Slavs and Slaves.

Victor Mair at the Log has a very useful roundup of the connection between the words slave and Slav; I’ll quote the section on Proto-Slavic slověninъ, from Wiktionary, and send you to the link for those on English slave and Ancient Greek Σκλάβος:

Roman Jakobson insists on this etymology: from *slovo (“word”); with link to Old East Slavic кличане (kličane, “hunters, who raise game by shout”) : кличь (kličʹ), and also on the opposition *slověne vs. *němьci.

Trubachev (Трубачёв): Jakobson’s etymology is promising, with the verb *slovǫ, *sluti (“to speak (understandably)”).

Vasmer: it has nothing to do with *slava (“glory, fame”) which influenced it in terms of folk etymology later. *slověne can’t be formed from *slovo because *-ěninъ, *-aninъ only occurs in derivations from place names, however local name *Slovy is not attested. Most likely it’s derived from a hydronym.

Compare Old East Slavic Словутичь (Slovutičĭ) ― Dnepr epithet, Russian Слуя (Sluja) ― affluent of Вазуза (Vazuza), Polish river names Sława, Sławica, Serbo-Croatian Славница and others which brings together with Ancient Greek κλύζω (klúzō, “I lave”), κλύζωει (klúzōei) · πλημμυρεῖ (plēmmureî), ῥέει (rhéei), βρύει (brúei), κλύδων (klúdōn, “surf”), Latin cluō (“I clean”), cloāca (“sewer pipe”). Other etymologies are less likely.

Otrębski brings up an interesting parallel ― the Lithuanian village name Šlavė́nai on river Šlavė̃ which is identical to Proto-Slavic slověne.

Бернштейн repeats this etymology: from Proto-Indo-European *slawos (“people, nation, folk”).

Maher agrees with Trubachev’s connection of it to *sluti (“to be known”), on the grounds that *slovo (“word”) is an s-stem, *sloves-, which would have led to an expected form *slovesěni (compare Russian слове́сность (slovésnostʹ)

Messy, but fun.


  1. Stephen C. Carlson says

    Not a fan of the “I lave” gloss for κλύζω. The word is no longer in common use and it unclear whether it is supposed to mean more than “wash.” “I wash/rinse out” would be better.

  2. I agree.

  3. January First-of-May says

    and also on the opposition *slověne vs. *němьci.

    Checked out the latter on Wiktionary, and was surprised to find out that it is apparently the origin of the modern Arabic name of Austria (and I do mean Austria, not Germany, which was probably the most surprising part).

  4. Father Jape says

    Well, the Ottoman Empire bordered the Austrian Empire, and the border peoples were Slavic speakers, so…

  5. David Marjanović says

    Makes sense, though – right next to the Ottoman empire!

  6. Ксёнѕ Фаўст says

    It’s increasingly speculated here that Polish state was built by the Piast dynasty in the 10th century on the foundation of slave trade. Certainly archeological findings of many tribal fortresses from that period razed to the ground aren’t in conflict with this theory. Basically this is what happened in other areas and periods (Africa, North America) when a less technological culture met a more advanced one (and slave trade wasn’t taboo), the more ambitious and ruthless leaders soon realized that people e.g. war prisoners are a resource that’s valued by their trade partners more that anything else they could offer so they tapped it to fund further expansion. Also, early Slavs had an ‘open’ slavery system as evidenced by chronicles and allowed the slaves to integrate quickly if they were willing to accept their customs.

    As for the name Slavs, Zbigniew Gołąb’s hypothesis I have a soft spot for is that the original form was *Svoběne (‘our own ones’, cf. Suebi), as evidenced by the ancient record Suobenoi, which turned into Sloběne by dissimilation (cf. the oscillation svoboda ~ sloboda in various Slavic dialects) and Slověni by folk etymology (the worded ones), Slavjane is a further development due to sound change and folk etymology (the glorious ones).

    A fun case is the Romanian one, where the Romance-speaking people’s autonym rumîn came to mean ‘serf’ at one point (I wonder what was the neutral self-designation then, Vlah perhaps?); also the Slavic loan rob was a word for ‘slave’. Şcheau/șchiau from Latin *sclavus meant ‘Bulgarian’ as far as I can find. Nowadays they also use sclav for ‘slave’ (due to 19C relatinization-westernization?).

  7. I am curious what happened to the thousands of Slavs that were supplied to the Arabs. Did they leave descendants and have they made a contribution to the gene pool?

  8. It is inconceivable in 2019 to use an ethnic name to describe a category of people in bondage and servitude. But that is exactly what happens when using the word “slave”. There is no excuse for it in English, because there exists a substitute word “thrall”.

  9. Ксёнѕ Фаўст says

    I am curious what happened to the thousands of Slavs that were supplied to the Arabs. Did they leave descendants and have they made a contribution to the gene pool?

    AFAIK this was minimized through widespread castration and closed slavery system (slaves were a closed caste with no hope of getting out, they were kept to work and as a status display; those countries have little arable land, people married and bred carefully).

    Some rebelled though, there are records about a Slavic fortress somewhere in North Africa.

  10. Canada’s Northwest territory is divided into five administrative regions, two of them are called North Slave Region and South Slave Region.

    The indigenous people of the area are called Slavey (or Slave, Slavé) people.

    The name was given by their enemies – the Cree Indians who liked to enslave them.

  11. xyz: There is no excuse for (using an ethnic name) in English, because there exists a substitute word “thrall”.

    I agree, cf “to jew” etc., but are you using thrall now instead of slave? Thralary, the thrall trade, a thrall unit etc?

  12. @ Bathrobe:
    I am curious what happened to the thousands of Slavs that were supplied to the Arabs. Did they leave descendants and have they made a contribution to the gene pool?

    Undoubtedly. Throughout much of the Islamic period, a majority of slaves traded from different regions were female, and some of these were used concubines/sex slaves, whose children could grow up as free, becoming members of the general society. A rather extreme later example is Roxelana, originally a Ruthenian woman kidnapped into slavery by Crimean raiders, who became the favorite wife of the Ottoman sultan Süleyman, and mother of his successor.

    In addition, male slaves could become freedmen if they were manumitted by their owners (something encouraged by Islam).

  13. Stu Clayton says

    Grace Jones might have sung “Thrall to the music”, but I doubt with equal success. It’s more of an Edith Sitwell title.

    Or rather a word postulatoried by the thought of her:

    # … contrast Waldron with two other poets — Edith Sitwell and Stevie Smith. You might find these strange bedfellows but both poets are sui generis writers and developed a linguistic world: percussive, postulatory and in thrall to the syntactic music of the poem #

  14. Thralary

    Naah. Salary is connected with salt, but we don’t know exactly how. In Roman times salarium, which is formally the neuter form of the adjective meaning ‘salt-related’, may have meant ‘salt given as an equivalent of wages’ or ‘money paid to buy salt’, or who knows what.

  15. Undoubtedly
    Population genetics usually provides good answers not only on the extent of these ethnic admixtures, but also on timing (by segment size distribution), most likely subregional sources (by scoring best fitting matches) and sex biases of admixture (both by classical uniparental markers and by X-to-autosome ratios).
    But I don’t recall thorough studies of the Middle Eastern Muslim countries by this methodology. The easiest to observe admixtures there are well known though, and attributed to slave trade, specifically from Subsaharan Africa. They are widespread in Muslims, especially in Egypt and Yemen, but absent in non-Muslim groups (Coptic, Yemeni Jewish) or in pre-islamic samples.

  16. John Cowan says

    It is inconceivable in 2019 to use an ethnic name to describe a category of people in bondage and servitude. But that is exactly what happens when using the word “slave”.

    It’s purely a historical fact, not present in the minds of English-speakers, who see no connection between slave and Slav, any more than between insignia and ensign, or attach and attack, or cream and grime, or carton and cartoon, or cloak and clock, or domain and dungeon, or sovereign and soprano, or tradition and treason, or chakra, cycle, wheel, or price, prize, praise, pry, or dish, desk, disk, discus, dais, disco, or chief, chef, cape, capo, caput, head, or any of hundreds of others. If we called slaves slavs, you’d have a case; the argument against Slavey is quite reasonable. In any case, to enthrall someone with your voice, or beauty, or guitar playing, is by no means to enslave them.

    (An Irish doublet just for the Hat: cuid and píosa , both ‘part, portion’. The first is native, whereas the second is from Gaulish > Late Latin pettia > Normand piece > Middle English > Middle Irish pissa.)

  17. In 1914, 26-year-old Sitwell moved to a small, shabby flat in Pembridge Mansions, Bayswater, which she shared with Helen Rootham (1875–1938), her governess since 1903.

    There is nothing SHABBY about Pembridge Mansions. It’s one block from Kensington Gardens. In 1917, the architect of the National Theatre Denys Lasdun was born at 17 Pembridge Place, around the corner. I spent my childhood at number 18. Why would a Wikipedia article writer have a grudge against Pembridge Mansions?

    You may choose whatever spelling you like for thralary, John. It’s all yours. I remembered slave & slav have totally different meanings. Never mind the number of Ls, thral- is nutty. I don’t know what came over me.

  18. No need to invent “thrallary” when we have “thralldom” already.

  19. slovo (word) and slava (fame, glory) come from the same PIE *ḱlēw-. Russian slang word клёвый /klyovyj/ (cool, awesome) probably has nothing to do with them and comes directly from клевать (peck, bite)

  20. David Marjanović says


    So the *slověne – *němьci opposition would be real, but only as folk etymology! I like that. How old is this Suobenoi, though?

  21. David Marjanović says


    Short vowel – the length ( > -a-) is attributed to Balto-Slavic word derivation by vrddhi.

  22. I think zyxt is baiting us.

  23. “South Thrall languages like Servian…”

    Damn, never realized how awful these names actually sounded

  24. So do I. Mea culpa.

  25. Hello bathrobe
    I just don’t think that it’s OK that Slavs should be regarded with inferiority, or associated with or linked to human trafficking & exploitation. That’s what happens when their name is used in such a context.

  26. But is it what happens? That’s an empirical claim for which you have provided no evidence at all. Do you have any examples of complaints by Slavs (who certainly know how to complain) about the inappropriateness of skllav, esclavu, sklav, slave, esclau, slaaf, sklavo, esclave, σκλάβος sklavos, sclábhaí, schiavo, sclavus, sleab, escravo, esclavo, sciavo, שקלאַף shklaf, or even Swedish slav, which actually does mean both ‘Slav’ and ‘slave’ (though differentiated in the plural, slaver vs. slavar)?

    Rumania and Servia, the serf lands.

  27. David Marjanović says

    …Just in case, srb(-) < *sьrb > serb- has been the self-designation of the Serbs and the Sorbians “since ever”. It has nothing to do with servus or its descendants, except perhaps by occasional folk etymology of especially the Greek version, which has been spelled with β for traditional reasons and accordingly pronounced with [v] for something like 1400 years now.

  28. John Cowan, is cuid cognate to quantum?

  29. John Cowan says

    Wiktionary says no, that Proto-Celtic *kʷesdis > cuid has no known cognates, and conjectures it is a substrate word. In the insular P-Celtic languages, the sense shifted to ‘thing’, giving Breton pezh, Welsh peth, Cornish peth/pyth.

    However, the Late Latin form pettia has spread widely: besides its direct descendants Italian pezzo, Spanish pieza, Portuguese and Occitan and Catalan peça, there is also Albanian pjesë direct from Latin, and Danish pjece ‘gun; booklet’, Swedish pjäs ‘(performance of a) play; playing piece (in games)’, Russian пьеса pʹjesa ‘performance of a play, musical composition’, Kurdish piyes ‘theatrical performance’, all from French. All these meanings are specific senses of French pièce and/or English piece.

    On the other hand, English fit in the sense ‘part of a musical composition or poem’ (Lewis Carroll subtitled “The Hunting of the Snark” as “An Agony in Eight Fits”, punning on both nouns) is a good semantic and formal match for *kʷesd-is (Icelandic fit ‘web’ and older German Fitze ‘skein’ are good formal matches though less good semantically), so it may have spread from Celtic to Germanic or vice versa before Grimm’s Law, or alternatively we are dealing with an irregular descendant of PIE *pedjo-. “Infinite are the arguments of mages.”

    Note: The Gaulish form is reconstructed; I should have written it *pettiā above.

  30. Ксёнѕ Фаўст says

    How old is this Suobenoi, though?

    It’s in Ptolemy (2nd century).

  31. Zeleny Drak says

    @Ксёнѕ Фаўст

    The term rumîn came to mean serf just in one part of the country (ironically enough in the region called Țara Românească=Wallachia). In the principality of Moldova the name for serfs was “vecin” (same origin as the Spanish word vecino). Rumîn seems to have expanded the sense to mean man then from there to peasant. Once almost all peasant lost their freedom and became tied to the land (end of the 16th century) the term got the meaning of serf, as there was no longer any meaningful separation between peasants. I’m not sure if there was another neutral term used. For sure it was not Vlah as this was never a word in Romanian (only used in the Slavonic documents). I would imagine it was something like Creștin (Christian), but I might need to check that. Keep in mind that starting from the early 18th century most of the nobility was of Greek origin.

  32. David Marjanović says

    or alternatively we are dealing with an irregular descendant of PIE *pedjo-.

    That would give the Germanic forms directly: Grimm, West Germanic consonant lengthening, umlaut. It would not explain a Celtic or Latin -tt-, however.

    It’s in Ptolemy (2nd century).


  33. John Cowan says

    Yes, I meant that the Celtic forms, not the Germanic ones, were irregular, which is why Wikt says conservatively that they are not IE.

    The words for ‘thing’ in IE have an extraordinary variety of origins: ‘assembly’ in Germanic, ‘piece’ in Brittonic (and Icelandic, where þing retains its original sense), ’cause’ in Romance, ‘what is spoken of’ in Slavic, ‘what is done’ in Hellenic, ‘substance’ (I think) in Indic, I have no idea what in Baltic or Iranian.

  34. The semantic space for “thing” just seems like it attracts new coinages and repurposings. Note that there is still the phenomenon of using nonsense words to mean “thing.” Some of these stick around: doohickey, doodad, thingimajig, etc.; however, most if them do not persist. Billy Crystal had a joke about how his family did not actually know much Yiddish, but they were good at making up Yiddish-sounding nonce words for things.

  35. @zyxt

    Apart from etymologists, I don’t think anyone connects ‘Slav’ with ‘slave’. Your well-meaning suggestion can only lead to another convulsion of politically-correct word replacement. (Incidentally, have you always felt this way or did you only get this idea after you read this thread?) ‘Serf’, ‘slave’, ‘peasant’ all have their etymologies. Are you seriously suggesting that swathes of familiar terminology should be replaced because someone doesn’t like the etymology?

    Do you feel the same about ‘buggery’, which is based on the word ‘Bulgar’?

  36. Yeah, but other language groups got much prettier names, it’s so unfair!

    Like Romance, for example. 😉

  37. @Bathrobe

    When I was a child and first encountered the word “Slav”, the similarity to “slave” immediately struck me.
    More people would make the connection, if this were an aspect of history that was more generally taught.

  38. I don’t suggest for a moment that any contributors to this forum are prejudiced against Slavs. Indeed our gracious host LH has, if anything, a genuine pro-slav attitude, as evidenced in his numerous posts on Russian literature and related topics.

    As to empirical research, I can’t point you to any. It’s a matter for professional researchers to do that. In all likelihood, 60 years ago there was no empirical research on gender-biased language either.

    However, I can mention this off the top of my head:
    – My dictionary lists these words in this order: slav, slave, slaver, slavery, slavic, slavish. Plus there is a statement that “slave” comes “from the reduction to slavery of many Slavic peoples”. Enough to make anyone think of a connection.
    – There is a controversial play called “Slav” performed in Montreal at the moment. The controversy is NOT about the lack of Slav actors in the play.
    – It would be trite to list numerous examples of negative attitudes towards Slavs, prejudice, and negation or surpression of their role in history.

    I’m not interested in getting into disputes about this. The use of the words serf, peasant, etc is beside the point. The point is that it’s not OK to use the name of an ethnic group to describe a state of inferiority or bondage, and there is an alternative word to use. There is an opportunity to right a wrong. No need to shoot the messenger if you disagree with the message.

  39. John Cowan says

    There is a controversial play called “Slav” performed in Montreal at the moment.

    Red herring: the controversy is all about white actors playing black slaves and consequent cultural appropriation[*], and nothing to do with either Slavs or the word slave. I got to wondering why the play is called that, and turned up the Mop and Pail article, which styles it SLĀV, presumably North American Dictionary Respelling for slave.

    it’s not OK to use the name of an ethnic group to describe a state of inferiority or bondage

    Agreed. But slave/Slav is simply not a case of this, except possibly in Swedish. The only difference between this and the notion that picnic is a slur on African Americans is that the etymology in that case is false, whereas in this case it’s true. But it’s still an etymological connection only.

    there is an alternative word to use

    Unfortunately, there isn’t. Thrall ‘slave’ is simply obsolete: the last use (other than in historical or poetic contexts) in this meaning that the OED records is in 1612, where unredeemed human beings are described as “thralls to Satan”. You might as well try to revive waugh (although the Welsh might not like it).

    There is an opportunity to right a wrong.

    Nor that either. The Slavs who were enslaved are beyond our help, and their descendants and relatives do not suffer a continuing disadvantage because of it, nor are they touchy on the subject.

    [*] I spent yesterday evening listening to the music of Stan Rogers and singing along with it, thus appropriating an enthralling product of white Anglo-Canadian culture. Why, I have even learned some by heart and sung them all by myself, though admittedly not for money.

  40. In other news, the term “terrorism” now considered across the Galactic Federation unfair against Terrans.

  41. I wonder if Montenegro, Nigeria and especially Niger are considered racist by someone…

  42. The US states and territories with offensive names: Colorado (offends Colored people), Indiana (offends Indians), Virginia, West Virginia and US Virgin islands (offends women), Puerto Rico (offends poor people).

  43. John Cowan says

    If Geraint is any indication, however, New Jersey does not offend people from Jersey: indeed, they consider it a colony of theirs.

  44. J.W. Brewer says

    I think the fact that Slav and slave are neither homographs nor homophones in (most varieties of? or is it all varieties of?) modern English is what means that the distant historical connection is at least as opaque as the Bulgar/bugger one (i.e. pretty opaque for the overwhelming majority of speakers with no interest in curious etymological trivia). To take some parallels, I doubt most Anglophones would assume a connection, however vague, between “pop” and “pope” or “mat” and “mate.” The words in those pairs have different vowels, just as Slav and slave do. That some random theatrical folks in Canada have chosen a non-standard spelling of “slave” for their production does not change that.

    To take another obvious example, perhaps “cretin” should be considered an offensive word on other grounds, but not because of its opaque etymological connection to “Christian.”

    One perhaps cutting the other way — in Latin fourteen-odd centuries ago the words for “English” and “angel” were close enough to enable a famous bit of wordplay (“non Angli sed Angeli”), but in modern English us Persons of Anglo-Saxonness are rarely taken, even in jest, to be angelic.

  45. J.W. Brewer says

    To the extent it’s true as John Cowan says (which it certainly is afaik) that it’s difficult to turn up Slavs who are touchy on this point, I wonder to what extent there’s a systematic difference between an endonym which gave rise to pejorative derived words in other languages (as is the case here) and an exonym which “came with” pejorative baggage from the outsiders who imposed the name? Slavs (or rather the minority with any predisposition to care about that stuff) may be less likely to be touchy if they are more focused on the positive vibe of the endonym in its own language-of-origin. Not to mention the fact that in most Anglophone societies, slavery is a purely historical phenomenon — if it were ongoing it would no doubt have already gone through a couple cycles on the euphemism treadmill and now be called something else.

  46. I think that the main reason people in general don’t connect the two words is that for the average English speaker, “slavery” is practically synonymous with the trans-Atlantic slave trade and slavery in the antebellum South. That’s how people are educated. If it weren’t for that, it would be very easy to see a link, the two words are very similar and you don’t need to know any laws of phonetics to see the similarity.

    This hasn’t been an issue because the Slavs haven’t generally been in antagonistic contact with English-speaking communities and there isn’t any political capital in this to be made by anyone. If the German word for “slave” was “Slave” instead of “Sklave” I expect that would be seen as a problem (Slav being Slawe).

    I bet a lot of people in Eastern Europe wonder about it though, when they learn English.

  47. J.W. Brewer says

    The fictitious nation of “Lower Slobbovia” in the old Li’l Abner comic strip is perhaps some evidence that if AmEng speakers are looking for a pejorative word that evokes or resembles “Slav,” it turns out that “slob” is a closer fit than “slave.” (FWIW, the first kinda-plausible etymology for “slob” I googled up is kinda hilarious because it says it came from a borrowing from Irish, but the relevant Irish word had itself been a loanword from an earlier era of English, with that English word having fallen out of use in English during the intervening centuries.)

  48. January First-of-May says

    I wonder if Montenegro, Nigeria and especially Niger are considered racist by someone…

    When the Everything Wrong With Yakko’s World video got to the Niger line, it did, in fact, respond with “that’s racist!”
    So when the inevitable Everything Wrong With Everything Wrong With Yakko’s World came out, the response to that line was “it’s not racist, it’s the country’s name!”

    (The “negro” part of Montenegro is at least of the same origin as the respective racist epithet; Niger, however, is probably entirely unrelated – though IIRC the etymology in question is still somewhat disputed.
    Nigeria, of course, is of the same origin as Niger, whatever that might be, both countries having been named for the Niger river.)

  49. David Marjanović says

    In other news, the term “terrorism” now considered across the Galactic Federation unfair against Terrans.

    Fearless Flightsuit did consistently dissimilate terror to Terra, and War on Terror to Warren Terra.

  50. J.W. Brewer says

    I’m just happy that the Montenegrins apparently remain content to be known in English (and numerous other non-Slavic languages) by a calque (into Venetian, sez one source, which I guess retained the “g” that standard Italian lost) of their autonym and aren’t demanding we call the place Crna Gora in English.* Let’s not encourage any move in that direction. They may be the only UN member, but other toponyms with “negro” as a component are very common in parts of the world where speakers of Spanish or Portuguese did much of the naming, and that’s a lot of different places.

    *Fun fact: as of a few years ago I used to see a car with the N.Y. vanity license plate CRNAGORA with some frequency within a one-or-two-mile radius of my house, but not recently, so the owner must have either moved out of the neighborhood or gotten a different license plate.

  51. John Cowan says

    Quoth Wikipedia:

    The earliest use of the name “Niger” for the river is by Leo Africanus in his Della descrittione dell’Africa et delle cose notabili che iui sono published in Italian in 1550. The name may come from Berber phrase ger-n-ger meaning “river of rivers”. As Timbuktu was the southern end of the principal Trans-Saharan trade route to the western Mediterranean, it was the source of most European knowledge of the region.

    Medieval European maps applied the name Niger to the middle reaches of the river, in modern Mali, but Quorra (Kworra) to the lower reaches in modern Nigeria, as these were not recognized at the time as being the same river. When European colonial powers began to send ships along the west coast of Africa in the 16th and 17th centuries, the Senegal River was often postulated to be the seaward end of the Niger. The Niger Delta, pouring into the Atlantic through mangrove swamps and thousands of distributaries along more than 160 kilometres (100 miles), was thought to be no more than coastal wetlands. It was only with the 18th-century visits of Mungo Park, who travelled down the Niger River and visited the great Sahelian empires of his day, that Europeans correctly identified the course of the Niger and extended the name to its entire course.


  52. Jack De Manio was famous in Britain during the 1960s as the presenter of Today, the BBC radio news programme from 6 – 9 am. He had a Woosterlike accent, was pretty conservative and looked like a frog, but I enjoyed listening at the time. Some loathed him (my great-uncle). His career nearly crashed in 1956 when he was duty announcer for the BBC’s Home Service. A major radio feature, The Land of the Niger, was broadcast worldwide to mark a Royal visit to Nigeria. Carelessly, he back-announced it as ‘The Land of the Nigger’.[3][4] There was outrage; he was immediately suspended and then returned to the General Overseas Service.

  53. @J.W. Brewer: Slob is also nicely complemented by the Yiddish borrowing zhlub or s(c)hlub, which has a somewhat similar meaning but is apparently unrelated (with Polish żłób being a likely source). I usually hear it with /ʃ/, but with my dad being a native(ish) Yiddish speaker, I acquired it with /ʒ/.

  54. Perhaps Fergie finds New York offensive…

  55. The rule of thumb with the potentially-offensive words is, etymology and the past use don’t matter. What matters is a) the negative feeling of today, b) the availability of nicer synonyms, and c) the political clout of the offended group. Since b and c aren’t there, we are free to discuss the (otherwise irrelevant) etymological stuff.
    Calling Russians the russkis OTOH may be etymologically fine, but there is a) and b) so it’s out.

  56. Reminds me of the wonderful note in Urban dictionary:

    Definition: Slang for russian.
    (not racist because russians are not a race)

  57. I feel like there should be a term for pejorated endonyms.

  58. January First-of-May says

    Medieval European maps applied the name Niger to the middle reaches of the river, in modern Mali, but Quorra (Kworra) to the lower reaches in modern Nigeria, as these were not recognized at the time as being the same river.

    Ironically enough, one of the early 19th century British expeditions to find the Niger actually started from the Quorra (and never found anything it could identify as the Niger).

    I feel like there should be a term for pejorated endonyms.

    I definitely agree that there should be a term, and wouldn’t be very surprised if there actually is one already.

  59. Topical, if not yet completely pejorated: the word for the United Kingdom of Great Britain and Northern Ireland leaving the EU ought to be Brexeunt. That would leave ‘Brexit’ for Great Britain leaving without taking N. Ireland with it (see ‘backstop’).

  60. Despite ‘Negro’ being etymologically unexceptionable it has been deprecated and is no longer commonly used. ‘Niggardly’ is also etymologically unexceptionable but is avoided by many because of its resemblance to another word.

    In both cases the problem is that people take issue with their perceived discriminatory overtones.

    It is conceivable that one day some people will take issue with the word ‘slave’, but I doubt it will have anything to do the Slavs. It is more likely to be an objection, for whatever reason, to the characterisation of people as ‘slaves’.

  61. California is not just offensive, but outright extremist – because it is named after the title of ISIS leader….

  62. Wow, I just found out that the term nègre littéraire (“literary Negro”) is a still-current way to say “ghostwriter” in French. Apparently some replacements have been mooted, though, like prête-plume and fantôme écrivain.

  63. David Eddyshaw says

    On the other hand, âme damnée is so much better than “catspaw”, and moreover is not offensively antifelinist. Anglophones are so insensitive.

  64. Thrall to mean “slave” is sufficiently obscure these days that there was this question was asked about whether thrull from Magic: the Gathering was a play on it. (I was reminded of this today because my answer suddenly got a new flurry of up votes.). Of course, there is the further complication that the Order of the Ebon Hand’s thrulls were slaves intended specifically for dual use as servitors and sacrifices.

  65. John Cowan says

    It’s interesting how the Thrulls have kept an oppressive name and made it their own. “Kaffir” is deeply insulting in South and East Africa and in Mauritius, neutral in Reunion, and a term of pride in Sri Lanka for the African-descended (etymologically it is Arabic for ‘unbeliever’).

  66. “catspaw”

    Then monkey shoulder is positively outrageous.

    The name Monkey Shoulder originates from an injury that the maltmen (those that hand turn the grain in a malthouse) would get as it would leave one arm hanging.

  67. I feel names like White Sea, Black Sea, Red Sea and especially Yellow Sea are kind of racist too.

  68. David Marjanović says

    Fun fact: in Turkish, it’s the Mediterranean that is called White Sea (Akdeniz), evidently simply in contrast to the Black Sea (Karadeniz).

  69. For what it’s worth, my personal preference is for the de-hydronymic origin of *slověne, since the the suffix *-ěn- otherwise occurs only in “habitative” ethnonyms — names of tribes or communities based on their place of origin (either a toponym or a common noun such as ‘wood’, ‘meadowland’, ‘lake’ etc.). It has the same function as the much more frequent suffix *-jan-. Gołąb’s *svoběne fails on this count.

  70. Ксёнѕ Фаўст says

    In earliest sources, the form словѣни occurs, though. There’s also the -ěnъ in adjectives and nouns of quality like *maslěnъ ‘butter _’, *kostěnъ ‘bone _’, *molděnъ ‘young man’. So maybe the original form was *Svoběnъ, not *Svoběninъ.

  71. David Marjanović: Careful! In Turkish color terms were used once to refer to directions: “black” stood for North and White for “West”, so the Turkish name for the Mediterranean may have nothing to do with its being less dark than the Black Sea…

    Lazar: “nègre littéraire” is certainly no longer current in French today, the noun now being as stigmatized as its English equivalent,

  72. David Marjanović says



    Still, Akdeniz refers to the whole Mediterranean, not just to the Aegean, so it’s mostly south rather than west of Turkey, and the Black Sea was called “black/dark” in some Iranian language even before the Quite Ancient Greeks Indeed misinterpreted that as axeinos “inhospitable” and then changed that to euxeinos.

  73. So maybe the original form was *Svoběnъ, not *Svoběninъ.

    Which gives us two irregular substitutions arbitrarily blamed on folk etymology plus an unmotivated change from a plain o-stem to a consonantal stem of the kind associated with “habitative” ethnonyms. And why should the allegedly original ethnonym have been just an adjective without any noun-forming suffix? It strikes me as unusual in the Slavic context. After all, we have *němьcь with the *-ьcь < *-iko- extension, not *němъ (= the basic adjective).

    The derivation doesn’t work without several ad hoc assumptions.

  74. >Still, Akdeniz refers to the whole Mediterranean, not just to the Aegean, so it’s mostly south rather than west of Turkey, and the Black Sea was called “black/dark” in some Iranian language even before the Quite Ancient Greeks Indeed misinterpreted that as axeinos “inhospitable” and then changed that to euxeinos.

    In Bulgarian the White Sea is specifically the Aegean. I would love to know if this is post-Ottoman or pre-Ottoman. I should research that.

  75. John Cowan says

    You may choose whatever spelling you like for thralary, John.

    Sorry for being thick, AJP. I genuinely didn’t read that as thrall-ery.

  76. Lars (the original one) says

    the problem is that people take issue — if it was only that. Some words are inextricably bound up with actual, (hopefully) historical crimes against humanity, and if the victims or their descendants tell you that those or similar words make them uncomfortable or worse, the only courteous thing to do is to believe them and try to change your usage.

    Then there are people who seem to delight in taking offense to words on behalf of others, or pretending that innocent use of certain words is ipso facto proof of bad intentions. I have no patience with that.

  77. Same here.

  78. I genuinely didn’t read that as thrall-ery.

    But like salary! Nice one, John. It’s my silly affected accent (‘thrawl’).

  79. John Cowan says

    I say “thrawl” too, but of course when I say it, it’s just normal for a person from the Northeast, or indeed the South (like my wife). 🙂

    It’s the /l/ that pushed the vowel over to aw, as in all, stall, gall etc., but then in BrE it changed Americans who merge THOUGHT with PALM of course say “thrahl”. I doubt if anyone says “thral” to rhyme with “Sal” any more (maybe in Scotland or Ireland). Are there non-North-Americans who say “throll” to rhyme with “doll”?

  80. Eli Nelson says

    Some accents of Scottish English are supposed to merge LOT and THOUGHT (as [ɔ]), but I don’t know whether “thrall” is a THOUGHT word in such accents.

  81. John Cowan says

    It’s more that they never split TRAP and PALM in the first place: psalm rhymes with Sam, and both are [a], as in Middle English. TRAP is the normal reflex of ME short /a/, but in a few classes of words such those with compensatory lengthening (as in palm itself) and lengthening before /r/, plus the anomalous word father, a new low back vowel PALM arose. PALM then added onomatopoeic words and foreign borrowings, and in TRAP-BATH-splitting varieties all the BATH words as well. Wells himself said that PALM is the least satisfactory of his 24 keywords, partly because of its heterogeneous origins: any keyword you pick might not belong to PALM in some areas.

  82. Slavonian means slave. They were slaves of germanic tribes. Copied many germanic words too.

  83. There otta be a law that you aren’t allowed to add a comment until you have read the other comments, or most of them. But I can’t blame foreigners (that is, non-Hattics) for not knowing that this is a conversation, not a place where people air their views in total isolation from everyone else’s views, like YouTube comments.

    Somebody said that the Greeks said that you didn’t have an actual opinion until you had rubbed up against other people’s opinions, only a dokei-moi ‘it-seems-to-me’. My memory of all this is obviously vague beyond all bearing.

  84. David Eddyshaw says

    like YouTube comments


  85. @David Eddyshaw: I think that by 2008, when that comic came out, YouTube had already shown itself to have the worst comments of any major World-Wide Web site. “Never, ever read YouTube comments,” was a standard piece of Internet advice for a long time. Ironically, the comments on YouTube have probably gotten a bit better since then—and they certainly look better in comparison to the ones at some other sites—because a lot of the really vile posters have migrated to other platforms in the intervening years.

  86. jack morava says

    @ JWB 3 years ago : …. it’s difficult to turn up Slavs who are touchy on this point …

    I have a kind of itchy grudge against Franz Boas [bless his heart as my momma would say] for chasing strange cultures in strange places with his dueling scars while AFAICT ignoring the Slavs all around him. But alas everyone has blind spots…

  87. David Eddyshaw says

    Just not exotic enough. Now, had he been surrounded by Welshmen … but, alas, it was not to be, and he had to make do with Kwakiutl.

  88. J.W. Brewer says

    Perhaps (the currently-living members of) some of the ethnolinguistic groups Boas did focus on wish he’d diverted his attentions elsewhere?

    @Brett: Of course that stereotype was what led the proprietor of the “Sad Youtube” blog (floruit 2012-2015) to take the contrarian view that there was gold in them thar comment threads if you knew how to look for it, and I found his results quite impressive. “Moments of melancholy, sadness and saudade from the lives of strangers, gleaned from the unfairly maligned ocean of YouTube comments.” I assume that once the contrarian proprietor got rolling he figured out how to find the stuff he was looking for pretty efficiently, although I don’t know how many man-hours he needed to invest before he got the knack of it.

  89. ktschwarz says

    Proto-Slavic slověninъ … Messy, but fun.

    Somebody on Wiktionary has since swept aside that messy list into a box labeled “obsolete and other etymologies”, keeping only the hydronym theory, which was the one favored by Vasmer and Piotr Gąsiorowski.

    The Language Log post also copy-pasted AHD’s word history box for slave, which surprisingly doesn’t hint at any mess or controversy but just states confidently that Slav means “famous people” — even though that was dismissed by Vasmer and wasn’t supported by anybody else in the messy list. Is AHD perpetuating a folk etymology? That would be very uncharacteristic, but no dictionary is perfect.

    Proto-Slavic *slověn- is one of the few starred reconstructions that the OED3 is willing to publish, in the etymology of Slovene (2013), which cites the reasoning about the suffix given by Vasmer and Piotr:

    Existence of a common Slavonic self-designation with the base *slověn- is suggested by Old Church Slavonic slověninŭ (plural slověne) and cognate forms … *Slověn- is perhaps a derivative of the Slavonic base meaning ‘word, speech’ (compare Old Church Slavonic slovo (see loud adj.) …). However, derivation from an unattested place name (compare Old Russian Slovutič′ the Dnieper) is often considered more likely, since the cognates of the Slavonic suffix *-ěn- in various Slavonic languages occur almost exclusively in place-name derivatives.

  90. Is AHD perpetuating a folk etymology? That would be very uncharacteristic, but no dictionary is perfect.

    Indeed. A bit of a shock, but a useful reminder to always keep one’s guard up, even with a beloved reference work!

  91. The substantial weakness of Piotr’s etymology is that it basically says that any name of a space-faring species most likely comes from the name of a planet.

    The river is hypothetical. Yes, it totaly can be a hypothetical river name.

    Yes, demonyms are usually derived from places (not sure about demonyms associated with sich large groups). Yes, names of smaller groups of Slavs are frequently derived from rivers, so it is not impossible that the name of a relatively small group became the name of a larger group (or this smaller group became more numerous).

  92. David Marjanović says

    Словутичь is not hypothetical, it’s in the OP (and also two comments above yours), and it’s in the right general region…

  93. Also:

    the cognates of the Slavonic suffix *-ěn- in various Slavonic languages occur almost exclusively in place-name derivatives.

  94. @DM, I do not think that Piotr means that Slavs are called after an epithet of Dnepr.
    Словутный/славутный means “famous”.

    We can consider this suggestion, but this is your suggestion:)

  95. I think “epithet” here is being used to mean “alternate name” or the like.

  96. P.S. the epithet is derived from a name Slovuta.

  97. Also:

    the cognates of the Slavonic suffix *-ěn- in various Slavonic languages occur almost exclusively in place-name derivatives.

    Or in other words, most demonyms are derived from places.

  98. LH, don’t you see how “Tunisian”

    (1) results from a convention that involves not only Tunisians but also other people: a Tunisian can say “I’m a Tunisian” to a Pole or Algerian in English or Arabic and be understood.

    (2) designates people from a region.

    I think 2 and 1 are connected.
    A Pole of course can call a Tunisian “an Arab”, but there is no point in telling “I’m an Arab” to an Algerian.

    (1) a convention among Slavs but NOT others.
    (2) does not designate people from a region, but a linguistical community.

    If you propose that ONCE upon a time “Slavs” functioned within a convention between geographical rather than linguistical Slavs and other (close, likely linguistically) people, you really can’t use as an argument that most names of the former (“Tunisia”) sort are derived from names of regions.

    These names (like “a Tunisian”) of course can extend to something else. E.g. Tunis extended to the whole country, similarly “Algiers” extended to mean a large city (beyond the “islands”) and then to the country. They can extend to something which is not even a geographical unity.
    But (1) is still important.

  99. This is about semantics.

    About the “nisba” *-ěn- itself… I assume when people name people the meaning comes first.
    So yes, when you need a demonym derived from a place, *-ěn- works.

    To demonstrate that it does NOT work for demonyms based on abstract qualities you need…. demonyms based on abstract qualities. If you have a plenty of those and all are derived differently, that means something.

  100. David Eddyshaw says

    most demonyms are derived from places

    In Kusaal, no demonyms whatsoever are derived from place names. This is pretty typical for West African languages in general, I think.

    In Europe, there are any number of instances of place names derived from demonyms, rather than the reverse: Britain, Cymru/Wales, Powys (a Latin loan!), Scotland, Dumbarton, England, Devon, Cumberland, Essex, Suffolk, France, Paris, Italia, Deutschland, Sachsen, München (if you call monks “people”), Belgique, Sverige, Polska, България … Россия …

  101. …Yisra’el, Filasṭīn…

  102. DE, do you mean that Kusaasi simply do not refer to “people from the city X”?

  103. …X-stan…

  104. David Eddyshaw says

    DE, do you mean that Kusaasi simply do not refer to “people from the city X”?

    Not at all. They say “people from city X”, e.g. Bɔk dim “Bawku people.” Most Bɔk dim are, of course, Kʋsaas, but there are quite a few Nwampuris and Baris, and in the Muslim quarter most people are (as you’d expect) Zangbɛɛd. There have never been many Nasaarnam*, but there are quite a few Kambʋmis these days. (They come from Ya-dagɔbʋg, of course; or as we used to call it, Zuoya.)

    * They come from Nasaateŋ, but of course that’s named after them, not the other way about. It’s pretty cold there by all accounts, and everything is very expensive. They can cope because they’re all very rich.

  105. Wow!
    Kusaal Wikipedia!

    (I googled Kambʋmis)

  106. David Eddyshaw says

    So there is!

    (I didn’t know.)

    I got a mysterious email not long ago asking me to asses the bona fides of a potential Mooré WP proposal; no idea how they got hold of my contact details. (I declined: I know more about it than many, but that’s not difficult, and there are literal millions of people who would be more suitable than me. There seemed to be internal evidence in the email that they had mixed me up with someone else, too. I haven’t heard anything further.) It looks like there is a real push going on to have a Western Oti-Volta WP presence, though. Excellent.

    (There doesn’t actually seem to be a Kusaal version of the WP page for Ashanti/Kambʋmis, even though a Google search on “Kambʋmis” found it. The same etymon appears in Dagbani, though: maybe that was enough.)

  107. “literal millions of people who would be more suitable than me. ”
    Yep:) I agree with this reasoning.

  108. @Y I don’t know about Filasṭīn but Russian fylistim-lyane has almost the required suffix.

    Not quite -ěn- but -jan.

    So it must be a place name. Alongside with Christ and Hagar.

  109. David Eddyshaw says

    I think this is a question of timescales. “France”, for example, is derived from the name of the Franks, and the the toponym – entirely uncontroversially – derives from an ethnonym; but that doesn’t mean that you can’t nowadays be French (quite a different thing from being Frankish.)

    But most of the demonyms-derived-from-toponyms (like “French”) reflect relatively modern, if not very modern concepts revolving around modern ethnostates. In fact, I’d say “French” isn’t actually a demonym at all, in the sense that “Frankish” or “Kusaasi” is; it’s more like Bɔk dim “Bawku people.” (In Kusaal, I think you’d actually have to say something like Farans teŋ dim “France-land people” for “the French.”)

    So projecting back the demonyms-from-toponyms thing probably gets less and less plausible the farther back in time you go.

    Having said that, exonyms meaning things like “the river people”, “the mountain people” go back a long way. Autonyms, too: the self-designation of the Solla, pì-Yɔ́bɛ, seems to mean “Mountain People.” (I got interested in the Solla because their language, Miyobe (? “Mountainspeak”) looks like a pretty good candidate for the closest relative of Oti-Volta.)

    I don’t think those are precisely based on placenames, though; the Piyobe (for example) aren’t the people of some particular (named) mountain: they are “the mountain(eer) people.” (Much like their direct neighbours to the north are the Oti-Volta “Builder People”, the Batammariba.)

  110. I remember seeing the argument from Georg Korth in “Zur Etymologie des Wortes ‘Slavus’ (Sklave)” (1970) that Medieval Latin sclavus “slave” derived from σκῦλον or σκύλον “war booty” via the corresponding verbs σκυλεύω or σκυλάω being Latinized as *scylavus and had nothing to do with the Slavic tribes.

    He connects this to 10th- and 11th-century Andalusian Arabic sikláb, pl. sakálibah which he supposes is a metathesized version of skiláb from *scylavus. I assume this is the same lexical item as صقلبي ṣaqlabī, pl. صقالبة ṣaqālibah but I’m not sure where he got the forms he used.

    The comments on this post reminded me of this, but when I tried to look up the article, I also found a savage rebuttal by Hans Ditten from 1972 criticizing the many bold claims made by Korth and accusing him of being ignorant of the research on the area. I don’t read German so I’m struggling to make sense of what he is saying exactly, but he seems merely to defend the view that Slověnecorresponds to Σκλαβηνοί, Σθλαβηνοί, Σκλάβοι, Sclaveni, and Sclavi (something about Slavic ă > o before 800 CE, the insertion of k (c) or th in Greek and Latin for the unusual cluster sl-). So he’s probably appealing to Occam’s razor and negating the need for an alternative derivation.

  111. DE, there is a long list of various names of people in early Slavic texts with -ěn- or -jan which either translate various Biblical words (e.g. John 1:47 “Israelite” is izrailit-ěn-inъ where -inъ is singulative) or designate various Slavs.

    Cf., 1086: Termini autem eius occidentem versus hii sunt. Tugust, que tendit ad medium fluminis Chub, Zediza et Lusane et Dazana, Liutomerici, Lemuzi usque ad mediam silvam, qua Boemia limitatur. Deinde ad aquilonem in sunt termini: Pssouane, Chrouati et altera Chrowati, Zlasane, Trebouane, Pobarane, Dedosize, usque ad mediam silvam, qua Milcianorum occurrunt termini.

    So taking Pobarane as an exmaple, they are assumed to be derived from the name of a river Bobr.
    This is the only text where they appear, but in a version of the same text in Chronica Boemorum they appear as Boborane. Historians call these entities “tribes” because they don’t want to call them “entities”.

    This suffix also can appear in names of people of some city, or in names of mountaineer people (from “mountain” accordingly) etc.

    I don’t understand the relationship between -ěn- and -jan, and in modern Russian they merge… and are used in the manner of -ian (as in Tunis-ian).

  112. David Eddyshaw says

    Good point about Biblical names. And certainly the Greek city-states named their citizens after the cities: to say nothing of the Rom-ans. So I’d have to bring my definition of “modern” back a couple of millennia to save my thesis …

    And “Babylonian” (the language) was actually called after Babylon even by Akkadians, even when they didn’t call it after (third-millennium) Akkad …

    The Hebrew Bible names in -i are often based on place-names which the text itself says are themselves the personal names of supposed ancestors, so a Canaanite is, in-universe, a descendant of Canaan rather than someone from Canaan. I’m not sure whether that helps or hinders my original proposal …

    Thinking about it, the Oti-Volta usage must reflect the fact that towns are a relatively recent phenomenon in those parts; thus Kusaal ethnonyms are all unanalysable within Kusaal, whereas almost all local placenames have evident meanings. (Same for Oti-Volta in general: “Batammariba” is very much the exception to prove the rule.) And it would hardly be astonishing if this modern custom of naming ethnicities after places was started off by this new-fangled fad for living in cities. (Hausa uses terms for its dialects based on the original Hausa city-states, too, from before the jihad which created the Sokoto Caliphate.)

  113. David Eddyshaw says

    Come to think of it, “Agolle” and “Toende” are placenames, and the terms are used in Kusaal itself to distinguish the dialects: O pian’ad Agɔl “She speaks Agolle Kusaal.” That is not taken to be an ethnic division, though: speakers of both dialects are clear that they’re all Kusaasi.

    It’s just occurred to me that Biblical locutions like “Land of Canaan”, which one tends to parse as similar to “City of London”, are actually more like “City of David”: the second element is still a personal name rather than a toponym.

  114. Good point about Biblical names.

    Actually Biblical names often are like “Israelite”: more like Slavic -ichi and Arabic banu names than Sahrawi (or Tunsi to that matter) or Slavic -ěn- names.

    izrailit-ěn-inъ looks like a derivation from a demonym meant to make it recognisable.
    Same for fylistim-ljan-e.

  115. David Eddyshaw says

    Actually Biblical names often are like “Israelite”: more like Slavic -ichi and Arabic banu names

    Yes, that’s what I’m saying. And indeed, more so than one sometimes realises, reading the text with modern notions of ethnicity.

    It occurs to me that names of languages/dialects may quite often be based on toponyms even when ethnonyms aren’t; things like the Kusaal usage of “Agolle” and “Toende” are probably not unusual. (This reminds me, too, of the self-designation Koyra Chiini “Cityspeak” for Timbuktu Songhay. And at the other end of the Songhay world, “Dendi”, I gather, means “Downriver.”)

  116. Zangbɛɛd

    Is this “Hausa”? (Like Mooré zãngoeere “Hausa homeland, Hausa language, zãngoeoogo “Hausa person”, from the online Mooré dictionary here.)

    Do you think these are related to the regional Arabic جنجاويد janjāwīd, جنجويد janjawīd “Janjaweed”? Last time I looked there was no etymological progress on this word, either, and I don’t have time at the moment to look again.

    I assume that the phonetic resemblance to Zangbeto is just purely coincidental, but what do I know?

  117. David Eddyshaw says

    Is this “Hausa”?

    Yes. It’s the standard Western Oti-Volta name for them. I’ve long wondered about its origin. (Mooré has lost the labial-velars kp gb, replacing them throughout with plain velars.)

    It’s not just that it’s opaque etymologically (which is actually normal for local ethnonyms); it’s also got odd class membership (the Kusaal singular is Zangbɛog.) Human-reference nouns in this class pairing/”gender” are normally pejorative (like dabiog “coward”, zɔlʋg “fool”), and this is the only ethnonym that I know of in this class.

    It looks like it should have something to do with the Fulɓe personal name “Jambeedu”; it’s possible that Oti-Volta speakers didn’t make fine distinctions among vaguely-Muslim outsiders at one point. (The usual word for “Fulɓe”, seen in e.g. Mooré Silmiisi, has sometimes ended up used for “Europeans”, too. I suppose the French and British invasions could be expressed as “The Attack of the Fulboids.”)

    I have a definite feeling that we discussed Zangbeto before somewhere.
    Yes, we did:

  118. David Eddyshaw says

    Incidentally, although the Mooré Zãngoeere as “Hausa language” is regular (cf “Mooré” itself), the word for “Hausaland” ought to be Zãngoeoogo (cf Moogo “Mossi country.”) Presumably the unexpected class membership of the word for “Hausa person” has screwed up the system; the relationship between ethnonyms, toponyms and glottonyms is normally very regular in WOV languages (I’ve got a whole table of them in my Kusaal grammar.)

    I couldn’t elicit a toponym for “Hausaland” in Kusaal at all. (I did try.) It’s just now occurred to me that this may actually be the reason for the gap.

    (Actual Hausaland is of course far away from the Kusaasi area, but I should explain that these ethnonym-derived place names need not refer to any formal state-like thing at all: Kʋsaʋg could be anywhere where a lot of Kʋsaas live, for example. So the Muslim quarter (“Zongo”*) of a town like Bawku would certainly qualify to be called “Hausaland” if there were such a word in Kusaal.)

    * From Hausa zango “camping-place.” I’ve wondered if that might somehow be the origin of the WOV name for the Hausa, but I can;t really see a way of making it work.

  119. David Eddyshaw says

    Just to complicate matters further, Gulmancema has janjiagu, pl janjiadi “Hausa person” and also gbangbiagu pl gbangbiadi “Hausa person” (there’s no /z/ in Gulmancema.) The tones are mostly not given in the dictionary, unfortunately, but one of the examples for “Hausa language” marks them as low throughout; that’s the same as in WOV, which is wrong for a cognate and suggests that the words are actually borrowed from WOV. That would probably explain the variation in the forms, too.

    (In any case, it’s hardly plausible that there was a proto-Oti-Volta word for “Hausa”, of course. The Hausa Drang nach Westen must have happened much too late for that.)

    Buli has Zanggbiok, pl Zanggbaata. That actually looks like a loan too; the peculiar vowel change is actually regular in Buli, but in echt Buli vocabulary such words go back to stems ending in aa- (i.e. it’s the plural that shows the original vowel.) So it looks like a word remodelled by analogy.

  120. David Eddyshaw says

    Boringly, Mbelime just has Hɔɔsɔ, plural Hɔɔsibɛ. The Hɔɔsibɛ speak Hɔɔsimɛ, naturally. (In fact, the dictionary defines Hɔɔsɔ as “nom d’une personne qui parle le hausa”, which is interesting; “Hausa” is unusual as being primarily the name of a language rather than an ethnic group.)

    Nawdm is similarly unenterprising; one Ĥawusa, two Ĥawusmba, all speaking ɦawusm together.

  121. I just posted this on Facebook but probably nobody will see it there, so I’ll ask it here (DE may know):

    I own two books on Swahili by D. V. (Daisy Valerie) Perrott — Teach Yourself Swahili (1956) and the Teach Yourself Swahili Dictionary (1965) — and it’s frustrating to me that I can find zero information about her online: no birth/death dates, no educational experience, not even nationality. Does anybody happen to know anything about her?

  122. David Eddyshaw says

    I only found out she was a girl, even, from this very site. I forget who it was who pointed this out: but out there there is certainly a Hatter Who Knows.

  123. From your keyboard to Akismet’s All-Seeing Eye!

  124. Here is a Mrs. Perrott associated with African Studies in some snipped way.

  125. Nice find! Alas, “Mrs. Perrott” doesn’t seem to lead to any further information.

  126. David Eddyshaw says

    That’s got to be the right one. (Searching “Perrott Snoxall Polome” just took me to the WALS page on Swahili.)

    So we have established that (like Ethel Ashton) she was married. Well, it’s a start …

  127. And while we were chatting, Dmitry Pruss on FB has found she was b. 1886 Edmonton, Middlesex, England, d. 1971 Hertfordshire, England, was an elementary school teacher in 1911, and was one of the British POWs interned in German East Africa (January 1916). Holy cow, he’s good.

  128. And Steven Green found “D. Valerie Perrott taught in the Universities Mission to Central Africa and was a member of the International Institute of African Languages and Cultures.”

  129. Dmitry Pruss says

    I don’t know why the crazy Fb put your plea on top of my feed 🙂
    It’s actually harder for me to find interesting stuff here on LH, because thread titles often have no relation to what’s being discussed, and only a handful of the most recent comments are easily visible anyway.
    I wondered what earned DV Perrott a place among the POWs, but the document is paywalled, although I think I know some UK genealogists who should have a subscription…

    (I was playing with another delicious word in the meantime, salmiak, and thought I might need LH wisdom but it was an easier puzzle that I feared…)

  130. David Eddyshaw says

    Ah. She seems to have been a UMCA missionary:

    That would account for the Canon of Zanzibar link.'_Mission_to_Central_Africa

    [Comprehensively Ninja’d by Dmitry Pruss]

  131. durlng her 30 years in East Africa.

    It’s not credited by the library, but apparently this Bible reader is her work, too.

  132. David Eddyshaw says

    She was presumably actually baptised “Daisy” rather than “Margaret”, interestingly (unlike my mother’s mother, who was a couple of decades younger.)

  133. David Eddyshaw says

    Heh. I’m not the only person to have been taken aback by Canon Hellier’s remarks about how easy Swahili is:

    Quote of the week:

    Swahili is an easy language, its use is widespread, and it may be that there is no easier language to learn.

    The Late A. B. Hellier, Canon and Chancellor of Zanzibar


    Acholi is even easier.

    Adonga Moses and Otto Lucy, Language Writers

  134. “It’s a sweet language,” he says. “It flows and it’s relatively easy. Tafudhali-‘please, please.’ Isn’t that a nice word? I had a lot of fun with it in Africa, learning just enough to communicate with people. Then one day, the tour crew brought me the Nairobi Times, like I could read the goddamned thing. They thought I was that proficient, but I wasn’t, of course. I had to con my way through it, like I’m enjoying my newspaper. But it was a nice experience.”

    Johnny Carson

  135. David Eddyshaw says

    Well, it’s easier than Navajo. So it is relatively easy.

  136. David Marjanović says

    a savage rebuttal by Hans Ditten from 1972

    The numerous typos make it look like Ditten was seriously angry when he wrote it. It doesn’t actually contain any argument, it just presents Korth’s “rather bold theses” – they’re bold indeed – and asks if the whole thing is “intended as a satirical play to entertain the readers”, helpfully noting that it’s at the end of the fascicle.

  137. ktschwarz says

    I’m glad Jongseong Park brought up the ‘war booty’ theory; I’d wanted to ask if anyone took that seriously, since it was in the Wiktionary entry for Σκλάβος when Victor Mair copied it, but was later deleted. (Mair cheerfully copy-pasted a few tertiary sources and called it a day, without noticing or caring that they contradicted each other.)

  138. David Marjanović says

    Mair’s reaction to contradictions is excitement: “The debate begins.”

  139. Either that or deletion.

  140. David Eddyshaw says

    zyxt should have a few words with the Italians about how they go round insensitively saying “ciao” all the time.

    I would think that this is the fault of that pretty young thing Giorgia Meloni. (She has been led into bad habits by associating with disreputable elder men. It would never have happened of she had fulfilled her God-given role of staying at home to produce more of those much-needed Italian babies.)

  141. In the biblical genealogies, ethnonyms have been reinterpreted as the names of progenitors. Surely there is a name for this phenomenon.

    Did Israelites of that time think of כְּנַעֲנִי kǝnaʿănî as simply the ethnonym ‘Canaanite’, or as ‘descendant of (the ancestor) Canaan’?

  142. DE, I remind that production of babies is the role of both sexes.

    And all should stay at home and produce more babies.

  143. David Eddyshaw says


    That’s very much what I’m wondering.

    I think that “Canaan” is not actually used by itself as a place name in the Bible. Come to that, when “Israel” is used collectively, it refers to the supposed descendants of Israel/Jacob, not to the “Land of Israel.”

    “Hear, O Israel” is clearly not addressed to the landscape. It can’t hear you: it hasn’t any ears.

    Cities and the like certainly get personified in poetry in the Bible, but that doesn’t really help either way. (E.g. Ezekiel 26, where Tyre is in for it; addressed as 2nd singular feminine, as befits a city.)

    This issue must surely have provoked reams of commentary previously, if one knows where to look.

    And all should stay at home and produce more babies

    Ideally, yes: but then who will deliver the Amazon parcels?

  144. but then who will deliver the Amazon parcels?


  145. Definitely, the Kingdom of Israel and its offshoot, the Kingdom of Judah, are named after a progenitor with a human name; whereas Canaan, Cush, Javan (= Ionians) are foreign ethnonyms, personified wholesale.

    who will deliver the Amazon parcels?

    The selfsame stork will be sub-contracted.

  146. DE, well, my comment about Slavic names is meant to clarify why I object to

    …the cognates of the Slavonic suffix *-ěn- in various Slavonic languages occur almost exclusively in place-name derivatives.

    Whoever wrote this implies that semantics of *-ěn- itself is narrow and does not allow it to work with anything but places. Taking literally this makes no sense to me: to say if it works with abstract qualities you need names based on abstract qualities.

    When all names are like Pobarane (presumably Bobrane) ALL you can say is “most demonyms are derived from places”.

  147. David Eddyshaw says

    Ah. So I was supporting your argument all along. Such fiendish ingenuity!

  148. DE, one possibility is simply that we should pay more to people who deliver parcels.

    Another option is that young people do that.
    Either both boys and girls, or only boys do that and girls don’t do that and then boys buy presents for girls. This is called Traditional Society. They everyone gets married and starts making babies.

  149. David Eddyshaw says

    whereas Canaan, Cush, Javan (= Ionians) are foreign ethnonyms, personified wholesale

    Well, you know that, and I know that, but did the actual writers think of those personifications as mere literary tropes, or as quite literal? The very existence of such aetiological myths suggests that contemporaries found “explanations” of that kind completely plausible, after all. Even necessary.

    I think you’d have to look at the actual grammatical usage in prose passages to see whether such names can function as locative without further ado, are whether they can only do so with a preceding construct “land of ..”, “city of …” and so forth. Somebody must surely have done that already.

    Possibly one might even find that usage changed over time within the Bible text itself, though unfortunately that would lead you down the rabbit-hole of trying to date the different parts of the Bible in ways that didn’t just lead to circular arguments … (we’ve been there, done that on LH …)

  150. David Eddyshaw says

    מִצְרַיִם “Egypt” might be a good test case (or counterexample), what with being formally dual and all. Though even there, Mizraim gets to be a son of Ham … there is no escape …

    Lots of occurrences to look at the grammar of, though. And, lo and behold, Exodus 13:3 (first case I thought of) has the name “bare.” So, yes … no further epicycles will be constructed.

    [On the other hand, it has been proven by science that the entire Torah was written by Greek-speaking Jews in Alexandria, as we recently learnt here … clearly the entire passage is calqued on Greek … can I have my medication now, please, Nurse?]

  151. PlasticPaddy says

    When are we talking about? Nomads don’t have places, they have ancestors, gods or totems. They gather to get drunk and have faction fights at particular times and insult one another’s ancestors and praise their own ancestors.

    This lists 17 tribes, the names are
    Darini-worshipper of a god or divine ancestor Dáire Barrach (see Brigantes)
    Voluntii = wolf people?
    Epdani = horse people
    Cauci = Gaelicised German pirates?
    Manapii = more Gaelicised pirates retrofitted with a divine ancestor called super-foreign Manach
    Coriondi= not a real tribal name, just means “army”
    Brigantes = more Darini (or would-be Darini) descended from Dáire Barrach
    Usdiae = the beyond the beyond, so sort of a place
    Iverni = the yewpeople or yeomen
    Vellabori = the braggarts
    Luceni = Lughs people (this is Orosius, not Ptolemy)
    Galengani = descended from Cormac Galleng
    Auteini = children of Ith (a god)
    Nagnatae = real name not known to Ptolemy
    Erpeditani = oak people or glorious people
    Vennicnii = the family
    Robogdii = the wreckers

  152. David Eddyshaw says

    Well, yes: that’s my overall (somewhat vague) hypothesis: ethnonyms based on toponyms came about when people founded cities. (This hypothesis actually has the useful property that, if sufficient care is taken with the wording, it can probably be made effectively tautological without obviously seeming to be so.)

  153. David Marjanović says

    production of babies is the role of both sexes

    Meloni’s party is unusual among extreme right-wing parties, even within Italy, in not being for Putin. In all other respects it is depressingly predictable.

    This lists 17 tribes, the names are

    …What are all these p doing in Ireland!?!

  154. Per the linked article, the names with p were of a Continental Celtic or Germanic origin.

  155. “Nomads don’t have places…”
    “when people founded cities. …”

    Let’s add two things. First names of Slavic entities based on rivers. I don’t know (and no one does) to what extent this implies a sense of identity or a polity, but they are neither nomads nor people from cities (here I remind of the Middle Eastern dialectological tri-partite system: urban, rural, nomadic [dialects] – something similar you find in Central Asia)

    And secondly, in the North Africa a city of some clan founded by a certain Tunisian man who once married a certain Italian woman is something quite unremarkable. And I think both the man and the woman are perfectly historical.

  156. David Eddyshaw says

    Well, there’s a certain Italian city supposedly founded by the descendants of of a man who brutally jilted a Tunisian woman. (She is supposed to have taken it very badly.)

  157. “Gaelicised German pirates? …. more Gaelicised pirates retrofitted with a divine ancestor called super-foreign Manach”

    As a random association: when I tried to read books from Hugo and Nebula list, I read Arab fantasy by Saladin Ahmed (just because it is about the ME). Well, it is fantasy, and it is about the ME.

    One of this year’s nominations is “The Adventures of Amina al-Sirafi”, about a female (Persian?) pirate where some of the action takes place in Socotra. I would definitely read it if it were about a modern female Somali pirate (but not Houthi, this would be too politicised):) But no, it is 12 century – when Socotra was still Christian.
    I think I won’t read it, but note it here, because Socotra.
    The author is a lady who converted to Islam and wanted to study the ME history.

  158. Might Perrott Snoxall Polome have been related to Edgar C. Polomé, one of whose interests was Bantu languages in general and Swahili in particular (

  159. David Eddyshaw says
  160. Just noticed that the funny etymology of Slav (Gothic slawan “to be mute”) is not mentioned in the post.

  161. jack morava says
  162. >Slavic entities based on rivers.

    This doesn’t change things, since we know most ancient rivers were named for people, in a strange way:

  163. David Marjanović says

    Gothic slawan “to be mute”

    Unfortunately not related to northern German schlau “smart, cunning”.

  164. A friend is an EU academic (literally, at an EU-sponsored think tank) from the US and early on tried to understand the German being spoken at her by asking people to speak schlaue.

    She got quzzical looks several times before someone realized she must be thinking it meant slowly and set her straight.

    I may be missing an adverb suffix. I don’t speak German and heard the story years ago. But the image of her asking people to speak more cunningly has stuck with me.

  165. David Eddyshaw says

    I once inadvertently offended a Hausa lady by saying of her son Yana da wayo, intending “He is clever.” Unfortunately wayo has more of the associations of “cunning” then “cleverness.”

    (Actually, what I said was in fact a pretty valid assessment of the lad. But I hadn’t meant to say as much to his mother …)

  166. @Ryan: German has no separate suffix for adverbs, but uses the same form both as adverb and as predicate; that English distinguishes both is one of the things we have to learn and which can trip us up. German employs the basic, citation form for both uses, so in this case that would be schlau sprechen.

  167. From Hausa zango “camping-place.” I’ve wondered if that might somehow be the origin of the WOV name for the Hausa

    Thank you for your full answers, David Eddyshaw.


    I wonder what the meaning of that is in Fulɓe. Not an alteration of an Arabic name?

  168. David Eddyshaw says

    I don’t know much about Fulɓe onomastics. They have a lot of Arabic-derived names, as you’d expect, but also plenty of pre-Islamic ones still.

    Jam means “well, healthy”, but I’ve no idea whether that is anything but coincidence. (The autonym “Jamsay”, used by the largest single Dogon group, is actually from Fulfulde “just health”, from its use as a greeting: same principle as “Farefare”, the name used for the Gurense by their neighbours in Ghana. Pokemon exonyms.)

    It would be nice to derive Kusaal Zangbɛog and its relatives from Hausa zango. The semantics works fine, and the labial-velar is not a big deal (WOV languages borrow labialised velars as labial-velars, and have no velar/labial-velar contrast before rounded vowels.) But I can think of no parallel morphological development with a loanword, and the tones are wrong (zango is HL, whereas Zangbɛog is LL.)

  169. David Eddyshaw says

    There’s a paper on Fulɓe names by Mariame Sy, but it doesn’t seem to be accessible.

  170. “that English distinguishes both”

    Is strange actually.
    I mean, English so comfortably treats nouns as verbs as nouns – and yet.

  171. @Hans,

    Yes, except in this case the American academic probably did say “schlauer sprechen” since she was apparently trying to say “speak slower”.

  172. David Eddyshaw says


    On Hausa zango versus Kusaal/WOV Zangbɛog: I wonder whether one could derive the WOV forms from a Hausa derivative *Bazangwe, plural *Zangwawa “Zongo person”? (The gw when an unrounded vowel replaces o is entirely in line with how Hausa operates.)

    Compare Bahaushe/Hausawa “Hausa person” and Bakatsine/Katsinawa “Katsina person”, Ba’amirke/Amirkawa “American.” The formation also appears with common nouns as base, e,g, bafaɗe/faɗawa “courtier” and baƙauye/ƙauyawa “villager.”

    I can’t actually confirm the existence of Bazangwe/Zangwawa as a genuine Hausa form, but the derivational process involved is clearly productive in Hausa.

    Adding class suffixes to a borrowed stem *Zangwe:- would be very much in line with WOV borrowing strategies (compare e.g. Kusaal lɔmbɔn’ɔg “garden”, from Hausa lambu.) Borrowing Hausa gw as gb is normal (cf Kusaal bakpae “week” from Hausa bakwai “seven”) and ignoring the prefixed ba- would not be unprecedented either, especially as it’s dropped in the Hausa plural.

    Kusaal avoids the usual “human” singular noun class assignment with stems ending in vowels. So you might have expected the word to end up like “Mossi”, “Kusaasi”, with the ga/se class suffix pairing/”gender”, as Zangbɛɛg, plural Zangbɛɛs. However, there are a number of stems in final ɛɛ which unexpectedly do take the singular suffix -go instead of -ga, probably reflecting the fact that the vowel has historically been monophthongised from *ew: a very common one turns up in Kusaal pɛ’og pl pɛ’ɛs “sheep.” So even the unexpected class membership can probably be explained away without adding the (unlikely) assumption that WOV speakers had a particularly negative opinion of the Hausa which led to them using a pejorative form.

    The weakest spot in this hypothesis is that the presumed Hausa original is not attested in any materials I’ve got; on the other hand, it definitely is a possible formation in Hausa, and the relevant derivational process is clearly very productive.

Speak Your Mind