Your link does not work for me Martin. How many vowels did the winner ultimately have?

(and how are your little ones in a complete aside)

Perhaps enough to want to make the others cry "vowel!" ?

Wie kann man die vowel variety untersuchen und ausgerechnet die Nasale nicht gelten lassen? Ist nasal keine variety?

Bei mir funktioniert der Link übrigens auch nicht.

versucht's ohne Klammer und Punkt:

http://wals.info/feature/description/2

http://wals.info/feature/description/2

Noch ein Versuch mit dem Link.

Only one language in the sample, German, uses 14 vowel qualities and only 2 make use of 13, namely the variety of British English included here and Bété ( Kru, Niger-Congo; Côte d'Ivoire).

Aus dem Text.

In the linked text it says German had "No glottalized consonants" - can someone give me an example for a glottalized consonant?

No one has made clear whether we are talking just about phonemes. If we are, I think English and German both have 20 vowel phonemes . It does of course depend on the variety of German/English under discussion, and quite what you define as a phoneme.

Birgila: it is difficult to do this in writing. What it (usually) means is, a consonant pronounced with a simultaneous glottal stop. In this sense, German certainly has no glottalized consonants.

Incidentally, the exclusions used in the original list are totally arbitrary, and misguided. Length alone is rarely a distinctive feature, and certainly not in either English or German.

Even more reason for the others to cry "Vowel!" :-)

@escoville: We are clearly not talking about phonemes. The text that Martin referred to explicitly states that it discusses vowel quality alone.

I disagree with your statement that it is arbitrary (or even misguided) to exclude vowel quantity and nasalisation. The discussion of vowel quality does not deny the obvious fact that, for example, in some languages there is an opposition between nasalised and non-nasalised vowels of the same quality (like in French 'bon' and 'beau'), which makes them different phonemes.

It is possible without difficulty to count German [a] in 'Stadt' and [a:] in 'Staat' as the same vowel with a view to quality. But you couldn't say that [ɪ] in 'Bitte' and [ʏ] in 'Hütte' are the same vowel. In German, there is only short [ɔ] (as in 'floss') and long [o:] (as in 'Floß'), but in Italian, both constitute different phonemes, cf. http://en.wikipedia.org/wiki/Italian_language... ('botte').

In German, as a rule, short and long vowels have different qualities. You can argue that quantity is the primary feature and quality the secondary feature, but that approach is contestable, cf. http://de.wikipedia.org/wiki/Aussprache_der_d.... And Czech is an example of a language where quantity alone is a distinctive feature, cf. http://en.wikipedia.org/wiki/Czech_language#Vowels. (Interestingly, Czech and closely related Slovak are the only Slavic languages that distinguish short and long vowels.)

The bottom line is: There is nothing wrong with choosing quality as the criterion of identity for vowels.

I'm not sure I understand your point, Frank.

Unless perhaps you mean that many of the German vowels called 'long' and 'short' actually do differ in quality, so the names 'long' and 'short' don't actually refer to quantity in the sense of time anyway, they're just arbitrary names. (And in fact some textbooks call them 'open' and 'closed' vowels, especially for E and O, which makes more sense to me.)

That is, for most German vowels, the 'short' versions are indeed formed differently in the mouth than the 'long' versions; they're not just the exact same vowel sustained for a longer time.

I didn't click through to the study, sorry, but if their count is 14, it looks to me as if they must be counting so-called 'short' and so-called 'long' as different qualities anyway, except in the case of Ä, which really differs from short E not at all, except that long Ä is sustained longer. So:

Long A, short A,
long E, short E (= short/long Ä),
long I, short I,
long O, short O,
long U, short U (10),
+ short Ö, long Ö,
short Ü, long Ü (4).

Maybe actually it's only the names 'short' and 'long' that are confusing here, because those terms don't actually refer primarily to duration in the sense of time (except in the case of long Ä).

Not knowing Czech or Slovak, the language that came to my mind as an example where vowel quantity in the sense of duration really constitutes a phonemic difference is Latin. I may well be totally wrong about this, because I never really studied Latin. But as I very tentatively understand it, there are really only 5 vowels in classical Latin (or is it medieval/church Latin I'm thinking of?), but in things like poetry they really were sustained different lengths of time. Like beats of music, two beats for long vowels and one beat for short vowels, or something like that. (To hear this, you have to use the Italianate pronunciation, not the German one, which uses German vowels. But it may not work even then if it's too Italianate, because modern Italian also has more vowel sounds.)

As for why they don't want to use the word phoneme, I'm not up on linguistics any more than I am on Latin. At first I thought the same as escoville: if they form minimal pairs, they're phonemes.

But I wonder if it's because phonemes would also include diphthongs, which aren't single vowels.

Sorry for the broken link. Here is the address of the home page:
http://wals.info/
To see the text with the statement about the number of distinct vowels in various languages, click on "Features", then click on Section 2 "Vowel Quantity Inventories".

@escoville, unfortunately the site did not count the languages in which vowel lenght is a distinctive feature, but Wikipedia has some words to say on this issue: http://en.wikipedia.org/wiki/Vowel_length

@Birgila, Wikipedia also has something to say about glottalized consonants as well: http://en.wikipedia.org/wiki/Glottal_stop.
As they point out, in many varieties of English (especially BE), glottal stops, although not necessarily "standard", are frequently heard. For example, the /T/ in "whaT are you doing?" is often rendered as a glottal stop.

So Martin, I guess what I'm saying is that I don't think they did ignore 'Staat' vs. 'Stadt,' because that's also a difference in quality, not only in length.

What they ignored is the difference between the vowel in Segge (short), Säge (long), and Säcke (short). They're counting all of those as just one vowel, and indeed, the quality of that vowel (how it's formed in the mouth) doesn't change, only its duration does.

(Sorry, a 3-way minimal pair was beyond me, maybe someone else can think of one. (-: )

But all the rest have different qualities.

*f5* Oops, hi, Martin. Good to see you around again. (-:

Sorry, one more thought: I still don't understand why they didn't count nasalization as a quality, though. Sure seems like one to me, and that would raise the count for both French and Portuguese pretty sharply.

@hm-us. Responding to post #12:

I arrived at a count of 14 in a different way. Your list omitted the schwa, which I thought had a different "quality" than the others.

I assumed that they included the schwa, and treated your "long A" and "short A" as differing only (or primarily) in length, and therefore counted them as a single vowel, as far as vowel "quality" is concerned. (All the other members of your long-short pairs differ more markedly in "quality"; i.e., in the position of the tongue, jaw, and lips.)

@hm-us (#15): ...and if they counted tone, it would raise the count of Cantonese vowels very dramatically.

What I guess is that the author considered the positions of the mouth (tongue, jaw, lips) as being more essential, and that vowel length, tone, and nasalization were "accidental" modifications of the basic mouth shape.

But certainly, as he clearly admits, there are other ways to count vowels.

@hm-us Re #14): Trying for a three-way minimal trio with
Fälle / Felle / fehle. But I think in most German dialects Fälle and Felle are pronounced the same. and were not counted separately in the WALS article. (The goal could be a four-way minimal quartet with a rhyme for *Fähle.)

By the way, what do you think are the 13 distinct English vowels? (I only counted 12 in my N.Y. dialect, and I make the merry/Mary/marry distinction.)

Yes, I too didn't count long and short Ä as separate; in fact, I counted them both as the same as short E, which had already been counted.

But you're absolutely right, I didn't count the schwa at all. So wouldn't that be 15? What would you leave out? Or have I just miscounted somewhere? Math is not my strong point these days ...

You're also right about pitch; I too would consider that a quality, at least, in languages that define it as such.

But I suppose part of the question here is who gets to choose the definitions. (-:

Re #18: Nope, 'fehle' doesn't cut it because long E is definitely a different quality, so it doesn't fit with the other 3.

But if 'Fähle' existed, it would be perfect. Are we sure it isn't a really really obscure subjunctive? For example, 'Er sagte, dass ihm noch ein Vokal fähle.'

*unschuldigguck* Works for me. (-;

Re 13 English vowels:

1 long A bait
2 short A bat
3 long E beat
4 short E bet
5 long I bite
6 short I bit
7 long O boat
8 short O bot (= other long A BE bath, AE father)
9 rounded O bought (for those with cot/caught distinction, i.e., not me (-; )
10 long U boot
11 short U but
12 other short U book (damn, I was doing pretty well with the minimal dodecatuplets [?] till then -- can I count northern BE but as in Liverpool? (-; )
13 schwa sherbet (also not brilliant, oh well ...)

Sorry, I really ought to be going to bed. But it's been fun. Don't be a stranger. (-:

... but some of those, upon belated reflection, are probably in fact diphthongs, in which case I give up. Sorry, over and out. (-:

My count:

IE of diese
I of ich bin
EE of Beet; E of bete (the first E); EH of Fehle; etc.
E of Bett; Ä of Männer; etc.

UH of Uhr
U of Kummer
OH of ohne
O of Opfer

ÜH of Hühner
Ü of Hülle
ÖH of Höhle
Ö of Hölle

A of Stadt and AA of Staat
E of Hülle, Höhle, Hölle, etc.

RECOUNT: With hm-us's hints, I counted once again, and I now think my variety of English has 14 vowels as well...

(Anyone from New York to confirm my count?)

beet / heed /
bit / hid / -
bait / hayed / -
bet / head / merry

boot / who'd / moory (unknown to me until just now)
- / hood / -
boat / hoed (past tense to "to hoe"/ -
bought / hawed (as in "hemmed and hawed") / -

but / hud /Murray
bot / (pod) / (Ari; sari)
bat / had / marry
- / (bad) / Mary
Bart / hard / (sorry)

Plus schwa.

Ich verfolge diesen Faden fasziniert und zuweilen leicht überfordert...ist aber sehr interessant!

Was bitte ist "schwa"??? Könnte mir das mal jemand mit Beispiel erklären?

Very interesting. They didn't count phonemes (minimal pairs) but just quality, defined as height (how far your jaw drops), tongue position (back-middle-forward), and lip position (open/relaxed to closed). This may be debatable (the Wiki-link "Aussprache der deutschen Sprache") provided by Frank actually discusses the merits or shortcomings of a quality/quantity approach).

I actually arrive at 15 different vowel qualities in German (after some pondering), and I think that hm--us's count is not correct, even with Martin--cal's substitution of schwa for a/aa. Then, in #23, Martin--cal counted a/aa as different qualities and arrived at 14, but I think that's quantity alone (which doesn't count).

Here is what hm-us had, with possible corrections:

Long A, short A, (no difference in quality, IMO)
long E, short E (= short/long Ä), (something missing here!)
long I, short I,
long O, short O,
long U, short U (10),
+ short Ö, long Ö,
short Ü, long Ü (4).
(Add the schwa.)

Accounts for 13. Martin--cal's list in #22 amounts to 13 as well (the a/aa is not a distinction in quality). It took me a while to find the missing one because it's not a phoneme, as far as I can say, but still a different quality according to the above-mentioned 3 criteria.
(I'll split this)
1/2

2/2The e/ä sounds are the most difficult, so I'll put them at the end. Sorry I can't get the IPA-characters in, but I'll try my best.1. [a], [a:]. Stadt, Staat. These are two phonemes, but the formation of the vowel is the same in both cases. The distictive feature is length (quantity). One vowel.2. (without the dot in IPA), [i:]. Irre, ihre. Different phonemes, but also different vowels.3. Same for the "o"s in offen, Ofen. Note that the "O" [o:] of Ofen has a short variant in unstressed syllables, [o] as in Roman, tomistisch, modern (engl. "modern", not "to rot away"). The difference is length, not quality.4. Same with "u". Krume, krumme.(adj. fem. nom.) (Or, "Uhr" and "Kummer" from Martin--cal, but I wanted a minimal pair.)5. Two more vowels with Höhle, Hölle.6. And two more with fühle (v. 1st sg.) and Fülle.Do you agree that makes 11 vowels which are distinct in terms of lip/tongue/jaw-position (or rounded/unrounded, forward/back, open/closed)?7. The "e" and "ä" in spelling are difficult to describe in pronunciation. There is a long "ä" such as in Ähre, and an [e:] in Ehre, but the short variants amount to the same vowel. Or don't they? a) [e], [e:]. One vowel, the only distinction is quantity. The short [e] appears in unstressed syllables. The initial "e" in Genom and in genommen is pronounced differently, and the first one is a shortened variant of the "e" in gehen. It is my suspicion that the shortened closed "e" [e] as well as the "o" [o] as in Romantik are limited to latin or romanic words which found a place in the German language but that's just a guess.b) The sound which distinguishes Messe and mäße (v., Konj. Imp.). Again, same formation, different quantity, and one vowel. The sound can be spelled "ä" or "e" in the short variant. Words: Käse, essen, Messer, erblich, ändern, ärgerlich, lässig, mäßig. In IPA, its a Greek epsilon, followed by : for length.c) Now this is a contest. Is the last one the "schwa"? It must be. It sounds like a vowel which differs from other e-sounds. In formation, it is very much like the "ö" in "öffnen", just slightly less stressed. And the authors compared more than 500 languages, so English or German preferences didn't play a predominant part. The "schwa" has got an extra character in the phonetic alphabet, so it is a sound on its own. I'm not sure, but if they counted the "schwa", they missed something else, and that is the presence of the sound "ae combined" in the German language. It's deeper than the short "e" in 7b, and it occurs in words like sterben. Compare that to sterblich (or ärgerlich); these are clearly different sounds. But the authors of the map may be justified in their opinion that the sound in sterblich is, in terms of vowel formation, just the short variant of the first vowel sound in Käse.Did I miss anything?

@ Goldammer, #25: Als "schwa" oder "schwa-Laut" wird ein unbetonter Vokal bezeichnet (meist ein E). In der internationalen Lautschrift wird er durch ein auf den Kopf gestelltes E gekennzeichnet, und so ist das auch in Wörterbüchern, die Angaben zur Lautschrift enthalten. Beispiele (engl) above, alone, (dt.) Erde, lassen. Die fett hervorgehobenen Laute sind schwas (falls ich die Formatierung hingekriegt habe). Das "schwa" ist eher ein ziemlich tonlos hingeseufzter Vokal als ein wirklich tontragender Vokal, und er taucht im Deutschen meist beim E auf. In anderen Sprachen können auch andere Vokale zu diesem Seufzlaut abgeschwächt werden.
Soweit klar?

I apologize for #27. This is unreadable. I had clear paragraphs in what I posted and I don't know why LEO ignored it. I can try to post it again, but can you figure it out?

Hi Sebastian,
danke für die Erklärung von schwa. Kapiert. Das umgekehrte e aus der Lautschrift ist mir vertraut.
Noch eine neugierige Frage hinterdrein: warum um Himmels Willen heißt das Ding "Schwa"?? Ich finde das eine nette Bezeichnung, hatte sie tatsächlich bis heute noch nie gehört - aber woher kommt das Wort?

re #27:
Mich hat die absatzlose Formatierung abgeschreckt, alles zu lesen ... vielleicht keine schlechte Idee, es nochmal mit Absätzen zu schicken.

Was mich noch bewegt: in meiner dialektgefärbten Aussprache gibt es keinen Unterschied zwischen dem kurzen E und dem schwa. "Ende" z.B. hört sich für mich vorne und hinten gleich an...hier könnte ich noch Hochdeutsch lernen, wenn ich diesen Faden noch lange genug mitverfolge...

PS OT
*neugierig sei*
Sebastian, ist eine der Sprachen im Profil deine Muttersprache? Da steht alles, was du sprichst, aber nirgends Muttersprache...

sebastianW, das ist mir in der letzen Woche auch zwei-, dreimal passiert, dass LEO mir die Absatzformatierung geklaut hat! :-( Besonders ärgerlich ist es bei so ausgefeilten Beiträgen, wie es deine immer sind.

Zum geschlossenen /i/ noch eine Ergänzung, obwohl es sich für das WALS um denselben Vokal handelt: Es gibt es eine kurze und eine lange Variante:

Wien /vi:n/ - Vitamin /vi .../ (meine mich zu erinnern, dass das kurze, geschlossene /i/ nur in Fremdwörtern vorkommt.)

Oben hatte schon jemand erwähnt, dass das schwa nur "meist ein 'e'" sei. Ich würde noch das a-schwa als eigenen Vokal zählen (das 'a' auf dem Kopf, der Zungenrücken ist höher als beim /a/, das wir z.B. am Wortende für '-er' sprechen) ():

Butter, meiner (im Gegensatz zum e-schwa: meine), etc.

@Goldammer: The word "schwa" comes from the designation of one of the vowels in Hebrew.

@sebastienW: Since I'm not a native German speaker, I don't think it makes sense for me to comment on questions of German pronunciation. Thanks for your contribution, though.

Here's the IPA designation for schwa: ə. (I hope this works!)

Assuming that that worked, you can get IPA symbols into Leo by cutting and pasting from a list at this URL: http://www.theiling.de/ipa/

I don't want to be a nuisance, but I'll try to post #27 again (with a couple of corrections, epecially at the end). I changed the brackets to (), maybe they were interpreted as formatting. If it doesn't work this time I'll give up.

The e/ä sounds are the most difficult, so I'll put them at the end. Sorry I can't get the IPA-characters in, but I'll try my best.

1. (a), (a:). Stadt, Staat. These are two phonemes, but the formation of the vowel is the same in both cases. The distictive feature is length (quantity). One vowel.

2. (i) (without the dot in IPA), (i:). Irre, ihre. Different phonemes, but also different vowels.
Please add what Ingeborg wrote in #32: Vitamin has a short variant of the vowel quality in ihre.

3. Same for the "o"s in offen, Ofen. Note that the "O" (o:) of Ofen has a short variant in unstressed syllables, (o) as in Roman, tomistisch, modern (engl. "modern", not "to rot away"). The difference is length, not quality.

4. Same with "u". Krume, krumme.(adj. fem. nom.) (Or, "Uhr" and "Kummer" from Martin--cal, but I wanted a minimal pair.)

5. Two more vowels with Höhle, Hölle.

6. And two more with fühle (v. 1st sg.) and Fülle.

Do you agree that makes 11 vowels which are distinct in terms of lip/tongue/jaw-position (or rounded/unrounded, forward/back, open/closed)?

7. The "e" and "ä" in spelling are difficult to describe in pronunciation. There is a long "ä" such as in Ähre, and an (e:) in Ehre, but the short variants amount to the same vowel. Or don't they?

a) (e), (e:). One vowel, the only distinction is quantity. The short (e) appears in unstressed syllables. The initial "e" in Genom and in gelten is pronounced differently, and the first one is a shortened variant of the "e" in gehen. It is my suspicion that the shortened closed "e" (e) as well as the "o" (o) as in Romantik are limited to latin or romanic words which found a place in the German language but that's just a guess.

b) The sound which distinguishes Messe and mäße (v., Konj. Imp.). Again, same formation, different quantity, and one vowel. The sound can be spelled "ä" or "e" in the short variant. Words: Käse, essen, Messer, erblich, ändern, ärgerlich, lässig, mäßig. In IPA, its a Greek epsilon, followed by : for length.

c) Now this is a contest. Is the last one the "schwa"? It must be. It sounds like a vowel which differs from other e-sounds. In formation, it is very much like the "ö" in "öffnen", just slightly less stressed. The "schwa" has got an extra character in the phonetic alphabet, so it is a sound on its own.

I'm not sure, but if they counted the "schwa", they missed something else, and that is the presence of the sound "ae combined" in the German language. It's deeper than the short "e" in 7b, and it occurs in words like Ärger. If the e in kess is the short variant of the ä in Käse, the ä in Ärger can't be, and if "Ärger" and "Käse" have the same quality, "Bett" is different. Compare Bett and Ärger -- these are clearly different sounds. And it is not the presence of the "r" which colours the sound. Compare Erbe, Erde, and vergessen.

'ere we go...

*wow, worked*
@ Martin--cal: "Since I'm not a native German speaker, I don't think it makes sense for me to comment on questions of German pronunciation" -- why that? Both you and hm--us did it in your contributions, and why not?

Darf ich trotzdem einen Versuch machen, die englischen Vokalqualitäten zu zählen???

Ich gehe von der britischen Aussprache aus. Diphthonge zählen nicht, nach den Wals-Kriterien ("diphthongs (and triphthongs) are considered combinations of simpler vowels"), also ist bait kein langes A und bite kein langes I. Ich komme trotzdem auf 13.

Spricht man die folgenden Wörter nacheinander aus, stellt man fest, dass sich von einem zum anderen mindestens eines der Formierungsmerkmale ändert (Zungenposition, Lippenstellung, Kieferstellung offen/geschlossen):

dot -- door -- but -- barm -- mist -- meet -- foot -- mood
girl
get -- fat -- fair
Plus schwa: misery.

@SebastianW: yes, "girl"; good catch! If that counts as a separate vowel (and I think it does for me, as I don't pronounce post-vocalic R), then it adds another one to my list of (NYC) English phonemes.

I think you are wrong to exclude the vowels of "boat" and "bait" as diphthongs. It's true, there is a "glide" at the end of each of these vowels, but the starting point of the vowel is not otherwise in the list, and so should be counted.

But "Bite" should be omitted - that, I think, is a genuine diphthong.

So that would give us a count of 15 vowels. I have essentially the same set of vowels as you do in your variety of English.

Using hm-us's tabulation as a starting point, (because of its almost complete set of minimal pairs) and adding the two new distinct vowels (substituting one of them for #5 bite), we now have a count of 15 "vowels" for our varieties of English:

1 long A bait
2 short A bat
3 long E beat
4 short E bet
5 NEW: "flavored A" Bart (same as farther, father, cart,...)
6 short I bit
7 long O boat
8 short O bot (distinct from father/farther for me)
9 rounded O bought
10 long U boot
11 short U but
12 other short U book (or put)
13 schwa ("between" keeps the B--T pattern)
14 MARY vowel (also used in bad, bag, etc.,but not in bat, back,...)
15 GIRL vowel Burt

Hurray for the new champion in the vowel contest! (Unless someone complains about the vowels we're counting...)

1. German a/aa - You almost had me convinced, but I'm sitting here making a and aa sounds (fortunately I'm all by myself!), and I don't quite agree. They're incredibly close, yes, but when I say an "a" sound a few times, I actually do drop my jaw a bit to make it an "aa" sound. Shouldn't that be a difference in quality?

2. German "ää" - let's not forget that some German speakers pronounce it like "ee". I'm still trying to find out whether I pronounce some "ää" as "e" and the rest as "ee"... or whether I'm even using some middle ground for some. But I guess that wouldn't count because it's influenced by a dialect?

3. I'm having trouble with Messe/mäße - I feel as if "Messe" was formed a hint higher up in my mouth. "Mäße" can, but doesn't have to be formed there (as a default, I don't form it there). Okay, even if true, that may not qualify as a "difference"... but is it true? Do you experience the same thing? Or am I just losing it after having been thinking about vowels too much? :o)

4. Genom/gelten: Hm. I think I pronounce the "e" in Genom either as a short (not in the grammatical sense) "ee" or simply as an "e"... or don't I?

5. English: fair? At least according to pons, it seems to be a diphtong in BE, and an "e" as in "get" in AE.

6. Ärger/Bett: not much of a difference for me... but if there is one, it seems to me like (3) Messe/mäße, only that "mäße" is long. Except that I tend to use a hint of a diphtong for "är" (sorry, I forgot what to call that thing for the "r")... (I'm not sure about your Erde/Erbe/vergessen example. Eerde has a long "ee", but Erbe und vergessen are the same for me.)

7. Schwa: As ever so often, just try Wikipedia - http://www.wikipedia.org/wiki/Schwa

@Femwa, regarding your point #5, for me, the vowel of "fair" is quite distinct from that of "get". The first rhymes with Mary and hairy, and the second rhymes with merry and berry. (And then there's marry, Barry and Harry with yet another vowel.) I detect a slight glide at the end of "fair" to a schwa, but the sound before the glide is, I think, unique. (But of course I admit a trained phonetician might analyze it differently.)

Was mir beim Studium dieses Fadens (Lektüre greift irgendwie zu kurz... :-) ) bewusst wurde:

Einer der gravierendsten Unterschiede meines Dialekts (Schwäbisch) zum Hochdeutschen ist die andere Färbung der Vokale. Es gibt natürlich auch Unterschiede im Vokabular und in der Grammatik (z.B. kein Imperfekt), aber das, was es so schwierig macht, den Dialekt nicht durchkommen zu lassen ("wir können alles außer Hochdeutsch"), sind die Vokale.
Beispiele: für mich ist der Vokal in "Bett" derselbe wie in "Kette" und nicht wie der in "nett"

Das o in "von" "doch" ist von der Klangfärbung dasselbe wie in "oder" und nicht wie in "morgen"

Was mich zu einer anderen Frage bringt:
Gibt es für diese Art sprachwissenschaftlicher Betrachtungen ein definiertes "Hochdeutsch"? Etwas wie ein "Tagesschau-Deutsch" analog zum BBC-Englisch? Sind nicht alle gesprochen Varianten irgendwie dialektgefärbt und damit - s.o. - insbesondere in der Aussprache der Vokale regional unterschiedlich?

Goldammer, genau das überleg ich auch schon die ganze Zeit: wie klingt denn ein offizieller deutscher Vokal? Auch bei manchen Tageschausprechern kann man an der Färbung der Vokale heraushören, ob sie aus dem Norden oder Süden stammen. Etwas verallgemeinernd gesagt, ein helles "ei" verrät oft den Schwaben, ein dunkles "a" den Bayern und viele helle "a"s den Österreicher.

Bei einer gefühlten Mehrzahl der Leute klingt "Das Mädchen näht und ist Käse" wie "Das Meedchen neeht und isst Keese". (Das "äh" stirbt aus - meine Hypothese)

Ich nehme an, dass die überraschend hohe Zahl unterschiedlicher deutscher Vokale auch aus dem Unterschied von deutsch-deutscher, österreichischer und Schweizer Klangfärbung resultiert, oder?

Nein, die große Anzahl der deutschen Vokale resultiert nicht aus der Einbeziehung von dialektalen Varianten, sondern vor allem daher, dass wir für jeden Vokal (fast jeden) zwei Versionen haben, die durch verschiedene Mundstellungen (Stellung der Zunge, des Kiefers und der Lippen) gebildet werden und, das ist wegen der unterschiedlichen Querschnittsänderungen des Ansatzrohrs die unweigerliche Folge, auch verschieden klingen. Diese beiden Varianten heißen "offen" und "geschlossen", wobei es eigentlich "offener" und "geschlossener" heißen müsste, denn geschlossen ist das Ansatzrohr, sprich: der Atemweg bei Vokalen nie.

Bei 8 (geschriebenen) Vokalen (incl. Umlaute) käme man auf insgesamt 16 Vokale, mit dem Schwa-Vokal wären es 17. Da aber einige Vokale, z.B. das offene und geschlossene "a" "zusammenfallen", d.h. sich, wenn überhaupt, nur durch die Länge unterscheiden lassen, wie in "Sache" und "Aachen", ergibt sich eine Gesamtzahl die unter 17 liegt. Andere Beispiele für das "Zusammenfallen" sind das offene "e" und das offene "ä" (wie in "Stengel" und "Stängel"), während die geschlossenen Versionen, z.B. "Säle" und "Seele" in den meisten deutschen Dialekten deutlich unterschieden werden.

Falls einiges von dem, was ich hier darstelle, schon diskutiert worden ist, bitte ich um Verzeichnung. Der Faden ist ja sehr schnell sehr lang geworden ...

Verzeichnung -> Verzeihung.

Die Tagesschau hat meines Wissens keinerlei Anspruch auf Hochdeutsch. Sie bemüht sich, genau wie andere auch.

Als "offizielles" Hochdeutsch hätte ich immer die Bühnensprache (-> Theater) angesehen. Dort wird zwar bestimmt nicht immer das Ziel einer definierten, einheitlichen Aussprache erreicht, aber es wird zumindest verfolgt.

http://de.wikipedia.org/wiki/Bühnendeutsch

Wikipedia schreibt allerdings, daß all unsere Seifenopern-Sternchen und das Privatfernsehen (oder: des Privatfernsehens?) das Bühnendeutsch derzeit unterminieren. Und da eh nur noch alle Fersehen gucken und (fast) keiner ins Theater geht, wird die Tagesschau sich vielleicht tatsächlich einmal in der Rolle der Aussprachenwahrerin finden. O tempora, o mores...

I've been reading along but I can't really take time to think this through at greater length today, and maybe not for several days (by which time maybe it will all be over anyway (-; ). And like Martin, I hesitate to say much about German, since obviously I'm not a native speaker.

But a couple of tentative interim observations anyway:

1) Surely a lot of this, as AndreasS suggests, is a matter not of quality but of degree. At what point along the spectrum from one mouth position to the next do you define a sound as being officially a different vowel? At what point does an on- or off-glide or a vowel with a slight regional coloration become officially a diphthong?

2) I think Ingeborg was right to bring up the other neutral vowel in German, the sound of E in final -er as in Butter. It's definitely different from a schwa and is represented with a different IPA symbol (the upside-down V). In fact, on the spectrum, it actually approaches the 'ah' of 'father.'

3) I think Femwa may well be right to be skeptical that the A's of Stadt and Staat are exactly the same vowel. To my ears also, the short A is definitely also more closed. In fact, for Germans learning English, along the spectrum it actually approaches the short U of English 'but' very closely (even though English speakers perceive a difference). Remember the joke someone told about (IIRC) their little boy learning the German word 'Bagger' at an early age and then traveling to the UK and cheerfully announcing 'bugger!' at every construction site?

So basically, where do we draw the line between

(en) sherbet* > (dt) Batzen > (en) but > (dt) Butter > (dt) bat

?

(If I even have those in the right order; vowel charts were one of the things I sort of skimmed over in linguistics.)

*Martin, I chose that one because it's a final syllable; between often goes more toward a short I. But I think variant schwas are probably beyond the scope of this discussion.

4) I don't understand how 'gelten' got into sebastianW's group 7a (long E); doesn't it belong in 7b (short E = Ä)?

5) I don't understand why 'fair' should be considered different from 'bait.' For me, those are both long A. Yes, in strict phonetic analysis that's probably considered a diphthong. But I think Martin was right to point out that we still have to account for the base vowel in the diphthong if it's not represented anywhere else in the list. The same goes for English long O, I think; even though it has a glide to U, it still needs to be accounted for somehow.

6) In general, I don't think we should use words with R in them as examples at all, as that's another whole can of worms (American vs. British R, R-tinged vowels, etc.) that I suspect is beyond the scope of this discussion.

words with R in them —> i.e., words in which the vowel in question is followed by R

@hm-us: Speaking just for my variety of English, I certainly believe that "fair" and "bait" have different vowels. But I agree, the R-coloration confuses the issue. Trying to take the R out of the equation, for my speech the difference can be heard in the contrast (R-tinged) "lair" /læə/ vs. "layer" /leə/, or (not R-tinged) "fad" /fæd/ vs. fade /fed/.

Meine verspäteten Beobachtungen:Der Endlaut in Butter, Bagger, Mutter ist definitiv ein separater Vokal. Was übrigens daran zu erkennen ist, dass englischen Muttersprachlern große Probleme bereitet, obwohl diese das Schwa ja aus der eigenen Sprache kennen: Er von diesen oft zu offen gesprochen, sodass die Ausssprache Richtung Buttah, Baggah, Muttah tendiert.@ SebastianW:Danke für deine detaillierte Aufstellung, der ich auch weitgehend zustimmen würde. Lediglich beim letzten Punkt wundere ich mich und gebe femwa (# 38) Recht:vergessen, Erbe, sterben, sterblich, ärgerlich, Ärger - das ist für mich jedes Mal derselbe Vokal wie auch in Bett, dort aber nicht durch das nachfolgende R gefärbt.Zumindest von "clearly a difference" kann ich da nichts hören.Danach wurde eventuell doch das Schwa mitgezählt. Trotzdem fehlt mir immer noch der oben angesprochene Laut am Ende von Mutter.

If anybody is still interested in this topic, I just ran across the following English sentence:

"Who would know aught of art must act, learn and then take his ease."

(From John Wells's blog: http://www.phon.ucl.ac.uk/home/wells/blog0802b.htm)

Wells notes that this "is a sentence that not merely contains 14 vowel sounds, but has the sounds in a particular order (going clockwise more or less round the periphery of the vowel area: i.e. with a gradually rising second formant)."

Anyone want to try to construct the German equivalent? (I would wager that it can't be done because of the lack of enough monosyllabic words in German.)

@AndreasS: Jetzt, wo wir in diesem Faden auch einen Fachmann unter uns haben -- falls hier sonst noch jemand sich professionell mit Phonetik/Phonologie/Phonemik/Phonematik (hab ich was vergessen?) beschäftigt, bitte ich um Verzeihung. ;-)

Wie würdest Du das unbetonte "-er" in "besser" einordnen, wenn der r-Laut, nicht -- wie in betonter Silbe ("Erbe") -- als [r], [ʀ] oder [ʁ] deutlich hörbar ist, sondern das Schwa um einen schwachen uvularen Beiklang bereichert und/oder annähernd als [ɐ] realisiert wird? Gehört das zu Kategorie Qualität oder zu einer ähnlichen Kategorie wie nasal/nicht-nasal und glottal/nicht-glottal?

Jetzt hat sich doch noch einiges angesammelt. Ich konnte zwischendurch nicht antworten, möchte aber doch auf ein paar Missverständnisse hinweisen, die sich eingeschlichen haben, und ein paar Dinge klären, zu denen ich angesprochen worden bin (# 35).
Mittlerweile haben alle so viel gesagt, dass es unübersichtlich wird. Femwa z.B. hat in #38 gleich mehrere interessante Punkte zugleich angesprochen, aber ich möchte lieber von Punkt zu Punkt gehen statt auf einzelne Beiträge zu antworten.

Die Klärungen betreffen jetzt die deutsche Seite wie auch die englische. Ich mache es in Raten, um ein eventuelles erneutes Formatierungsdesaster in Grenzen zu halten.

1. Die deutsche Vokalliste.

Hier gab es drei Unklarheiten, von denen ich eine selbst provoziert habe: Stadt/Staat; Bett - Ärger - Käse; das Schwa.

a) Stadt/Staat dürfte geklärt sein. Im Wiki-Artikel (s. #11, Aussprache deutsche Sprache) wird hier ein Unterschied nur in der Quantität, nicht der Qualität gesehen, und das waren ja die Wals-Kriterien. In der deskriptiven Phonetik, die auch dialektale und interindividuelle Artikulationsunterschiede beschreibt, werden hier kleine Unterschiede gesehen, aber sonst ist der Unterschied (im Gegensatz zu lang/kurz bei anderen Vokalen) nicht relevant. Dem hat auch AndreasS hier noch einmal zugestimmt. Der kleine Unterschied in der Praxis dürfte ganz einfach daran liegen, dass man bei einem langen Vokal mehr Zeit hat, den Kiefer fallen zu lassen, bevor man ihn für das /t/ wieder zuklappen muss. Deshalb fällt das kurze "a" eine Spur weniger offen aus.

b) Den "Ärger" mit dem "Bett" nehme ich als "Käse" zurück. Der Gedanke dahinter war: Wenn die beiden kurzen Laute in Ärger und Bett sich qualitativ unterscheiden, dann kann nur einer von ihnen der Vokalqualität des langen Lautes in Käse entsprechen. Und dann hätten wir einen 15. Vokal.
Für meine Begriffe unterscheiden sich die Vokale tatsächlich ein wenig, aber in der Lautschrift werden sie offenbar mit dem gleichen Zeichen notiert. Offensichtlich stehen auch hier individuelle Sprechgewohnheiten im Vordergrund. Für mich sind die drei Laute in "verärgern" unterschiedlich (wenn man nicht jede Silbe betont, sondern nur langsam mit der Betonung auf der zweiten spricht) Der erste ist der Laut aus "Bett", der zweite deutlich weiter im Rachen und offener (Kiefer gesenkt), der letzte ein Schwa. Ich spreche ein deutliches "R" und dachte, dass das der Grund dafür sein könnte, deshalb das Gegenbeispiel "Erde", das zeigt, dass vor dem "r" auch ein anderer /e/-Klang stehen kann. Also, nehmen wir an, dass "Bett" und "Ärger" dieselbe Vokalqualität haben. Dann landen wir bei den 14 Vokalqualitäten, die Wals gefunden hat.

c) Es gibt deutsche Mundarten, in denen die Buchstabenkombination e-r leicht diphtongiert ausgesprochen wird, und manche, in denen nur ein Diphtong und gar kein "r" gesprochen wird ("Mama, ea hat mich ge-EA-gat!!" = Mama, er hat mich geärgert). Solche regionalen Varianten waren für die Wals-Liste nicht maßgeblich, deshalb zählen sie nur ein "Schwa". Es gibt das e-Schwa, das a-Schwa und, wie hm-us erwähnt hat, auch das i-Schwa ("between"). In allen Fällen ist das ein abgetönter Vokal, dessen Herkunft aus einem bestimmten Vokal nicht so entscheidend ist wie seine Neutralität. Die Qualität des Vokals besteht gerade in seiner Neutralität. Im Englischen wimmelt es von Schwas, im Spanischen glänzen sie durch Abwesenheit, im Deutschen mag das a-Schwa eine gewisse Bedeutung haben, da es auch die Abschwächung eines Konsonanten anzeigt (Auslaut-"er") (Wäre das auch eine Antwort an Rheiner #48?). Aber gerade wegen dieser Funktion wird es wohl gerade nicht als eigene Vokalqualität gezählt worden sein. (Gibt es Schwa-Spezialisten unter Euch?)

Also, ich halte die Wals-Liste für plausibel. Wenn man die Voraussetzung akzeptiert, dass weder bedeutungstragende Unterschiede durch Länge und Kürze bei gleicher Vokalqualität (Phoneme) noch Diphthonge noch der Ort des Luftaustritts (nasal/nasaliert/oral) berücksichtigt sind, sondern nur die Vokalqualität nach horizontaler und vertikaler Beschaffenheit des Artikulationsraums mit dem ergänzenden Kriterium der Lippenstellung (das bei sehr offenen Vokalen wegfällt), dann dürften die 14 Qualitäten wohl in den folgenden Wörtern zu finden sein:

Staat -- Ehre - Ähre -- ihre - irre -- Höhle - Hölle -- fühle - fülle -- ohne - Sonne -- Krume - krumme -- /schwa/.


2. Ein paar Kleinigkeiten.

@ hm-us 45: The "V" upside down is not the sign for the "a-schwa" but for the sound in "but". As others have noted (see 32), what you probably meant was the upside down "a". And your point #4: Yes, of course, "gelten" is a short "e", as in "Bett" and "Ärger", but my point was a difference between two short forms of "e" ("The initial "e" in Genom and in gelten is pronounced differently". I admit that short forms of "o", "e", and "i" as equivalents of their long variants are extremely rare in German (and some have argued in this thread that they don't notice them). For all practical purposes, the shorter forms are not variants, but distinct vocal qualities. "Genom" is actually mentioned in the Wiki-article, but "Roman" (as opposed to "Rock" but similar in quality to "Rot" ) wasn't, and the "Vitamin"-example was given by Ingeborg.

@Goldammer: Sprichst du die beiden "e" in "Ende" wirklich gleich? Ich mag ja Süddeutsch, aber wie süddeutsch ist es denn? So süddeutsch, dass es schon schweizerisch ist? Und sprichst du "von" so wie die erste Silbe von "Phonetik", auch so lang? Und "doch" hat nicht das gleiche "o" wie morgen? Mundarten sind vielfältig! @ Dein OT neugierig: Deutsch.

@Martin-cal: I agree with your #39 which I'm sure will pass any phonetician's critical eye but your # 47 doesn't. For the "layer" and "fade" examples you should have added an "i" in the transcription (the one without the dot in IPA): /feid/. It's a diphthong, same as "bait", but this must wait ... (Your transcription is the one for "feed" in the past tense, "fed").

Wenn die Nerven es mitmachen, möchte ich mich noch mit den transatlantischen Leoniden anlegen, aber jetzt erst mal Ruhe.

@SebastianW: You say that the vowel in fade is a diphthong. I may have misinterpreted you, but I think that you were arguing that it therefore should not be included in the tabulation of English vowel qualities in the sense of the article.

Now everyone agrees that there is a glide in it to /ɪ/. I think that this glide is shown in a strict phonetic transcription as /feɪd/, but nevertheless (I think everyone agrees) that vowel is a single phoneme in English (conventionally transcribed /e/).

Its first element is not otherwise accounted for in the WALS tabulation, and indeed Madieson does explicitly account for that vowel (in the English word "sate") in his brief introductory statement.

Und wenn du dich mit uns transatlantischen Leoniden anlegen willst, leg denn doch an! :-) (But now its time for me to call it a day...)


P.S. I'm sorry, but I didn't quite understand your point about my transcription being the one for "fed". (The phonemic transcription of feed is /fid/ and that of fed is /fɛd/, as I'm sure you know).

Oh, duh, sorry about confusing the IPA symbols, I was doing that from memory and didn't actually check -- and now I'm not at home with my dictionaries and I can't check.

Anyway, I don't think getting the IPA symbol wrong affects the point that it does have a different IPA symbol. Whether it's considered a variant or subcategory of something else, though, like a variant A vowel or a variant schwa, I couldn't say.

I would also be interested to see what some of the people in the forum who are more up on phonetics have to say about any of this. I realize the issues are fast multiplying out of control, but still, the discussion has been interesting. Thanks to everyone who's participated so far. (-:

@ Martin--cal: "...mit uns transatlantischen Leoniden anlegen willst" -- I should have added a smiley! Alas, it's too late and bloodshed cannot be averted...

3. The English list of vowel qualities

As Martin--cal observed in #37, his list, which took hints from hm--us, is not far away from mine in #36, except for the fact that two (debatable) diphthongs, as in bait and boat, are included and except for some confusion in terms of what a vowel is and how we call it. We were basically all trying to find out what might be the rationale behind the WALS project, that is, which 13 English and 14 German vowels they actually counted and which they discarded.

They didn't count phonemes (as shown by the fact that they counted German /a/ and /a:/ as one vowel although they are phonemes); they counted vowel qualities in terms of horizontal and vertical position plus lip movement. But obviously all those which remain are phonemes.

a) Letters, names, and vowels. I don't want to annoy you by starting from scratch but I have to make a few basic remarks. In phonetics, vowels are traditionally described as variations of the vowel letters in the Latin alphabet and in that order, a - e - i - o - u; with the mixed forms somewhere in between. Some languages use additional letters to describe additional sounds such as å, ø, æ, or the German ä, ö, ü; some languages use additional rules regarding the position of a vowel letter in a word as an indication of how it is pronounced. Spanish has, according to WALS, five vowel qualities which conveniently correspond to the five vowel letters, and their IPA symbols are /a/, /e/, /i/, /o/, /u/. Deceptively simple. German uses its additional umlaut letters plus a set of rules of position, e.g. "a double consonant at the end of a syllable shortens the vowel" or "the letter h after a vowel lengthens it" or "an e in an unstressed syllable is a schwa". English uses the Latin alphabet without any extra letters although it has more vowels than letters for them. The pronunciation is exclusively determined by rules of position (or combination) plus tradition. Furthermore, English tends to diphthongize vowels to a high degree. An indication of this is the fact that three of the five names of vowel letters are themselves diphthongs (a, i, o,). The name of the letter E is the way it is pronounced more often than not in a stressed syllable, i.e. the sound which is in IPA /i:/ and is usually associated with the latin letter I. In German, Spanish, and others, the names of the vowels agree with their most common sound value.

b) Describing vowel qualities. It is therefore very confusing to use the letter names in a description of English vowels. It is not opnly confusing to German readers but also gives rise to misunderstandings in English readers. As seems to be the case here.

It is quite natural to call the vowel sound in bait a "long A", as both hm--us and Martin--cal have done in their lists, but from a phonetician's point of view it is useless and misleading. Leaving the diphthongs aside for a moment, it leads to funny descriptions such as
"short U: but"
"other short U: put (or book)".
How can the clearly different sounds in "but" and "put" both be a "short U"? Only in writing. And "book" and "put"? Not even in writing, but this time the sounds agree.

The fact that the sounds in "bite", "boat", or "bait" actually correspond to the English names of certain letters should not deceive one into calling them "long I" etc. In terms of phonetics they are diphthongs (even if they are clearly phonemes). More, the diphthong doesn't even start with the given letter: /baɪt/, /bəut/, /beɪt/.

I think that we should discuss sound quality in terms of variants of the IPA and Latin vowels a-e-i-o-u.

c) English vowel list. If we now rearrange the vowel list on the basis of Martin--cal's last updated list, we arrive at the following (I put my suggestions from #36 in parenthesis):

/a/ (1) father (barm, or, hm--us, if you prefer one without an "r", for fear of worms (-:, "balm"). Note: This "flavored" a-sound is the only long /a/ in English and is clearly distict from German or Spanish /a:/; so distinct in fact that it has an own IPA-symbol (can't reproduce it here, it looks like a handwritten "a").

(2) but (but). The v upside down in IPA. No German equivalent.

(3) bat (fat). IPA /æ/, no German equivalent.

/e/ (4) bet (get). German equivalent only in unstressed syllables ("Genom", "Etage") and not counted there as it is the same vowel quality as /e:/ in "Ehre" which, in turn, does not have an English equivalent.

?(5) "Mary vowel" (fair); IPA /ɛ/. This one is difficult. German equivalent in "Bett" or "lässig". In English only in the diphthong /ɛə/ (and I can't find examples where it doesn't precede a "r" as in fare, share, lair). Clearly a distinctive vowel quality, and a phoneme as well. Did they count it? BTW, it is not the vowel in "bag" or "bad" as you have stated, Martin--cal, at least not in BE. But I confess I have very little experience with AE.

(6) Schwa. According to some sources, the most prevalent vocal sound in English. Equivalent in German is the unstressed "e" in word endings. If an "r" follows, it is transformed into the "a-schwa" which earned some discussion in this thread.

(7) "Girl vowel", Burt (girl). Yes, how would you call this vowel? ....uh.....erm... Yes, exactly. That's what it sounds like. It can be spelled "er" (germ), "ir" (girl), "or" (worse), "ur" (curl), and even "ea" (learn). Remind me if I missed a few. Phonetically, it's half way between the schwa and German short open "ö", and apparently it forms a symbiosis with the "r" (no example without r found). Again, no German equivalent.

/i/ (8) bit (mist). One of the few vowel sounds which English and German share. IPA /ɪ/.

(9) beat (meet). /i:/ in IPA. Shared again. No short Variant in English, short variant in German extremely rare ("Vitamin").

/o/ (10) bought (door).

(11) bot (dot). Short open variant. It's not quite the equivalent of German "offen"; IPA has different symbols for them.

/u/ (12) put, book (foot). Same as in German "Mutter". Some argue that book, hood, or foot have a slightly different sound from "put" and that's certainly the case in some Northern English dialects. I don't know about US. The IPA symbol looks a bit like a rounded vase, it's different from the long one which follows.

(13) boot (mood). IPA /u:/ and another shared one.

So, if WALS didn't count the "swear"-sound in (5) above, they may have counted the "bait"-sound (but where's the "boat" then?). And, finally, to make matters worse, they might count the diphthong in "around" which is usually transcribed in IPA with an initial /a/ in it. And that is actually the only occurrence of this sound in English.

4. Very, very last remarks.

Please forgive me for totally ignoring possible differences between American and British English. This is due to my ignorance of American regional pronunciation (I may have heard some, but can't really talk about it, let alone use it myself). It is maybe justified by the fact that WALS based their assessment of English on a preferred British dialect and I strongly presume they used Daniel Jones and his well-researched Received Pronunciation. Strange enough, their maps don't have the appropriate red symbol (for many-vowelled languages) for the US or Australia (see link).

But even if there are differences between AE and BE, I don't know where to look them up. We could have saved much time and avoided a couple of misunderstandings if we had consulted the IPA transcriptions and convenient tables in bilingual dictionaries such as Pons. I don't know an American dictionary which uses IPA.

Having said that I must address a last remark to Martin--cal: "The phonemic transcription of feed is /fid/ and that of fed is /fɛd/, as I'm sure you know". Sorry, I don't. My BE-based Pons at least says /fi:d/ and /fed/ and I believe them. How elso could you distinguish "bet" and "bait"? Or would you say that "bet" is /bɛt/? Enough...

This has been very time-consuming, both your time and mine, but I hope it was not completely wasted.

P.S.: Take any of those, they're left-overs
ɛ ɛ ɛ ɛ ɪ ɪ ɪ ɪ ə ə ə ə

@Frank FMH:
> Wie würdest Du das unbetonte "-er" in "besser" einordnen, wenn der r-Laut, nicht -- wie in betonter Silbe ("Erbe") -- als [r], [ʀ] oder [ʁ] deutlich hörbar ist, sondern das Schwa um einen schwachen uvularen Beiklang bereichert und/oder annähernd als [ɐ] realisiert wird? Gehört das zu Kategorie Qualität oder zu einer ähnlichen Kategorie wie nasal/nicht-nasal und glottal/nicht-glottal?

"Uvular" und "glottal" sind für mich keine Kategorien, in die ich Vokal einordne. Die Vokale unterscheiden sich durch den Querschnittsverlauf des Ansatzrohres, d.h. durch die Stellung von Zunge, Kiefer und Lippen, wobei die Stellung des Gaumensegels (offen oder zu) den Unterschied zwischen "nasal" (offen) und "nicht-nasal" (geschlossen) bestimmt. Die Bildung der Vokale ist einfach. Der (Primär-)Ton, der in der Kehle entsteht, wird durch das Ansatzrohr geschickt.

Der Querschnittsverlauf des Ansatzrohres bestimmt nun die Filterwirkung des Ansatzrohres. Der Primärton ist (als Dreiecksschwingung) sehr obertonreich, und der Filter bewirkt, die Dämpfung und relative Heraushebung (was ist das Gegenteil von "Dämpfung"?) gewisser Obertöne. Die Frequenzbereiche, in denen Obertöne ungedämpft durchgelassen werden, heißen "Formanten". Die Lage der Formanten (die von der Tonhöhe unabhängig ist) bestimmt, welchen Vokal wir hören.

Nun zu Deiner Frage:
Das unbetonte "-er" in "besser" entspricht in den meisten deutschen Lautungen dem Vokal in der ersten (unbetonten) Silbe von "kariert", ist also ein kurzes, offenes "a". Das "er-" (auch unbetont) in "Verkehr" enthält eine leichte Diphthongisierung, d.h. dass während der Vokalbildung eine leichte Stellungsverschiebung, die unweigerlich eine Änderung der Formanten und damit eine Änderung des Klangs nach sich zieht, stattfindet.

Das sind aber sehr subtile Differenzen, die sich von Lautung zu Lautung, d.h. von Person zu Person, und nicht nur von Dialekt zu Dialekt, sich auch anders darstellen können.

Beantwortet das Deine Frage?

Übrigens: DER Schwa, abgeleitet von "der Schwa-Laut" oder "der neutrale Vokal".

Meine Frage zum Suffix "-er" (wie in "besser" oder "Lehrer") hast Du klar beantwortet. An die vor dem betonten Wortstamm stehenden unbetonten Präfixe/Infixe "er"/"ver"/"zer" als weitere "Problemfälle" hatte ich gar nicht gedacht. Danke für Deinen Hinweis -- die leichte Diphthongisierung kann ich durch Selbstbeobachtung ohne Probleme nachvollziehen.

Der Hintergrund meiner Frage war, dass ich mich zu erinnern glaube, auch schon [ɚ] statt [ɐ] (oder [ər] als phonematisch orientierte Umschrift) zur Wiedergabe des "-er" in "bessser", "Lehrer" etc. gesehen zu haben, wobei das Schwänzchen am Schwa andeuten soll, dass der stimmhafte uvulare Frikativ mitklingt. Lustigerweise bringt der Wikipedia-Artikel http://de.wikipedia.org/wiki/R auch noch [ɘ̯] als weitere Variante.

Jedenfalls: Wenn ich mich nicht täusche, spreche ich persönlich das Suffix "-er" nicht wie das unbetonte kurze "a" in "kariert" aus (das wäre m.E. auch -- jedenfalls bei langsamer, sorgfältiger Aussprache -- eher ein halblanges [a] als ein flüchtiges [ɐ]), sondern eher zentral, also in Richtung Schwa -- aber auch eindeutig anders als ein schlichtes unbetontes "e", also [ə]. Wenn ich "Pita", "bitter", "bitte" nacheinander ausspreche, höre ich nach meinem Empfinden drei verschiedene Laute und nicht nur zwei verschiedene. Dabei bin mir natürlich bewusst, dass das unbetonte a vor dem betonten Stammvokal wie in "kariert" gut möglich ein anderer Laut ist als das unbetonte "a" in einer anderen Position und dass "Pita" (oder "Mama", "Oma", "Opa") keine besonders aussagekräftigen Beispiele sind, weil das keine rein deutschen Wörter sind und sie wahrscheinlich eher ein halblanges [a] enthalten, das irgendwo zwischen dem "a" in "Stadt" und dem "a" in "Staat" einzuordnen wäre.

Wie dem auch sei -- mir ist jetzt noch nicht ganz klar, ob Du nasal/nichtnasal ebenfalls zur Vokalqualität rechnen würdest. Gehört das Gaumensegel in die Reihe Zunge, Kiefer, Lippen oder nicht? Falls ja, warum bzw. falls nein, warum nicht?

Abschließend auch noch ein "à propos" von mir: http://www.duden.de/suche/index.php?begriff=s... Mit maskulinem Genus ist mir das Schwa vor Deinem Posting noch nie begegnet, sondern analog zu "das A, das Alpha, das Aleph" -- sozusagen nicht nur ein neutraler Vokal, sondern auch ein grammatisches Neutrum. ;-)

@sebastienW: I certainly understood an implied smiley when you talked about mit transatlantischen Leoniden anzulegen...

Just a few words: (a) I agree completely with what you said about the need to avoid confusing the names of letters with the sounds. Terms like "long I" are non-sensical (or worse) from the linguistic point of view, and are only justified by tradition. The trouble with IPA is that not everybody knows it, and even among those that do, or think they do, there appears to be some disagreement. Perhaps the best approach for a discussion such as this is just to use the name of a common example: e.g. the HAT-vowel, the HATE vowel, the HIT-vowel, the HEAT-vowel, etc.

(b) I really don't know how WALS arrived at their list of 13 English vowel qualities. Your tabulation certainly appears plausible, though.

(c) You said that the MARY-vowel is not the vowel of bag and bad. I don't know about BE, but -- sorry -- in my speech (Bronx, NY) it surely is. (And, by the way, for me back and bat have occurrences of the MARRY-wowel.)

(e) Yes, I would transcribe "bet" as /bɛt/. (Your Pons is giving you a phonetic transcription, it would seem. And transcribing the vowel of fed as /fed/ puzzles me, as /e/ is phonetically the first member of the gliding diphtohong in fade, and is (for me at least) quite distinct from the vowle sound in fed. )

Now maybe I've been wrong all these years, in which case I have to apologize, but this is the way I would transcribe the phonemes in the English words you cited:

feed = /fid/; beat = /bit/
(no example); bit = /bɪt/
fade = /fed/; bait = /bet/ (phonetically a diphthong)
fed = /fɛd/ ; bet = /bɛt/

Thank you very much, Martin--cal, for your reply.

I quite agree that the Marry-vowel is also found in back and bat, but I also find it in bag and bad. This is obviously one of the many existing differences between AE and BE.

Another difference is introduced by the dictionaries themselves. Bilingual dictionaries provide a phonetic description which allows learners of a language to reproduce the sounds of a language which is not their own. They have to decide on a transcription system and, also, on a specific variety of the language in question. They'd probably not pick The Bronx ;-) but rather something which they call "general American". Most D-E/E-D dictionaries use the British Received Pronunciation set, but even monolingual dictionaries in Britain may use slightly different sets. American dictionaries, as far as I know, tend to use homemade concoctions based on "rhymes with...". There is a good reason for that (they don't have to explain about regional differences in a country with many regional variants). (I think this was discussed in some other thread, but I can't remember which one that was.) The tricky thing about it is that you can't use "rhyming" definitions such as "the hat-vowel" if you want to describe the quality of the vowel in question. And this is what WALS was about. Even with IPA, the thing is not altogether clear, as the present discussion shows.

And, of course, phoneticians quarrel about phonemes and the best way to describe a spoken language. There is a list of English phonemes in Wiki (I should have been smart enough to look it up earlier), and while some articles in Wiki must be taken with caution this one seems to be the work of specialists. It includes a nice comparison of phonemes for the variants of English spoken in different countries and has some telling footnotes regarding the words/sounds which have been controversial between us.

Here's the link:
http://en.wikipedia.org/wiki/International_Ph...

@Frank FHM:
Ja, das Gaumensegel gehört absolut in die Reihe der vier beweglichen Teile des Ansatzrohres (die anderen sind Zunge, Kiefer und Lippen). Es ist schwer vorzustellen, aber wahr, dass die gesamte Sprechbewegung oberhalb der Kehle von diesen vier Körperteilen bewirkt wird. "Nasal / denasal" ist natürlich eine Vokalqualität, da ja der Resonanzraum durch die Öffnung / Schliessung hochgradig verändert wird.

Zum Schwänzchen am Schwa: Das amerikanische "r" wird ja, ebenso wie das deutsche "norddeutsche Schwach-r" traditionell als Konsonsant eingeordnet, obwohl keine Obstruktion des Ansatzrohres vorliegt. Das aber wiederum macht beides zum Vokal. Die Kombination von "e" und "r" (ich habe leider keine IPA-Zeichen zur Verfügung) läuft deswegen (wenn überhaupt) auf eine Diphthongisierung hinaus. Dasselbe gilt übrigens für das engl. "w", wie in "water" oder "well". Dem Inhalt des Restes des Absatzes, Beispiel:
"Pita", "bitter", "bitte", ein gutes Beispiel übrigens :-), stimme ich voll zu. Wir sehen, dass wir bei enger Transkription auf noch eine Menge mehr Vokale im Deutschen stoßen als auf meine 13 bis 15.

"Das / der Schwa": Ich muss Dir Recht geben, auch in der phonetischen Literatur findet man oft "das Schwa", obwohl es MIR völlig absurd vorkommt. Ich habe allerdings in den USA Phonetik gelernt, wo es das Problem natürlich nicht gibt. Die meisten Autoren drücken sich um die Wahl des Artikels, indem sie "der Schwa-Laut" schreiben, z.B. Pompino-Marschall.

@AndreasS:

Dein Kriterium "bewegliche Teile des Ansatzrohres" leuchtet mir ein. Aber ist deswegen der Ausschluss der Kategorie Nasalität in dem Text, den Martin--cal verlinkt hat, tatsächlich schlicht willkürlich?

Der Hinweis, dass es zu einem Nasal ein nichtnaseles Gegenstück gibt, ist sicherlich nicht überzeugend (das könnte man ebenso auf gerundete und nicht gerundete Vokale anwenden). In der Wikipedia heißt es allerdings zum Thema Nasalvokale: "Nasale weisen allerdings eine gewisse Behinderung des Phonationsstromes auf, nämlich die, die sie zu Nasalen macht." ( http://de.wikipedia.org/wiki/Vokal ) Das wäre dann doch ein Grund, die Stellung des Gaumensegels anders zu behandeln als die Stellung der Lippen, oder nicht?

Der zitierte Wikipedia-Artikel geht übrigens von 17 deutschen Vokalen aus. (Ich höre 17, wer bietet mehr?)

Die Unterscheidung Vokal/Konsonant hängt wohl, wenn man nicht rein phonetisch/akustisch (also naturwissenschaftlich) an die Sache herangeht, von einem gewissen Vorverständnis ab, das durch das Anschauungsmaterial geprägt ist. In den romanischen Sprachen Französich und Portugiesisch liegt es sowohl bei diachroner/historischer Betrachtung als beim Vergleich mit den verwandten Sprachen, die keine Nasale haben, nahe, einen Nasalvokal dadurch zu erklären, dass der "reine" Vokal durch einen folgenden nasalen Konsonanten entsteht (bzw. entstanden ist). Bei frz. "bon/bonne" kann man das deutlich daran sehen, dass das nasale o in "bon" in bestimmten Positionen wieder zum oralen o wird, sodass "bon ami" und "bonne amie" gleich klingen. (Ähnlich ist es bei "beau" und "belle", die aus lateinisch bellum und bella entstanden sind, wobei hier das l erst zu u und dann der Diphthong aus o und und u zu einem geschlossenen o geworden sind. Da wird die unterschiedliche Aussprache aber auch in der Schreibung markiert, deswegen "bel ami".)

PS. Die IPA-Zeichen übernehme ich einfach mit "copy & paste". Da LEO kein Problem mit Unicode hat, funktioniert das wunderbar. :-)

Ich war jetzt ein paar Tage weg zwecks Arbeit, aber ich komme darauf zurück, sobald ich Zeit finde ...