Translating Smartphone Technology.

A short but interesting Economist piece about efforts to create technological terms for speakers of smaller languages:

Ousmane sweats under a tin roof as he thumbs through a Chinese smartphone that he is selling at the technology market in Bamako, Mali. Words in French, Mali’s official language, scroll down the screen. “A ka nyi?” (Is it good?) a customer asks him in Bambara, Mali’s most widely used tongue.

Mozilla, the foundation behind Firefox, an open-source web browser, wants Ousmane’s customers to have the option of a device that speaks their language. Smartphones with its operating system (OS) are already on sale in 24 countries, including Bangladesh, India, Indonesia and Mexico, for as little as $33. Other countries will be added as it makes more deals with handset manufacturers. And Bambara is one of dozens of languages into which volunteer “localisers” are translating the OS. […]

Ibrahima Sarr, a Senegalese coder, led the translation of Firefox into Fulah, which is spoken by 20m people from Senegal to Nigeria. “Crash” became hookii (a cow falling over but not dying); “timeout” became a honaama (your fish has got away). “Aspect ratio” became jeendondiral, a rebuke from elders when a fishing net is wrongly woven. In Malawi’s Chichewa language, which has 10m speakers, “cached pages” became mfutso wa tsamba, or bits of leftover food. The windowless houses of the 440,000 speakers of Zapotec, a family of indigenous languages in Mexico, meant that computer “windows” became “eyes”. […]

As a non-profit, Mozilla can put effort into languages that offer no prospect of a quick return. Songhai and Fulah, recently made available in Firefox, are spoken mainly by poor, illiterate herders and farmers in the Sahel, who do not have smartphones. But when such people eventually get online, they will benefit more if they can do so in their own tongues.

As more languages are added, the Firefox OS will create a sort of global Rosetta stone. It uses all parts of speech, and older, colourful words are pressed into service. Mozilla has created a statistical tool for linguistic analyses. And though 40,000 words is not a whole vocabulary, it is a significant part. As well as bringing the linguistically excluded online, localisation may keep small languages alive.

Incidentally, what they call Fulah is also known as Fula, Fulani, Fulfulde, Pulaar, and Peul. With some 25 million speakers, I wonder if it’s the most-spoken language for which there is no settled term?


  1. David Eddyshaw says

    I met someone in the nineties who did this for Mampruli (one of the dozens of languages of Northern Ghana.)
    Mind you, that was as a joke at the time. He himself had the only laptop I ever saw in the region, and in those days there were no telephones locally, let alone internet connexions.

  2. David Marjanović says

    40,000 words?!? If nothing else, the corpus linguists will be all over it in no time!

  3. I wouldn’t say it’s exactly unsettled. The Pulaar/Fulfulde split is dialectal, and the other names are exonyms.

  4. George Gibbard says

    Also note Pular (with a short vowel) in Guinée. In Sudan Fulbe are called Fallata or Takarir. The latter is the plural of Takruri, which is the same word as “Toucouleur” in Senegal. Attested variants are Takarin and Takarna.

  5. I found the piece interesting too, but was nonplussed by the reference to “languages shaped by livestock, farming and fishing.” That’s all languages, surely?

  6. Mildly OT: About five years ago I was asked by an American colleague to help him with the Hebrew database for the Android O/S predictive spelling utility. He had been approached by the Android team because of his earlier work with other language corpora, but knowing not a word of Hebrew he turned to me.

    He had collected several gazillion Hebrew words (OK, maybe only a few million) from the web and then run them through the MS Office spellchecker. It accepted most, but rejected perhaps 10,000. For example, Word rejected individual letters — but Hebrew can use letters to designate six of the seven days of the week, and of course they’re used as abbreviations of peoples’ names. Others were legitimate but uncommon words and neologisms that Microsoft hadn’t yet incorporated into the Office spellchecker.

    A keen eye was needed to examine these rejects and determine whether they should be included in the database. I turned to an Israeli colleague, then the recently retired editor for many years of the internal magazine of the Israeli police, a fellow who had also taught Hebrew in the U.S. for a few years. He spent a few days going over the list, pitched the misspellings and typos, and Android’s your uncle.

  7. Oh yeah, I forgot about Toucouleur, that should be on the list too.

    I wouldn’t say it’s exactly unsettled. The Pulaar/Fulfulde split is dialectal, and the other names are exonyms.

    Dialectal shmialectal. Lots of languages have dialects; that doesn’t stop there from being a standard name for the language. It’s easy-peasy to say “Do you speak French?” or “…Russian?” or “…Hausa?” But if you want to ask about this language, you have to pick one of half a dozen names and hope for the best. I call that unsettled.

  8. —In Malawi’s Chichewa language, which has 10m speakers, “cached pages” became mfutso wa tsamba, or bits of leftover food.

    Etymology of cache in English:

    cache (n.)1797, “hiding place,” from French Canadian trappers’ slang, “hiding place for stores” (1660s),

    Close enough

  9. Well, it’s easy to say “Do you speak Wu?” But the person you ask will stare at you incomprehendingly, even if they do, unless they happen to be a linguist. There are 80 million speakers of Wu, but they aren’t conscious of it.

  10. David Marjanović says


    This dialect family (and especially Southern Wu) is well-known among linguists and sinologists as being one of the most internally diverse among the spoken Chinese language families with very little mutual intelligibility among varieties within the family. […] some Wu varieties like Wenzhounese have gained notoriety for their incomprehensibility to both Wu and non-Wu speakers alike, so much so that Wenzhounese was used during the Second World War to avoid Japanese interception.


    Reputation for Eccentricity

    Due to its high degree of eccentricity, the language is reputed to have been used during the Second Sino-Japanese War during wartime communication and in Sino-Vietnamese War for programming military cipher(code)[5][6][7][8] Due to its unique grammar, vocabulary, and pronunciation, the language is basically impossible for any non-local to understand.

    There is a common “fearless” rhymed saying in China that reflect this comprehension difficulty: “Fear not the Heavens, fear not the Earth, but fear the Wenzhou man speaking Wenzhounese.” (天不怕,地不怕,就怕温州人说温州话)

  11. There’s no settled term for Spanish. It’s both Español and Castellano. There’s no settled term for Serbo-Croatian either: it’s Serbian, Croatian and Bosnian depending the location.

  12. Ladino, Spanyol (Spanyolit in Israel), Judesmo…

  13. By any honest standard, Wu cannot be considered one single language. England-style or Iberia-style mutual intelligibility is usually only possible somewhere between the level of 片 and 小片 in Chinese dialectology. Now, of 片 there are six, and of 小片 a dozen.

  14. There’s no settled term for Spanish. It’s both Español and Castellano.

    There is a settled term in English, which is what I’m talking about. And in Spanish, there is a settled term in any given country.

  15. England-style or Iberia-style mutual intelligibility

    Those things weren’t available in England or Iberia either before modern education, and the only education in Wuland has been in either Classical Chinese or Mandarin, ever. If that’s the criterion of languagehood, there are huge areas where there are no languages spoken at allses.

  16. in Spanish, there is a settled term [for Spanish] in any given country

    Wikipedia shows that the name is far from settled in Spain itself, with castellano popular among most regionalists but español popular among most nationalists and some regionalists. And despite official uses of one name or the other in Latin American countries, the popular usage does not always agree.

  17. Fair enough, but that still doesn’t cause the kind of problem I’m talking about, because everybody understands both terms and it’s just a matter of which one is used on a given occasion. Few English-speakers have even heard of The Language Known to Wikipedia as Fula, and it would be a lot easier to talk about it if it had a settled name. Come to think of it, I guess the name Wikipedia uses is likely to become as much of a settled name as it has; chalk one up for the cultural influence of Wikipedia!

  18. Fula is a Manding name for the language and Fulani is a Hausa term.

    In the name of political correctness, we should use one of the names used by the speakers themselves – which is either Fulfulde or Pulaar (sometimes spelled as Pular)

    Fulfulde Wikipedia gives preference to Fulfulde:

    “Fulfulde walla pulaar maa pular ko ɗemngal Fulɓe”

    But I foresee a potential problem here, since Fulfulde and Pulaar are names for different dialects of this language, so by choosing Fulfulde over Pulaar we are risking to offend speakers of the latter dialect.

    Very difficult case

  19. “And in Spanish, there is a settled term in any given country.”

    Castellano is used in Latin America in everyday speech all the time. Along with Español, naturally. They seem utterly interchangeable there. In Spain itself I’d expect them to carry slightly different political political implications, while still describing the same language.

  20. Songhai and Fulah, recently made available in Firefox, are spoken mainly by poor, illiterate herders and farmers in the Sahel, who do not have smartphones. But when such people eventually get online, they will benefit more if they can do so in their own tongues.

    (Emphasis added.)

    In my extensive experience with illiterate smartphone users, the quality of translation of error messages isn’t their most pressing concern. (My children routinely change the language of various apps to Russian or Chinese and then ask me what it says. “Shut up and herd your camels”, I tell them.)

  21. @Glossy:

    Castellano is used in Latin America in everyday speech all the time. Along with Español, naturally.

    I second this point; the names are used interchangeably, at least in Argentina and Uruguay, and a quick look at CREA (damn its clunky interface and lack of export capabilities!) suggests that’s the case elsewhere. It’s only in Spain that the distinction reflects the speaker’s politics.

    (Minor nitpick: language names, like demonyms in general, do not take an initial capital in Spanish.)

  22. Lars Mathiesen (he/him/his) says

    The way things are working right now, translating error messages is both good and bad. If my herding app tells me that Din kamel mangler vand, it’s useful. But if it’s the solar panel control app that does it, I need to google Your camel is out of water to get useful hits.

    (Joking aside, setting the locale to some kind of English is routine in my line of work for getting searchable error messages from programs. Or even to get intelligible ones if the localization was done by someone lacking in domain knowledge. If the programmer is Hungarian and only the Hungarian messages are useful, I’m up the creek).

  23. Stu Clayton says

    It is a truth universally acknowledged, that error messages are misleading at best, and frustrating in general. No matter in what language they are conveyed, the user is at a loss as to what to do. That’s because the people who wrote the messages also wrote the software.

    I am currently pondering an error message masquerading as a helpful hint. In my phone Chrome since yesterday is a little red icon next to the text “update available”. I press it, and land in Google Play for Chrome. There are two buttons there: “uninstall” and “open”. Pushing “open” takes me back to Chrome. If I pushed “uninstall”, I’m afraid my bookmarks will be lost. There is no information about backup.

  24. The latter is the plural of Takruri…” – and what is the etymology of “Takrur”?

  25. Good question. Wikipedia doesn’t attempt an etymology; French Wikipedia has an extraordinarily unconvincing one:

    Selon Niang, l’origine comme la formation du terme “Tekrur” ou “Takrur” associe en réalité les noms des deux plus anciennes provinces du Fuuta : « Law » + « Toor », car, de leur association, on obtient : « Takroor » / Tak-roor /, avec la forme « Tak » qui est variante de « Law », tandis que « roor » est la base qui évoque le nom de l’ancien royaume « Tooro » (Toor-o) .

  26. WIktionary تكروني,
    1. (dated) Having the characteristics of a تكروني‎. 2. (ethnic slur, by extension) Extremely black.
    1. (dated) A pilgrim of Mecca of a West African ancestry, especially when becomes naturalized or a permanent resident. 2. (ethnic slur, by extension) An extremely black person.

    ‘Umar Al-Naqar, Takrur the History of a Name, The Journal of African History, Vol. 10, No. 3 (1969), pp. 365-374, (, pdf on sci-hub):

    Delafosse suggested, first, that with the expansion of the Fulani from Futa to Darfur, all this region became known to the Arabs as Takrūr: ‘par suite d’une extension de sens facile a comprendre, a toute cette vaste region ou il avait introduit avec lui la langue Tekrourienne.’ [21] In another place, he suggested that the name was given ‘specialement a l’ensemble des pays soudanais conquis ou organises par des princes originaires du Fouta’. [22] Arkell suggested a common root in the names of Kuri, of Koara (an old name for the Niger), of the oasis of Kawar, near Bilma, and of the name Takrūri, ‘which still today at Mecca is practically equivalent to Sudanese’. This common root, taken from Berber languages, usually took the form Kuri (plural Kawar). In Arabic Kawar is thus the exact equivalent of the word ‘Sudan’. [23] Burckhardt suggested that Takrūri was derived from the Arabic ‘takarar, meaning to purify, i.e. their faith, by pilgrimage etc.’ [24]

    A major shortcoming in the above explanations seems to be their common failure to recognize the fact that the term ‘Bilad al-Takrūr’ was essentially a popular concept of the Middle East. The primary requisite for the popularization of the name Takrūri was not knowledge of the history of the state of Takrūr or of the Islamic exploits of its rulers and people: these could anyway have been available only to the very few. It was rather
    the fact that the Takarīr themselves were seen in sufficient numbers in Middle Eastern countries to attract attention.

    Some early references suggest that a number of Takarīr had found their way to the Middle East, where they took up residence. …

    [21] M. Delafosse. op. cit., I, 234.
    [22] M. Delafosse and O. Houdas’s translation of Tarikh al-fettash of Maḥmūd Kaʿti (1964), p. ii and note.
    [23] A. J. Arkell, ‘The history of Darfur, 1200-1700 A.D.’, Sudan Notes and Records, XXXII, pt. I (1957), 55.
    [24] J. L. Burckhardt, Travels in Nubia (1819), 404

    op. cit. : M. Delafosse, Haut-Sénégal-Niger, II (Paris, 1912), 353 and note, and article ‘Takrūr’, Encyclopaedia of Islam (Houtsma). The site of Takrur is taken to be near the present position of Podor.

  27. Koara the old name for the Niger – I didn’t know.


    … When in 1324 Mansa Mūsa of Mali made his celebrated pilgrimage, al-ʿUmari was prompted to make the cautionary remark that the king of Mali was angered by the appellation of ‘Malik al-Takrūr‘, because Takrūr was only one of the provinces of his vast empire.[28] It is possible that the name Takarīr first came into use as the comprehensive name for all West Africans not in Egypt, but in the Hijaz.

    By the eleventh century the world of Islam had incorporated peoples from three continents whose differences were linguistic, ethnic and cultural. During the pilgrimage season in the Hijaz, one spoke in terms of heterogeneous groups like the Syrian, Egyptian, Maghribian Ḥajj caravans and so on, which all collected a diversity of peoples en route to Mecca. The cities of Mecca and Madina, where many Muslims take up permanent residence in mujawara, have become a microcosm of the wider world of Islam. Their resident population is consequently classified under large but convenient groupings. Under the name Jawah, according to C. Snouck-Hurgronje, ‘are included in Arabia all the people of Malay race, in the fullest meaning of the term; the geographical boundary is perhaps from Siam and Malacca to New Guinea’.[29] The name Shanaqiṭ[30] (singular Shinqiṭi), no less well known than Takarīr and used to refer to people from the Western Sahara, is another instance and, possibly, a clue to the origin of the name Takarīr. Mauritanian traditions explain the relationship between the town of Shinqiṭ, the Middle Eastern generic Shanaqiṭ, and the pilgrimage in the following manner: ‘Pilgrimage caravans used to depart annually from the town of Shinqiṭ. All those from the outlying districts who wanted to make the pilgrimage went with this caravan. Thus the people from all these regions-I mean from al-Saghia al-Ḥamra to the Sudan-if seen in the East are known only as Shanaqiṭah [Shanaqiṭ]’.[31]

    Although our earliest records of pilgrimage from West Africa do not include Takrūr,[32] it seems none the less plausible to suggest that the earliest West African pilgrims, given the facts about Muslim Takrfir and its people, may have been Takarīr….

    [28] S. al-Munajid, Mamlakat Mali ‘ind al-jughrafiin al-JMuslimin, I, texts (Beirut, 1963), 44. C. Gaudefroy-Demombynes, op. cit. (translation of Masalik), 53-4.
    [29] C. Snouck-Hurgronje, Mekka in the Latter Part of the Nineteenth Century, trans. I. H. Monhan (1931), 215.
    [30] Also spelt Shinjiṭi (plural Shanajiṭ or Shanajiṭah).
    [31] Ahmed ibn al-Amin al-Shinqiṭi, al-Wasiṭ fi tarajim udabā Shinqiṭ (Cairo, 1911), 413. Also see H. T. Norris, ‘The History of Shinjit according to the Idaw ‘Ali Tradition’, Bulletin de l’I.F.A.N., I, sér. B., nos. 3-4 (1962), and by the same author, Shinqiti Folk Literature and Song (1968), 3. Mr Norris gives alternative dates for the foundation of the town of Shinqiṭ, the earliest of which and the most acceptable is ca 1300. On the Shinqiṭi pilgrimage caravan itself there is remarkably little information. It is possible to suggest, from references in al-Wasiṭ supported by oral information obtained from present-day Mauritanian scholars, that such organization may have developed in the seventeenth, eighteenth and nineteenth centuries and was necessitated by increasing insecurity of the west Saharan routes.

  28. David Eddyshaw says

    I imagine that the word is related to

    However, Tokolor seems to be an exonym, and its own origin seems to be equally obscure.

  29. Yes, obcure: “The vision of Almamy Abdul Qadir did persist, however, as the ideal of a Muslim leader and state. It also emerged in the word Tukulor, as a way of referring to the inhabitants of the middle valley, or at least many of them, who made Islamic culture into a kind of badge of ethnic identity. At least, so thought the Wolof, and the French after them.” (D. Robinson, Encyclopaedia of Islam 2)

    (that’s of course everyone, including D.R. derives it from Takrūr, but…)

  30. David Eddyshaw says

    Spencer Trimingham’s Islam in West Africa (from 1959, but still very useful) says

    The term Tukolor (or (Tokoror) is used by Wolof for those Negroes with whom they are in contact who speak the same language as the Fulbe. Arab geographers adopted the word under the form Takrūr from which they formed the ethnic Takrūrī (pl. Takārīr), a word which came to designate any west Sudan Negro. They have no name for themselves, for they are not so much a people as a hierarchized class society. When pressed, they call themselves Hal-pulāren “Pular-speakers” or Fūtankōbē (sing. Fūtankē) from their centre in Senagalese Futa.

  31. « Tak » qui est variante de « Law », tandis que « roor » est la base qui évoque le nom de l’ancien royaume « Tooro » – Sounds suspicious indeed. IF Tak is attested in the same meaning, one can of course imagine taktur~takrur.
    I seriously understimated Mauretanian phonology. The map shows the region of Trarza and the city of Mederdra on the other bank of the Senegal-река. I know adrar (which happens to be a name of another Mauretanian region), but… (but it is accidental: other toponyms are less crory-drory)
    It reminds me what I felt when I first saw the medieval map of the FYLOSC area.
    Нерентия/Неретвия/Погания, Захумле, Травуния и Дукля. And Рашка.

    To a Russian it is a collection of fairy-tale names (though Захумле reminscent of translated tales: (1) the troll Komle (Кумле) from the Russian translation of Trollkrittet who impressed my exaclty because of the peculiar sound of -umle (2) LotR, in one of translations, Зусумки) that do NOT sound as names from our fairy-tales. Rather like names a modern Russian would invent for a fairy tale…

  32. David Eddyshaw says

    I suppose the thing to do would be to see if “Tukolor” has an obvious explanation within Wolof.
    If it did, that would probably settle the issue, though if it was farfetched-sounding you’d still have to wonder about folk etymology.

    In the Oti-Volta area, practically no group seems to have an ethnonym that actually means anything in any known language. (Unlike place names, the great majority of which are pretty transparent.) It’s not at all like all those New World and Siberian folk who just call themselves the Real People and have done with it.

    Not sure if the Wolof were as far north as that at the time that Arab geographers would have first appeared on the scene, though. Maybe a Mande origin?

  33. But why such a concentration of rhotics in the region? (Trarza, Mederdra)

    And where does the apostrophe in R’Kiz comes from, Klingon?

  34. And where does the apostrophe in R’Kiz come from

    Just throwing this out there, to jumpstart the conversation… Adrian Room, African Placenames (2008) p. 160, says this:

    Rkîz (village, southwestern Mauritania) The village takes its name from the lake here. Its own name means “end,” “extremity,” as all of the local rivers flow into it.

    But he doesn’t say in what language… Fulɓe? Wolof? Soninke? However, it sort of reminds me of the Arabic root rkz, having to do with fixing (something) in the ground or in a certain place, fixing one’s place of abode, as in the noun of place مركز markaz ‘center’. But more specifically, some derivatives of rkz have to do with gold and silver placed in the earth by God and wealth buried in the earth by men: ركاز rikāz ‘vein or lump of gold or silver; precious ore’; ركيزة rakīza ‘vein of gold, silver, or other metal’, etc. I have no idea how current these words are in these senses in Hassaniya Arabic. Was gold panned from the sediments of Lake Rkiz and the streams that feed the lake and connect it to the Senegal River (the ‘River of Gold’)?

  35. I’m fairly sure Rkiz is Arabic; I think Arabic is the only language in the region that even allows final -z. (Except maybe Zenaga, and this doesn’t look like Zenaga.) The apostrophe comes from the same place as the one in N’Djaména (an Arabic toponym, btw) and N’Guigmi: a French surveyor’s rear end reluctance to transcribe a sonorant-initial onset as such.

    Trarza looks like a broken plural of Terrāzi or something like that; I should look it up. I assume Mederdra is “scattered” or the like.

  36. Logically, the area was populated by Berber speakers (and possibly some other languages) before Char_Bouba_war. But I don’t have Taine-Cheikh’s Zenaga dictionary:(


  37. David Eddyshaw says

    I was wondering whether some Berber form might lie at the back of Takrur/Tukolor in some way, on the highly scientific grounds that ultimately-Berber loans seem to get in everywhere in West Africa.

  38. rkz̩

    “Be muddy (undrinkable)” seems very promising as a starting point!

  39. Tacoronte is a toponym of Guanche origin, believed to be derived from Tagoror, meaning “place where the Council of Elders meets”.

  40. @DE, k-r-r is something you expect to find in Arabic and Berber dictionaries even without knowing them (and indeed the Tuareg dictionary quoted above has a page of krr-s). And both have t-.

    A folk etymologist has a generous choice of takrīr, “purification”, takrār, takarrur* “repetition” in Arabic (or t-a-kruri-t-t “ball” in Tuareg…)

    Many Russian tr/kr-words are like трещать, дребезжать, кричать, грохотать – sound production or violent action or violent sound production or sound production with a forceful action – and -r-r- is usually a prefix pro- or pere- attached to r- root: prorok “prophet”.

    Takrur does look Arabised and Berberised and if it is not Berber, then I wonder if there is a sound law where [something] in some language > r in Arabic/Berber (and a language where [something] is normal).

    This was my logic and then I saw Trarza and Mederdra. Having this said, of course Adrar is a normal Berber word for “mountain” and Mederdra does looks Arabic as Lameen said.

    * if one tries really hard, she will also remember Latin recurrō …

  41. N’Djaména

    There is a Tunisian village Takrouna (known from the collection of dialectal texts Textes arabes de Takroûna, presumably somewhat less important for locals than for pre-Internet Russian Arabists) claimed to be either Berber – or Andalusian, named after Takoronna in Muslim Spain.

    And there is Takoronna, variously claimed to be either a Berber name or a Berberised Roman name. Spelled as Takoronna and Ta Kurunna in Spanish. This spelling found its way in WP Takrouna (the Andalusian family mentioned in the WP article is Gmach (Gomez)).

    (it is when I was googling this Takoronna I saw Takoronte above)

  42. @Xerîb, yes, as I don’t have a dictionary, I tried Touareg at random, but semantically it is excellent. I also found the reference in African Placenames, but Google stubbornly was showing me preceding entries but not this one.
    I finally obtained the same text as you, but it took quite a while…

    @Lameen, thank you!

  43. January First-of-May says

    Rather like names a modern Russian would invent for a fairy tale…

    I agree, the names Захумле and (especially) Травуния sound like something I’d see in one of them fairy-tale-themed fantasy books – the likes of Profession Witch.

    Russian Wikipedia has the first place under Захумье, which makes it, if anything, sound more like one of those modern-fairy-tale-ish places.
    (“некоторые сведения по истории славянских племен Рашки, Дукли, Травунии, Захумья, Пагании и Боснии” – that familiar “Bosnia” at the end makes it sound like Gulliver’s Third.)

    I gather that it’s etymologically **Захолмье – essentially “over-the-Hill land”. (I’m capitalizing Hill because it’s referring to a specific mountain.)

    It’s not at all like all those New World and Siberian folk who just call themselves the Real People and have done with it.

    Also, reportedly, the Nuer (according to Harari’s Sapiens, anyway). I guess they had to do something simple.


    And where does the apostrophe in R’Kiz comes from, Klingon?

    I’ve been wondering about that! Apparently it does seem to be a French convention for writing sonorant onsets (previously on LH, kind of), but from what I’ve seen in geo-databases it feels like toponyms in Rk- and the like are in fact strongly concentrated in Morocco/Mauritania. Must be a Berber phonology thing.

  44. > The name Shanaqiṭ[30] (singular Shinqiṭi), no less well known than Takarīr and used to refer to people from the Western Sahara, is another instance and, possibly, a clue to the origin of the name Takarīr.

    Earlier discussiion of Shinqit and other related topics was here.

  45. Yes, I saw it in the the form Захумле in a map, but I think another possible Russification is Захолмие rather than Захумье. As for Погания instead of Пагания, this was my adaptation (Slavic pogan “pagan”).

    sonorant onsets

    Well, we have Рцы [rt͡sɨ]. The name of the letter R, now obsolete:) Should I write R’TSy?

    Earlier discussiion of Shinqit and other related topics was was here.

    I quoted the part about Mecca because I simply did not think about that before: Meccans need names for various parts of the world.

  46. Exporing Russian onsets further:

    Rvi “tear!” is unsportive, for /v/ (< /w/) is not a stop. So are:
    Ržatʹ “neigh”.
    Rža[včina] “rust” is one not associated with violent action or sound production but English “rust” (cf. rusp, lisp…) makes me doubt… Lust and list are innocent.

    Nnndaaa… and Mmmdaaa… mean “hmmm”. From da “yes”? Cf. spinžak, spindžak, colloquial < pidžak “jacket”.

    prorok “prophet”, Rcy:

    This root (rok/rek “say”) is onomatopoetic itself. From Wiktionary:

    Per Vasmer, cognate with Lithuanian rẽkti (“to shout, to roar”), 1sg. rėkiù, Lithuanian rékauti (“to shout, to roar”), 1sg. rékauju, Latvian rèkt (“to shout, to roar, to howl”), rȩ̃kôt (“to talk, to chat”), as well as Sanskrit रचयति (racáyati, “to work, to construct”), Gothic 𐍂𐌰𐌷𐌽𐌾𐌰𐌽 (rahnjan, “to calculate”), Tocharian B reki, Tocharian A rake (“speech, word”), Latin raccō (“to roar (like a tiger)”), Old Irish réimm (“shout”), from Proto-Indo-European *rekmen.
    Per Derksen, possibly cognate with Sanskrit रचयति (racáyati, “to work, to construct”) and Welsh rhegi (“curse”).

  47. Old Irish réimm (“shout”)

    A rare word of apparently uncertain meaning; eDIL: “n a shout, cry ? […] Possibly same as 2 réim(m), one of the feats of a `druth’ being shouting.” There’s a much more common word réim(m) with an interesting set of meanings: “`going’, movement, advance, course; course, manner, way; series, succession, list; an inflected case of a noun.”

  48. January First-of-May says

    sonorant onsets

    The closest I can think of to a sonorant-unvoiced stop onset Russian toponym is Ртищево, and it sounds weird even by Russian standards. The same onset also shows up in the paradigm of рот “mouth”, but nowhere else I can think of offhand.

  49. On the other hand, roaring like a tiger is a very appropriate thing to do for a prophet…

    The same onset also shows up in the paradigm of рот “mouth”,

    Wow! I did not think about it…
    Of course it is a new cluster (after the fall of short u’s and i’s), but yet.

  50. Ah, and rdet’ “to be red” – not unvoced, but a stop.

  51. PlasticPaddy says

    In that long list the Gothic reflex has meaning “calculate”, so perhaps (a) the “roar” and “list/calculate” words are unrelated homonyms or (b) the “roar” sense is derived from something like [ list > list grievances > give someone a “ticking off” > roar at someone ] or [list > recite from memory > declaim > roar ]

  52. Roar sounds as a good name for a dialect of Lisp…

  53. R’K

    For many years editors of Russian WP kept complaining that Khoisan names “violate the rules of Russian orthography” for “according to the rules of Russian orthography” words can’t contain signs like “!”.

    One of Russian editors works in the Institute of Linguistics (and is, accordingly, a linguist). He also writes articles about languages for the Great Russian Encyclopaedia. In Wikipedia he adds maps.

    He listened to the complains and…. published a proposal for a Khoisan orthography in GRE (noting that people complain at !’s). Then he added a link in WP to the system used in GRE. Just added a link (and noted it).
    When people who complained saw it, they were happy!

    Now every Russian will think that Khoisan languages sound like Abkhasian and will be able to pronounce – with great effort and numerous glottal stops – Khoisan words….
    Unfortunately those difficult sequences have Absolutely nothing to do with click consonants they designate.

  54. David Eddyshaw says

    Zulu orthography manages OK, by co-opting ordinary Latin-alphabet symbols not needed for non-click consonants and using a lot of digraphs and trigraphs:

  55. Speakers can use even logographic systems.
    Transcribing names for people who don’t know the language is a different story.

    I believe in the context of Yuri’s interests (cartography. The linguist in question is Yuri Koryakov) digraphs make sense, because when two generals are using a map they need to be able to say the name somehow.
    A Russian name would totally do the job.

    From the point of view of WP editor, we need a Russian name, but there is not an established Russian name. So we need some formalised way to create one.

  56. The need of being able to read it disappears in texts.
    I mean, well, we all can read ǃXóõ (in Yuri’s spelling Къхонг) somehow. /gzu/ or /ksoo/ or… whatever. But this ! draws our attention, maybe we say “/ksoo/ with an exclamation mark”. Maybe for generals it is worse than it is for readers of a text.

    Trigraphs act as fig leaves that hide the pudendum (click consonants). !’s draw attention to them.
    You either hide them, or you draw attention to them, the option “neither hide nor draw attention” does not exists until people get used to the offensive sight. And I believe it is a good idea of an article in encyclopaedia to exactly draw attention to unusual facts (click consonants in this case).

  57. “Западный къхонг



  58. David Marjanović says

    Abaza, rather. That’s the two-vowel language with heaps of consonants but only the Russian letter inventory (because it’s spoken on the Russian side of the mountains), so most consonants are written with di- or trigraphs.

  59. David Eddyshaw says

    The highly aberrant Indo-European language English represents its cross-linguistically rare phonemes* /θ ð/ by digraphs. Well, digraph; the traditional writing system is a rather makeshift adaption of a system originally devised for one of the more mainstream Indo-European languages.

    * Or “pudenda”, as they are called in Russian.

  60. Aren’t various Cyrillic digraphs used to represent different sounds in different languages? If so, wouldn’t an educated Russian not jump to conclusions about interpreting them?

    As in, many English speakers pronounce <xh> in both Xhosa and Hoxha as /ks/ or /ksh/. Others might be more cautious if they recognize that a foreign orthography might be odd.

  61. @DE, we arew accustomed to the offensive sight of interdentals. You can stick your tongue out in public and no one blushes.

  62. David Eddyshaw says

    According to Sander Steeman’s grammar of Sandawe, the Usandawe actually do avoid speaking their language in front of other Tanzanians because the click sounds are stigmatised.

  63. I honestly don’t know what will happen if two people will start speaking a langauge full of click consonants in Moscow. Maybe they will gather a crowd around (for this you need to make it clear that they don’t object: instincts make everyone pretend that they don’t notice anyone:)).

  64. @Y, yes, but people, educated or not, will still try imagine it. If you have never seen an airplane or a motorcycle and I tell you that X is a vehicle you imagine all sorts of horse-driven or donkey-driven or… vehicles.

    I think click consonants are beyond the limits of Europan imagination (or to a lesser extent tonal language for people unfamiliar with them). Cf. languages in sci-fiction series.

    When reading such Khoisan transcription you will
    (1) try to pronounce it
    (2) hope that your pronunciation approximates it
    (3) understand that it only approximates it very crudely.
    (4) believe that this language is more like languages of Caucasus, than, say, Chinese. (a “guttural” language:)) because those in Cyrillic transcription look very similar.

    Even though you do think (3), you still believe it is an unknown language that does not need a serious update to your ideas of language typology.

  65. Surely there were Zulu and Xhosa students in Moscow back in the day? Or did SA discourage those would-be communists from going there?

  66. PlasticPaddy says
  67. An article on the site of the Peoples’ Friendship University (walking distance from me…) for I don’t know what exact year says they have 139 students from SA (and that it is 5th largest group). “Medicine 65, veterinary medicine and agronomy 51, the rest study languages and foreign relations” But their database of alumni does not list this country for some reason.

    (a more stupid article for people who love the genre. I know no one loves it… But who and why writes them?:( “Students from SA sang to Lavrov about the motherland” it informs. ).

  68. David Eddyshaw says

    I’ve known some Ghanaian doctors who qualified from Moscow (mostly about my own age.)
    No clicks in Ghana, though.

  69. PlasticPaddy says

    I did not want to complicate this, but Kotane’s parents were L1 seTswana speakers (Bantu language with 3 click consonants, so still OK for drasvi).

  70. Actually I meant a language with high rate of clicking… I suspect their concentration (per second) is lower for Khosa or Tswana compared to Khoisan. Totally enough to make classmates curious, but maybe less likely to attract attention and curiousity of strangers.

  71. And I don’t know how many students from SA came to study here in Apartheid times.

Speak Your Mind