I just discovered a very interesting etymology that has apparently been developed only recently. The OED says the origin of calypso (the name of an Afro-Caribbean style of satirical song) is unknown; Webster’s Third New International (1961) says “probably after Calypso, island nymph.” But Richard Allsopp’s superb Dictionary of Caribbean English Usage
[ < Efik ka isu ‘go on!’, also KID Ibibio kaa iso ‘continue, go on’, a common phrase used in urging sb on or in backing a contestant. The Efik-Ibibio being the established middlemen in the slave trade (ex at Calabar) the slaves of other ethnic groups would have brought this item (as they did BAKRA) to the CarA as part of the private vocabulary of slave life. In the context of CarA plantation-life, crowds backed creole teasing-songs against MASSA shouting ‘Ka iso!’ wh gradually lost its original meaning. Kaiso is still the regular ECar folk name, not calypso. The phonological development /ka-iso > kariso > kaliso/ is attested by KARISO (Dmca, etc), KARUSO (USVI), and KALISO (StLu) this last recognized as ‘another form ‘Calisseaux’ … in use at the same time as ‘Carisseaux’ ‘ —(Espinet & Pitts, Land of the Calypso, 1944, p. 47). The development > calypso is through corruption (through folk etym) by English writers in the 1930s, influenced by the name (Calypso) of the amorous island nymph of Greek mythology, plus an anglicized shift in pitch pattern /1’12/ > /1’21/]
Most of the abbreviations should be self-explanatory, but CarA is “Caribbean area” and KID in the first line is Elaine Kaufman’s Ibibio Dictionary (Leiden, 1985). The key to the etymology is the recognition that the original form is kaiso; I love the fact that the transmogrification to the highfalutin “calypso” is called, quite properly, folk etymology—the ignorant “folk” aren’t always poor and unlettered! And the careful discussion of how the African form would have reached the Caribbean and been preserved should be a model for such things (there are far too many silly African pseudo-etymologies floating around out there).
The entry in Merriam-Webster’s Collegiate says “Trinidad English, alteration of kaiso, perhaps ultimately of Afr origin”; the caution is probably appropriate for a general-usage dictionary, but personally, I’m provisionally convinced by the Efik-Ibibio etymology.
Nice to see an African slave etymology that’s actually backed up with some evidence.
I just sent this to oed3 at oup dot com.
Kaufman’s dictionary was first published in 1972 and reprinted unchanged in 1985.
The relevant sublemma (including tone marks — this is a pet peeve of mine, but these really should be regarded just as essential as consonants and vowels in most African languages) is as follows:
kàá ísó ‘go forward, continue; go on (take place, occur)’
And the form is a phrasal verb based on kàá ‘go to, go for’ (Kaufman 1972:216)
The tone marks (L-H HH) provide evidence for the anglicized shift in pitch pattern. Assuming that the tone hasn’t changed substantially in Ibibio, that is.
It is indeed refreshing to see such a detailed and carefully argued etymology.
Thanks for the info, mark; I’ve emended my Allsopp accordingly.
I find the Allsopp pretty convincing — and not only because it’s a lifetime project, I know Allsopp was working on it at Cave Hill when I was an undergraduate at the sister campus at Mona in the 70s. Literary references to ‘cariso’ and ‘caliso’ go back quite a way, and Trinidadians to this day say ‘kaiso’ rather than ‘calypso’ which is a term used only in formal English and by foreigners. That argues, by itself, against the OED etymology.
That argues, by itself, against the OED etymology.
Eleven years later: you mean the Webster’s Third etymology.
The OED revised the entry in 2021; the etymology now reads:
And the etymology for kaiso now reads:
I guess JC’s e-mail got through!
Elaine Kaufman’s Ibibio dictionary (which has kàá ísó “go forward, continue, carry on”, as mentioned by @mark in 2008) can be found online, as also can her very good Ibibio grammar (her 1968 PhD thesis)
The former shows some nice Volta-Congo etyma, like dóm “bite” (proto-Oti-Volta *dʊ̀m-, proto-Bantu *dʊ́m-; Kusaal dum, Swahili uma.)
Ibibio íkít “tortoise” looks like a reflex of that damn Volta-Congo invasive tortoise yet again (proto-Oti-Volta *kúd-dɪ́, proto-Bantu *-kʊ́dù.) Beastie gets in everywhere.
Proto-Oti-Volta *kúd-dɪ́, proto-Bantu *-kʊ́dù
The group of Mandinka kuto, Malinke kuta ‘tortoise’, and cognate forms elsewhere of Manding is usually cited as the etymon for regional American English cooter ‘turtle’. What would account for the resemblance beween the Manding forms and the Volta-Congo forms like Mooré kùrí? Borrowing into Manding? Areal word of the entire Niger basin? Sheer coincidence? Or even inheritance from a common ancestor? But I gather that there has it has never been established that Mande is even part of any putative larger Niger-Congo, whatever else one thinks might be included there—and the forms for ‘tortoise’ look very alike considering the many millennia that must be involved.
As an aside, the old (old!) entry in the OED has a rather odd etymology for cooter, associating it with a verb coot ‘(of tortoises) to copulate’, origin unknown, here. (According to Green, slang cooter is quite recent, so much so that it is hard to imagine a direct connection to coot ‘(of tortoises) to copulate’.)
The first cite in the OED for coot ‘(of tortoises) to copulate’ from Henry Stubbe (1667) ‘Observations Made by a Curious and Learned Person, sailing from England, to the Caribe-Islands’, Philosophical transactions of the Royal Society of London, vol. 2, p. 499 here:
Initially, I would have thought a typesetter’s error for coit, but the verb is used twice… and then Dampier repeats it serveral times in A New Voyage Round the World, as here, vol. ii, part iii, chapter 1, ‘A discourse of Winds, Breezes, Storms, Tides and Currents’, page 5 here:
Also Griffith Hughes (1750) The natural history of Barbados. In ten books, p. 309 here:
There coition is right beside cooting. Are Dampier and Hughes both reliant upon Stubbe?
Dutch kut?
What would account for the resemblance beween the Manding forms and the Volta-Congo forms like Mooré kùrí?
Actually, the r in Mooré kùrí belongs to the noun-class suffix, not the stem (cf plural kuya.) The word is the regular reflex of proto-Oti-Volta *kúd-dɪ́, but root-final POV *d in Western Oti-Volta became /j/, and then mostly disappeared altogether.
I’ve wondered about the suspiciously universal Volta-Congo “tortoise” elsewhere. While shared descent is certainly involved within V-C branches (like Oti-Volta or Bantu) it just seems such an unlikely word to have been conserved all the way back from proto-Volta-Congo when not even “man” or “woman” were.
The truly reconstructable pV-C vocabulary is much smaller than the numbers that Greenbergian lumpers arrive at by happily counting lookalikes and not bothering with regular correspondences. Setting aside pronouns and lower numbers, my own count is only around thirty; though further research will undoubtedly add to this, I doubt that it will add to the number by an order of magnitude.
The evidence that Mande is related to Volta-Congo is just not there; these days that seems to be the consensus even among people much less splittery than me. So I think one really has to conclude that the resemblance to Manding kòrókàrá etc cannot be due to common inheritance.
Borrowing is always a possibility in polyglot West Africa, and most of the area really is demonstrably a Sprachbund:
https://www.researchgate.net/publication/300471822_Proto-Bantu_and_Proto-Niger-Congo_Macro-areal_Typology_and_Linguistic_Reconstruction
However, my own best guess is that the “tortoise” word is phonaesthetic, along the lines of “round creature.” I think there is a lot more phonaesthetic stuff going on in West Africa than is generally realised, and many attempts at comparative work have been undermined by failure to realise this (one of the supposed cognates between Mande and V-C is *pu/*fu- “blow”, to take a particularly egregious example.)
This would be hardly surprising in view of the prominent role that ideophones play in so many West African languages. In Kusaal, many words that correspond to adjectives in SAE are ideophones, not adjectives (and some of them are definitely borrowed.)
And in fact, I ejected “round” from my Swadesh lists when doing (primitive) lexicostatistics on Oti-Volta, for the very reason that language after language has a form like kpil/kpul/gbil/gbul, but it is impossible to set up regular sound correspondences between them (for example, the regular correspondence of WOV/Buli/Nawdm root-final l is with n in Gurma and non-Waama Eastern Oti-Volta, not l.)
I’ve wondered about the suspiciously universal Volta-Congo “tortoise” elsewhere.
You took it back to proto-World here.
PIE *kʷel-…
Indeed.
Sound symbolism is difficult to study in a systematic and rigorous way, and I suspect this leads to it being generally underrated as linguistic feature. It’s not helped by the fact that ideophones still get treated in grammars as a somewhat quirky marginal feature, a sort of optional extra, even in languages where they are very common.
We’ve discussed this before in relation to some papers that it was easy to pick holes in. But it does really seem to be the case that even with ordinary vocabulary sound symbolism seems detectable: for example, cross-linguistically, words for “tongue” seem much more likely than chance to begin with an alveolar or dental.
@de
Maybe you explained this before, but how do such forms survive regular sound change, e.g., in case of the tongue word, Latin lingua (no dental) from original form with dental? If you are saying they are resistant, this would show, not always. If you are saying, they are reinvented with dental, Romance would not have seen to have done this (yet) in this case.
Lingua starts with an alveolar … [and the change from dingua was not regular.]
https://en.wikipedia.org/wiki/Voiced_dental,_alveolar_and_postalveolar_lateral_approximants
I did say alveolar or dental.
But I certainly wouldn’t maintain that either there were no exceptions to my generalisation (there are a great many), or that historical developments might not turn an alveolar or dental into a consonant at another position of articulation (athough that actually seems quite uncommon in practice with alveolars/dentals.)
Hausa harshe “tongue” is actually from Afroasiatic *lVs-, but has acquired the “body part” prefix ha- at some point. So sure, such things happen.
Oti-Volta has Gulimancema lánbū, Byali dyān̄fə̄, Ditammari (fānán-)dīānfà, Nateni dēn̄fá, Mbelime dɛ̄ndè, POV *lém-fʊ́, cognate with Miyobe tì‐lénbí (where ti- is a class prefix); looks as if it’s related to proto-Bantu *-dímè, but the vowels don’t match properly.
But Oti-Volta also has Nõotre cèrìfò, Kusaal zilim, Mooré zɩ̀lemdé, Buli gīnggēlūng, Konni jɪ̀lʊ́ꜜkʊ́, Yom jèrēŋà, Nawdm gélmgá, Waama déránfā …
A Bantuist I was corresponding with suggested that the initial CV- of these forms might be a fossilised noun-class prefix, but I think that’s extremely unlikely – there are no parallels for that in Oti-Volta.
The “tortoise” word turns up in Mbelime as hūdìkɛ̀, where the h is basically just a placeholder consonant for a completely lost original initial; so if this really was a phonaesthetic word in the first place, that’s certainly a case where regular historical sound changes have obscured the picture (as with the loss of the original root-final *d in Western Oti-Volta.)
A clearer example: Kusaal bìlìm “roll” beside Gulimancema bílíní “roll.” Here, correspondence is quite regular, except for one thing: the initial consonants, vowels and tones all match regularly, and the Gurma n for POV non-initial *m is regular too.
But many good series show that Kusaal non-initial l corresponds to Gulimancema n, not l (and that *l is the original consonant in such cases), so either the forms are independent phonaesthetic creations, or Gurma has resisted the change of non-initial POV *l > n because the word is phonaesthetic.
I’m not sure how one could actually tell these scenarios apart.
Significantly (I expect), forms resembling bilim for “roll” are very common in the area, including in languages much less closely related to Kusaal than Gulimancema is.
David Eddyshaw: A clearer example: Kusaal bìlìm “roll” beside Gulimancema bílíní “roll.”
Slavi-Congo!
Nah, it’s just Scandi-Congo. The culinary term originated in Kievan Rus.
Not that it matters, but most “dentals” worldwide are laminal alveolars: the tongue is held flat and articulates against the alveolar ridge. The alveolars of English and northernmost German, among not many others, are apical: the tip of the tongue articulates against the alveolar ridge (except /s z/, which are usually laminal – but they are apical in Dutch where there’s no /ʃ/ to be confused with). Apical dentals seem to be a thing in India, because in the Dravidian languages they contrast with apical alveolars, but very rare elsewhere.