Gbagbo

« previous post | next post »

In a comment on yesterday’s “Mele Kalikimaka” post, Eric asked:

When is a foreign sound so alien to a language that it’s “disallowed”? When does a linguist–or just a transcriptionist–decide to throw her hands up and say: “these people will never get this”?

I apologize for misleading Eric by using the word “disallowed”, which he seems to have taken to mean that some authorities — linguists or “transcriptionists” — have made a conscious decision to ban certain sound-patterns, or at least to stop trying to get people to say them “correctly”.  And I also need to make it clear that this has no necessary connection to what people can or can’t “get”.

Each language has its own phonological system, which provides an inventory of basic sound-elements and principles of sound-combination.  Like most other aspects of language, this is what Hayek calls a “grown order” rather than a “made order”: it develops organically through the communicative interactions of a speech community. No ruler or government agency (or linguist!) decides what it should be.

So what happens when members of one speech community borrow a word from another? As a rule, they map the alien sound-pattern onto (what seems to them to be) the most similar sound-pattern available in their own system. In some cases, if enough borrowings motivate it, they may adapt their sound-system a bit by introducing a new sound or a new pattern of sounds; but this is less common. Even when most speakers are bilingual, there’s less adaptation of this kind than you might expect. [In modern times, the factor of spelling also enters into the process; but I’ll ignore that aspect of things for now.]

It’s easy to see examples of this process in English.  One example that’s been in the news recently is the name of the embattled quasi-president of Côte d’Ivoire, Laurent Gbagbo. The first name is French, and therefore various familiar adaptations are needed to render it in English, which we’ll pass over in silence for today. The last name, Gbagbo, is presumably from his native language, which is (said to be) Bété. He was born in the town of Gagnoa, so the variety is presumably what Ethonologue calls Ganoa Bété, which is a language in the Eastern Kru family. (See this Ethnologue language map for more information.)

The Eastern Kru languages generally have only open syllables — vowel, consonant + vowel, or consonant + /r/ + vowel. However, like many languages in Africa and a few languages elsewhere, their inventory of consonants includes doubly-articulated labial-velar stops, generally written in IPA as /k͡p/ or /ɡ͡b/, where the tie over the top of the two consonants indicates that this is a doubly-articulated consonant, not a cluster. For English speakers, one way to think of these sounds is to start from English /kw/ and /gw/, which in generally are also phonetically doubly-articulated labio-velars, in that the lip-rounding for the /w/ starts from the beginning of the /k/ or /g/, and is released synchonously with the velar occlusion. To turn this into /k͡p/ or / ɡ͡b/, just make the labial part a complete closure rather than the partial closure of /w/.

(See William Everett Welmers, African Language Structures, for more information. And note for the record that Eastern Kru languages generally have labialized velar stops /kʷ/ /ɡʷ/ as well.)

It’s not very hard to learn to make these doubly-articulated labio-velar stops — but that doesn’t mean that English speakers should necessarily do so when pronouncing Dr. Gbagbo’s name in English.  The second /ɡ͡b/ fits easily into the sound patterns of English, as a syllable-final /g/ followed by a syllable-initial /b/, as Digby or Dogbert. For the first one, though, there are three basic options:

  1. Insert an epenthetic vowel to break up the apparent “cluster”, as in the pronunciation of Ptah here;
  2. Omit the velar (/g/) part entirely;
  3. Try to approximate the Kru labiovelar stop.

Among broadcast news, option (3) hardly ever happens — at least I’ve never heard it — and option (1) is pretty rare. The result, unsurprisingly, is a name that sounds pretty normal in English, as if it were spelled “Bagbo”. (Kru languages also generally have a complex system of lexical tone, which we’re ignoring completely.)

Here’s one set of examples, from NPR’s Tell Me More, broadcast on 12/22/2010 (“Political Tensions Continue in Ivory Coast“, NPR . Michelle Martin (the program’s host), Marco Chown Oved (an Associated Press reporter in Abidjan), and Y.J. Choi (the head of the U.N.’s operations in Côte d’Ivoire).

Note that Ms. Martin starts with a somewhat disfluent attempt at (1), and then goes with (2) the rest of the way. Mr. Oved sticks with (1), and also adapts the first vowels to an American /æ/ rather than Ms. Martin’s backer (and thus more stereotypically foreign-sounding) /ɐ/. She accomodates a bit to his pronunciation of the ‘a’, and then reverts to the educated American “‘a’ in foreign words” norm.

(I’ve omitted all but one of Mr. Choi’s renditions of Gbagbo, since the case of a speaker of Korean adapting a Kru word to English norms is more complicated…)

[audio: http://languagelog.ldc.upenn.edu/myl/GbagboS.mp3]

Martin:  There, incumbent president Laurent Gbagbo lost his bid for reelection […]
Martin:  And most of the international community is standing firm against Gbagbo […]
Martin: But Mr. Gbagbo says […]
Oved:  If we’re not openly supporting Gbagbo we’re being portrayed as […]
Martin: Now Gbagbo last night announced that he was ready to meet […]
Oved: ((Mis))ter Gbagbo ((who)) lost […]
Oved: What we’re seeing now is Gbagbo’s camp […]
Oved: So I think Gbagbo’s trying to […]
Choi: But since the declaration made by president Gbagbo yesterday evening […]
Martin: It’s our understanding that president uh Gbagbo has told […]
Martin: Mr. Gbagbo seems to be […]
Oved:  A virtually unanimous uh international opinion against mister Gbagbo […]
Oved: uh resources of Gbagbo’s government […]
Oved: Mister Gbagbo’s not receiving any development money […]
Oved: … to cut off all funds to mister Gbagbo and this would literally mean that Gbagbo would not have a treasury.

This kind of adaptation is normal and reasonable, even for people who speak the source language to one extent or another. My favorite personal example of this is a memory of the British phonetician Dennis Fry offering a glass of Dubonnet. Dennis was bilingual in French and English, but he unselfconsciously anglicized French words when he was speaking English, and so in his English, Dubonnet was pronounced as “dew-bunny”.

[Dr. Gbago himself speaks many languages: this BBC profile explains that his high-school nickname was “Cicero” due to his fondness for Latin. He also has a doctorate in history from the Université Paris 7 – Denis Diderot, and lived for a decade or so in France.  And I would guess that like many other West Africans, he speaks one or more other local languages as well as his mother tongue and the colonial language.]



61 Comments

  1. Charles Gaulke said,

    December 26, 2010 @ 11:32 am

    I’ve always felt like this kind of adaptation, while reasonable and indeed unavoidable with loanwords, ought to be minimized to the extent possible with names, if only out of a kind of basic respect. (If I’m honest, though, that feeling largely stems from my own frustration at bilingual Quebecois who could pronounce the word “charcoal” insisting on calling me “Shahl” no matter how I introduced myself.)

    That said, I’m no better at mastering foreign phonological systems than anyone else, and I always go along when someone gives me an “adapted” version of their name to begin with. No matter how impolite it might feel I suppose it is still necessary.

  2. Ryan said,

    December 26, 2010 @ 11:45 am

    Martin’s first attempt sounds to me less like a disfluent attempt at a trisyllabic rendition of Gbagbo than a false start.

    [(myl) Maybe, but it’s a false start that starts with /g/.]

  3. Mr Fnortner said,

    December 26, 2010 @ 1:11 pm

    English has some doubly-articulated initial consonant clusters that we seem to honor in the breech. The first to come to mind is the initial x, as in Xavier and xylophone. Is the difficulty in pronouncing “gs” the reason for ex-zavier? And then there is duh-wayne for Dwayne. Is this a reading skills issue or a genuine pronunciation difficulty?

    [(myl) In the case of xylophone and Xavier, English doesn’t allow /ks/ or /gz/ as onset clusters, or as doubly-articulated consonants either. So the two possibilities (assuming a proximate source with initial /gz/, which I very much doubt in both cases) are to add an epenthetic vowel, or to drop the stop.

    As for Dwayne, English obviously has no trouble with /dw/ as an onset. But there are usually pronunciation doublets when the reduction of an initial unstressed consonant + schwa creates a possible onset sequence. Thus there’s not much acoustic difference between (e.g.) “suppose” as two-syllable /səˈpoʊz/ and as one-syllable /səˈpoʊz/, so a lot of people (as pre-literate children) lexicalize the one-syllable version, at least as an alternate. Similarly, when you hear /dweɪn/, it’s hard to know whether the target might have been /dəˈweɪn/.

    I don’t think that either reading skills or pronunciation difficulties have anything to do with any of these cases.

    In the flow of speech, pretty much every articulatory gesture is phonetically co-articulated with its neighbors. But phonotactically, some multiple articulations are treated as a unit, and some aren’t. (And sometimes you can’t tell…)]

  4. mike said,

    December 26, 2010 @ 1:32 pm

    Is there an example recording of (something approaching) the Bété pronunciation for comparison?

  5. Erik Zyman Carrasco said,

    December 26, 2010 @ 2:02 pm

    A little tangential, but Anglicizing French final stressed -et as an unstressed [i] (or so) seems to me like a pretty British thing to do (am I right?). By contrast, I (19yo, NYC) would Anglicize Dubonnet as something like [d¨ɪuw.b¨ɪ.ˈneɪ],* if I ever had occasion to use the word (which I don’t).

    *Not sure how to precisely transcribe my pronunciation of the first two vowels (especially the first, which I realize as a diphthong or triphthong), but the basic idea is /dubəˈne/.

  6. Henning Makholm said,

    December 26, 2010 @ 2:10 pm

    Help a non-linguist out here: How does one actually manage to do (3) rather than (1)?

    [(myl) There are actually lots of different ways to do this. As Welmers writes in the cited book,

    The double closure required to produce these consonants results, obviously, in a space between the two closures. This space can be characterized during the double closure by suction, so that air goes into the mouth when the consonant is released; or it can be characterized by pressure, with a resultant local (not lung air-stream) aspiration; or it may be neutral. In addition, or courses, the bilabial and velar articulations may be quite independent as far as the degree of fortisness is concerned. The result is that there are just about as many kp sounds as there are languages which have one, and a similar variety of voiced counterparts.

    I have no idea what the varieties of gb used in of Gagnoa Bété are. But if you do something along the lines of the manoeuver that I describe, you should approximate the version used *somewhere* in the region.]

    My own (completely introspective and probably scientifically wrong in seventeen ways) impression of how to say /g/ is that I block the airflow with tongue against back palate. Then I blow air against the blockade while relaxing the tongue, and at some definite point in time the air breaks through, making a noise. That single, discrete point in time is the /g/. I distinguish between /g/ and /k/ by some subtle variation in when the voice sets in (which I can do now, but distinctly remember having to struggle to learn as a child; even so I’m not sure now exactly what the difference is).

    I think of /b/ as the same thing, only blocking the airflow with lower lip against upper lip. Again, there is a definite, discrete point in time when the air breaks through and makes the /b/ noise.

    If I try to say “gbV”, there has to be some period of time after the /g/-breakthrough has happened but before the /b/ one happens. (Or should I try to make them coincide exactly?) During that interval, the voice must be on (otherwise it would be “kb-“) which by definition results in some kind of vowel sound, thus stranding me with mispronunciation (1) gVbV.

    Is there some trick related to when one closes the lips to prepare for the /b/, relative to the /g/?

    “Gw-” (or “ks-“, “kn-“, “pf-“, etc), is much easier to imagine for me, because /w/ “takes up time”, so I can start saying /w/ as soon as air begins to flow after the /g/ stop. All the initial consonant clusters that I can think of in languages I speak follow this pattern, with at most one stop extended with continuants before and/or after.

    [(myl) But /b/ takes up at least as much time as /w/ does, as a matter of articulatory fact. (And /w/ itself is a doubly-articulated labio-velar approximant, though never mind that for now!)

    In fact you’re already deeply familiar with doubly- or triply- articulated sounds (quite apart from the inevitable coarticulation of neighboring sound-elements): every voiceless consonant involves a co-ordinated articulation of tongue/lips and larynx; every nasal consonant or vowel involves carefully-synchronized articulations of tongue/lips and velum; every rounded vowel involves simultaneous tongue and lip gestures; and so forth.]

  7. Bloix said,

    December 26, 2010 @ 2:59 pm

    It’s been my observation that the English anglicize foreign words as a matter of course, so that to say “Du-bohn-AY” while speaking English would be positively incorrect.

    [(myl) As Erik Z.C. suggests above, I think there’s a trans-Atlantic difference in the manner and degree of anglicization of such words. Most Americans, in my limited experience, would say something like /ˌdu.bəˈneɪ/, using American vowels (including a reduced vowel in the second syllable). Dennis Fry is the only Englishman I can recall pronouncing the word, but if an American used (an Americanized take on) his version (/ˌduˈbʌ.ni/) it would be a matter of making fun of imagined lower-class hick pronunciation, or something like that.]

  8. Henning Makholm said,

    December 26, 2010 @ 3:36 pm

    I have trouble with “/b/ takes up at least as much time as /w/ does, as a matter of articulatory fact”.

    In my mouth, /b/ takes up zero time; it is an idealized point in time. Before this instant, no air flows and the “b” has not been said. After this instant the air flows, “b” has already been said and cannot be unsaid. On the other hand “w” has a beginning and an end (if both sliding, gradual ones), so it takes more than zero time.

    I can say a “w” fast or slow, stretch it out as much as I want to. But if I try to say “b” slowly, I end up with, at best, an ordinary instantaneous “b” followed by some amount of a strange labial fricative that I have no name for.

    Myl’s remark about double and triple articluation seems from my perspective to miss the mark. Of course I can coordinate things that happen at different places in the speech apparatus. That does not help me understand what is meant by doing two “instant-like” movements, not at the same time, yet not allowing anything “state-like” to come between them.

    [(myl) You have a right to your own phenomenological intuitions, but not to your own articulatory facts. When any human, including you, makes a bilabial stop, the lips must close, stay together for a period of time, and then open again. For an English /b/ in the onset of a stressed syllable, this whole process is likely to take 200 msec. or more. You can time it yourself by saying “ba ba ba ba ba ba …” as fast as you can, and estimating your rate of syllable production, which is likely to max out at about 6 syllables per second. If you watch yourself in the mirror, you’ll see that your lips are moving up and down constantly during the production. Thus each open-close or close-open cycle is about 1/6 = 0.167 sec. long. This is pretty much as fast you you can do it — for more leisurely (and more normal) rates of speech, the period would be longer: a comfortable “ba ba” rate is more like 3-4 Hz., corresponding to a close/open cycle time of around 250-300 msec.

    For the labial part of a labiovelar approximant like /w/, the process is essentially the same except that the lips don’t actually ever close. The overall duration of the gesture is similar, as shown by the fact that your maximum rate of “wa wa” production is also about 6 Hz.

    Your perception of [b] as a point-time event apparently means that you’re identifying it with the point of stop release, which is indeed short (though hardly instantaneous). But the bilabial gesture as a whole is much more than that — you can’t open your mouth unless you’ve closed it first, and once you start opening it, you have to keep going for a while if you want to make a large enough hole in your head to say an unrounded vowel. The available muscle forces, the mass of the jaw and lips, and f=ma do the rest.]

  9. Robert Coren said,

    December 26, 2010 @ 4:48 pm

    I think /ˌdu.bəˈneɪ/ is not far off from what I (a native of New York) say, although I might make a bit of an effort in the direction of /ɔ/ for the second syllable. (That’s if I’m not feeling pretentious and going for a more-or-less French pronunciation, complete with /y/ for the first vowel.) /diu/ or /dju/ sounds weird to me, as does stressing the second syllable.

    Unlike Erik Zyman Carrasco, I have occasion to say it fairly often.

  10. Erik Zyman Carrasco said,

    December 26, 2010 @ 5:50 pm

    @ Robert Coren: yes, I (also a New Yorker) leave out “optional” [j]s almost as often as possible, so for due I have /du/, not /dju/. What I was going for with the [¨ɪuw] in the first syllable of Dubonnet was not the glide [j]—I was trying to capture the fact that I realize the phoneme /u/ as what’s probably a triphthong. Specifically, I don’t go to the vowel quality of [u] directly (unlike some people I know); rather, I get there by way of an introductory vowel with a different quality, which I represented above as [¨ɪ] but am not actually certain how to transcribe.

  11. J Lee said,

    December 26, 2010 @ 6:44 pm

    ” Dennis Fry is the only Englishman I can recall pronouncing the word, but if an American used (an Americanized take on) his version (/ˌduˈbʌ.ni/) it would be a matter of making fun of imagined lower-class hick pronunciation, or something like that.”

    There is a famous precedent, of course, from Sir Anthony Hopkins: http://www.youtube.com/watch?v=BjGpcEA-FyE

    Concerning ‘Dwayne,’ to me the disyllabic form has everything to do with the ultimate-stressed pattern typical of black American names (both real and pseudo-African).

  12. Ryan said,

    December 26, 2010 @ 7:15 pm

    @Henning Malkolm: The thing about measuring articulatory gestures is that you have to measure the whole thing. A /b/ isn’t just the moment when you open your lips. It’s the build-up of pressure in your mouth, the prevoicing (if it’s really a [b], and maybe it’s not), and then there’s the puff of air that counts as a release.

    Get a decent microphone and record yourself saying pairs like wheat/beet, weed/beat, etc., and then look at the recordings on Praat or something.

  13. Ryan said,

    December 26, 2010 @ 8:04 pm

    That second pair should have been weed/bead.

  14. Keith Brinton said,

    December 26, 2010 @ 8:33 pm

    The discussion of “Gbagbo” makes me think of two language: Brazilian Portuguese, and Vietnamese.

    It is standard pronunciation of Brazilian to insert a vowel sound between two adjacent consonants like the “dv” in “advogado.” That ends up being a five-syllable word. One humorous linguist then pointed out that we thus have a syllable consisting of the letter “d.”

    In Vietnamese, some sounds are created by closing the glottis and the lips at the same time, then exerting pressure with the glottis so as to puff the cheeks out. Example: “hoc.” Because of the closed lips, the “c” sounds like a “p” but learners who pronouce it as only “c” or “p” haven’t got it down yet.

  15. Mr. Fnortner said,

    December 26, 2010 @ 10:21 pm

    Yet it is interesting that dwarf, Dwight, and Dweezil do not seem to produce the initial duh syllable. Any thoughts why?

  16. Mark Young said,

    December 26, 2010 @ 11:48 pm

    I not sure I understand the answer to Henning Makholm’s question. From the original description, I /think/ the answer is that (3) differs from (1) in having the lips already closed when the g-like release happens. What else happens at/near that release depends (from my reading of the follow-up) on the particular variety of the consonant. Probably the easiest for an English-speaking person would be to let the released puff of air force open the lips to make a b-like sound hard on the tail of the g-like sound.

    By contrast, in (1) the lips would be open for the g-release — which results in the intervening vowel sound HM mentions.

    Am I close?

  17. Keith M Ellis said,

    December 26, 2010 @ 11:59 pm

    Charles Gaulke wrote,

    I’ve always felt like this kind of adaptation, while reasonable and indeed unavoidable with loanwords, ought to be minimized to the extent possible with names, if only out of a kind of basic respect.

    I’ve long felt that peoples’ names and place names are the semi-prescriptive exception to descriptivism. That is to say, there is an authority or authorities on these words–the person whose name it is, and the people who live in the place. I agree that especially in the case of people, one should make a good-faith effort to pronounce it as close as possible to how he/she does, out of simple respect.

    And I feel the same about place names, though I know that far fewer people will agree with me in that case. Sure, anglophones have their own names for places we call “France” and “Germany”, &tc. But to some degree, that’s a result of the inertia of historical cultural chauvinism that we rightly try to avoid these days.

    Similarly, I find it interesting and suggestive that in BE there’s a much stronger tendency to anglicize non-English words than there is in AE. It demonstrates that although the British are more European and cosmopolitan than Americans (as non-Americans normally suppose/assume), it’s also the case that Americans are intrinsically more multi-cultural as a result of the culture being built on large and sustained immigration from several different non-Anglophone cultures.

    Even so, there’s certainly class/educational/regional variation in American tendencies to anglicize non-English names and other words.

    I suspect that I’m particularly sensitive to it for a very similar reason to Mr. Gaulke’s: I’m a(n anglo) native New Mexican, which is about as close a US parallel to Quebec as there is (unless Puerto Rico becomes a state, in which case there will be a true similarity). Anglophones in (central and northern) New Mexico (that is, the parts which aren’t culturally Texan and are, instead, culturally Hispanic for the last 400 years) assimilate a great deal of Hispanic culture and Spanish pronunciation without necessarily learning any Spanish. There’s still varying degrees of anglicization, but much less so that one finds elsewhere in the US.

    This was among my biggest pet-peeves when I lived in Austin, Texas for eight years. Guadalupe is a main street bordering University of Texas and is mentioned often. And it’s pronounced “Gwad-a-loop”. Which I absolutely, positively refused to accommodate. But you’ll notice that even among the natives who are anglos here in Albuquerque, there’s a distinct minority (or majority?) who elide the final “e” in Rio Grande. It’s hit-or-miss, but generally closer to Spanish than you’ll find elsewhere. (Sante Fe anglos are, in general, much more faithful to the Spanish.)

    Frankly, it seems to me that Texans willfully anglicize Spanish person and place names. There’s a lot of Hispanics in Texas, of course, but Texans “remember the Alamo” and there’s a cultural hostility there that, I think, still exists in, at least, a subterranean sense.

  18. J. Goard said,

    December 27, 2010 @ 12:01 am

    one way to think of these sounds is to start from English /kw/ and /gw/, which in generally are also phonetically doubly-articulated labio-velars, in that the lip-rounding for the /w/ starts from the beginning of the /k/ or /g/, and is released synchonously with the velar occlusion.

    Really? From where I’m sitting, a significant problem for Korean EFL learners is moving away from such double articulation, towards the cluster that I, at least, use with /kw/.

  19. iching said,

    December 27, 2010 @ 1:04 am

    Lou Reed says/sings “Dubonnet” a couple of times in the song “Berlin” from the eponymous album in this live performance (Lou Reed & John Cale, Bataclan 1972).

    It’s also the way I pronounce it (as a native AuE speaker). But maybe I was influenced by listening to the album (a favourite of mine back in the day).

  20. Nathan Myers said,

    December 27, 2010 @ 3:53 am

    My understanding was that the English strangle French words deliberately, as the echo of a show of independence from their French-speaking rulers. No?

  21. Alon Lischinsky said,

    December 27, 2010 @ 5:02 am

    @Keith Brinton:

    One humorous linguist then pointed out that we thus have a syllable consisting of the letter “d.”

    Syllabic consonants are not that rare. English prominently features [n̩] (think the last syllable of ⟨lemon⟩) and [m̩] (⟨blossom⟩.

    The Slavic languages have several, and they are common as well in the Sinitic languages. I’ve always heard Mandarin 四 (⟨sì⟩) as a vocalic [s] with descending tone.

  22. John Cowan said,

    December 27, 2010 @ 5:12 am

    Though /dw/ is a legal English syllable onset, it’s a very rare one: dwarf, dwell, dwindle, the neologism dweeb, and some pronunciations of boudoir about cover it.

  23. LDavidH said,

    December 27, 2010 @ 5:27 am

    I think accommodating foreign names & words to your own language patterns is inevitable, even if you disapprove of it. Nothing to do with cultural superiority (although often caused by linguistic ignorance)! Virtually every language has “unusual” sounds, and you can’t really expect non-native speakers to be able to reproduce them properly – just as I, a native Swede married to a Brit, still cannot reproduce all the sounds and intonations of the Queen’s English properly, despite living in the UK. And I have yet to meet a foreigner who can pronounce the Swedish town of Jönköping correctly!

  24. Bob Ladd said,

    December 27, 2010 @ 6:59 am

    There’s a very clear difference between British and American patterns of adapting French words. In most two-syllable words and many three-syllable words American English uses final stress and British English uses penultimate stress. Pretty much everything else – in particular vowel quality – follows from the stress difference. That covers the case of Dubonnet that Mark started with. It also covers the strikingly different pronunciations of two-syllable words like buffet and gateau.

  25. PeterL said,

    December 27, 2010 @ 7:53 am

    A counter-example to Dubonnet/dew-bunny — Canadian ice hockey and news announcers (in both official languages). I have no idea how they pronounce “Gbagbo”.

  26. Leonardo Boiko said,

    December 27, 2010 @ 8:37 am

    Here’s an example I found curious about the kind of “disallowing” that arises spontaneously in languages. In Portuguese, words ending in /s/ or /z/ usually become plural by adding the syllable /es/: 1 GÁS2 GASes, 1 inGLÊS2 inGLEses. These root words both had stress in the last syllable (marked here by capitals). But when a word is stressed earlier, such as in Ônibus “bus”, it gets no plural at all; 1 ônibus, 2 ônibus. What gives?

    As it happens, all Portuguese words, without exception, are stressed in one of the last three syllables (the “three-syllable window”), with the last being most common and 3rd-to-last, such as Ô·ni·bus, the rarest. If you added another syllable to build a plural, you’d end up with *Ô·ni·bu·ses, which stress in the 4th-to-last—outside the window. That simply cannot happen. To make it feels like Portuguese you’d have to change the stress and get *o-ni-BU-ses. But we native speakers strongly reject a plural that changes stress patterns. The plural of ca·RÁ·cter was ca·ra·CTE·res; this was felt to be so weird that a new word ca·ra·CTE·re was coined to match the 2nd-to-last stress scheme (or, rather, the plural form made people think the new singular form already existed, before it was even dictionarized).

    Since Portuguese disallows *Ô·ni·bu·ses and dislikes the pair Ô·ni·bus→*o·ni·BU·ses, ônibus has no plural. This “disallowing” or “disliking” is not something decided by language authorities; rather, any native speaker will agree these words just sound too strange.

  27. Henning Makholm said,

    December 27, 2010 @ 9:49 am

    Yes, I count only the release. That’s what makes a sound, and the rest is just preparing to make the sound. When I say “banana” from a standing start, I don’t begin by closing my lips; they are already closed. Would you argue that the “b” I say then lasted half an hour?

    [(myl) From a perceptual point of view, the formant transitions into and out of the /b/ are generally just as important as the burst spectrum is. As for your “standing start” example, it’s childish: you could hold a kissy-face expression for half an hour before releasing a /w/, if you wanted to make that particular point as well. The fact is that in normal running speech, articulator positions don’t jump instantaneously from one position to another, but move smoothly as governed by the laws of physics and physiology, and that’s exactly as true for the realization of /b/ as for the realization of /w/.]

  28. Michael said,

    December 27, 2010 @ 10:45 am

    If the video quality isn’t too poor, here is ECOWAS President Victor Gbeho pronouncing “Gbagbo” twice. Blindly guessing from Gbeho’s name, I’d imagine he has some authority in properly pronouncing a “gb”.

    Democracy Now, “Real Video Stream” at 0:54 and 1:11.

  29. Megs said,

    December 27, 2010 @ 11:36 am

    @Mr Fnortner

    Just wanted to throw out there that I think the word initial X varies by American dialect. I’ve definitely heard it pronounced as Z and occasionally, when dealing with multilingual speakers, an actual x (voiced or unvoiced), but only rarely (generally from Californians) with an E on the beginning. Almost always, I hear Z: Zavier, zylophone, zenobiologist.

  30. Jon said,

    December 27, 2010 @ 1:06 pm

    Keith M Ellis said,
    “I’ve long felt that peoples’ names and place names are the semi-prescriptive exception to descriptivism. That is to say, there is an authority or authorities on these words–the person whose name it is, and the people who live in the place.”

    I agree, and for placenames I feel that English, as the international language, has a particular responsibility here. Why should we teach Japanese and Estonians to say Florence instead of Firenze?

    My impression (backed up by a radio discussion including someone from the BBC pronunciation unit) is that aggressive anglicisation has steadily decreased in the UK over the past century. European places that used to have uniquely British pronunciations (Lyon, Ypres, Mallorca) are now commonly pronounced with an attempt at the local version. But there is still a long way to go.

  31. language hat said,

    December 27, 2010 @ 3:33 pm

    Why should we teach Japanese and Estonians to say Florence instead of Firenze?

    Because that is the name of the city in English. Let’s turn it around: if you were learning Japanese, would you insist on saying Zhongguo for “China” rather than Chugoku, out of respect for the Chinese? If you were learning Estonian, would you insist on saying Rossiya for “Russia” rather than Venemaa? I think not. The fact is that English is the only language I know of whose chattering classes have adopted this self-flagellating embarrassment about having (oh the horror! the shame!) their own forms of foreign names rather than somehow managing to learn every toponym in the world as spoken by its inhabitants, no matter how remote. The French make no attempt to pronounce Chicago as a Chicagoan would, nor are the Russians embarrassed about calling the Rocky Mountains Скалистые горы [Skalistye gory]. That is the only sensible approach, and I can’t think what quirk of sociohistory has made English speakers (primarily Americans, I’m guessing) so batty on the subject. In the immortal words of Elvis Costello, I used to be disgusted, now I try to be amused.

  32. Robert Coren said,

    December 27, 2010 @ 3:41 pm

    Mr Fnortner on the absence of an interpolated vowel in “dwarf”, etc., reminds me of a Minneapolis bartender who pronounced the surname of our very own Arnold Zwicky with at least half a vowel between the z and the w.

  33. Ellen K. said,

    December 27, 2010 @ 6:05 pm

    Megs, do you have any evidence that any words in English beginning with Z besides Xavier have other pronunciations besides the X being pronounced like a Z? Do you have any evidence that the variation for Xavier is dialectal, rather than some other sort of variation? Seems to me the variation is more likely random (not connected to anything) and/or by person referred to rather than dialectal.

  34. ohwilleke said,

    December 27, 2010 @ 6:43 pm

    In Kate Elliott’s Jaran series (whose heroine is a linguist by training) the heroine’s older brother Charles has, by an odd turn of events, suddenly been award status as the Duke of several territories. One of the running jokes of the series (loosely tracking the spread of the Indo-European languages) is that one of those territories is always referred to as the “unpronounceable one” in the dialog of the characters who just collectively give up on transliterating it into any sound-system with which these multilingual cosmopolitians are familiar.

    In real life, I’ve encounted unpronounceability issues most often with vowel deprived words in Welsh, and Central American indigeneous languages, where I am not familiar with the way words in those languages are converted into sounds.

  35. John Cowan said,

    December 27, 2010 @ 7:30 pm

    Welsh is not a bit vowel-deprived. Serbian, maybe. Nuxalk, most definitely. But the fact that Welsh uses w and y as vowel letters (and also as consonant letters) doesn’t make it vowel-deprived.

  36. J Lee said,

    December 28, 2010 @ 12:05 am

    I have to agree with language hat about Anglicization in general — they are simply English words, and the only reason I see to single them out is for snobbery (i.e. as shibboleths for sophistication).

    Often the distinction is useful and necessary — /noʊtərˈdeɪm/ has nothing to do with France.

  37. Keith M Ellis said,

    December 28, 2010 @ 12:21 am

    Language Hat’s point applies equally well to personal names of which there are common anglophone equivalents. So, when speaking to someone with one of these names, does he refer to them by the anglophone equivalent, or by how they refer to themselves? For that matter, why bother to attempt a non-English name of which there is not a common English equivalent? Just make one up. You’re speaking English, after all. Right?

  38. B.Ma said,

    December 28, 2010 @ 12:24 am

    Jon / language hat:

    Not sure whether I understand your points exactly, but there is some inconsistency in whether foreign languages adopt the local pronunciation or the English name. For example, in their own languages Japanese and Estonians already say “Firenze” (or at least “フィレンツェ”), as based on the Wikipedia translations, but when learning English they should obviously say Florence until such time as native English speakers start to say Firenze (which is when the English wikipedia article will get renamed). The thing is, a Japanese tourist to Italy is more likely to speak English than Italian..

    On the other hand, only Estonians will say something related to “Suomi” for Finland whereas the Japanese term is “Finrando”. I wonder whether that was borrowed from the Swedish name directly or via English – and if the former, why not Helsingfors?

  39. Coby Lubliner said,

    December 28, 2010 @ 12:44 am

    What happens to Dubonnet (French /dybɔnɛ/) in English is what happens to any foreign word with a word-final /ɛ/, whether stressed or not: the vowel becomes /e/ (usually realized as [eɪ] or [ɛɪ]) or /i/. English-speakers are perfectly capable of articulating /ɛ/, but the language doesn’t tolerate it at the end of a word. In a comment on a recent post I noted that an actress named Karen LoGiudice atypically pronounced her surname as an Italian would, except for the final vowel.

    As regards Gbagbo, I am curious about how Bavarian or Austrian newsreaders pronounce the name. In Bavarian the initial cluster gb- is not uncommon; German geboren, for example, is Bavarian gbora.

  40. Jon said,

    December 28, 2010 @ 4:36 am

    “if you were learning Japanese, would you insist on saying Zhongguo for “China” rather than Chugoku, out of respect for the Chinese?”

    No, because the Japanese would not understand it. But I think it would be far better if we Anglophones and the Japanese used the native name, out of respect. And some people certainly are offended. I knew a Greek woman who was upset at her country being called Greece.

    But all that is missing the point that I was making. Because English is the de facto international language, Japanese and Estonians and most people from other countries will, when visiting Italy, use English when talking to the natives. And when a Brazilian businessman is talking to an Indonesian colleague, they will use English. It is daft and disrespectful for most of the world to ignore what the locals like to call their place, and pretend it is called something else. As Keith M Ellis said, it is like refusing to call someone by the name they like to go by. I would not address someone as Billy if he prefers to be called William.

    As to whether the trend to using the correct names is mainly American, I doubt it. As the BBC has found, it is a long term trend in the UK.

  41. language hat said,

    December 28, 2010 @ 10:22 am

    But I think it would be far better if we Anglophones and the Japanese used the native name, out of respect.

    It would be far better if there were no war and no poverty and the lion lay down with the lamb.

    I knew a Greek woman who was upset at her country being called Greece.

    That woman was a fool.

  42. Alces said,

    December 28, 2010 @ 11:24 am

    It is daft and disrespectful for most of the world to ignore what the locals like to call their place, and pretend it is called something else. As Keith M Ellis said, it is like refusing to call someone by the name they like to go by.

    For you, maybe, but this is just a rule of politeness and rules of politeness have no rational basis to back them up. Most people lack this rule, so why impose it upon them?

    I’d also say it’s a bit different from refusing to call someone by their name, because someone’s name is the only way to identify them. If you called them something different they wouldn’t be sure who you were talking to.

  43. Jon said,

    December 28, 2010 @ 12:10 pm

    Language hat and Alces:

    Names are not just words. They are part of people’s identity: they care about them, and feel insulted if they are misused. The same applies to names of places that people are attached to, though of course the effect is weaker. Calling people fools for caring is a pathetic, feeble argument.

  44. J. W. Brewer said,

    December 28, 2010 @ 1:43 pm

    Compare this vintage Dubonnet commercial made for (I believe) the UK market http://www.youtube.com/watch?v=W5d4Xv020P4 with this one of similar vintage made for the US market http://www.youtube.com/watch?v=tGPCABCGGo8.

    In terms of what “politeness” suggests, one might observe that different ethnic groups of originally non-Anglophone immigrants in the U.S. display markedly different patterns in how they name their American-born kids in terms of the prevalence of “American-sounding” versus “old-country” names. It’s almost as if non-native-Anglophones don’t form a single homogenous mass with homogenous tastes and preferences in subjects relating to language use.

    In terms of when Anglophones do and do not substitute the English equivalent for a personal name, I don’t think the patterns (to the extent there are any) indicate that using (possibly with mangled pronunciation) original language name is more polite. We may use the original language name to emphasize the fact that its bearer is a foreigner and thus uncivilized. So, e.g., it’s always Ivan the Terrible but also always St. John of Kronstadt (Ivan to his fellow Russophones). And it’s almost always Kaiser Wilhelm (not William) but not infrequently Emperor Francis Joseph (not Franz Josef), because the Hapsburg enemy was less vilified than the Hohenzollern.

  45. Ellen K. said,

    December 28, 2010 @ 2:12 pm

    I think there’s a difference between having a different name for something (Germany vs. Deutschland) and having a varient pronunciation of the same name, which we also have within a single language do to different accents. When I met someone from Germany named Dörte, I wasn’t expected to pronounce the German ö sound nor the German R sound.

    Of course, it’s not always straightforward one or the other, either a different name or a different way of saying the same way. It’s a continuum. If we agree that dropping into a different accent in the middle of speaking American English (or whatever variety one is speaking) is not required, then the question becomes, how far along the continuum should we go? And I don’t think there’s a single answer that always applies.

  46. Keith M Ellis said,

    December 28, 2010 @ 2:19 pm

    In terms of what “politeness” suggests, one might observe that different ethnic groups of originally non-Anglophone immigrants in the U.S. display markedly different patterns in how they name their American-born kids in terms of the prevalence of “American-sounding” versus “old-country” names. It’s almost as if non-native-Anglophones don’t form a single homogenous mass with homogenous tastes and preferences in subjects relating to language use.

    That’s not a direct comparison. A direct comparison would be those who anglicize their given names upon immigrating to the US. I don’t doubt that a fair portion of those Ellis Island immigrants who had their names involuntarily anglicized didn’t mind it so much, but I know that quite a few did…and quite a few of their descendants have reverted back to the non-anglophone names.

    This isn’t really that big of a deal; yes, it’s easier to go along with convention rather than bucking it (something that one might consider when one is among a community where the convention is along my lines, rather than LH’s). But it seems to me that Language Hat’s claim that his is a principled objection to my point if view is clearly suspect, considering that I highly doubt he anglicizes the names of the non-anglophones he knows. Yet the principle is exactly the same in the case of personal names as it is in the case of place names. (The claim by a commenter above that the active principle is “no one would know to answer to another name” is clearly specious, given that there’s a long history of things like the Ellis Island example above, as well as very common contemporary examples of purely-for-convenience-sake anglophone names for people with names that are very unfamiliar to anglophone ears, such as Chinese.)

    Furthermore, it seems to me that there’s a clear trend in Europe and North America to accept and use foreign names of foreigners, rather than on insisting on localizing them, as was usually done historically. The increasing use of foreign place names in the US is, I think, part of the same trend, done for the same reasons. Only rarely, I think, as “shibboleths for sophistication”. Certainly that’s not true in my case. In my case, this dates back to my childhood when I was quite shocked to discover that the English names I knew for foreign places often had no relationship at all to how the natives referred to them. It made no sense to me, other than as a sort of arrogance. Also, as I mentioned, I would notice an unflattering correlation between those (mostly Texans) who heavily anglicize Spanish place names in the SW US, particularly NM, and a cluster of other habits and ways of thought which are, in my opinion, offensive.

    In any case, unless you are directly affected by this–it’s your name or your hometown–then whether someone does or does not pronounce it as you do is not really something to get upset and insulting about.

  47. Keith M Ellis said,

    December 28, 2010 @ 2:25 pm

    If we agree that dropping into a different accent in the middle of speaking American English (or whatever variety one is speaking) is not required, then the question becomes, how far along the continuum should we go? And I don’t think there’s a single answer that always applies.

    You make good points.

    My answer would be that one should just make a good faith effort to approximate the name as one is able in one’s native language (assuming that’s the language one is speaking at the time). Different people will manage differently.

    From the other direction, what one shouldn’t do is to simply assert that one absolutely won’t say foreign names and, instead, use a native substitute. You mention Deutschland…really, that’s just not that difficult for an anglophone to closely approximate.

  48. J. W. Brewer said,

    December 28, 2010 @ 4:32 pm

    I think it may not be helpful to blur together how one treats the personal names of particular individuals with how one treats toponyms (without even getting into the possible application of different norms for direct address versus third-party reference). But on the latter subject, one of Gbagbo’s predecessors as dictator of the Ivory Coast launched a campaign (wikipedia says in 1985) to get the country referred to as “Cote d’Ivoire” (I’m not going to figure out where to get that diacritical mark on my keyboard) in English-language sources (indeed in all sources in all languages), which has met with some but apparently not uniform success. I don’t know whether U.S. or U.K. radio/tv announcers pronounce “Cote d’Ivoire” in a way that closely approximates how it’s pronounced in Paris or (if different) Abidjan, but I would be dubious.

    By contrast, the current government and populace of the Bundesrepublik Deutschland both seem fine with their country being named a dizzying array of things in other languages (German- in a lot; Alleman- in most Romance languages; Nem- in most Slavic languages, Saksa in Finnish, which reputedly derives from Sachsen; the etymology of the Lithuanian “Vokietija” seems speculative and disputed; etc etc). I would not infer from this that Anglophones (or Francophones, or Russophones, etc etc etc) have a higher degree of regard and respect for the Ivory Coast than they do for Germany.

  49. Anthony said,

    December 29, 2010 @ 1:51 am

    “I knew a Greek woman who was upset at her country being called Greece.”

    For a Greek person with perpetual-outrage syndrome, this is a really clever thing to chose to be outraged by, because pretty much every language in the world has a name for Greece (and Greeks) that is not “Hellas” (and “Hellenes”). And it saves her from getting in trouble for burning down banks.

  50. J. Goard said,

    December 29, 2010 @ 1:52 am

    Another aspect of this argument (the difference between personal and place names) is that, for an ordinary person, there’s generally not an established history of the alternative name. An etymological connection, as between “Peter” and “Pedro”, is not grounds for adaptation by contemporary English speakers (although it apparently was in the past, and apparently is by speakers of other languages; Chinese names certainly tend to be adapted to their cognates in Korean, and the Chinese speakers of Korean seem quite happy with this.)

    If there is an established history, however, it seems to me that one cannot reasonably claim a strong privilege over reference to oneself by other people. I am called “Jeremy” by many groups, “J” by many others, typically “Jer” by my family — and variably “Jeremy”, “J” or “Jae-jin” in Korea. Now, I might start preferring one name over others, and I could certainly let my friends know as much, but to think I ought to be able to quell three of my names, even with people who are talking about my outside my presence, just seems ludicrous.

    It’s only when there are implications beyond conventional reference that we ought to have a problem with variant names. It’s the racism behind the immigrant name changes that irks, as it is extremely offensive to refuse to call a transsexual person according to their identified gender. No such issue exists when we use “Aristotle” instead of “Aristoteles” or “Prague” instead of “Praha”.

  51. J. Goard said,

    December 29, 2010 @ 1:55 am

    I clearly meant “transgender” rather than “transsexual” there!

  52. Bob Violence said,

    December 29, 2010 @ 5:31 am

    Chinese names certainly tend to be adapted to their cognates in Korean, and the Chinese speakers of Korean seem quite happy with this.

    I’m not very familiar with Korean and haven’t spent much time in the country, but the impression I get is that Korea (at least the south, I dunno about the north) is increasingly shifting towards transcription of Chinese names and away from cognates. Judging from Google results (not the most scientific approach, I know), personal names seem to depend on how contemporary they are: Sino-Korean 장예모 Jang Yemo (Zhang Yimou) is more common than 장이머우 Jang Imeou, but Hu Jintao (whose fame is more recent) is usually 후진타오 Hu Jintao instead of 호금도 Ho Geumdo. The process appears more advanced for city names, since 베이징 Beijing and 상하이 Sanghai get far more hits than 북경 Bukgyeong and 상해 Sanghae.I notice when I visit Korea that nobody has any trouble understanding “Qingdao”, where I spend most of my time, even though it doesn’t sound much like the Sino-Korean cognate 청도 Cheongdo.

    Vietnamese, which uses the Latin alphabet, has some interesting variations in the treatment of both personal names and toponyms. Traditionally people and places were given “Vietnamized” names,* but recently the trend has been towards writing foreign personal names in their original Latin forms, or the most common Latinized version for figures from Russia, Japan, etc. Sometimes you even seen pinyin for Chinese names, although the Sino-Vietnamese forms (Hu Jintao = Hồ Cẩm Đào) are still more common. Place names, especially sub-national entities, increasingly lean towards local spellings or standard transliterations, so that Seoul is now usually Seoul instead of Sơ-un or Xê-un, Moscow is Moskva instead of Mát-xcơ-va, Île-de-France is just Île-de-France (even though “France” itself is still normally Pháp), etc. In some cases a pronunciation guide (using more “authentic” Vietnamese orthography) is provided. There is also apparently a geographical split involved, with northern sources tending to be more traditional than southern counterparts.

    *Sometimes the older Vietnamized versions of foreign names still break the standard rules of Vietnamese phonology — for example, Tolstoy traditionally becomes Tôn-xtôi, with an xt (/st/) consonant cluster; as far as I can tell, modern standard Vietnamese has no consonant clusters.

  53. minus273 said,

    December 29, 2010 @ 12:12 pm

    My personal variety of Chinese chauvinism actually prefers that Chinese names are rendered into cognate pronunciations in Japanese/Korean/Vietnamese, for the simple reason that they are pronunciation-of-Chinese-characters-wise no more different from Standard Mandarin than other Chinese “dialects” (except for the lack of tones in J/K), and speakers of one dialect gladly substitute their name into an equivalent form in another dialect.

  54. Peter said,

    December 30, 2010 @ 1:35 pm

    The issue of approximating “foreign” pronunciations comes up even between different dialects of the same language. I’m British, so I pronounce my name /piːtə/; most Americans I’ve known have called me by the American pronunciation /piːɾəɹ/, with a flap in the middle, and an articulated ‘r’ at the end. But some have imitated my own pronunciation: most often in friendly teasing, sometimes out of sincere politeness, a few (I think) because they simply didn’t recognise my pronunciation as the familiar name “Peter”.

    Hearing them mimic (with varying success) the British pronunciation is always jarring, however well-intentioned it may be. The reverse, if I met an American Peter and tried to follow his pronunciation, would be even worse: word-final /r/ (in isolation) is completely foreign to my phonetic inventory, though I can certainly produce it.

    Since meeting this, I’ve felt much more comfortable approximating foreign names by more familiar phonemes; so I’d call a French Charles /ʃɑː(ɹ)l/, not either /ʃɑʁl(ə)/ or /tʃɑː(ɹ)lz/. [Apologies for any inaccuracies in the transcriptions: I’m trying to notate some subtleties a bit beyond my competence with IPA.]

  55. Bloix said,

    December 31, 2010 @ 10:00 am

    These issues arise even within North America. For example, people from Wisconsin call their state Wis-GON-sin, people from Oregon say “AR-ih-g’n,” and Anglophones from Montreal say “Mun-tree-ALL.”

    But if you live anywhere else, you say Wis-CON-sin, OAR-ih-gawn, and MON-tree-ALL, and if you mimicked the pronunciation of the local people you’d sound ridiculous.

  56. M. Heinz said,

    December 31, 2010 @ 2:46 pm

    Is there a technical term for “disallowing” that does not have the same unfortunate alternate denotations as the word “disallow” itself? This phenomenon seems pretty common in language, and it would be useful to have a term that could be used without qualification or quote characters.

  57. Alan Shaw said,

    January 7, 2011 @ 3:15 am

    @Alon Lischinsky
    > I’ve always heard Mandarin 四 (⟨sì⟩) as a vocalic [s] with descending tone.

    In fact that’s what it is. As in zi, ci, zhi, chi, and shi, the i is an othographic convention: must provide a vowel letter to support the tone mark.

  58. How Do You Pronounce “Gbagbo”? | AcheFor.com said,

    April 6, 2011 @ 12:09 am

    […] professor Mark Liberman offers a slightly more complex discussion on his Language Log blog. Gbagbo is a native of Gagnoa, which would mean that his last name […]

  59. Paul Shaddick said,

    April 6, 2011 @ 6:14 pm

    Laurent Gbagbo is indeed a speaker of Gagnoa Bete (An implosive ‘b’ by the way).
    I have done some (unpublished) work on Gagnoa Bete phonology whilst living in Gagnoa and just want to clarify that /gb/ in the language does not usually feature lip rounding. I understand that you are trying to explain how to achieve the double articulation by your reference to English /gw/

  60. Katherine Philippe said,

    April 11, 2011 @ 11:38 am

    Fascinating. Never having heard the name before your helpful insert, I had considered the “gb” to be a blend and had divided the name “gba/gbo.”

    Thank you for the information.

  61. Nutella said,

    April 13, 2011 @ 3:08 pm

    @Ellen:

    When I met someone from Germany named Dörte, I wasn’t expected to pronounce the German ö sound nor the German R sound.

    She’s probably used to be called Dirty.

RSS feed for comments on this post