Monday, 20 June 2011

errata typographica

The Deutsche Aussprachewörterbuch that I was writing about on Thursday and Friday last week is — as far as I can tell — pretty free from typos and careless errors. (I speak from experience when I say how difficult this is to achieve in a book such as this one, full of complicated phonetic symbols.)

It is all the more dismaying, then, to have to report two phonetic symbols that are wrongly set, and not just once. Fortunately they are used in the preliminary matter (which is nearly three hundred pages long), not in the body of the dictionary.

One error concerns [ɤ], the ‘ram’s-horns’ symbol which the IPA prescribes for the mid-close back unrounded vowel, secondary cardinal 7. This is erroneously written “[ɣ] “ passim in the discussion of Chinese vowels, p. 130-132 — with exactly the same symbol as the book uses (correctly) for the representation of the voiced velar fricative, Greek γ, and of the Spanish intervocalic g in the discussion of those languages, p. 165 and p. 209. This is an error rather often found in printed material (my blog, 24 Sep 2009)

The other error concerns the familiar symbol ʊ, the lax close back rounded vowel of English foot fʊt and German Bucht bʊxt. This is fine when on its own. The corresponding non-syllabic symbol is not required in dictionary’s transcription of the standard German of Germany, since — as we have seen — the diphthong of Haus is transcribed not as aʊ̯ but as aɔ̯. In the discussion of Austrian pronunciation, however, the dictionary reports (p. 238) that the diphthong of Haus can be [ao̯] or “[aʋ̯]”. That is, it represents the second part of the latter diphthong with the symbol that properly stands for the voiced labiodental approximant, ʋ, plus the non-syllabic diacritic. This “[aʋ̯]” reported for Styria and western Austria, and optionally for die gehobene Standardaussprache (elevated standard pronunciation), ought to be [aʊ̯].

Where the same symbol is required in the discussion of Swiss diphthongs (p. 263), it is correctly set as [aʊ̯]. It is also correct in the body of the dictionary, where Eusebio is given (without discussion) as “span. ɛʊ̯sˈeːbi̯oː” and Eusébio as “port. ɛʊ̯zˈɛːbi̯uː”. So it seems likely that the error is down to the contributor responsible for the Austrian section, named as P. Wiesinger. But why didn’t some editor pick it up?

For the avoidance of confusion:
[ɤ] and [ʊ] stand for vowels, [ɣ] and [ʋ] for consonants.

As you would expect, the dictionary reflects the 1996 German spelling reform. You will find dass rather than daß, Tipp rather than Tip, Balletttänzer rather than Ballettänzer. In this connection I was also going to check on the spelling of the word for ‘nut’, which used to be Nuß and is now Nuss. But I found neither spelling! Somehow that word has got omitted entirely, along with such compounds as Nussbaum, Nussknacker and Nussschale, all of which presumably ought to have been included. (But I did find the compounds Haselnuss and Walnuss.)

People who live in glass houses… the first edition of my LPD somehow didn’t include the word marathon. I’m sure the DA will get its nuts sorted out when there’s a new edition.


  1. I am surprised that the Dictionary interprets (German?) long vowels (in 'Eusebio') into languages that simply don't have vowel quantity, Spanish and Portuguese. Germans usually imagine Italian has vowel quantity, which it has not (due to the habit of most Italians of lenghtening stressed vowels when under logical or emotional emphasis) but that now even the Iberic languages should have it... .

  2. Italian doesn't have vowel quantity?

    This German dictionary is a good book and a laudable project. But there are a few but's about it. Like the already mentioned fussiness, these new diphthong transcription choices and the fact that, unlike Max Mangold, it only has germanized pronunciations – which is a big minus in my book.

  3. Ad Dinora

    Italian has no vowel quantity. It has consonant quantity, a.k.a. consonant gemination, but no vowel quantity. Neither does Spanish nor Portuguese. This does not mean that in these languages all vowels have identically the same length, because much depends on emphasis (logical or emotional), for instance, Italians often lengthen stressed vowels, especially in open syllables, when the word comes last in a sentence, or before a comma, and/or is the bearer of some sort of emphasis. But this is a phenomenon different from what standardly goes under the name of 'vowel quantity'.

    People whose native language has a trait, such as vowel quantity, have naturally the tendency to 'hear it into' other languages which do not. I have known Chinese persons who believed that German had word-tones, like Chinese... . Well, in a sense it does; depending on speech-context, Germans pronounce their words high, low, rising or falling or undulating... true enough, but this is not what Chinese tones are about. German, in its turn, has vowel-quantity, so Germans tend to 'hear' short and long vowels in Italian, and when they speak Italian they apply the imagined 'rules', making their Italian sound funny. I have no problem with that; what I think is mistaken is making a principle of such mis-interpretations and building them into API-transcriptions.

  4. Wojciech, you're so right. I just like to analyse the foreign accents of unsuspecting people around me to see how many traits I can identify from their native language affecting their unique expressions of English. If you pack a sandwich and drink, you can whisk off downtown to make a fun afternoon of it. LOL! (No seriously, my life isn't that boring, I swear!)

    But seriously, on the main topic now, it seems like computers have to get to the next level to iron out these typographical problems. We take for granted spell-check programs but should we not also have IPA-check programs designed to identify potential goof-ups in IPA? The example of [aʋ̯] seems to be an obvious and easy example that could be weeded out by such a program since naturally consonants shouldn't need the non-syllabic diacritic.

  5. Hold on, just realized that maybe /ʋ/ may be considered syllabic in some languages. Duh! I'm sleepy today, sorry! At any rate, an IPA-check program would still be handy if it could be coded somehow.

  6. Of course Italian vowels can be lenghtened for all kinds of paralinguistic reasons (as I guess they can in most other languages), but the lenghtening of vowels in stressed open syllables is an honest-to-goodness allophonic phenomomenon which occurs even in completely ordinary speech. IMO in tufo ['] / tuffo ['] the difference in vowel duration is almost as obvious as that in consonant duration, and ['] with a short vowel would sound ‘weird’ to many speakers (though not many of them would misinterpret it as tuffo). I wouldn't omit the length mark except in broadest transcriptions (i.e., those where [ŋk] is transcribed as /nk/).

  7. Ad Army1987

    I am not a linguist so I can't argue with you on a truly scholarly level; my opinion on the matter is based on years of observation, talking over the matter with Italians, including linguists, plus familiarity with languages which beyond doubt do have vowel quantity.

    No doubt Italians have a tendency to lengthen their stressed vowels, when they utter the word separately, for illustration purposes, or when they lay logical or emotional emphasis on it, and such. This varies from individual to individual, I know e.g. a guy who can't say 'noi' (we) without drawawawling it: nawawawi. But there is nothing incorrect in saying simply 'noi'. Or imagine a sentence like: 'In Italia ci sono moltissimi partiti, uno tra loro e` il partito communista' with EVERY stressed vowel lengthened, not just 'partiti' (before a pause) and 'communista' (final). It would sound ridiculous, or rather like mocking Italian, holding it up to ridicule (in fact, this _is_ done sometimes for this purpose). In normal speech, 'partiti' and 'communista' would get their stressed vowel slightly lengthened (my friend would say 'parteeeeeti', 'communeeeeeesta'), but even without lengthening the sentence would sound perfectly OK.

    Re tufo/tuffo --- in the Milanese dialect there is in fact a distinction like that, tufo has a long vowel and tuffo has a non-geminated consonant before the vowel.

    Italians also sometimes 'sing' their words, quite often in fact. Shall we say they have tonal accents, like the Serbs or the Swedes? Clearly, no.

  8. the first edition of my LPD somehow didn’t include the word marathon. I’m sure the DA will get its nuts sorted out

    So between you, you got your Snickers in a twist?

  9. I meant stress as in intonational-group-level primary stress, not as in word-level stress, Of course partito would receive no primary stress in such a sentence, but neither would English Party in many parties, amongh which the Communist Party.

    As for the lack of lenghtening in comunista, that's one of the reasons why I'd syllabify it as /'nis.ta/ rather than /'ni.sta/ as the hyphenation rules suggest (and I know I'm not the only one).

    even without lengthening the sentence would sound perfectly OK.
    Huh, that also depends on what you mean by OK. Would English feel pronounced an unvelarized L sound “OK” to you?

    Italians also sometimes 'sing' their words, quite often in fact. Shall we say they have tonal accents, like the Serbs or the Swedes? Clearly, no.
    Well, that depends only on the general intonation of the sentence, not on the word itself. OTOH tufo would have a longer vowel than tuffo even in othewise identical sentences with otherwise identical prosody. (BTW, of course I'm not claiming that Italian has phonemic vowel length, only that is not an exclusively paralinguistic phenomenon any more than, say, de-aspiration of stops after /s/ is in English.)

  10. Ad Army

    Thank you.

    Again, I must explain that I am no linguist.

    'Communista' would, according to my observation, get its vowel lengthened before a pausa, or when placed emphasis on, for most speakers, most of the time. The 'theory' of those who read vowel-quantity into Italian is that stressed vowels get lengthened only if in an open syllable. But according to my observation, this is not true. I have recently cooperated with an Italian partner, this went on for months, we were communicating in Italian and for some subject-matter-related reasons he had often to use the word 'norma'. It struck me how often he lengthened the vowel: 'nawawrma', but only before a pausa or with emphasis on it.

    'Feel' with a 'light' l, if that is what you mean, would of course sound foreign, non-English (at least in the accents of English I am familiar with). But I insist---again, based on observation---that if you 'just' stress 'i' in 'partito' or 'o' in 'norma' without lengthening it, you don't sound foreign. (Unless for native speakers of northern dialects, perhaps.)

    There so many awesome parties in Britain, amongst these the Conservative Party---in this sentence, methinks, no matter how carelessly pronounced, you'll always hear the _relative_ vowel length in 'awesome' and 'parties', and 'these' and 'Conservative', in the latter I mean the 'er'-vowel. But saying 'In Itaalia ci soono molti partiiti, tra cuui anche il ceelebre Partiito Communista' is ridiculous (or making Italian appear ridiculous, caricature-like, plus italien que le Pape, mocking it ... or maybe something for a publicitity spot). I know an Italian, though, a 'big shot' in the academic world, who (because of his self-importance?) drawls the stressed the vowel of every second word, so he might pronounce that exactly that way....

    As I said, I am not a linguist, so I cannot really say if the factual lengthening of stressed vowels in Italian is para- or something-other-linguistic in this or that sense. In my perception, though, it lies on a different level of language (speech?) than the real vowel-length in, say, English, or Czech, or Finnish, or German. It's rather like speaking in a 'whiney' tone of voice which (it is sometimes said) many Russians are in the habit of using for their native idiom (and sometimes for others). You can speak non-whineily and thereby speak not exactly like many or most Russians, but speak perfect Russian nonetheless, without sounding 'weird' or 'foreign' or sumpin'....