Comments: PRONOUNCED LIKE B.

This calls to mind the difficulties faced by the marketers of the Chevy Nova vehicle in Mexico: No va... If only they could have emphasized the intervocalic (and non-word-initial) nature of the v!

Posted by Susan at November 12, 2007 02:25 PM

Allophonic variation or no, the "Nova" story is just an urban legend.

Posted by Ben Zimmer at November 12, 2007 02:47 PM

So does that mean I was misinformed when someone told me the Bay of Biscay was in fact pronounced /vis'kaja/ (roughly)?

Posted by Sili at November 12, 2007 04:12 PM

Yeah, in Spanish it starts with a /b/.

Posted by language hat at November 12, 2007 04:38 PM

"The letters b and v are pronounced identically in Spanish, as a bilabial stop (/b/) at the start of a word"

Actually the relevant parameter isn't word but utterance.

the word vaca (cow) begins with a [b] but in la vaca (the cow) it becomes [B] (the fricative).

Similarly, Biscaya will begin with either the stop [b] at the beginning of an utterance (or after /n/ and [B] otherwise.

Posted by michael farris at November 12, 2007 04:49 PM

Well, yes and no. Viskaya in isolation would begin with [b]; but in the form Golfo de Viscaya, which would be normal, you'd get a [β] there.

Ironically, the underlying Basque word is Bizkaia, where the written "b" firmly represents [b].

Posted by John Cowan at November 12, 2007 05:02 PM

Re: "The fact is that the closest thing to an accurate pronunciation of the Spanish word, unless you're a Spanish speaker, is re-ven-TOHN.": I'd say that's less a fact than an opinion; to me the [β] sounds somewhere between an English [b], an English [v], and an English [w], and I'm not at all sure the [v] is the closest. (It also seems to vary with accent/dialect; my impression is that speakers from the northern parts of Latin America pronounce it more [w]-like than others, but I'm not sure of this.)

Posted by Ran at November 12, 2007 05:52 PM

The fact is that the closest thing to an accurate pronunciation of the Spanish word, unless you're a Spanish speaker, is re-ven-TOHN.

Once again I note the American choice of the vowel in toe to represent most non-English "o"-vowels (even by highly proficient linguists, when they write informally). This is quite striking to the attentive non-American, who would normally give "re-ven-TON", in the present case.

No va y Reventón? Surely we must add Pajero to this dishonourable list. I rather prefer the names drawn from music: Concerto, Sonata. I'm holding out for next season's Passacaglia, or the sporty Doppio Movimento that so far remains under cover of sepulchral secrecy in Turin. (Be warned: avoid the Klangfarbenmelodie. It steers like a moose.)

Fascinating to look through such efforts as Virginia Tech's online music dictionary, to see how the uniform Italian "o" is represented four different ways: "ah" (octave), "oh" (first one in con sordini), "oe" (dopo, ossia), and "o" (first two in oratorio). Three observations: First, who should cast the first stohne, anyway? Many of our representations here are lax, and I'm among the worst offenders. Second, we note that VT, typically, does not give the most informative pronunciations at all. There is nothing at the entries for a piacere or à la hausse, for example, yet these are the sorts of things that give students most trouble. Third, several intentions compete, in dictionaries like VT's. They seem to want to show the original pronunciation, but also an accepted anglicised one. Add this to English spelling, regional differences, linguistic ignorance, and general insouciance, and of course you get inconsistency. (I say nothing here about Australia's much vaunted Macquarie Dictionary – but I could.)

Posted by Noetica at November 12, 2007 08:31 PM

Once again I note the American choice of the vowel in toe to represent most non-English "o"-vowels (even by highly proficient linguists, when they write informally). This is quite striking to the attentive non-American, who would normally give "re-ven-TON", in the present case.

Any American would pronounce "TON" like the word ton, which is totally useless. The vowel in toe is the closest most Americans can come to the Spanish o.

the uniform Italian "o" is represented four different ways: "ah" (octave)...

Huh? That's not "Italion o," that's an English o, and in America it is pronounced "ah."

Posted by language hat at November 12, 2007 09:09 PM

O sorry, a wrong example with the octave. I should have shown one from their Italian words: opera buffa, say, which is given with "ah-". (Of course, in British and Australian English octave has nothing like that "ah" of which you speak. This led me into carelessness.)

As for "tohn" and the English word ton, do you mean that you would represent obbligato with two instances of "o", but that in the syllable "ton" you prefer to use "oh" for the same sound? Fair enough! But perhaps misleading, like most of the options. And even the syllable "to" would cause problems, on the principle in question, since it might be read as having the sound of the English word to. Incidentally, VT gives "oe-blee-GAH-toe". The "GAH", at least, is confusing, if elsewhere "ah" represents any sort of an "o"-sound, isn't it? LH, how would you informally represent an good and accurate pronunciation of obbligato, for students? For Australian students I would have "ob-bli-GAH-to". They would still pronounce each "o" differently, but for my part I would have done the best I could.

The broad conclusion remains unassailed: pop pronunciation guides are all mixed up, and even established dictionaries do these things imperfectly. So do even the most sophisticated experts linguistics, when they set aside the formal apparatus for managing such things. It's hard to see a solution to this difficulty: the formal apparatus exist for a good reason.

Posted by Noetica at November 12, 2007 10:22 PM


...that you would represent obbligato with two instances of "o"...

Or of "oe", perhaps I should have added.

Posted by Noetica at November 12, 2007 10:25 PM

O[h], and VT gives "OE-prah SAIR-ree-uh" for opera seria (note elision of "e", and two ways for uniform "a"), though it gives "ah-" for opera buffa. Goe figure, as they say in the States.

Posted by Noetica at November 12, 2007 10:44 PM


[Sc. "experts in linguistics".]

Posted by Noetica at November 12, 2007 10:46 PM
Incidentally, VT gives "oe-blee-GAH-toe". The "GAH", at least, is confusing, if elsewhere "ah" represents any sort of an "o"-sound, isn't it?

In American English, the "o" in "octave" and the "a" in "obbligato" have the same sound, so what's confusing?

Posted by KCinDC at November 12, 2007 10:56 PM

In American English, the "o" in "octave" and the "a" in "obbligato" have the same sound, so what's confusing?

Confining ourselves to some sort of standard American, we are to consider two words: octave and obbligato. Now, Merriam-Webster dictionaries agree that the vowels in oc-, ob-, and -ga- are the same for most Americans: /ä/, in their usage. They gloss like this: "bother, cot; most American speakers have the same vowel in father, cart". Right. Good enough.

On the other hand, I wanted a way to give a "good and accurate pronunciation of obbligato"; the way that I gave, in accord with what the Virginia Tech people seemed to be aiming at, was something more like the Italian than a fully americo-anglicised version in which the vowels in "ob-" and "-ga-" are identical, and the final "-o" differs. But attempts to make the necessary distinctions in representing such Italian words are all mixed up. In evidence I tendered examples from Virginia Tech. Take the two "opera" examples, and add some similar ones, including French (same for present purposes), from that site:

opera [duly marked "It."]
AH-prah

opera buffa
AH-prah BOO-fah

opera seria
OE-prah SAIR-ree-uh

operetta
oe-per-RET-tuh

opéra comique
OE-prah koe-MEEK

Now, I could simply allow these examples – from academics presumably conscientious and competent in their area, who have put together a reasonably useful little dictionary – to argue without further articulation the case that all of this is mighty confusing, and generally poorly managed. (The VT dictionary was just a random first hit, as I checked the currency of doppio movimento – "DAH-pee-(y)oe", in VTsprache.) Let's just say that sometimes all instances of "o" are rendered the same, but sometimes instances of "o" and "a" are rendered the same.

How are they arriving at these pronunciations? Purely haphazardly? Perhaps: but there seems to be some element of method. Take their "oe-blee-GAH-toe". First there might be the "Italianising" decision that the first instance of "o" should be rendered the same as the last one, which itself doesn't look or sound right as an "ah". So "ah-blee-GAH-toh" is rejected. But then, "oh-blee-GAH-toh" is also rejected, because of the prevalence of "h", which looks un-Italian. (Ich weiss? I'm speculating!)

Anyway, the whole thing is a mess, around the web. Look at Wikipedia, in which pronunciations cannot simply be given in American English, since the English-language version serves for the whole English-speaking world. So IPA is pushed. But then, it is badly implemented, often with bitter disputes between the High-German purists, the Austrian-German purists, the American purists, and the British purists haggling bitterly over their favourite Germanic composer's slightest allophonic curlicue. ˈlʊdvɪg væn ˈbeɪtoʊvən versus ˈluːtvɪç fan ˈbeːthoːfn; ˈvɔlfgaŋ amaˈdeus ˈmoːtsart, ˈvɔlfgaŋk amaˈdɔʏs ˈmotsaɐt, ˈvɔlfgaŋ amaˈdeːʊs ˈmotsaɐt... and so on, with little enlightenment for anyone seeking simple practical guidance.

Posted by Noetica at November 13, 2007 07:19 AM

in Spanish reventón means 'burst,' or—in an automotive context—'blowout, flat tire.'

Isn't that exactly the translation of 'Jeremy Clarkson' ?

Posted by Richard Parker at November 13, 2007 07:49 AM

Noetica: "I rather prefer the names drawn from music: Concerto, Sonata."

I used to drive a Fugue, but now I have no idea where it's parked.

Posted by HP at November 13, 2007 05:34 PM

Nice one, HP. How about the new Canon Cancrizans? Parks sideways.

Posted by Noetica at November 13, 2007 06:03 PM
Ironically, the underlying Basque word is Bizkaia, where the written "b" firmly represents [b].

Does it? I've read Basque does the exact same thing to intervocalic stops as Spanish. Some even think the modern Basque /b/ is a merger of earlier /b/ and /w/, just like the Spanish one is a merger of Latin /b/ and /v/.

And while I am at it, what is is this [fn] nonsense? [ˌluːd̥vi̽g̊ˌfanb̥etˈhoːfm̩] (or -[ˈb̥eːthofm̩], and never mind [nb]-to-[mb] assimilation), [ˌvɔlfg̊aŋamaˌd̥eːʊsˈmoːt͡saˑt]. Turning -/ɪg/ into -[ɪç] is a regional thing of northern Germany, and the ability to pronounce an [aɐ̯] diphthong is restricted to... maybe Dortmund. I really can't imagine anyone sees an eu in Amadeus in terms of German phonology...

Posted by David Marjanović at November 13, 2007 07:48 PM

Sure, David. As I say, these things are a mess. IPA has been no solution at Wikipedia, for all sorts of reasons.

Incidentally, if in American "o" in "top" is regularly pronounced the same as "a" in "father", "obbligato" could be glossed this way, chez les vespuciens: oeb-bli-GO-toe. Why not? (No big problem with this "-GO-" being read as sounding like the word "go", surely. It could be given as a strict rule that no syllables are rendered in such a way, so that an "o" unaccompanied by any other vowel letter could efficiently represent the sound that it most often has in use.)

Posted by Noetica at November 13, 2007 08:02 PM
IPA has been no solution at Wikipedia, for all sorts of reasons.

One is, I suppose, that you can't look at a transcription in brackets and tell how "narrow" it is. (I have been very pedantic above.) Another is sheer lack of knowledge about how IPA works; no surprise there, it simply isn't taught anywhere except in linguistics courses at universities. I had to teach myself online.

Posted by David Marjanović at November 13, 2007 08:12 PM

And then there are university courses that teach mistakes. There are audio files online that will teach you a slightly fronted velar stop, like French /k g/, is "palatal".

And then there are traditions of "broad phonetic transcription" that... ah, whatever, it's a quarter past 2 at night, I've lamented enough.

Posted by David Marjanović at November 13, 2007 08:15 PM

[incidentally: in mexican spanish reventón means "big party", maybe even an all-nighter... eg: anoche me fui de reventón: i partied hard last night!]

Posted by alonso ruvalcaba at November 15, 2007 03:18 PM

Thanks -- I love the wide regional differences in Spanish!

Posted by language hat at November 15, 2007 03:58 PM

"Blowout" can refer to a big party (or a big sale) in English as well, at least in the US.

Posted by KCinDC at November 16, 2007 11:38 AM

And then there's French: la boum...

Posted by David Marjanović at November 16, 2007 02:48 PM

Two comments: Spanish gets its fricative pronunciation (under some theories) of B's, G's, and D's from Basque, and in modern Basque, there is only B (no V's natively), which sounds just like the Spanish bilabial fricative, with all the same restrictions ([b] at the beginning of utterances and before M, fricative version between vowels, etc.).

Incidentally, this pronunciation is old: note the Latin saying "Beati Hispani quibus bibere vivere est" - "Happy are the Iberians (or whatever, let's not quibble) for whom to drink is to live."

Second comment:
A lot of people regularly pronounce the letter V as [b] in English in one word - Mojave as [mohabi] (rightly or wrongly).

Posted by Homo Œconomicus at November 18, 2007 01:16 AM

Interesting—I didn't know about that saying.

Posted by language hat at November 18, 2007 07:50 AM
Beati Hispani quibus bibere vivere est

:-o Wow. Wow! How old is that?

(I fear it's medieval. Many Latin proverbs are -- but usually they are easy to spot, because they rhyme, like plenus venter non studet libenter.)

Posted by David Marjanović at November 18, 2007 07:19 PM

I have seen this observation credited to Scaliger. E.g., 1911 Encyclopedia s.v. Spanish Language. The version I remember (well, after a little refreshing from GB) is:

Haud temere antiquas mutat Vasconia voces,
Cui nihil est aliud vivere quam bibere
from Burton in Brasil, here. Likewise Max Müller. A search does turn up a variety of other forms attributed to him, though, some more similar to the one Homo Œconomicus gives above.

In a discussion of this pronunciation in De Causis Linguae Latinae i. x., he refers to an epigram. I can't figure out how to deep-link to a page on the UCM site, but the book is here: Ir a pagina 43; and the image inside the frame is here, but no navigation.

And indeed that one is quoted in the posthumous Poetices: iii. cxxvi., but with non not haud.

I think Neo-Latin is the only language with a comprehensive directory of most every digitized book, The Philological Museum, so finding copies of these books online was a comparative breeze.

Maybe the name is why the internets attribute “Beati Hispani quibus bibere vivere est.” to the other Julio César. Of course, it's possible, even likely, that it was said many times before in different forms.

Posted by MMcM at November 20, 2007 06:03 PM

Wow, thanks a lot!

BTW, is the quote an elegic distichon...?

Posted by David Marjanović at November 20, 2007 07:23 PM

Yup, and I add my thanks.

Posted by language hat at November 20, 2007 08:09 PM