suggested as the origin both of Spanish hasta and of the extremely widespread West African Wanderwort which turns up as e.g. Hausa har and Kusaal hali “as far as, until.”

    The latter is Jeffrey Heath’s suggestion; for the former


    I must say that both look pretty farfetched to me, but stranger things have happened …

    “Money” words must be prime candidates for borrowing on first principles. The Oti-Volta words for “money’ nearly all look related on the face of it, but they show an umlaut of the root vowel which only really makes sense language-internally in Western Oti-Volta, so most are probably really intra-Oti-Volta loans. Given the power of the (Western-Oti-Volta-speaking) Mossi-Dagomba kingdoms in that area in precolonial times that seems quite plausible.

  87. The cowrie words are mysterious to me too; proto-Songhay is *nogru, which I guess might be connected somehow but can hardly be the source for the Hausa form.

    The hasta etymology for Spanish looks okay to me, though not compelling, but the har/hal etymology is wrong; that whole complex of West African forms comes from Berber *har, which definitely does not come from Arabic ḥattā.

  88. David Eddyshaw says

    Actually, I can make a stab at connecting proto-Songhay *nogru with the Oti-Volta forms.

    Although the Oti-Volta words look reconstructable to POV, I think there is good reason to think that this is an illusion, and that in reality the “cognates” outside Western Oti-Volta are borrowed from WOV.

    The proto-WOV root would be *làg-. The n/l correspondence is not insuperable: at some stage, perhaps prior to POV itself, *l and *n were actually in complementary distribution, and in fact in WOV this is still the case for initial l/n: a following vowel is always nasalised after /n/, but never nasalised after /l/.

    There is also good evidence for a stem-final alveolar suffix of some kind after the root: cf e.g. Kusaal ligidi “money.” The /i/ for /a/ is an expected thing in Western Oti-Volta, where the plural class suffix -i causes umlaut: Kusaal waaf “snake”, wiigi “snakes,” The Kusaal singular of ligidi is in fact la’af *lagfu “cowry.”

    There is also a proto-WOV *làg-. “item of goods”, which I have always assumed has the same root as in the “cowry/money” word, but it’s not an obvious match semantically, after all, and it might even have provided a basis for remodelling of a Songhay loanword.

    It’s all a bit of a stretch, admittedly, but I don’t think it’s completely off the wall.

    The time-depth of proto-WOV, is, as ever, guesswork: the WOV languages are about as diverse as the Romance languages, which suggests something like 1500 YBP, but it seems difficult to believe that their current wide spread is unconnected with the rise of the Mamprussi kingdom and its offshoots in about the thirteenth century.

    And anyway, if we’re talking about a loanword, it could easily have spread from one WOV language to another: at the very least, the WOV languages must still all have been very similar to one another in the thirteenth century.

  89. ḥattā seems to be more borrowable than its European translations, same must be true for har… for whatever pragmatical reason. But I don’t understand how they are going to obtain har from it. And Heath couldn’t miss Berber har(~ar~al).

  90. David Eddyshaw says

    Heath actually floats this suggestion in the context of Songhay: I got it from his book on Tondi Songway Kiini. He only makes it very tentatively.

    The har word is all over in the savanna/Sahel bits of West Africa. Its manifestation hali is actually the best reason for claiming that Kusaal has a /h/ phoneme at all: most [h] is either just an allophone of /s/ or turns up only in “aha!” “hmph” type words, but hali is not only a very common word, but nobody seems to have any trouble at all pronouncing the initial consomant.

    On borrowabilty: most words in Kusaal (and its relatives) which translate as English conjunctions or prepositions are either outright loans or calques. What seems to have happened is that the whole syntactic category has been borrowed. You can actually express the same meanings perfectly well using other constructions, and presumably that was once the only way of doing it. But people just decided they wanted some of that sweet conjunction/preposition action for themselves. Why should Berbers have all the fun?

  91. John Cowan says

    If it comes to that, the Romance languages separated 1500 years ago, but their current wide expanse (i.e. the New World) is only a third as old, and New World French, Spanish, and Portuguese remain very close to their Old World co-dialects compared to the differences between the Old World languages. Other than creoles, no new Romance languages have appeared in the New World.

  92. David Eddyshaw says

    But I don’t understand how they are going to obtain har from it

    I should say that in Hausa, at least, deriving har from something like ḥattā seems entirely natural: syllable-closing /t/ in Hausa regularly turns into /r/. If Hausa had been the only language involved, I think Heath’s idea would look pretty reasonable,

    Lots of Oti-Volta languages make no distinction between d/r except word-initially, too. (In fact it’s easier to list the few that do make such a distinction.)

    So it’s not as far-fetched phonologically as you might think.

  93. David Eddyshaw says

    the Romance languages separated 1500 years ago, but their current wide expanse (i.e. the New World) is only a third as old

    Quite possibly a very good analogy. I think I’ve tended to operate with a rather simplistic Stammbaum mental model of Western Oti-Volta. The WOV homeland must have been pretty much where Kusaal is now, and perhaps not hugely more extensive, but there may very well have been a good bit of language differentiation already there in situ before the Mossi-Dagomba imperial expansions, but with the languages still in geographical contact and with diffusion of features between the languages still very possible.

    Dagaare is a good example. Traditional history suggests that the Dagaaba have only been in their current position, geographically well isolated from the rest of WOV, for a two or three centuries; on the other hand, Dagaare is much more different from what would presumably have been its neighbours, Farefare and Mampruli, than a couple of hundred years would account for. So it must already have been well differentiated from its neighbours before the Dagaaba got fed up with Dagomba suzerainty and set out to conquer the West for themselves.

  94. John Cowan says

    Dagaare is a good example.

    So I thought I would look at the WP article, as I generally do in order to get a quick overview, until I fell over the third sentence: “Dagaare language varies in dialect stemming from other family languages including: Dagbane, Waale, Mabia, Gurene, Mampruli, Kusaal, Buli, Niger-Congo, and many other sub languages resulting in around 3 million Dagaare speakers.” That is Not Even Wrong.

    But could it be that the reason Dagaare is so different from the rest of WOV is simply that it has been mugged by its immediate neighbors (like the Philadelphian theory of Germanic as originally Balto-Slavic-adjacent but with most of its vocabulary replaced by Italo-Celtic stuff)?

  95. David Eddyshaw says

    Yes, the WP article seems to have been rendered out the the original Gibberish by Google Translate.

    It’s trying to say that Dagaare is a dialect continuum, that the French call it Dagara, and that Wali is part of the continuum linguistically though the Wali themselves tend to regard it as a different language.

    Actually, Dagaare is not that different (that wasn’t what I had in mind in wittering about Stambaums.) If it were still geographically adjacent to the rest of WOV it wouldn’t stand out especially.

    It looks rather different, but it’s pretty much all phonology: it’s got more lenition of word-internal stops than most WOV languages, and also rhotacism (of the Latin kind), e.g. bie “child”, plural biiri, corresponding to e.g. Mooré biiga, pl biisi, Kusaal biig, pl biis. This has led to some remodelling of noun flexion (I once came across an exceptionally silly article about how significant it was psycholinguistically that one noun class in Dagaare has a plural suffix which is the same as the singular suffix of another class: not a glimmering of recognition that this was simply the outcome of a sound change regular enough to please any good Neogrammarian. People really ought to be made to study comparative linguistics before they are allowed to write linguistics articles.)

    Traditionally, WOV is divided into Southwestern and Northwestern, but one of the slightly unexpected things I discovered in looking into it is that although Southwestern is real, there isn’t a corresponding Northwestern group: its supposed members, Dagaare, Farefare, Mooré and Boulba are as different from one another as each is from Southwestern. In a nutshell, Dagaare is not really more aberrant than Farefare or Mooré.

    What I did have in mind was Boulba, the other WOV stray, way off in Benin. Boulba is aberrant, and probably does represent its own primary WOV branch, but again it’s not that aberrant. The other WOV languages share a notable sound change which Boulba lacks (POV *c *ɟ -> s z) and a morphological change (the peaceful winding up of the original “tree” noun class), but if the non-Boulba WOV languages were all adjacent to one another until fairly recently, as seems to be the case, those changes may actually have been areal, and not a common innovation. (The neighbouring Gurma language, Moba, also shares the “tree” thing.) Much of the rest of Boulba oddness can be plausibly ascribed to it having wandered into the Atakora Sparchbund, where they don’t like voiced plosives, and the fact that all the speakers seem to be bilingual in Byali.

    Mind you, in some ways this is a bit of a distinction without a difference: the non-Boulba WOV languages do share some non-trivial common innovations: the only question is to what extent those languages were already distinct from one another when those changes happened, and that isn’t a question that even necessarily has a neat answer (what is a “dialect”? what is a “language”?)

  96. David Eddyshaw says

    I should expand my treatment of WOV a bit in my magnum opus, and spell out these issues a bit better. I’ve tended to scoot over questions of the internal relationships within WOV, as the WOV languages are all rather boringly similar anyway, so that the differences aren’t very important in reconstructing proto-Oti-Volta. But they raise some interesting theoretical points (and there is some potentially relevant material from traditional history which bears on them too.)

  97. Stu Clayton says

    (that wasn’t what I had in mind in wittering about Stambaums.)

    Stammbaums. The rustic plural is no biggy, but single eyebrows may be raised.

    Witter, but witter with care!

  98. David Eddyshaw says

    Sorry. That should, of course, have read starmbormen.

  99. Stu Clayton says

    Google search throws up its hands in despair at starmbormen. There is, however, a Dutch word which goes by the name of stamboom. By coincidence (I hesitate to make a stronger claim) it means the same as Ger. Stammbaum. I am not sure of the significance of the capital S.

  100. David Eddyshaw says

    The capital S is just the Neogrammarians showing off.

  101. @DE, tt>r can’t be uncommon, because it happens the most common langauge, the English langauge!

  102. David Eddyshaw says

    Actually, I misspoke on this above: hardly any Oti-Volta language makes a distinction between d and r in initial position either: Nawdm is the only one. The only ones that make this distinction anywhere at all are Mooré, Agolle Kusaal (but not Toende Kusaal) and Nawdm.

    The details of just how the distinction has got lost vary quite a bit from language to language, so I suppose it’s some sort of areal thing. Hausa doesn’t participate though: not content with one distinct rhotic consonant, it’s got two, a flap and a trill. Oddly (it seems to me) it’s the trill that represents original syllable-final t, not the flap. I’d have expected it to be other way round. (The trill has got promoted to contrastive status because it also turns up in loanwords, of which Hausa has plenty.)

  103. John Cowan says

    The capital S is just the Neogrammarians showing off.

    They were overreacting to the grimms and their kleinschreibung.

    tt>r can’t be uncommon, because it happens the most common langauge, the English langauge!

    It’s quite remarkable how often linguists overlook English, perhaps because it is their metalanguage and therefore backgrounded. At Academia I often find myself saying “Y’know, your exotic feature is in English too”, and there is a colorable argument that English is of all things a mixed language, or at least only quantitatively different from one (half a million words, just 1800 native roots).

  104. David Marjanović says

    T-flapping never goes all the way to a merger with /r/, except in porridge (apparently from pottage). But it gets close.

  105. @DE it seems nearly everyone (everyone I cheched) who writes about this word mentions the Arabic etymology, not just Heath. Except I think Kossmann (“Berber loanwords in Hausa” does not look like a title that will deal with Arabic loanwords in Hausa…)

    As for t → r̃ , my impression from a few google snippets of Paul Newman (2004). Klingenheben’s Law in Hausa. (Chadic Linguistics 2.) Cologne: Rüdiger Köppe Verlag. Pp. ix+103. and a review [sci-hub link] of it is that everything is complicated.

  106. Your “review” link is missing a URL.

  107. [The review says that] Newman says Klingenheben’s Law is not applicable to geminates and that all words were -V back then, so not before the word boundary. Then [the review says that it says] in Eastern Hausa it works for coronals before the word boundary. It also says [this time a google snippet] that before coronals and again the word boundary (“where it mostly occurs in loanwords“) r is always a trill.

  108. David Eddyshaw says

    Yeah, there’s a lot of uncertainty about the details.

  109. Oh, it seems in Riffian there is a polyploid ḥtalami~ḥtarami (with Arabic ḥta and Berber al/ar)

    And also there seems to be a hatta in Dagbani.

    Function Words of Arabic Origin in Hausa, Patryk Zając (sci-hub)

    …West Africa is har̃ “until, even” derived from Ar. ḥattà “until”. It has been identified in 26 (!) languages (Baldi 2008: 131), i.a. in Koyraboro Senni (Heath 1999: 144–45) and Kanuri (Ziegelmeyer 2008: 23–24). As to the question of areality, this particle was presumably borrowed into Kanuri via Hausa (Löhr 2009: 29), and it seems to be borrowed by Hausa via Tamasheq. However, the trail of this loan is uncertain as har̃ is highly frequent in West African languages which historically were in continuous contact (Baldi 2008: 131; Gouffé 1971–72: 161–162). The same Arabic etymon emanates also as hàttā “including, even” but in this form it functions only in Hausa, Dagbani and Kanuri (Baldi 2008: 131). As a variant of har̃, this form is usually lesser used (Löhr 2009: 29). Other grammatical items borrowed from Arabic surfacing across West African languages (including Hausa) are: ammā (in 22 languages); immā (3); aw (4) (Baldi 2008: 52; 58) just to mention a few.
    Baldi, S. (2008) Dictionnaire des emprunts arabes dans les langues del’Afrique del’Ouest et en suahili. Paris.
    Gouffé, C. (1971–72) Notes de lexicographie et d’étymologie soudanaises, III. Contacts de vocabulaire entre le haussa et le berbère. Comptes rendus du G.L.E.C.S XVI, 155–73.
    Löhr, D. (2009) Reduction of Dialectal Features in Kanuri as Outcome of Language Contact, in N. Cyffer, G. Ziegelmeyer (eds.), When Languages Meet. Language Contact and Change in West Africa, 23–42. Köln.
    Ziegelmeyer, G. (2008) Aspects of Adverbial Subordination in Kanuri. MAJOLLS X,

    (I mostly quote it because of Dagbani and 26 languages) Ammā sounds like a good argument for massive borrowing of this sort of vocabulary from Arabic specifically.

  110. David Eddyshaw says

    Ammā is actually rather an odd one: it means “but” in all the languages I know of, including Hausa and e.g. Kusaal (amaa.) But in Arabic it means “as for …” Maybe it’s a regional Arabic thing (at least two Hatters will actually know.)

    Not all of these very-widespread words are from Arabic by any means. There’s also, for example Hausa sai, Kusaal asɛɛ “unless, except”, which also turns up all over. (For example, the endonym of the biggest Dogon language, Jamsay, contains it: neither part of this name is actually of Dogon origin, the first bit being Fulfulde jam “health.” The language name is taken from a common greeting – as with “Farefare”, too, incidentally. It’s like calling Cockney “Wotcha.”)

    Paradoxically, I think words like amma are actually not a good argument for massive borrowing from Arabic, at least in Oti-Volta, because the actual construction they appear in is actually an innovation. In Kusaal, for example, amaa is not a clause-linking particle in the sense that ka “and” and ye “that” are: formally, it’s a sentence-modifying adverb and it can actually co-occur with the “real” clause-linking particles:

    amaa ka ba pʋ siak o nɔɔrɛ
    but and they not agree his mouth.NEGATIVE
    “but they didn’t obey him.”

    And unlike the real clause-linking particles, amaa and asɛɛ don’t alter the tone of the following main verb.

    So these words haven’t been borrowed into an existing niche and displaced indigenous words. In fact, the inherited ways of expressing the concepts still exist alongside them (“but” can be expressed by the particle lɛɛ, which follows the subject, instead of preceding it. Oti-Volta languages do that a lot.)

    There’s also the problem that the “as far as” word often has /l/, not /r/, as with Kusaal hali. That’s not something you can explain by flapping of /t/.

  111. David Eddyshaw says

    Incidentally, no Dagbani *hatta appears in the very extensive SIL dictionary: the particle in fact appears there as hal or hali.

    Tony Naden’s Mampruli dictionary, under the corresponding word halli, actually cites the presumably very same S Baldi as suggesting an origin “from Tamasheq ar“, alongside Naden’s own suggestion, which is the Arabic one again.)

    All the WOV languages have l in this word, as do Moba and Gulmancema. And NAwdm.

  112. David Eddyshaw says

    In the rest of Oti-Volta, non-initial l has fallen together with d and r anyway, so you can’t tell what it started out as. Mbelime has hadi (though that is in fact pronounced [halɪ] for what little it’s worth.)

    Despite the word being found throughout Oti-Volta, even if you didn’t know that it was found all over the languages outside the group as well, you’d have to conclude that it couldn’t go back to POV because the sound correspondences don’t work, even if you posited an otherwise unknown POV *h for the initial. In principle it could be reconstructed for proto-Western, and that might even be historically possible if you assume it’s from Songhay or something rather than Hausa. It seems an unnecessary hypothesis given the highly infective nature of the word, though.

  113. Sorry. That should, of course, have read starmbormen.


  114. David Eddyshaw says

    I’m afraid that the word is not to be found in any Human language, rozele (as Stu has rightly noted.)

  115. David Eddyshaw says

    The various WOV dictionaries mostly seem to flag up Hausa as the immediate source of the hali “up to/as far as” word, but that looks unlikely to me. In Kano Hausa it’s har and in Western Hausa it’s either har again or ha with gemination of a following consonant; neither looks at all likely to have got borrowed as hali. I think the word must have got to the WOV languages by some other route(s).

    I think Hausa often gets tagged as the source of loans just because it’s so well documented.
    (Niggli’s dictionary of Toende Kusaal actually says that kut “iron” is a Hausa loanword. No idea why. The word has an impeccable Oti-Volta etymology and isn’t really at all similar to Hausa ƙarfe “iron.”)

    Older Hausa dictionaries (like Bargery’s) similarly tend to foist Arabic origins on to Hausa words even when the evidence is really very thin. Arabic rather lends itself to hopeful word-origin-hunters, what with its enormous vocabulary and cool dialects. And it’s rather better documented than … well, pretty much any (other) West African language.

  116. David Eddyshaw says

    Looking at Heath’s grammar, I see that the Tamasheq “as far as” in in fact hɑr, not ɑr as cited in Naden’s Mampruli dictionary; there is also an ɑr, but that one means “except.”

    So the Tamasheq form is identical to the Hausa, basically. (And neither sheds any light on the Oti-Volta hali type.)

    However, Timbuktu Songhay has hal and Tondi Songway Kiini has hali. Gao Songhay, awesomely, has hala, hali, hara, har, hal, kala and kal. Talk about covering all your bases …

    (Heath at that point reiterates his Arabic-origin idea, but once again, only tentatively.)

  117. “That’s not something you can explain by flapping of /t/.”

    As I understand, in Western Hausa it becomes /l/ before consonants and remains /t/ otherwise (biyar~biyat).

    “and in Western Hausa it’s either har again or ha with gemination of a following consonant;”

    Seriously? Oh.

    It seems, descriptions of Hausa dialects are only published (and read) in Nigeria:(

  118. David Eddyshaw says

    There’s a good grammar of the Hausa of Ader by Bernard Caron.

    (It has the TOC at the front, despite being in French, presumably because it is part of the series Sprache und Oralität in Afrika of the Frankfurter Studien zur Afrikanistik.)

    But there are indeed quite a few studies of Hausa dialects out there. Hausaists are spoilt. (It’s not even as if the dialects are all that different: to say how large an area Hausa is spoken over, it’s surprisingly uniform. Relatively-recent-expansion thing, presumably. I once travelled from Ghana to Kano overland with a Hausa-speaking driver from the Kumaasi zongo in the south of Ghana, and he had no trouble communicating with local Hausaphones at any point on the trip.)

  119. ‘there is also an ɑr, but that one means “except.”’

    “For dialects with ɑr, this ‘until, all the way to’ conjunction is homophonous with the ‘except’ particle (§11.3.1). However, T-ka distinguishes hɑr ‘until, all the way to’ from ɑr ‘except’.”

    Christiansen-Bolli gives the same translation for har in Tadaksahak:

    “ ….

    har ‘until’ is identical with the particle translatable as ‘except’ (see 4.6.3). “

  120. David Eddyshaw says

    It is odd how all these languages, belonging to quite unrelated families and often structurally very different, have nevertheless ended up sharing this little subset of function words.

    As I said above, in Kusaal (and probably elsewhere in Oti-Volta, though I don’t know the syntax of any other languages well enough to be sure) words like hali “until” and asɛɛ “except”, for all that they are very common and everyday words, have been shoehorned into the grammar (as it were): they’re not quite fully naturalised syntactically. For example, when linking sentences, semantically, they are actually construed syntactically like adverbs. In some ways that seems to make the phenomenon even weirder. Why would you even do that? (The language is quite able to express the same concepts in other ways.)

    I suppose I’m overthinking it really. As often, the answer to “Why would you even do that?” is probably just “Because I can. Wouldn’t you?”

    Come to think of it, all this is not a million miles away from the Japanese za that we were discussing the other day: borrowing of a whole grammatical construction along with its characteristic marker particle(s), and adapting it into something new along the way.

  121. @DE:

    in my experience, many trees are shtum. but it is true that the metaphorical ones are pretty damn talkative.

  122. @DE, if in WH har̃ is har, -tC- → -lC- and no change to -t in WH (and also that Klingenheben does not affect geminates in any Hausa) do not really support ḥattā > har̃. But I don’t understand where I’m supposed to look if I want to learn about Western Hausa:(

  123. David Eddyshaw says

    As far as printed books go, the Caron book is the most detailed account. It’s a full-dress reference grammar with texts and a vocabulary.

    As I say, though, the differences from “Standard” (i.e. Kano) Hausa are not huge (much less than Agolle versus Toende Kusaal, for example.)

    There’s a nice sketch (also by Caron) of Zaria Hausa here


    but I’m afraid that’s another Eastern dialect. (Its main peculiarity is that it hasn’t got grammatical gender, which is usual for L2 Hausa but not for L1 Hausa.)

    There’s a lot of published stuff on Hausa dialects, but I’m not sure how much is online. But I expect your Google-fu is greater than mine anyhow. Sokoto dialect would be the thing to look for.

    Ekkehard Wolff’s Referenzgrammatik des Hausa has a nice chart of the major dialect isoglosses in the introduction, but the grammar itself only makes occasional references to non-Kano dialects.

  124. I have seen it! Introduced thusly:

    Hausa Grammatical Sketch – HAL-SHS
    https://shs.hal.science › index › docid › filename
    Автор: G Sketch — The conjunction har means ‘as far as; up to, until, even, etc.’ and denotes the continuation of an action until a qualitative degree is reached, …


    (Автор “Author”)

  125. @DE, thank you.
    It seems the book is not immediately accessible, but maybe I’ll find something by the same author and I’ll keep it in mind.
    Thus far I only find “unpublished theses”. Titles in Hausa are particularly intriguing. If someone told “I’m a comparative Hausa grammar in Hausa” I’d read her, it’s an easy language. But they too are in libraries in Nigeria.

  126. David Eddyshaw says

    Probably the single commonest use of hali in Kusaal As She Is Spoke is actually just to mean “very, lots”, e.g.

    Li tɔe hali.
    it be.bitter as.far.as
    “It’s really hard.”

    Fʋ tʋm hali.
    you work as.far.as
    “You’ve really made an effort.”

    Titles in Hausa are particularly intriguing.

    There’s quite a tradition of modern Hausa grammatical study by L1 speakers, including stuff actually in Hausa, as you say. You can do degree-level work in Hausa in Nigerian universities, It probably helps that you can freely nick technical terminology from Arabic without it doing violence to the spirit of the language at all.

    Unfortunately it is indeed not easy to obtain works in West African languages from abroad. Almost all of what I’ve got in Kusaal or Mooré I got when I was in Ghana or Burkina Faso myself.

  127. There is Russian ot i do.

    ot (do) mean from (to) limits of something as opposed to iz (v) “from (to) inside”, iz-za (za) “from (to) behind”, ot (k) “from (to) vicinity of something”.
    Do is contrasted to k in that it points at limits, while k can be used with persons (I went k Masha, to Masha’s place), in the sense of “towards” with places etc. It can be a prefix: chitat’ “to read, dochitat’ “to complete reading [what you began reading]”.

    So ot i do means thoroughly.

  128. Unfortunately it is indeed not easy to obtain works in West African languages from abroad.

    I see some titles on Academia (Hausa. I doubt they have Kusaal) but without knowing what they are about, I’m not ready to start deciphering (and thus start my practical aquaitance with the langauge there too):-) Yet they are promicing.

  129. David Marjanović says

    Come to think of it, all this is not a million miles away from the Japanese za that we were discussing the other day: borrowing of a whole grammatical construction along with its characteristic marker particle(s), and adapting it into something new along the way.

    That’s what I think: a construction that can’t be calqued, so it can only be borrowed if its marker is borrowed along with it.

    If you borrow the concept of articles, for example, you can probably always find a demonstrative pronoun or a relative pronoun or some kind of singulative/individuating derivational ending or whatever to use for that. Not so, it seems, in these cases (the Japanese za isn’t “an article”, it’s much more specific than that).

    It probably helps that you can freely nick technical terminology from Arabic without it doing violence to the spirit of the language at all.

    I’m sure it does!

  130. David Eddyshaw says


    I think Google Translate does a fair job with Hausa, but I’ve never actually tried. Technical linguistics papers might be too much for its tiny non-brain, too.

    I’m not sure that I would agree that Hausa is easy, exactly, though I think that (like English) it’s comparatively easy to pick up enough to be of some practical use. And people are used to hearing and deciphering bad Hausa, so you don’t have to sound like a L1 speaker before people can make out what you’re trying to say.

  131. David Marjanović says

    Ekkehard Wolff

    A name that immediately makes you think “that man is meant seriously”.

  132. David Eddyshaw says

    Yes. It’s an excellent name.

    You instinctively don’t want to mess with an Ekkehard Wolff.

    I wouldn’t have the personality to carry it off, myself.

  133. I don’t know. “hard” without geminates….

  134. @DE, easy:
    There is a strategy that let’s you feel you’re playing wiht it rather taking an effort and yet keep feeling like you are learning a lot.

    But it also may depend on who I am, what language I speak (trivially, like “unrelated” and not so trivially) and of course depends on my approach. And all of this is only true for your first steps in Hausa, I’m not sure it will be also true for advanced Hausa – though being a person mostly interested in reading/listening* to I’m spared of many difficulties encountered by people who strive for accuracy.

    Also “there is a strategy” means: a natural strategy, that is, the strategy you try. Which of course depends on who I am: the girl who said “Russian is easy” has no previous experience with languages and approached Russian differently from those experienced learners who say it is a nightmare.

    *Actually what I want is to feel how native speakers feel, but this is not about easy/hard. It is about associations with words…)

  135. David Marjanović says

    I don’t know. “hard” without geminates….

    Depending on where he’s from, he likely comes with aspiration or a geminate. Definitely not both, though.

    It is about associations with words…

    And with the harder grammatical subtleties, like aspect.

  136. “And with the harder grammatical subtleties, like aspect.”

    Those too. My main source of frustration with langauges like English (that’s the languages I know the best but mostly from reading) is that in Russian I remember when I learned what and feel things. In English it is like walking in rubber boots across a grassy field as opposed to being a tiny bug in that grass. I still get from A to B, and that’s all.

    As for accuracy… I mostly read. When I need, I can make myself understood. Yes, maybe native speakers would be more comfortable if my English (or other L2) were less sloppy, but specifically with English it’s an experiment. I consciously avoid serious efforts* because seeing how my English reacts at change in input is too exciting (I think i mentioned how I once learned to understand spoken TV English – for I could not udnerstand even that before while being a very fluent reader – “overnight” with literal meaning of “overnight”. It was exactly “woke up and discovered that I can understand it”).
    So it grows like a tree, even though these 2-3-4 years it seems to “deteriorate” (my brain used to thinking in broken English began developing drasvish out of it. Quotemarks because maybe it becomes better drasvish…).
    * actually, I’m extremely curious about teaching methods and do like the idea of hiring a teacher (online) just to learn how she teaches. But yet. And I think I could do many think about it without teachers. With other langauges I don’t avoid it, but still English is my best foreign langauge.

  137. There was a song (by a group who writes lyrics by somewhat surrealist methods, I mean authomatic writing) clearly inspired by a foreign name that sounds brutal for Russian ear, Rogan Born. Contains many short words with r. https://www.youtube.com/watch?v=D2JrBKvdtvU

    (not a song I like, just spotted it somewhere. Even though I sympathise to the group, I hardly know 1/100th of their songs).

  138. David Marjanović says

    Contains many short words with r.

    I expected something similar to RRRRRRRRRRRRRRRammstein, but that’s not at all the effect they’re going for. Their Rs are hard to hear (especially against the background) – they’re uvular, some are probably one-contact trills, the others are probably approximants.

  139. It’s an excellent name.

    You can’t judge. Someone noted that without prejudice, “Arnold Schwarzenegger” could be the name of the kid in high school who timidly asks you if you want to join the math club.

  140. David Eddyshaw says


    Still, “Ekkehard Wolff” sounds like a Bond villain.
    You don’t get more badass names than that. I can just picture his Supervillain Lair.

  141. I immediately thought that it would be a better name for Curd Jürgens’ character in The Spy Who Loved Me. However, a minute’s upon further reflection, I realized Wolff wouldn’t be right for Jürgens’ web-fingered, ocean-obsessed maniac. It would have been better for Hugo Drax in the next movie, but they wanted him to be French….

  142. @DM, yes.
    The first verse, bold for hard /r/, italics for soft /r’/ (next verses don’t have that many rs).

    Тень Рогана Борна, нет / Выбери небо мне / Не говори не так / Передо мною ночь / Сине-зелёная / Выбери утро мне / Волосы белые / Волосы чёрные / Волосы белые / Волосы чёрные

    (I also marked l‘s. And I think I should have marked n’s and rn’s but…)

    The problem is that the singer mispronounces both r’s:) I do, I have a flap for the soft r (common for Russians) and an uvular trill for hard r (it is not very noticeable, if it were an uvular trill for the soft r, my accent would sound much thickier to others).

    But here (2:44, concert – I’m less sure abotu the recording posted above) in rogana borna I hear just approximants.

    So all the brrrrutality that he and others hear in the name is imaginary.

  143. Hausa of Ader by Bernard Caron.

    Apparently based on his thesis (Description d’un parler haoussa de l’Ader (république du Niger), doctorat d’État, université Paris 7 (directeur : Antoine Culioli)). 745 p. – either it is thickier or less dencely printed.


    At least he published “A Linguist’s Field Notes” (online for free).

  144. David Eddyshaw says

    It’s pretty good.

    He adopts a slightly different approach to describing the language’s verb morphology from what seems now to be the standard (Newman) line, which is interesting. He takes the actual verbal noun as basic rather than some sort of abstract stem.

    Not really worth taking the trouble to get hold of unless you have a particular interest in the language, though, especially given the fact that it doesn’t differ all that much from Kano Hausa, which must easily be the most thoroughly documented of all African languages (unless you count Arabic.)

  145. John Cowan says

    flap for the soft r

    I was actually taught flap+/j/.

  146. proto-Songhay is *nogru

    The proto-WOV root would be *làg-… There is also good evidence for a stem-final alveolar suffix of some kind after the root: cf e.g. Kusaal ligidi “money.”

    Can this be Arabic نقد naqd ‘ready money, cash, cash payment’, plural نقود nuqūd ‘sums of cash’? Compare Hausa lakadan in Bargery (from this site):

    [lakada”n] {adv.}. (Ar. nak’dan.)
    1. Cash down. (= hannu da hannu.) e.g. lakadan ko ajalan ? cash or credit ?
    2. = hakika q.v.

    1. {adv.}. Undoubtedly. (cf. azimun-azimun; dahir; dimun; gaskiya; kibbu; kirbulle; lakadan; labu; lazu; lalle; tabbaci; hakk’an; tamman; tabbas; wajuban; yanke; labudda; lilla alai hin; nana alaihin.)

    I gather that cowries were apparently originally supplied to Sub-Saharan Africa via Morocco, so perhaps instances of an apparent */g/ in the word can be explained by the pronunciation of Classical ق q as /g/ in many varieties of Arabic of North Africa. (See especially the passage quoted from Ibn Sa’id al-Maghribi on page 345 of this study (Mervyn Hiskett (1966) “Materials Relating to the Cowry Currency of the Western Sudan–II: Reflections on the Provenance and Diffusion of the Cowry in the Sahara and the Sudan”, BSOAS 29.)) Apparently Egypt later became the main commercial center from which cowries entered West Africa.

    (I recalled this LH thread while looking at the Kurmanji word next ‘bride price’ (beside neqd ‘money’ in the earlier Kurdish poets), ultimately from Arabic naqd, and looking around for other examples of Kurmanji /x/ from Arabic q.)

  147. David Eddyshaw says

    Can this be Arabic نقد naqd ‘ready money, cash, cash payment’

    A very interesting thought. I don’t think it’s outright impossible, but there are difficulties with the hypothesis.

    The initial l for n is obviously awkward. POV *n and *l were nearly in complementary distribution (probably actually in complementary distribution at some pre-POV stage), but this must long antedate any plausible time scale for Arabic loans. I don’t think this is an insuperable obstacle though: loanwords often seem to do odd things, and spontaneous n/l variations have happened in Oti-Volta in some cases.

    The forms for “cowries/money” in all Oti-Volta languages that have this etymon show umlaut before the plural suffix (POV *-ɰi.) This umlaut is only consistently seen in Western Oti-Volta and its closest relative, Buli-Konni, which suggests that non-Western languages have actually borrowed it from WOV: this is not improbable, given that the various dominant languages of the Mossi-Dagomba states were all WOV, and these states dominated the area in precolonial times. That does mean that the word would have to have been borrowed into WOV early enough to be retrofitted into the relevant “animal” gender, which in WOV also includes “small round things” (notably, all words for seeds.)

    However, that doesn’t necessarily imply a great time depth: these languages even now are fond of fitting loans into existing patterns by analogy, my favourite example being Kusaal lɔr “motor car”, plural lɔɔm (beside the much less outré lɔya), modelled on the highly irregular Mɔr “Muslim”, plural Mɔɔm, where the formation, otherwise unique within Kusaal, probably reflects both singular and plural having been borrowed from Mooré More, plural Moeemba.

    A point in favour is that *lagd- is an unusual stem shape for a noun not transparently derived from a verb stem, but this is not conclusive: there are other synchronically unanalysable non-loan nouns with *d as a derivational suffix, often referring to animals, e.g. Kusaal yugudir “hedgehog”, where the *yug- root turns up with other suffixes elsewhere in Oti-Volta.

    I think the main difficulties are with time scale: as far as I can tell, there aren’t any other Arabic loans old enough to have undergone this kind of significant remodelling prior to intra-OV borrowing. On the other hand, the whole concept of “money” might easily have come with Islam, and it wouldn’t be surprising if “money” was an earlier loan concept than anything specifically relating to Muslim culture.
    (To this day, Islam is strongly associated with trade in West Africa: trade has probably had much more of an effect in spreading Islam than any jihad ever did.)

  148. while looking at the Kurmanji word next”

    The etymology given for qelen in Wiktionary is unexpected.

    1. “erebî قران‎ (qeran: zemawend, dawet, dîlan)” – what’s this?

    2. Russian has lost its original word for brideprice (I don’t even remember it, but if anyone is curious, věno), today it is understood as a Turkic thing, designated by a borrowed word kalım. So appearance of Turkic *kẹlin (here asterisks denote Wiktionary reconstructions, modern gelin etc. “bride”) in Kurdish etymology is not unexpected for me, absence of Turkic *kalïŋ (modrn kalın etc.: “thick”, also “brideprice” in a many Turkic languages) is.

    I guess the two must interact already in Turkic? Wiktionary also gives me Kazakh qalıñdıq “bride”, “thickness”.

    3. they mention Turkic kalın and Kurdish qalind “thick”…
    … to explain why there are qeleng and qeleng alongside qelen. That is, merely as example of /d/ after /n/:)

  149. David Eddyshaw says


    Sorry, I missed that you had cited Lameen’s proto-Songhay (and I’ve just reread the thread and realised how much I’ve been repeating myself: I can’t even remember what I said before myself …)

    I think Songhay would make a very plausible way for an Arabic original to have got to (Western) Oti-Volta. Quite apart from the fact that there are other unequivocal Songhay loans, it would make a lot of sense historically if both cowry money and the name for it came from Songhay.

    If Lameen thinks that the Songhay could plausibly be from Arabic, I think that would substantially increase the plausibilty of the WOV forms being of the same origin. In fact, thinking about it some more, I don’t think the umlaut thing is really an obstacle; it’s really only the initial l/n thing.

    The Hausa lakadan is interesting. Bargery is sometimes rather overenthusiastic in pegging Hausa words as Arabic loans, though, and I also think Hausa probably came on the WOV scene too late to be a likely source of what would probably have to constitute a loan into proto-WOV. There are plenty of Hausa loans in the modern WOV languages overall, but both the frequency and the individual word forms vary a lot between languages. There are many more in Dagbani and Mampruli, where cultural Muslim influence has been strong for centuries, than in Kusaal, Mooré, Farefare or Dagaare, where it hasn’t. (Adams Bodomo, who is himself Dagaaba, denies that Hausa is a widespread lingua franca in northern Ghana at all: he’s resoundingly wrong about that, but the fact that he himself comes from the far west of the Oti-Volta territory may well explain how he came to this odd conclusion.)

  150. I had never thought of the possible Arabic etymology for the Songhay form. It’s very tempting, but I don’t see how to make the vowels work: you’d expect a in the first syllable rather than o, and probably some epenthetic vowel before the d (if we dare to assume that the r comes from a d; the Northern Songhay data doesn’t really support that, but there are some odd r-d correspondences within NS so maybe that’s not a problem.) Still, I wouldn’t rule it out for the moment.

  151. David Eddyshaw says

    The WOV forms definitely speak to *lagd- rather than *lagr-, too, which is a point against an immediate origin in Songhay.

    (Only Agolle Kusaal and Mooré within WOV actually still have a r/d contrast, though it certainly goes back to proto-WOV: but the forms in other Oti-Volta languages, presumably borrowed from WOV, also confirm an earlier stop, not flap, e,g, Waama dikiti “money” = Kusaal ligidi, Mooré ligdi.)

  152. The initial l for n is obviously awkward

    Roy Clive Abraham (1962) Dictionary of the Hausa language offers, in addition to Hausa lakada̱ŋ ‘cash down’ (adverb) from Arabic adverbial accusative نقدًا naqdan, the following apparent cases of dissimilation of n- to l- in the environment of a following nasal in Arabic loanwords (I’ve edited and recombined his entries for presentation here but kept his tonal notation, with a sign ̝ (like the IPA sign for raised articulation) representing the falling tone, I believe):

    la̱dāma̱ = n. na̱dāma̱ (Ar.) f. (1) regret, remorse. (2) being worried, e.g. n nā̱ na̱dāma̱r wuriŋ kwānā I’m worried as to where to pass the night. yanā̱ na̱dāma̱r wuriŋ mā̱tassa̱ he’s worried about his wife.

    (Arabic ندامة nadāma.)

    li̱’ima̱ f. prosperity (= ni̱’ima̱ q.v.).

    ni̱’ima̱ f. (pl. ni̱’imōmī = ni̱’i̱mū) Ar. (1) (a) prosperity. (b) fertility. (c) wurī ma̱i ~ (i) fertile place. (ii) cool, shady place. (2) (a ~ tā saukō the rainy season has begun. (b) gōra̱n na̝n bā̝ shi da̱ ~ these kolas have no dampness in their wrappings.

    (Arabic نعمة niʿma.)

    La̱ma̱rūdu̱ Ar. the giant Nimrod.

    (Arabic نمرود Namrūd, Numrūd. Lamarudu seems to cut a large figure in the legendary histories of the Hausa and the Yoruba. Fun fact: a Fula word for ‘sugarcane’ is lamarudu, I suppose from *‘giant’s grass’ or the like.)

    I have no idea whether this dissimilation occurred within Hausa or is to be attributed to mediating languages.

  153. David Eddyshaw says


    I think there is really no prospect of deriving the WOV “money/cowry” word via Hausa, though. It’s not just the initial consonant mismatch, but the internal /k/ and the vowels of the non-initial syllables. The word would have to have been borrowed into proto-Oti-Volta itself to account for all that, and even then there would be phonological difficulties. And Hausa itself has a quite different word (presumably also a loan, but clearly from somewhere else) for “money/cowry.”

    WOV l/n are in fact in complementary distribution word-initially in languages which retain contrastive vowel nasalisation: /n/ appears before nasal vowels, /l/ before oral. The obvious explanation with /n/, viz that the nasalisation is caused by the initial nasal (and it is regular after /m/), doesn’t account for the fact that /l/ is never followed by a nasal vowel (other alveolars can be.) I suppose that does open up the possibility that /n/ in loanwords might get borrowed as /l/ if the following vowel was oral, but there seem to be no actual examples of this, and a good many counterexamples, e.g. Kusaal Nasaara “European.”

    The Atakora languages make no d/l distinction, and often have /n/ for POV initial *d or *l before a nasal vowel, as in Nateni naama “beer” (= Mooré rã́am); however, the “money” word cannot have spread from the Atakora languages to WOV, because the plural umlaut seen in e.g. Kusaal ligidi is not seen in the Atakora: as a regular feature it is more or less confined to WOV and Buli/Konni. In any case, this change works the opposite way to what is needed to tie the word to Arabic.

    Unfortunately, in the light of Lameen’s comments, a derivation of the WOV forms from Songhay looks difficult too. Maybe, as I speculated above, the word is after all, a WOV-internal coinage from the same stem (*lag-) as “item of goods”, inflected in the “animal” gender, which in WOV (alone, I think) also includes “small round things.” That kind of derivation by manipulation of the core meanings of the “genders” is entirely characteristic of how things work in Oti-Volta (and indeed, in Volta-Congo in general.)

    That means that the stem-final *d of *lagd- would need to be explained as due to reanalysis of an original plural class suffix *dɪ as part of the stem, but there are actually a good few examples of such stem-suffix resegmentation processes having taken place in Oti-Volta, so that is not nearly so much of a stretch as you might think (and “item of goods” itself, naturally enough, occurs in the plural much more often than the singular.)

    A more complicated theory might be that the WOV is (sorta) derived from the Songhay after all, but via phono-semantic matching:


  154. Interesting stuff in that Zuckermann paper, e.g.:

    Consider Israeli (a.k.a. ‘Modern Hebrew’) yerı´kho ‘Jericho’, synchronically speaking a derogatory form of יריחו, usually yerikho´, the oldest town known in Eretz Yisrael (Palestine). Israel’s late former Prime Minister, Yitzhak Rabin, pronounced יריחו as yerı´kho once it became clear that the area was going to be ceded to the Palestinians, as though attempting to create revulsion for this land by hinting that ‘it is not biblical [jərˆı¯ho¯ ] but rather alien yerı´kho’. Possibly, the stress was also induced by Arabic أريحا [ʔarri:ha:] ‘Jericho’.

    (Sorry about the transcription problems; I’m too lazy to fix them. To see it properly, go to p. 2 of the paper.)

  155. “small round things that usually come in large numbers”, I suppose that it’s conceivable that the primary meaning might have just been “seed-like things used in commerce.”

    DE, oh, thanks, I was going to ask about it! I don’t know if you are quoting [the usual Africanist line of thinking?] or presenting is your own reasoning, but:


    A potential hypothesis that the root may be *ɣó (“to be small”), referring to the root in Yoruboid languages meaning “seed,” “particle,” or “grain.” See Olukumi íghóró (“seed”), Yoruba wóró (“seed, particle”), Yoruba hóró, thus suggesting that the word ultimately comes from the form “seed,” which was later applied to the form cowry shell as they were introduced in later centuries. It is not clear if seeds/particles (perhaps something like beads) were historically also used as money, though it is likely. However, these roots have not been discovered to be found in other Volta-Niger languages who have this form, like Igbo, Nupe, and Edo.

  156. David Eddyshaw says

    I don’t know if you are quoting [the usual Africanist line of thinking?] or presenting is your own reasoning

    Well, both. To unpack it a bit …

    The degree to which noun-class-based “gender” in Oti-Volta (and in Niger-Congo generally) correlates with meaning is very variable, both between genders and between languages.

    The gender which in POV had the singular suffix *-βʊ and plural *-ɰi contains animals throughout Oti-Volta, and that seems to be the semantic core. In WOV it consists almost exclusively of words for animals on the one hand, and for “small round things” on the other; the only exceptions in e.g. Kusaal are “heart” and “gun”, where the latter is an Arabic loan that has obviously been fitted into this gender by formal analogy (malif, plural mali “gun”, like mɔlif “kob”, plural mɔli.)

    In other Oti-Volta languages a lot of words for seeds belong to this gender, though not as consistently as in WOV (I was wrong in implying above that it’s just a peculiarity of WOV, though.) This is an odd situation; probably two originally distinct genders have fallen together historically, though there is no Oti-Volta-internal support for this hypothesis as far as I can tell.

    Be that as it may, in WOV the semantic correlations for this particular gender are particularly strong, and all words for seeds that I’ve discovered belong to it. As the similarity between cowries and seeds is pretty obvious, it’s not surprising that the word for “cowries” would end up in this gender whether or not the word was a loan or a language-internal neologism (or a bit of both, if it is indeed a case of “phono-semantic matching.”)

  157. @DE, yes, but for Yoruba they suggest meaning “to be small” and cognates are “seeds”. I did not know: what if its a shared knowlege that “West Africans often name money after seeds”.

    I wodner if beads were used this way.

    In some ways that would be natural: beads were used so by Europeans in transactions with various people (though I remember, in an old – like 200 years old – Russian book about exploration of the north this practice was compared not to the similar Western practice but to… Ancient Greek practice. Apparently referring to some Greek source I’m unaware of), and beads are found in thousands by archaeologists in Africa.

    But then I suppose, there must be uniformity in beads (or not? After all even if relative value of two neclaces is not determined because one is green and the other is not, they are still very convenient for a merchant. Cf. hacksilver) and modern words that mean “money, beads”.

  158. David Eddyshaw says

    “West Africans often name money after seeds”

    Well, they don’t in any Oti-Volta language; or in Fulfulde; or in Twi; or in Hausa, or in Yoruba, or in Nupe …

    “Seedlike” (a thing which is surely obvious, whether you are West African or not) is not the same as “seeds.” Kama da Wane ba Wane ba, as they say in one of the languages in which money is not named after seeds …

    Similarly, seeds are generally small, but they are not called “smalls” in any language I know of.

    “Small” was POV *bí-. It’s the same as the stem of *bí-ga “child” (as in Kusaal biig), and goes right back to proto-Volta-Congo *bí-. It’s one of the better candidates for even longer-range cognates, e.g. Fulfulde ɓiɗɗo “child.” Kusaal bil “small” is the same root, with the *-la singular suffix of a “diminutive” gender which is otherwise obsolete in Kusaal.

    As it happens, Kusaal bi’ig “ripen” (as in seeds) looks rather similar, but that is just coincidental: it goes back to POV *bêɰ-, possibly cognate with proto-Bantu *-bégʊ́ “seed” (Swahili mbegu), but possibly not: the POV *-ɰ is probably a fused monactional derivational suffix, because other Oti-Volta languages show forms just reflecting POV *bê-.

    Wiktionary’s *ɣó “to be small” seems to be a figment. The only alleged descendants are the Yoruba and Nupe words for “cowry”, and the similarity there suggests borrowing to me, not cognacy. Borrowing, with West African words for “cowry”, is exceptionally probable, and should really be the first thing to be thought of to explain similar forms like these, not far-fetched etymologies.

    In yer actual Yoruba, “be small” is kéré, and in Nupe, “small” is téténgi.
    “Seed” is irugbin in Yoruba and edzò in Nupe.

  159. David Eddyshaw says

    There is, now I think of it, an example within WOV of unexpected initial /l/ for /n/ in inherited vocabulary, in the Mooré líuulá “bird” beside Farefare níilá, Kusaal niiŋ (plural niimis.)

    This is a somewhat problematic word family, though. The Mooré and Farefare forms must historically have been diminutives meaning “chick” (like English “bird”, for that matter), which is straightforward, but the root vowel varies between /o/ and /i/ even between closely related languages (Mampruli niiŋŋa, Dagbani nooŋa.) Probably this reflects umlaut generalised from a former “animal” gender plural *niimɰi corresponding to singular *noomβʊ.

    The stem-final /m/ seen in these forms is derivational, and in Western Oti-Volta the corresponding underived root-stem form *nooga has ousted the inherited Oti-Volta word for “hen”: Kusaal nua, Mooré núaagà “hen”, beside Nawdm kɔ̀rgá, Moba kólĝ, Mbelime kódíkɛ̀.

  160. @DE, “africans name money after seeds” was humorous (and it was not what I thought Africans may do, it was what Africanists may think in darkness of their halls… unbeknownst to me).

    And the author of Yoruba etymology tried rather hard not to sound confident (which does not mean it was wrong to note one more flaw which she/he did not mention).

    The fact that surprises me is that I see similar words in your comments an in this text (I read them successively and was very surprised)….

    “Seed” is irugbin in Yoruba and edzò in Nupe.

    I’ll quote it again: ‘“seed,” “particle,” or “grain.” See Olukumi íghóró (“seed”), Yoruba wóró (“seed, particle”), Yoruba hóró’ English too has distinct words: “seed” (what you sow, metaphorically also something capable of growing), “grain/corn/kernel”. (Wiktionary says “corn” can mean money in Jamaica)

    See particularly hóró, “small round objects” are mentioned there.

    Actually it does not ruin the etymology: if there are not specific reasons to think that a borrowing into other languages that have similar words is unlikely, then what we have is just Yoruba etymology. Perhaps a correct one (in other words, perhaps w/hóró is the same root), I can’t know.

  161. David Eddyshaw says

    Well, the common thread is “cowries are rather like seeds in some ways”, which is obvously true.

    That’s quite enough to account for the word for “cowries” (wherever it came from) ending up in a noun-class-based “gender” which characteristically contains words for seeds, and there is nothing in any way far-fetched or controversial about it. That’s how “genders” work in Niger-Congo, and the facts regarding this particular Oti-Volta gender are not really in dispute.

    However, deducing from it that a word meaning “cowries” actually derives from a word/root meaning “seeds” is a lot more speculative, especially given that there don’t actually seem to be any examples where that has definitely happened. The “small” aspect is a bit of red herring, I think.

    I’d be interested to know where this “*ɣó ‘to be small'” actually comes from; even though I doubt its relevance in this particular context, the author would scarcely have just made it up.

    Unfortunately I know practically nothing at all about comparative work on Yoruba. On the face of it, it doesn’t seem likely that *ɣó- could be derived from proto-Volta-Congo *bí-, but then Kusaal zu “steal” doesn’t look much like it could be related to Swahili iba “steal”, either. Who knows?



    shows distinctly fewer likely cognates with proto-Oti-Volta than proto-Bantu does. Similarly with


    I sometimes get the impression that Oti-Volta is actually genetically closer to Bantu than either is to “Volta-Niger”, though that may well just be an artefact of the protolanguages actually having usable reconstructions to compare (as opposed to cherry-picking lookalikes from modern languages) and both (probably) being comparatively conservative phonologically.

    On the other hand, I was talking to a Bantoid specialist not long ago who thought that the class-suffixing versus class-prefixing division within Volta-Congo was unlikely to really reflect anything very profound in terms of genetic subgrouping. I think he’s very likely right. (One reason why I’ve been looking at Miyobe lately: it looks to be really quite close to Oti-Volta lexically, but currently, at any rate, has class prefixes and not suffixes. But that may be due to phonologically-driven final syllable loss …)

  162. @DE, yes, I suppose the content of the OV gender (and assignment of money to it) is a solid fact.

    A shared root of íghóró/wóró/hóro and V/ghó/Vwó can be as obvious as Russian rod and urod (R. urod “an ugly person”, Polish uroda “beauty”…), I can’t know because I don’t know Yoruba.

    What was striking is exact repetition of the same words in your and his/her comments.

    Speaking of semantics, I spotted:
    “Jalkunan gɔ̀yì money (slang) / argent (argot) N “gravel”; see wár, nàfòlò

    (J. Heath, Jalkunan-English-French Lexicon) (I guess wár has to do with wárí in Bambara, Jula/)

    Actually there is Russian meloch’ “(coll.) small things, creatures etc.” which also means “change (coins)”. But it is not money yet.

  163. Just realised that languages who borrowed “copper” in the sense “money” (definitions in french look cute: kɔpɔr “argent”), could actually have borrowed “coppers” in the sense “small coins” – or even more specifically “pennies”.

    At first I didn’t think about that, because in others the word for money coincides with “iron”…

  164. PlasticPaddy says

    Re urod, the original meaning was «слабоумный, юродивый». I suppose the semantic path was something like feebleminded > abnormal > ugly.

  165. PP, wow, thanks.
    So these are two different u-‘s, ѫродъ rather than оуродъ!

    (I guess related yurodivuj does not need an introduction)

    I never checked these words, it is just a common subject of Russian jokes about semantic development of cognates in different languages.

  166. I don’t see how to make the vowels work

    I was wondering whether one might better account for the vowels of *nogru (and possibly the *-r-) if one starts from Arabic plural نقود nuqūd ‘cash; coins, change; monies (as for the establishment of a waqf)’.

    I don’t think the umlaut thing is really an obstacle; it’s really only the initial l/n thing.

    Jeffrey Heath (2005) Tondi Songway Kiini (Songhay, Mali): Reference Grammar and TSK– English–French Dictionary has an entry for an apparent reflex of *nogru ‘cowrie’ with *l-. I hope you can see it on Google Books here. Of course, I do not possess the expertise to evaluate this datum at all.

  167. David Eddyshaw says

    Thanks, Xerîb!

    I actually have a copy of the TSK grammar. Yup, it’s all true …

    I think that shows that l/n variation is not uncommon in the region, which is certainly true for Oti-Volta (in fact, non-initial l/n have merged in Gurma and in the Atakora languages other than Waama, but not in e.g. Western Oti-Volta.)

    Obviously TSK itself is not possible as a source for the WOV forms, though, and elsewhere the word seems to have initial /n/ consistently.

    It did occur to me that the fact that the WOV stem was definitely *lagd- and not *lagr- could be explained away fairly plausibly by the fact that proto-WOV did not actually have any stems ending in /r/ after a consonant*, whereas /d/ was and remains an extremely common and productive derivational suffix. So it’s not unreasonable to think that a loan might be remodelled into a more normal stem type.

    * Initial and stem-internal POV *ɹ > proto-WOV *y, and was generally subsequently lost altogether non-initially. Non-initial /r/ was seen only in the singular noun class suffix *-rɪ and as a result of the change of POV *dɹ > proto-WOV *ry.

    Kusaal does have two verbs in CVCr-, viz kabir “call out to ask admittance” and sugur “forgive, be patient with, forbear”, but they are pretty certainly back-formations from kabire, the actual word you call out to ask admittance, and suguru “forbearance”, which must be loanwords in Kusaal (because of the final vowels) and are certainly regionalisms within WOV in any case. The -rV probably represents the reflex of this same singular class suffix from Mooré or Mampruli.

  168. Would be great to see a map of l ↻ n in Africa (and it is not the first time I think about that).

  169. David Eddyshaw says

    The Atakora languages have merged /l/ /d/ and /r/, mostly as [l], in line with their age-old hostility to voiced plosives, but they keep /n/ distinct.

    Quite a few West African language make no l/r distinction; indeed, in all of Oti-Volta only Nawdm, Agolle Kusaal, Mooré and Talni have this distinction, but that is mostly because /r/ and /d/ have merged, so there is no distinct /r/ left to contrast with /l/.

    Proto-Bantu had no /r/ and made no d/l distinction, but proto-Bantu had a very stripped-down consonant system in general. It had a /n/ separate from d/l, though.

    Pre-proto-POV seems to have had no n/l distinction, but that is not typical of West Africa at all, and proto-POV already did have that contrast.

  170. J.W. Brewer says

    I am advised that the basic Ghanaian currency unit, the cedi (etymologically “cowrie”) is subdivided into 100 pesewas, although what with chronic inflation the pesewa may often be worth too little to pay attention to on its own. One online source gives the etymology of pesewa thusly: “In Akan … Pésewabo is a seed of a plant, formerly used as the smallest gold weight.”

  171. David Eddyshaw says

    Yup. Christaller’s dictionary says “a pennyworth of gold dust”, followed by “the dark blue seed of a leguminous plant, used for the smallest gold weight.”

    Good find. drasvi is vindicated.

    I used to be paid my salary (via the bank) to the last pesewa. At the time, a cedi was worth about a penny (and pesewa coins, naturally, did not exist.) In 2007 the ceid was revalued by 10000 …

    In Kusaal yolʋg “sack” (= Hausa jaka) is used for “two hundred cedis” (formerly, “one hundred pounds.”) That would be quite a small sack nowadays.

  172. David Eddyshaw says

    I see that I have unaccountably failed to inform the Hattery that proto-Oti-Volta had two words for “squirrel”, which turn up in e.g. Mooré as, respectively, kɩ́ɩgà “striped ground squirrel, Xerus erythropus” and lòandá “Gambian sun squirrel, Helioscurus gambianus.”

    My duty is done.

  173. Here I must insert that the Burmese word for squirrel is rhañ /ʃḭ̃/, written ရှဉ့်

  174. pέsεwa-bo has turned up before at LH.

  175. David Eddyshaw says

    I’d forgotten about that.

    The -bo component is likely to be ɛbo “stone or seed used for a weight”, I imagine.

    I wonder if it could be related to the POV *bɪ̂(ɰ)- “ripen” and proto-Bantu *-bégʊ́ “seed” that I cited above (inaccurately, in the case of the POV)? The Akan initial /b/ for proto-Volta-Congo *b is right, according to John Stewart’s Potou-Akanic stuff. Rounding of root vowels after labials is very common in Oti-Volta, though I don’t know about proto-Akanic.

    I wonder if the Yoruba owó “cowry” could be roped in after all? Yoruba seems to have initial /w/ for proto-Volta-Congo *b in wọn “them” and “be”; I don’t know enough about Yoruba comparative work to know if those are really good comparisons (though I strongly suspect that nobody else does either.) It looks more plausible on the face of it than this *ɣó ” be small” thing, at any rate.

  176. 1. “erebî قران‎ (qeran: zemawend, dawet, dîlan) – what’s this?

    I think the qeran here is an attempt at rendering Arabic qirān ‘marriage’ (using the standard Kurmanji orthography, with a for /ɑ/ and e for /æ/), perhaps influenced by colloquial Arabic pronunciation in Mesopotamia, and so:

    1. Arabic قران‎ (qirān: ‘wedding’)

    (For the benefit of general LH readers, Arabic قران‎ qirān is more like ‘union; marriage’, not ‘wedding’. My impression is that the term was used historically of the private ceremony attended only by the qadi, witnesses to the marriage, and representatives of the parties involved, but this may have varied over time. Financial negotiations, including the bride-price, would of course have been a part of the process leading up to this conclusion of the marriage contract. In form, qirān is a verbal noun of the verb qārana ‘associate closely with (someone), enter into a partnership with (someone), join forces with (someone), etc.’ (root qrn relating to notions of connecting, joining, putting together).)

    zemawend, ‘wedding (celebration), wedding dance’; this is from Sorani زەماوەند zemawend or its equivalent in Gorani, derived from the Gorani word zamā ‘bridegroom’; cf. Kurmanji zava, Sorani زاوا zawa, Persian داماد dāmād. (Gorani here refers to the literary koine in which lyric songs popular among Kurds throughout the Zagros region were composed. It is based on Hewrami, one of a group of Iranian dialects of the Zagros (Gorani in the larger sense).)

    dawet ‘invitation; convivial occasion; wedding’, from Arabic دعوة daʿwa ‘invitation’.

    dîlan ‘celebration, especially a wedding celebration; dance song (lively song to which people sing and dance); circle of dancers’, the most usual Kurmanji word. I have heard an interesting (folk-?)etymology for this word from a dear friend from Farqîn (Silvan), to whom I must now write… Perhaps more later.

  177. @DE, are these Dagaare nouns related to *lagd ?

    lìbíbìnzíé bank
    líbíbɔ́ quest for money
    lìbíbùgó safe for money
    líbídááná rich person
    líbídì the act of spending money
    líbímá capital
    lìbípɪ̀làá cowry
    lìbírì money
    líbízɔ́ŋ (1) gold dust (2) plentiful money

    P.S. ? in *lagd? reads as a glottal stop:/

  178. Xerîb, shame on me then. Everything is really stupid: I did not know qirān in this meaning (though I see how it can be derived from qārana)! I need to read more about marriages:((((

    Thank you for the explanation (and the background).

    Regarding the rest, I did not mean that anything is necessarily wrong with their derivation from gelin etc. (I said “unexpected”) but kalım is a well known word in Russia, so I was surprised not to see it… and then when kalın was mentioned to explain an entirely different thing, I was surprised again.

  179. Also when I looked these (Kurdish now) words up, I was surprised to see “dance”.

    Apparently dances and weddings are strongly connected in local culture, and that is unsurprising. But in Russia dances and partying were connected (nowadays young people don’t dance) while weddings and partying are distinct. It would not occur to me that a word can mean both.

  180. David Eddyshaw says

    are these Dagaare nouns related to *lagd


    Dagaare is phonologically weird. It’s the French of Western Oti-Volta.

    (The dictionary is nice, but I wouldn’t rely on the grammatical sketch at the beginning at all.)

    Where did you get “*lagd?” from?

    The Kusaal singular is la’af, where the a’a is phonologically a glottalised vowel; the realisation can be as creaky-voice, or with an actual glottal constriction after the first mora.

    AFAIK all other studies of Kusaal apart from mine have taken such sequences as /VʔV/. This is definitely incorrect phonemically: such sequences do not behave as /VCV/ in any context, but exactly like long vowels (there are, for example, tone realisation rules which distinguish the two cases.)

    I’m no Ladefoged, but I’m pretty sure that it’s phonetically wrong too. Quite apart from the possibility of realisation as creakiness, when there is a glottal constriction, it’s not actually a voiceless [ʔ].


    Interested Hatters can find New Testament readings in Kusaal online and see what they think …

    However, in cases like la’af the glottalisation is actually a recent development: the form was *laɣaf in Toende Kusaal as recently as Prost’s 1979 grammar. Prost writes x (for [ɣ]) in cases where the glottalisation has arisen from *g following a short open vowel, as here, and uses ‘ in all other cases: those represent the majority of Kusaal glottal vowels. In present-day Kusaal there is no longer any difference.

    Contrastively glottalised vowels were certainly a feature of proto-WOV, though they are now preserved only in Farefare, Kusaal, Nabit and Talni.

    The internal consonant alternation between sg la’af and pl ligidi is regular: the loss of *g and consequent development of a long glottal vowel does not take place after close vowels.

    There is no doubt at all but that the the proto-WOV stem was *lag(d)-. How Dagaare ended up with a medial /b/ I know not. Sheer wilful perversity …

  181. Where did you get “*lagd?” from?

    I just used it in my question. Then I realised that it looks as a glottal stop and inserted a space.
    Then I thought that the fact that ? in *lagd? reads as a glottal stop is interesting and added a comment.

  182. David Eddyshaw says

    I see.
    Or possibly Ahʔ,

  183. “How Dagaare ended up with a medial /b/ I know not.”


    Am I right at assuming that

    – rì in lìbírì (libiri [lìbírì] n. money sg: libiruu pl: libie 2pl: libiiree) can represent -d of *lagd and belong to the stem.
    – other things after lìbí- are morphemes ( pelaa “white”) that have assimilated our coronal consonant?
    – -f la’af is a suffix?

  184. This time [the French administration] ended up quickly with a much larger number of centime pieces than it had issued.

    Eh? Does this mean that most of them were forgeries?

    That should, of course, have read starmbormen.

    If it were starnbornen, it would be a matter of mis-minim-ization.

  185. David Eddyshaw says

    Am I right at assuming that

    – rì in lìbírì (libiri [lìbírì] n. money sg: libiruu pl: libie 2pl: libiiree) can represent -d of *lagd and belong to the stem.
    – other things after lìbí- are morphemes ( pelaa “white”) that have assimilated our coronal consonant?
    – -f la’af is a suffix?

    The way that nouns work in Oti-Volta is that there are a number of “genders”, each having a distinctive singular and plural suffix pair (or just a singular, in some genders which only contain mass nouns.) The term “gender” is (unfortunately) thoroughly established in the literature: it tends to be particularly inappropriate in Western Oti-Volta because (uniquely for Oti-Volta) the “genders” there no longer command agreement, so the “genders” are really just “declensions.”

    In addition, the bare stem is also an active part of the noun paradigm, because compounding is the normal way of construing a noun with an adjective or demonstrative; thus in Kusaal:

    biig “child”
    biis “children”
    bisʋŋ “good child”
    bisʋma “good children”
    bikaŋa “this child”
    bibamma “those children”

    However it is quite commmon for this combining form, though theoretically just a bare stem, to be remodelled by analogy with the singular (interestingly, this remodelling only ever affects the segmental form, never the tones.)

    The “cowry” word in Kusaal goes

    singular la’af < *lagfɔ (possibly itself < *lagdfɔ)
    plural ligidi < *ligdii (the umlaut, though no longer an active process, appears throughout in the plural of this gender.)

    The expected combining form is la’- < *lag-; this does appear, in the sense "cowry": la’akaŋa “this cowry.” It also occurs for “money.”

    However, in the sense “money” it has often been replaced by lig-, with umlaut from the plural, so you have e.g. ligsa’ʋŋ “extravagance” (“money-destruction.”)

    This is what is going on in the Dagaare too: libi- is the combining form of libiri “money.”
    The actual form of this word in Dagaare is based on the inherited plural form *ligdii; in Dagaare, medial *d *s and *s all fall together as /r/, and the word has been reinterpreted as a singular with a singular suffix derived from POV *ɹɪ. (Dagaare does that sort of thing a lot: the profound phonetic changes it has undergone have caused a lot of remodelling of noun flexion.)

    The -b- for -g- is just irregular: but then this is a language which has made the word for “fire”, Kusaal bugum, into vũu: it is the only Oti-Volta language with v- for the expected b- here. Like I say, wilful perversity. Conceivably the form has been affected by analogy with biri “seed”, but it is not a compound of biri, and the li- part is not an independent morpheme.*

    Libipelaa is indeed etymologically “white cowry” (as opposed to “coin”): this way of specifying that you do actually mean a literal cowry rather than a coin is common in Oti-Volta, cf Moba lígpīēnm̀ “cowry.”

    * A possible analogy is Kusaal lɔmbɔn’ɔg “garden”, which is actually a loan from Hausa lambu, remodelled after Kusaal bɔn’ɔg “swamp, paddyfield”, as if it were “lompaddy.”

  186. @DE, thank you!

    Am I (again) right that
    – *-d belongs to the stem and was not a class suffix
    – but was reinterpreted as such (in the form -r-) in Dagaare and thus simply dropped from combining forms

    (or it would be dropped from combining forms anyway, even if not reinterpreted?)
    (vũu is just bugum with mouth full. But libi<ligi? if it were <*lVgu or < *lVgBV…)

  187. David Eddyshaw says

    As will be clear from my ramblings above (though you could be forgiven for not ploughing through them) I don’t actually know where the d in e.g. Kusaal ligidi “money” comes from.

    If it is actually part of the stem, then its disappearance in the singular would be unsurprising: compare Kusaal wief “horse”, plural widi, combining form wid-, supporting a rule *df -> f.

    Against this is the fact that the d never turns up in combining forms, but that could be explained away as the result of analogy from the singular, a common process.

    Driven to choose, I would actually still plump for the idea that it originated from the d of the plural class suffix *dɪ of *lagdɪ “goods”, reinterpreted as stem-final; that kind of resegmentation is also a known process in Oti-Volta. However, if the word is a (more or less mangled) loan from Songhay, the d was there all along.

    The case in Dagaare is actually rather simpler: whatever the ultimate origin of the word, Dagaare has just reinterpreted the -ri as a singular noun class suffix. (It does that.) Dropping the suffix to make the combining form is then perfectly regular.

    I am absolutely certain that there was no *b in this stem in proto-WOV.
    Having thought about it, I do rather like the idea that *ligiri has become libiri in Dagaare via contamination with biri “seed”, though. It’s a sort of eggcorn.

  188. I do rather like the idea that *ligiri has become libiri in Dagaare via contamination with biri “seed”, though

    As I was scrolling down the dictionary, I stopped each time I saw X-biri “such and such seed” and did not know why I’m doing it. I think that’s because libiri is X-biri.

    (On the other hand we STILL have not established the pan-West-African connection money-seeds. Naming money after seeds was my jokular suggestion about Africanists and their theories:))

  189. David Eddyshaw says

    we STILL have not established the pan-West-African connection money-seeds

    Well, no. That’s because there isn’t one. However, the analogy seed/cowry is pretty obvious, and seems quite often to have ended up reflected in language in one way or another.

    Apart from the Dagaare eggcorn, there’s the gender assignment of “cowry” in Oti-Volta and the Akan “pesewa” thing, for a start. And personally, I like the look of Yoruba owó “cowry” being related to Akan ɛbo “stone or seed used for a weight”, though I must admit that that is extremely speculative. And of course it would have to be a case of parallel development, not straightforward shared inheritance: cowries haven’t been around in West Africa long enough for that.

  190. “That’s because there isn’t one. However, the analogy seed/cowry is pretty striking, and seems pretty often to have ended up reflected in language in one way or another.”

    Then there is a connection:/

  191. David Eddyshaw says

    Sure. Just not “pan.”

    (Incidentally, I just checked the other “literary” dialect of the Dagaare dialect continuum, Dagara: “money” is libie. So it’s not just Dagaare-in-the-narrow-sense.)

  192. the dark blue seed of a leguminous plant, used for the smallest gold weight.” – the etymology is mentioned in Russian but not English WP, they simplified it to “Translated from Akan [it] means ‘dark blue seed of plant'” Sounds very strange. On the other hand, Wiktionary’s definition (‘from Twi Akan pέsεwa (“penny, penny’s worth of gold dust”)‘) reminded me about

    líbízɔ́ŋ (1) gold dust (2) plentiful money
    so I looked up
    zɔŋe [zɔ́ŋ] n. (1) flour (2) pulp 2 pl: zɔnnɛɛ

    Also zombie is plural of “insult, abuse”:)

  193. I don’t actually know where the d in e.g. Kusaal ligidi “money” comes from.

    Yes, I remembered that, I just was not sure if it can possibly be a class suffix.
    This Dagaare dictionary is… tempting.

    dombɛroŋ [dʊ̀mbɛ́rʊ̂ŋ] n. (1) large animal (2) famous animal pl: dombɛrɛ 2pl: dombɛrɛɛ/dombɛrrɛɛ
    dompelaa [dʊ̀mpɪ̀làá] n. (1) white animal (2) an ordinary animal (believed not to have any spiritual powers) pl: dompeɛle 2pl: dompeɛlɛɛ
    donsɔgelaa [dʊ̀nsɔ́gláá] n. (1) black animal (2) an animal believed to have spiritual powers pl: donsɔgelɔ

    Aha, cool, so doN-. But then:
    doŋa [dʊ́ŋá] n. (1) animal (2) mosquito pl: donne 2pl: donnɛɛ Var. doŋɔ
    domɔ[dʊ̀mɔ́] n. totem; the sacred bird, animal, or insect of a clan pl: donne 2pl: domɛɛ Var. domɔ
    donɛɛ [dʊnɛɛ] n. humankind Var. andonɛɛ

    And then you want to know the suffixes used in doŋa and domɔ (and donɛɛ if it belongs here) and why labialisation etc.:

    bone [bʊ́nɪ̀] n. thing pl: boma 2pl: bonnɛɛ
    bone [bʊ́nɪ́] adj.worthless or useless pl: boma

  194. J.W. Brewer says

    Obviously there is an ancient connection (however odd it seems in these days of paper money with no intrinsic value) between units of currency and units of weight that might have been applied to relevant amounts of gold, silver, or other currency-suitable materials. It is not that unusual cross-culturally for the very smallest units of weight to be defined in terms of some number of grains or seeds from some locally-common plant that is thought to have grains or seeds of predictably uniform weight. So to this day a troy ounce (still standardly used to measure gold etc.) weighs 480 grains (20 pennyweights @ 24 grains each), The “grain” underlying traditional Anglo-American weights was eventually standardized at (the internet advises me) 64.79891 mg. For whatever reasons, the avoirdupois pound got standardized at 7,000 grains, which does not factor very well especially given the traditional intermediate subdivisions of a pound, so the avoirdupois ounce is 437 1/2 grains and the avoirdupois dram (1/16 oz.) is 27 11/32 grains. I suppose with a few exceptions (e.g. gunpowder, historically), the sorts of things where measurement to the nearest grain was material tended to be measured by troy weight or the troy-adjacent apothecaries’ weight, where the larger units all equated to integral numbers of grains.

    The carat, used in modern times solely to give weights of diamonds and other precious stones, descends etymologically from the Greek word for “carob seed.” The Anglo-American carat traditionally weighed a bit over 205 mg and a bit over 3 grains (not an integral number of grains), but at some point the Jacobins seized the diamond industry and made uniform a metricated carat of exactly 200 mg.

  195. David Eddyshaw says

    It’s a nice dictionary, not least in that it marks tone.

    Dagaare is not a good starting point for studying WOV flexion: the phonological changes have driven a lot of analogical remodelling, and also caused quite a few originally distinct forms to fall together. (Dagaare also goes all in on vowel harmony, which is pretty limited elsewhere in WOV.)

    For example: dʊ́ŋá “mosquito”, pl donne is cognate with Kusaal duŋ pl dumis “mosquito”; the root is the same as in e.g. Kusaal dum “bite”, an etymon of impressive antiquity, with cognates as far afield as Swahili uma; the m is, without question, original.

    On the other hand dʊnɛɛ “humankind” is ultimately from Arabic dunyā “world.” (This explains the variant andonɛɛ, which is from the Arabic form with the article.)

    I’ve previously mentioned an exceptionally stupid paper I came across that made a big deal of the (true) fact that some Dagaare singular suffixes are identical in form to other plural suffixes, and concocted elaborate semantic theories to “explain” the individual cases. There was no glimmering of understanding that the coincidence in form was the result of completely regular historical sound changes (like -ri for both the original sg *rɪ and pl *dɪ.) It was like inventing a theory that the French Weltanschauung did not have a concept of “several cats” on the basis that “chats” sounds identical to “chat.” Pitiful stuff.

  196. David Eddyshaw says

    I should perhaps explain that the Dagaare pl donne “mosquitoes” is historically (apart from the lax root vowel) identical to Kusaal dumis; both come from proto-WOV *dúmsɪ́. The Kusaal development is not to hard to understand; the Dagaare makes more sense when you know that non-initial /s/ in Dagaare has consistently become /r/, so *ms > *mr > *mn > nn.

  197. yes, I suspected that it is dunyā, though the article again shows strange transformation (or it is from -dd-?)

    @JWB, the initial hypothesis was that they all are prototypically cowries and are grouped with seeds in the class of small and round things….

    grain as a meausure is familiar, but this line is very different from the seed-cowry line.

  198. Stu Clayton says

    the French Weltanschauung did not have a concept of “several cats” on the basis that “chats” sounds identical to “chat.”

    Ah, but it does have such a concept: chateaux. I recently indulged myself with a bottle of chatte au Polignac. No, not that kind of chatte.

  199. J.W. Brewer says

    Cowries may be “small” in an absolute sense, but when compared to grains/seeds as units of weight/value surely the contrast rather than the similarity in their sizes would seem more salient?

  200. David Eddyshaw says

    I agree.

    There is, now I think of it, the Kusaal la’abielif “coin”, which literally is transparently “money-seed.”

    I remembered it as being used for the widow’s mite in the Bible translation, but I remembered wrong: that’s kɔba ayi “twopence”, from Hausa kwabo “penny”; I imagine this could be from English “copper.”

    In the Bible, la’abielif is used for “half-shekel”, which I don’t think was very mite-like.

    Of course, not all seeds are small. Recall the wise proverb

    Ba nyɛɛti mak nɛ gbin ka vʋl kʋkpar.
    they HABITUAL measure FOCUS backside and swallow palmnut
    “One should measure the anus before swallowing the palm nut.”

  201. David Eddyshaw says

    not that kind of chatte

    I should hope not. The tastes are quite incompatible.

  202. January First-of-May says

    and concocted elaborate semantic theories to “explain” the individual cases

    I wonder if there are elaborate semantic theories out there to explain why the Hebrew future tense has identical 2SG.M and 3SG.F forms. Apparently the identical forms go right back to Proto-Semitic – though admittedly it wasn’t a future tense for most of that way…

  203. I wonder if there are elaborate semantic theories out there to explain why the Hebrew future tense has identical 2SG.M and 3SG.F forms.

    I’d be surprised if there are. Formal syncretism is common everywhere.

  204. David Eddyshaw says

    A case in Hebrew that is a bit analogous (I think) is the Biblical Hebrew vav-consecutive, where I’ve encountered some quite elaborate semantically-orientated “explanations” as to why an “imperfective” should have a “perfective” meaning in this context. (To be fair, most of the more outré seem to predate the realisation that the forms actually continue an old prefix-conjugation preterite, still alive and well in Akkadian.)

    Most of the accounts I’ve seen of the fact that Hausa uses its so-called “relative perfective” as the normal main-clause past in narrative, even accounts by sophisticated linguists, tie themselves in knots trying to make up semantic explanations of the phenomenon.

    The Chadic language Miya uses its subjunctive as the usual narrative past. However, the admirable Russell Schuh was far too sensible to fall into the trap of trying to “explain” it semantically in his grammar.

    I think it’s partly the ever-popular confusion of form and function that has led many a young grammar-writer on the path to sin, and partly a failure to appreciate that it’s not at all rare for formerly distinct forms to end up looking identical from sheer historical wear and tear.

  205. Dang. I realize I should have called it “common or garden syncretism”.

  206. Stu Clayton says

    “One should measure the anus before swallowing the palm nut.”

    “Look before you leap” pales in comparison. Unfortunately I do not live in a land where the palm tree grows in abundance. Many English expressions have lost their vividness as experience is lost. “squeal like a stuck pig”, “stubborn as a mule” and so on.

  207. David Eddyshaw says

    Astute Hatters (i.e. all of you) will of course have noticed that kʋkpar “palm nut” does not end in -f, and wondered how this can be squared with my assertion that all names of seeds belong to the fɔ/ii “gender” in Kusaal.

    As I can understand that you will all have been afraid to ask, I shall reveal the mystery.
    “Palm nut” (though this is the accepted name in English) is, properly speaking, a fruit, and the superior botanical skills of the Kusaasi are here evident in the very grammar of the Kusaal language itself.

    The proverb remains valid. Oh, dear me, yes.

  208. PlasticPaddy says

    Can the Kusaal word for “swallow” also mean “insert”? (Asking for a friend…)

  209. David Eddyshaw says

    No. Though the question is a natural one. Very natural.


  210. Use the right tool for the job!

  211. David Eddyshaw says

    In this case, a palmnut.

  212. Individial explanations can be silly, but I’m not hostile to the genre as such.

    Most of etymologies published in journals have competing theories, which means, majority of suggestions must be wrong. And?
    Also physics: beyond a certain limit it is also speculative (but extremely complex).
    I don’t think speculations have no place in science.

  213. @drasvi: Speculation in science is fine, so long as it is reasonably well informed. It’s not useful to promote speculative ideas that can be straightforwardly shown to be inconsistent with current knowledge (unless there is some reason for thinking that “current knowledge” might be seriously in error). Penrose’s idea that brains are quantum computers, with the cubits represented by conformational states of the tubulin proteins making the microtubules running along the axons of nerve cells. It’s very easy to show that there is no way to maintain the required degree of coherence for such a system, immersed as it is in a bath of other sludge.

  214. @Brett, I simply skipped that part of the book:/
    I found it not very interesting.

    Most of the book is like: quantum mechanics for housewives, Gödel’s theorem for housewives etc.
    But unlike popular books about physics, this one is personal – so I read it as a book about Penrose (how a mathematician feels about various things) and found it very interesting.

    Back to his suggestion, I understood it as an attempt to make a testable prediction just to make some testable prediction – with full realisation that it is unlikely to be right (one should be an idiot to think that it is likely to be correct).

    As for impossibility of its correctness, likely Penrose has arguments against it (against everyone’s arguments for impossibility), but I don’t know them (perhaps you can see that they are wrong).

  215. Generally, my argument for why it is not impossible that some quantum effects play some role in thinking is that it should NOT be “a quantum computer but in the brain”.

    People say: “No, consicousness has NOTHING to do with quantum mechanics”.
    This is… idiotic.
    They mean “it can’t be a quantium computer” but they offer a crazy exaggerated formulation.
    Everything around us has something to do with quantum mechanics.
    It simply describes our building material.

  216. Perpaps one could mean: “yes, but everything around us has to do with boring quantum mechanics, and does not involve any ‘funny’ quantum mechanics”. But one can’t discuss ALL ‘funny’ effects which are possible.

  217. David Marjanović says

    I never checked these words, it is just a common subject of Russian jokes about semantic development of cognates in different languages.

    As an alternative, I suggest the exact cognates ужасный “horrible” ~ Czech úžasný “wonderful”.

    (I think this is about the noun varying between “dream in general” and “nightmare”. Though, actually, if the Czech meaning is mere sarcasm, that would explain where Wiener Schmäh comes from.)

    I’m no Ladefoged, but I’m pretty sure that it’s phonetically wrong too. Quite apart from the possibility of realisation as creakiness, when there is a glottal constriction, it’s not actually a voiceless [ʔ].

    Like this?

    (Preaspirate this valiant attempt at a “voiced glottal approximant”, or tap rather, and you get my /h/.)

    likely Penrose has arguments against it

    In his book he didn’t mention any. He seemed completely unaware of the issue. It is statistically impossible to maintain a superposition of quantum states inside a microtubulus because those things are way too warm and, relatively, too large – there’s even water inside them.

  218. I’m not ready to defend this idea. The main source of disagreement between us is that people are interested in the part which I skipped!

    Usually I find popular books boring or authors enthusiastically offer analogies and simplifications which, as I think, obscure the matter. It is easier for me to understand a less popular book (even if I know nothing about the subject and it takes a lot of time).

    In his book Penrose offers a similarly simple explanation of several basic theories, but not with the purpose of enlightening the reader or promoting science. Just to let her understand why he feels what he feels about these theories and mathematical ideas and consciusness. And I found this text interesting. I don’t know if this text would work in terms of “enlightening” people, that is if DE would be able to learn something about thermodinamics from it or not. I already know the material he is explaining. But It is interesting to learn what a very smart physicist and mathematician (his contribution in science is sufficient to say “smart”) feels and thinks about what. I love his book, but what I found interesting is his subjective view.

    Now coherence etc.

    As I said, I think he wrote that not because he believed “it works”, but rather to move the conversation to a different level, where people propose

    The problem is that I’m not a specialist. It will likely take less time for me to understand the math in quantum mechanics than for, say, DE (but more time than for actual mathematicians here).
    So what exactly I can do when specialists disagree?
    I can invest some time and form my own opinion – and the idea is great actually, it’s a beautiful idea to understand quantum mechanics better!

    But I can say, that generally I do not accept “intuitive” arguments for impossibility of something, and everything but a mathematical proof is “intuitive” for me.

    I can intuitively deside that a certain direction of search is “not interesting” (that’s I don’t feel an urge to search in this direction). But when several clever people call another clever person an idiot and I know that he is much smarter than I am (and most of specialists, actually) I simply can’t tell if he has a point.

  219. It is VERY difficult to prove “impossibility”. There are usually too many ways to do a thing. Usually when people say “impossible” they mean: “those of these many ways that I know about won’t work”.

  220. Imagine a world where people (1) know quantum mechanics at least as well as we knew it in its early days (2) do not know superconductivity.

    Superconductivity was first observed, then explained.
    As far as I know, the theory of [low-temperature] superconductivy is not the sort of a theory which could appear before it was observed. I heard, it involves suspicious mathematical simplifications (for which it is not obvious that we have the right to do that). Contribution of something is declared neglible (as it happens in physics) when one can’t be sure if it’s neglible or not – but the resulting formula does work… Just not for high-temperature conductivity. So when HTC was discovered, that was a surpise.

    Well, neglecting and simplifying without providing a proof is what physics always does:) I just heard that this theory is somewhat more suspicious..

    So what Brett’ would say about superconductivity in such a world?

    IF, based on his immense (in that world) experience with conductors would say: “no, that’s not how metals work, you can’t ‘make’ it conduct without losing some energy, man you can’t do anything without losing some energy” (and would be mistaken) and IF his argument above (no way to maintain the required level of coherence) is of the same nature – then I simply can’t tell if a specialist who disagrees with Brett has a point (sees something Brett is missing rather than misses something that Brett sees).

    I can follow this specialist’s argument, and if I find actual errors there, then there is a chance that he’s an idiot. But even then, there are people who can intuitvely arrive to a good idea, but can’t express it mathematically and offer erroneous arguments in support of a good idea.

    It is impossible in principle to prove that one’s idea is poorly motivated. It is possible to show that arguments on paper are erroneous, but the actual motivation is in your head.

    What I do is just: I determine for myself what is interesting and what is not, and abstain from determining who is an idiot and who isn’t, especially when both are smarter than me:)

    This is what I do in general.
    But I haven’t spent a second thinking of how one can make a quantum computer work.

  221. “Usually I find popular books boring” – of course it is different, say, with Martin Gardner’s books. But those are books with problems

  222. January First-of-May says

    Perhaps one could mean: “yes, but everything around us has to do with boring quantum mechanics, and does not involve any ‘funny’ quantum mechanics”. But one can’t discuss ALL ‘funny’ effects which are possible.

    Previously on LH: “Apparently there’s a newfangled “CEMI theory” of consciousness that says it’s all about the EM field in the brain or something, and that we aren’t finding it in computers because computer EM fields work completely differently.”

    I guess that is one semi-plausible way of getting ‘funny’ quantum mechanic effects in your macroscopic system…
    (see also: On the Origin of Circuits, whereas an even more blatantly macroscopic programmable circuit ended up displaying supposed-to-be-impossible pattern recognition powers apparently due to weird EM field interactions)

  223. A silly way to define “boring” is “any effect that can be obtained from models that pre-date q.m.” Then of course discussing the space of everything that exists, but can’t be obtained from these models would be silly. Also without q.m. the universe would not exist, chemistry would not work and, I don’t think that the fact that we exist is “boring” no matter how you model it.

    A more meaningful way to define “funny” would be specifying the exact (quantum) behaviour would be funny to find in, say, a human body. Say, a certain way of calculating probabilities. I can’t readily do that.

  224. ужасный “horrible” ~ Czech úžasný “wonderful”.


  225. “West Africans often name money after seeds”

    This reminds me of Greek δανάκη, representing an Iranian *dānakā̆- ‘eighth of a Babylonian šiqlu/shekel, etc.’; cf. Middle Persian dānag ‘seed, grain’ (borrowed into Arabic as dānaq, a sixth of a dirham) and Classical Armenian դանգ dang, translating ἀσσάριον ‘a coin of small value’ (Matthew 10:29; Luke 12:6). This Iranian form is cognate with Vedic plural dhānā́s ‘roasted grain’, Lithuanian duona ‘bread’, Latvian dona ‘heel of a loaf of bread’, Tocharian B tāno ‘seed, grain’. The Iranian word is just one possibility (out of a few that have been proposed over the years!) for the family of Turkic words for various coins and monetary units that includes Tatar and Bashkir тәңкә, Kazakh теңге, Turkmen teňňä, Uzbek tanga, etc., and of course the loanword seen in Russian деньга, деньги. I won’t venture an opinion of the Turkic family of words. Y’all can check деньги in Anikin (page 207) for yourselves.

    The kissi penny reminds me typologically of the early Greek ὀβολός (see the top of pages 278–279 here).

  226. David Eddyshaw says

    A large part of the motivation for the ludicrous microtubule-consciousness thing seems to be a desire to have some sort of physical “explanation” for free will.

    Obviously, to a good Calvinist like myself, this is a solution in search of a problem, but it seems to me to miss the point pretty spectacularly even for the Arminian materialists amongst us.

    Assuming that mental processes are purely the result of physical processes of interaction between billions of neurons, each if which even individually operates in a way far more complex than any electronic gate, even if the underlying physics were wholly deterministic in principle*, the potential complexity of the resulting emergent behaviour would be far, far too great to be predictable by any conceivable constructable mechanism.

    Hell, even LLMs are not “predictable” in this sense, and they aren’t even intellligent. (The black-box impenetrability of such systems is a big issue when it comes to things like legal accountability, and the problems involved are, realisiically, insoluble. They are computationally completely intractable.)

    So in such a system, whether one’s will is “free” is such an abstract concept that it has no conceivable practical consequences anyway, and the issue is best left to metaphysicians, philosophers and theologians. Nobody else will see the point – and quite right too.

    * Which it isn’t, of course. But even if it were
    There is no need to bring in any Tao-of-Physics woo into this at all.

  227. @DE, you’re overcomplicating your opponents’ thinking.

    Either you have a reason to believe that consciousness can arise without “funny” q.m. or you do not have such a reason.
    If you do not (and I do not), then you explore this possibility.

  228. ѹжасъ is used in Mark 5:42

    καὶ ἐξέστησαν εὐθὺς ἐκστάσει μεγάλῃ .
    kai exestēsan euthys ekstasei megalē
    “And they were astonished with a great astonishment.”

  229. And I just discovered something Horrible.

    I thought that uzhast’ is a colloquialism, but it turns out it is Slavonic:/

    I remember, I was suprised that Vysotsky sings uzhasnyj instead of uzhastnyj( а тем временем зверюга ужасный коих ел, а коих в лес волочил) even though everywhere where he could he used archaisms, jokular forms and forms which educated speakers use when they want to sound like less educated people.
    E.g. princessa is of course prYncessa, (jokular “princess”, often when saying with disapproval that someone wants to be treated so: queen of Sheba).
    When I began helping a freind of mine with Russian, I wanted to share some songs with lirics and interlinear and asked my female Russian freinds (because this freind was female) what Vysotsky’s song they like most of all. Both named this song.
    It’s good indeed.

  230. I think this is about the noun varying between “dream in general” and “nightmare”.

    @DM, I’m unaware of the meaning “nightmare” (in Russian it is simply кошмар)….

  231. ужасный “horrible” ~ úžasný “wonderful”

    To make it appear less strange, one could translate both with English half cognates:

    ужасный “awful”
    úžasný “awesome”

    Feeling awe is apparently an emotion that is difficult to classify as good or bad.

  232. Hurrah, it’s working agian.

    “How and when did Cowrie money spread to West Africa?”

    @Trond, it seems people (I mean European people who write articles) mostly associate cowries with slave trade and then colonial times. There are medieval Arab accounts of their use as money (I don’t remember, maybe still not West Africa, just Sahel).

    Archaeology is ambigous first because they aren’t stones and their absence may mean they just were destroyed, second (this is my own thought) because how do you tell their use as a good from their use as money?

    Cowries are mentioned here (https://journals.openedition.org/afriques/1145, Initial Encounters: Seeking traces of ancient trade connections between West Africa and the wider world), “It has long been known that glass beads, copper alloys and cowries were important trade goods brought into West Africa during the Arab-driven trans-Saharan trade. However, there is now archaeological evidence showing that the same kind of artefacts started arriving in the region prior to the onset of that organised commerce.” but I don’t know in what article they date specifically cowries.

    She is discussing this site (Burkina Faso, Niger bassin)

    And there is also a later medieval caravanload of cowries in Mauretania (Ma’adin Ijafen).

  233. “this site”
    The nearest tributary must be Gorouol

  234. David Eddyshaw says

    Either you have a reason to believe that consciousness can arise without “funny” q.m. or you do not have such a reason

    There is no such dichotomy. In principle, there could be evidence of variable degrees of plausibility for either or both, or for something else entirely, such as consciousness being ontologically prior to physical reality.

    Whether the physics you deploy is classical or quantum-mechanical has no actual bearing on the fundamental philosophical issues involved: it doesn’t help you bridge the gap between an objective third-person account of the basis of consciousness and the first-person subjective experience of consciousness.

    The other motivation for the microtubule-consciousness thing, besides the notion that it might “explain free will”, seems to be the idea that what causes a wave-function to collapse has necessarily to be an observation by a conscious being. But there is no reason to think this at all. It’s a confusion resulting from overinterpretation of the poorly-chosen technical term “observer”:


  235. Well, my position with respect to quantum physics is this:

    1. I need less time than you to understand the math (does not mean I remember it at the moment or that all the math will be “easy”)
    2. I do not know what the fuck (sic!) “observer” means.
    3. I’m not a physicist (though I must admit there is a plenty fo them who don’t understand the math).
    So I’m mostly rather ignorant.

    So may them physicists talk about observers-[x]uyervers. It is a bit impractical to start from interpretations in my case. Eventually the word appears here not because physicists were inarticulate, they actually meant ourselves, those who carry out experiments.

  236. (Don’t miss Xerîb’s comment of October 15 at 7:19 pm, which had been languishing in moderation until I freed it just now.)

  237. David Eddyshaw says

    Seems like the essential seediness of money is an insight by no means limited to West Africa.

    Belatedly, I though of this, too:


    More shell money.

  238. And compare coins in (women’s) necklaces all over the world.

    I don’t know how exactly their work: how their value as a thing that can be sold (as coins – were coins with holes or attached ears still recognised? – as pendants reusbale for new neclaces, as metal – were they always made from valuable metals? – or as a whole thing as a necklace) interacted with their value as a thing that can’t be sold (but is pretty and makes sounds).

    Which of two mattered for women, was the former value much less than the latter, how easiy they were to sell (liguidity) and how willingly they were sold etc.

    Say, we very willingly “sell’ our money.
    We less willingly sell collectibles.
    We do not sell jewelry, but we can do it, and if the material is expensive, it is comforting that it can be sold.

  239. David Eddyshaw says

    Liguidity indeed (one of the few Oti-Volta loanwords in English: “ability to function as ligidi.”)

  240. PS *grivьna:
    1. necklace 2. monetary unit

    (grivna/hryvnia, modern Ukrainian money)

  241. Another perplexing word I just remembered monisto (by which I don’t mean Finnish word for “manifold” with a collective suffix sto…).

    It looks like a loan (what on earth is -sto?!), but it is old (it is even in Polabian). Presently Russians strongly associate it with moneta “coin”, so usually we imagine large neclaces made of coins, especially Gypsy.

    Or maybe worse, it seems now they imagine fake monistos (like what they sell in Egypt to female tourists for belly dance): https://arthobbycenter.ru/brand/monisto/

    But in dialects it means lots of things like (in pl.) “corals”, “garnets’, “(neck) chain”, “neckcloth”, “bracelet” (“arm/hand decoration” which I take to mean an article of jewelry worn around wrists but distinct from simple bracelets), “slack pockets”, “beluga jawbone” (beluga – a kind of sturgeon) and “bunch (of leaves or other things stringed together)” – but mostly “necklace”.
    These meanings don’t make one thing of coins.
    “A band of fabric with embroidery and pearl lace and decorated with buttons” does.

  242. January First-of-May says

    1. necklace 2. monetary unit

    …from грива, originally “neck”, which at least in Russian had long ago specialized as “mane”.

    As for монисто, I agree with drasvi that the association with монета is strong and I imagine jewelry made of coins or coin-like objects (probably a belt or even a dress rather than a necklace), and if I ever knew it wasn’t a loan I’ve forgotten that.
    (I asked my mom what she thinks it is, and she described it very similarly, except she said it’s either a necklace or a bracelet.)

  243. Very interesting Sprachgefühl stuff!

    Also, it turns out монисто is cognate with the maṇi of Om mani padme hum.

  244. @J1M, yes, initially I imagined a thing made of numerous coins linked together.

    Such a thing could be worn on the neck or on the waist or shoulders or even hanging from one’s headgear, but I did not think about wrists.
    And it were real coins.

    Then I saw pictures of women with such necklaces covering their chest, so now I think about a necklace.
    Also there appeared imitations.

  245. January First-of-May says

    And it were real coins.

    Apparently the 17th century “fishscale” silver kopeks were very popular for this in Mordovia, and after they were discontinued in the 1710s, the locals started making imitation kopeks specifically for jewelry. Reportedly some of those actually had legends in Mordvin languages…

    EDIT: turns out those were made since long before the 1710s; English Wikipedia has a weirdly slanted summary but there are probably better sources.

  246. It may be weirdly slanted but it seems to have been written by a Russian speaker — at least, it’s definitely not by a native English speaker.

  247. arafatka “kufiyyah” is not Yasser Arafat’s wife…

Speak Your Mind