I seem never to have mentioned the Nigerian anti-Western group Boko Haram here, and that’s a good thing, because if I had I would have spread the usual story that Hausa boko is from English book, and that turns out to be mistaken, according to “The Etymology of Hausa boko” (pdf) by Paul Newman, according to Wikipedia “the world’s leading authority on the Hausa language of Nigeria and on the Chadic language family.” Newman points out that:
1) If English book had been the source, it would have been adopted in Hausa as something like [búukùu] (he gives examples of such words).
2) The word boko “has a related morphological form marked by reduplication, short final vowels, and a set low-low-high-low tone pattern, namely bòokò-bóokò ‘deceptive, fraudulent’ … This pattern is found in Hausa with various other words… This reduplicated construction is unproductive and limited to a small set of words, many of which are now obsolete, thereby indicating that boko must be an old Hausa word with considerable ancestry in the language and not a recent loanword.”
3) It occurs with the word biri ‘monkey’ “as part of a fixed compound biri-boko (lit. monkey-fraud). … That biri-boko is found in Bargery’s dictionary… is a good indication that the compound is of considerable age in the language and hardly a recent creation…”
4) The order of definitions in old dictionaries suggests the original sense was ‘fraud.’
5) “It is perhaps worth pointing out that boko in the sense of something western or secular tended not to be used as an independent noun, like English book (as is now often done), but was almost always used as a modifier.”
6) Finally, “it would have been curious indeed for Hausa to have borrowed the English word book (in the form boko) and have it come to represent despised Western education. In the first place Hausa has long had its own word for book (littafi), which was borrowed at a very much earlier period from Arabic. This word was already well established and fully integrated in the language at a time considerably prior to the British takeover of northern Nigeria and the opening of colonial government schools in Kano at the beginning of the 20th century.”

His conclusion: “Hausa boko does not mean ‘book’ and it is not derived etymologically from the English word book. The phonetic and orthographic similarity between the two is purely coincidental. They are what the French call ‘faux amis’ (‘false friends’).” I was particularly impressed with this frank acceptance of responsibility for the error:

This is not a matter of an occasional reporter or amateur linguist going astray. This is a systematic error that we professional Hausa specialists have perpetuated over the years and thus we deserve real blame for having provided other scholars and the general public with misleading information.

(Thanks for the link, Paul!)


  1. The Tumbleweed Farm says:

    Interesting. Now I wish he had also given the true etymology of Hausa “boko”!

    Incidentally, the existence of an earlier word for some concept in a given language does not necessarily prevent the appearance of a loanword from a Western language later on. For example, there are quite a few Western (English or Dutch) / Arabic doublets in Malay. Among them are “sekolah” (Western type school) along with “maktab” (Islamic school), as well as “buku” (book) along with “kitap” ([usually religious] book).

    Incidentally, in Malay reduplication is fully productive (this is how plurals are formed), thus “buku-buku” etc.

  2. Now I wish he had also given the true etymology of Hausa “boko”!

    He certainly would have if it were known. Alas, Chadic etymologies are not as thoroughly researched (or even researchable) as IE ones, for various reasons.

    Incidentally, the existence of an earlier word for some concept in a given language does not necessarily prevent the appearance of a loanword from a Western language later on

    No, of course not, which is why he doesn’t give that as a knock-down argument, just one of a number of suggestive factors.

  3. If the original meaning was “fraud,” and it was frequently use in compounds, it may still have gotten its current meaning as a punning denigration on the English word “book.”

  4. marie-lucie says:

    the existence of an earlier word for some concept in a given language does not necessarily prevent the appearance of a loanword from a Western language later on

    This is not limited to indigenous-Western interaction. As an example, in French the generic word for ‘cake’ is gâteau, except for English-type fruit cake, which is known in France as cake (pronounced “kék”). What happened was that English cake, which is as generic as French gâteau, was borrowed to refer to a specific type of cake then unknown in France. Similarly, the examples given above for Indonesia (school, etc) show that the borrowed words do not replace indigenous ones for traditional things and concepts, but apply specifically to the new, imported Western-type ones.

  5. David Marjanović says:

    I had the choice between going to bed (it’s 1:10 in the morning) and learning something.

    I learned something. 🙂

  6. The etymology of boko is certainly interesting in its own right, but as far as reportage of Boko Haram goes, a synchronous description is probably more important.

    So — do most Hausa speakers know enough English to know the word book, and if so, do they perceive boko to be a borrowing of it? I find Dr. Newman’s comment very suggestive, that whereas boko formerly “tended not to be used as an independent noun, like English book“, it now often is. Is this just another coincidence, or does it imply influence from book?

  7. As I said to Hat previously on this subject:

    In fact I got rather sidetracked by the linguistics of the word. What I wanted to know was how the one word boko could mean “western education”, which [Newman] makes clear in the abstract:

    Rather, boko is an indigenous Hausa word originally connoting sham, fraud, deceit, or lack of authenticity. When the British colonial government imposed secular schools in northern Nigeria at the beginning of the 20th century, boko was applied in a pejorative sense to this new system. By semantic extension, boko came to acquire its current meaning of Hausa written in Roman script and Western education in general.

    Hat enlightened me that Hausa was originally written in Arabic script.

  8. David Eddyshaw says:

    I used to live in Northern Nigeria and can at least say that most people even in the big cities, let alone the countryside, have too little English to maintain a conversation in the language, at most; this doesn’t altogether answer the question though. Compare the fact that most Westerners effectively know no Japanese at all but are still well aware of the source of words like kimono, sushi, harakiri …

    However I’ve always felt that the supposed equation of “book = Western education” was peculiar, given that literacy in both Arabic and Hausa was well established centuries before that great bad man Lugard conquered the Sultanate of Sokoto (so recently, after all, that when I lived in West Africa there were still people alive who were born before the conquest.)

    There is no sense in which “book” itself could be a negative word for Muslims of any sort. I suppose there could have been a contrast between a “littafi” (Muslim book) and an infidel Western “book”, a bit like the Yiddish distinction between a “seyfer” and a “bikh.” Only trouble with this theory is that in fact there isn’t any such distinction in Hausa.

  9. David Eddyshaw says:

    Having said that, one of the dreadful legacies that my fellow-countrymen left in Northern Nigeria was that they agreed with the emirs who were left in place under Lugard’s pet scheme of indirect rule (he destroyed the career of a junior who had the temerity to question this policy, by the way) that Christian missions were not to be permitted in their fiefdoms. As mission schools were then about the only way for a Nigerian to acquire a modern technical or clerical education, this meant that the North lagged grievously in these areas when it came to providing African civil servants at independence; the posts even in the North were frequently filled by Southerners, especially Igbo, and this was a major factor in the lethal sequence of event that lead up to the Biafra war.

    There is a lot of back story in the Nigerian context to this idea that Western education is intrinsically anti-Muslim.

  10. Very interesting; thanks for adding that perspective!

  11. CuConnacht says:

    Similar to French “cake” is French “dogue”, which specifically means a mastiff, not any old chien.

  12. David Eddyshaw says:

    Likewise, Inuit “iglu” means “house” rather than “snow house”; the Algonquian original of “wigwam” also meant “house” in general, and certainly not a tepee; and the Japanese word for “kimono” is not “kimono.”

    There ought to be a name for this process whereby a word with a broad meaning in the original language gets associated with a narrower meaning in a borrowing language based on real or mistaken notions about the original speakers. Maybe there is, but I can’t think of it …

  13. …and “salsa” and “sombrero” and sometimes “canoe”.
    There should be a name for it, indeed. How about ‘exoticization’? It kind of works, but it’s somewhat pejorative, and looks stuffy.

  14. And also misleading, since nobody’s trying to exoticize anything: people are just taking a word and using it for the object at hand rather than in the full semantic range of the original language. But I agree, it’s a common phenomenon that should have a name.

  15. Well, aren’t ‘an odd kind of house which the Inuit use’ or ‘a funny Mexican hat’ exoticizing? Or ‘canoe’, which can mean the particular Canadian kind, or in general ‘any kind of traditional-style boat, made by people not well-known by Europeans before 1500 or so’.

  16. Unlike the other examples, dogue is not a case of semantic narrowing in the borrowing language, but of semantic widening in the source language.

    Exoticizing is used in cultural studies for the process by which culture A is seen by culture B through a filter that hides everything except what is extremely unusual about A by the standards of B. Orientalism in the sense of Said is a special case of exoticizing.

  17. David Eddyshaw says:

    Reverting to the orginal topic, it doesn’t seem unlikely to me that a preexisting Hausa “boko” basically meaning “bogus” could get associated with the English word “book” as a cross-linguistic joke at the expense of Western-style education in the first instance. That sort of linguistic play is pretty common in West Africa – hardly odd in an area where knowing several languages well is the norm rather than the exception.

  18. The usual BrE definition of “canoe” falls between Y’s suggested senses. It encompasses the AmE canoe (“Indian” or “Canadian canoe”), kayaks, and outriggers, but not coracles or dugouts.

  19. Jonathan D says:

    The French gâteau-cake example is mirrored in British English, where gateau refers to a subset of cakes. I can’t say I have a firm grasp on which subset this is, but I feel they generally involve several layers and cream, and a Black Forest Cake is definitely included.

    I was familiar with the generic French meaning (used in English by family members with some knowledge of French, probably at first to achieve avoid children’s comprehension, not that that worked for long) and not the British meaning until I lived in London. Somehow this particular word took more getting used to than most other particularly British vocabulary, and I was quite gratified when I witnessed an entertainingly lively conversation between an Englishman and similarly surprised Mauritian, focusing not only on the meaning, but also the pronunciation /ɡætəʊ/.

  20. marie-lucie says:

    JD: gateau refers to a subset of cakes. …they generally involve several layers and cream, and a Black Forest Cake is definitely included.

    Although such cakes would be called by the generic gâteau, these are by no means typical of French cakes! At least French cake does refer to an English type of cake.

  21. Interesting to learn the true etymology of “boko”, as I had shared the “book” hypothesis with several friends, including fellow linguists. Although now I’m disappointed that I won’t get to open my Barnes & Noble-style bookstore chain in Nigeria, “Book Harem” 😉

  22. mollymooly, in older books at least, any boat of the Pacific or of South America may be called a canoe, from small river dugouts to enormous Polynesian sailboats.
    The OED says “Extended to those of other societies and other construction, and used generally for any roughly-made craft used by American Indians, Malayo-Polynesians, etc.; most of these use paddles instead of oars, whence ‘canoe’ is sometimes understood to be any vessel propelled by paddles.” But I don’t think it goes far enough.

  23. I would certainly call a dugout a canoe; indeed, I would call it a dugout canoe unless context made it very clear it wasn’t a shelter.

  24. As an example, in French the generic word for ‘cake’ is gâteau, except for English-type fruit cake, which is known in France as cake (pronounced “kék”). What happened was that English cake, which is as generic as French gâteau, was borrowed to refer to a specific type of cake then unknown in France.

    A rare example of linguistic cookery-borrowing going south across the Channel. Normally it’s the other way round.

    Another one that comes to mind is Panzer; in German, a tank; in English, a German tank, almost exclusively a German WW2 tank. In fact, I should think that there are numerous examples from the military sphere; impi is a Zulu word that just means “military unit” but in English it means “a military unit made up of Zulus” for example. Similarly there are no doubt lots of words that in language X just mean “Y” but in English mean “the specific sort of Y used by speakers of language X”.

  25. Sombrero is my favorite example of this type.

    “What do you ask for in a Yanqui restaurant if you’re hungry?”

    Sombrero y botas [lit. ‘hat and boots’].”

  26. What would you say of the opposite? Like Greek “feta” being used in English for any kind of Balkan white brine cheese. Although some use “sirene” in English to differentiate the kinds of textures. But sirene can be a lot softer when fresh, and is a generic word for cheese in Bulgarian, so I guess this is partly the thing the blog post talks about.

  27. I wonder why сирене (sirene, stress on the first syllable) has that -ne extension? All the other Slavic languages have straightforward descendants of Proto-Slavic *surъ.

  28. I would point out that a borrowing need not result in a narrowing of meaning. It can also be a broadening. Example: Egyptian Arabic ‘tust’ (< toast) is sliced western-style bread.

  29. David Marjanović says:

    German As I Know It: Toast “toast”, Toastbrot “bad white bread specifically made for toasting”.

  30. @David Marjanović: Egyptian ‘tust’ need not be toasted. In fact, I don’t think it normally would be. I think it is mostly used to make sandwiches (and maybe display Western sophistication).

  31. marie-lucie says:

    GW: sliced western-style bread

    I think that by “western” you mean English/American, not the traditional bread of other countries. In France the pre-sliced, square-section bread, made to be toasted, is available as “pain américain”.

  32. marie-lucie: Yes, I am referring to American/English sliced bread. I said ‘Western’ so as to not inadvertently omit any other countries.

  33. (Dutch bread is mostly sold sliced. At the actual baker they keep whole loaves on the shelves and run them through the slicer and bag them as you order them, which is particularly cool.)

  34. marie-lucie says:

    des, in France too you can have your loaf sliced if you ask for it, but it is not sold pre-sliced.

  35. Stefan Holm says:

    Names on bread is a jungle. In my country we love a bread called pain riche and think it is the most French of all breads. And it is – but noone there knows what it is, they simply call it baguette.

    Another one is wienerbröd (the same in the other Nordic countries, “Viennese bread”). But in Vienna itself it’s called Kopenhagener Gebäck and in English usually Danish pastry. The explanation seems to be, that the dough/paste itself has its origin in Vienna but the actual bread in Copenhagen. We northeners feel comfortable though in knowing, that the masters of cuisine, the French, call it Viennoiserie.

  36. marie-lucie says:

    I wouldn’t call viennoiserie a type of ‘bread’ (pain), since it is sweet, made with pastry flour rather than the heavier bread flour.

  37. Stefan Holm says:

    Hey, Marie-Lucie, the flour (wheat) is essentially the same in a baguette as in a Danish pastry. It’s all about the amount of sugar and butter. When it comes to flour we way up north from climate reasons historically had to rely in rye, barley and oats for baking. (Still Scandinavia is north of the climate border for growing wheat. We need to use straw shorterners and should really leave it to others in more favourable climate conditions).

  38. Lars (the original one) says:

    This is the second time I encounter French cake in the Hattery this weekend, so here goes: the (plural of the) English word was also borrowed in Danish as kiks = (BrE) ‘biscuit,’ kex /kɛks/ or /ʃɛks/ in Swedish. (The Danish spelling is analogical: at the time of borrowing the normal pronunciation of old /i/ (so spelled) was mid-open (or so) but has since become closer under the influence of the spelling, pulling this word along).

    While of course F biscuit ended up as Danish beskøjt = ‘hardtack’ (via Dutch).

  39. I was looking into biscuits today in an attempt to explain what an American biscuit is in UK terms (if you want to know, it is a unflavored savory scone, or at most flavored with butter or buttermilk, and rather larger than an English scone; it is fluffy rather than crumbly), and as a byproduct I found out that biscuit is etymologically ‘twice-cooked’, like many other bread varieties, and originally applied to hardtack-like foods in the Anglosphere as well. It only shifted to ‘cookie’ and the thing I was just describing fairly recently.

  40. Trond Engen says:

    Norw. kjeks [ç] is strange in rendering English [k] as [ç]. Not sure if it came by nativization from Danish kiks or because initial [ke] was phonotactically unheard of at the time of borrowing. Borrowing [ei̯] as [e] is also unusual. [e:] used to be common, but I think it may have been shortened by the long coda [ks]. These days we have fewer qualms and would have borrowed it as keiks [kæi̯ks]

Speak Your Mind