Archive for the ‘Phonetics’ Category


Now I Know My [ei] [bi] [si]’s

Saturday, November 29, 2008

Today we finally made it for the first time on this trip to the Proveedora Escolar, a fantastic bookstore in the centro (It seems funny to me to refer to this area as downtown Oaxaca, although it is, since it doesn’t fit my conception of “downtown” for a city the size of Oaxaca. Instead of having modern skyscrapers and buildings, it is a beautiful colonial area; most buildings are only two, maybe three stories tall, with church domes and towers poking up a few stories higher). At the bookstore, we bought a couple books for Benjamin and a couple foam puzzles. One is really for me: a puzzle in the shape of the state of Oaxaca with the individual districts making up the puzzle pieces. The other one is an alphabet puzzle with foam pieces in the shape of letters. This is the second one of these we’ve bought for Benjamin-the last one, he ripped many of the letters into little pieces and we left behind what’s left of it in Macuil. But recently, he’s taken to singing the alphabet song-where he picked it up, I’m not sure-and seems more interested than ever in letters, so we decided to get him another alphabet puzzle.

These alphabet puzzles have a couple of interesting features. One, they only seem to come in the English alphabet, I haven’t seen a Spanish one yet. True, there’s not much difference in the English and Spanish alphabet. If ch, ll, and rr aren’t counted as distinct letters, that only leaves the existence of ñ in Spanish as the only difference between the alphabets. Still, I have yet to see a version that included ñ.

Even more interesting to me, though, is the pronunciations of the letter names that are stamped on each of the letters. They are intended to represent the English names of the letters, which is a fine goal I suppose. At first glance, I thought the names of the letters were being written according to Spanish writing rules-it would be like telling you to pronounce hola as you would pronounce something written owe-lah in English. This makes sense: a Spanish speaker who didn’t know English could buy the puzzle for their kid and make a stab at pronouncing the English names of the letters. This hypothesis works for the first six letters of the alphabet which are stamped with ei bi si di i and ef. Things get a little strange after that and I eventually realized the letters are stamped with rough, not entirely accurate, IPA (the International Phonetic Alphabet) representations of their pronunciations.

The first clue is the stamp on the letter g. It’s given as dyi (and similarly j is stamped with dyei). This is an odd choice to help Spanish speakers. A combination of d and y aren’t really used in Spanish. Of course, Spanish (and here and throughout I’m talking about Mexican Spanish) doesn’t exactly have this sound so something must be done. But the letter y often represents a sound that is close, though it sounds somewhere between an English y-sound (as in you) and the j-sound (of Jill)-these sounds aren’t distinct in Spanish, which can trip of those learning English when they are faced with pairs of words, like yellow and Jell-O, which are identical in pronunciation except for these consonants. Anyway, dy is not a totally unreasonable representation of this sound and maybe the d was added to make sure the y comes out more like English j than like English y.

But there’s some other odd choices on the stamps that this doesn’t explain. The alphabet elsewhere uses the letter j to represent the English y-sound. Thus, q is stamped with kju, u is stamped with ju and w with dobelju. This doesn’t really make any sense for Spanish. In Spanish, the letter j typically represents a velar fricative like the German ch in Bach or is often closer to English h in house. This would be a totally misleading way to stamp the names of the letters. But it makes sense if you are following IPA conventions, because in the IPA [j] represents this sound. So, in a broad transcription of English alphabet names [kju] and [ju] make perfect sense in IPA. (Dobelju has some problems, which I’ll come back to.)

The real clincher to me that this is supposed to be IPA is the name for h. It’s given as eift, which doesn’t work in English, Spanish, or the IPA. But it does look like the IPA with a few mistakes thrown in. If we were trying to spell the name of the letter h following Spanish conventions then it should have been given as eich, since Spanish represents that final sound with ch. The IPA representation of the name would be [eitʃ], which is starting to get us in the ballpark of what’s stamped on the letter. The IPA transcription involves a funky symbol [ʃ], which apparently the makers of the alphabet didn’t have access to. So they used (or their keyboards without the proper font installed used) something that’s visually kind of close: f. Then it looks like that working with symbols they weren’t familiar with, they accidentally switched the t and f, not that eitf would have been clear to anyone either.


This now explains g and j: the initial consonant in their names would be [dʒ] in IPA. Again, we don’t have the exact symbol, so something close (for some value of close) was put in, in this case y. Similarly, the name for w contains many sounds that aren’t in Spanish and are represented by special characters in IPA. With a broad transcription, the name would be rendered something like [dʌbəlju]. How this would get to what’s on the letter (dobelju) is not clear. Perhaps that initial vowel was represented with [ə] instead of [ʌ] or maybe the Spanish word doble, which is part of the letter name in Spanish influenced the choices. Finally, I guess it should be noted that the names of the letters f, l, m, n, s, and x would begin with [ɛ] in the IPA, but begin with e in the puzzle.

The surprising thing about all of this is that since there is no pronunciation guide offered on the puzzle mat, even if the manufacturer had gotten the IPA symbols right, it would require familiarity with the IPA, but it seems to me that familiarity with the IPA here must be as about as common here as is it is in the US, which is not very. And I don’t think it was intended for another audience, such as Spanish speakers from Spain, who might (I guess) be more familiar with the IPA, since it seems to be trying to capture standard American English pronunciations of the letter names, as evidenced by the r-ful pronunciation of r as ar and the inclusion of zi as the name for the letter z instead of zed.

As an interesting side-note, one of the books we picked up for Benjamin was an English book Times 1000 Word Dictionary by Myra Ellis which also includes IPA transcriptions of words. It, however, includes all the funky symbols and a pronunciation guide and seems generally accurate, although it is not based on an American standard dialect (it was published in Singapore) as it includes words like lorry, pronunciations like [mɛdsn] with two syllables for medicine, and non-rhotic variants as seen in [mɑ:tʃ] for march and [ˈfɪŋgə] for finger among many others (for a nice map on the distribution of r-lessness see here).

Overall, it was a very good outing with two linguistically interesting finds in one trip.