Ranking : The language with more vowels

[Saaropean]

Of course. But some might say a Portuguese [o~] is just a nasalized variant of the vowel [o].

It is indeed!

Oh, wow.
But isn't there a minimal pair that distinguishes [o~] and [o], so they are two different phonemes?

Diphthongs are two sounds combined, they are not a special sound which is realized as a single one... one of them is either what's is called the onglide or offglide, this means that either the first vowel is shorter (is not the target of the diphthong) or the second is the shorter of the two...

OK, then diphthongs don't count.
Although I would mention them when describing the pronunciation of a language.

The glottal stop is not written and not perceived as a phoneme, but it is usually pronounced in words starting with a vowel ("ich" [?IC]) and in vowel clusters ("naiv" [na"?i:f]).

Can a glottal stop distinguish meanng in German?

Sorry, I didn't express myself well enough. The glottal stop is not distinctive in German, and most people don't even know it's a phoneme, they just consider it the normal way to start a word that begins with a vowel.
Maybe that's why I have difficulties imagining a glottal stop at the end of a word.

It is indeed possible when you say: "in Standard (Hoch?) German there are such and such number of phonemes" or "in this other variety of German, the phonemes are such and such"... We just need to keep in mind that phonemes are segments that distinguish meaning ONLY... the rest are allophones of a phoneme!

The standard is not 100% homogenous. And allophones should at least be mentioned. [C] and [x] are allophones in German, but they are pronounced at two entirely different positions, and it is important to distinguish them. Pronouncing [ix] instead of [iC] just marks a foreign accent, but [aUC] instead of [aUx] wouldn't be understood at all.

[Pittsboy]

I have no idea of what the hell you're talking about. Do remember that not everyone around here studies Linguistics!
I just presented a simple list with portuguese vowel, diphthongs and triphthongs, either nasal or non-nasal sounds. Considering Strigo's question, I think he's more interested in a simple general answer than on this complicated discussion most people don't even understand...

Luis, Strigo asked: The language with the most vowels.... and I am telling you that when you say: "this language has more vowels than that other one", you do not count allophones, or vowels that appear due to a phonetic context... you just say how many vowel phonemes a language has!

You define vowel phonemes as those vowels that distinguish one word from the other, as in 'jogo' (game) and 'jogo' (I play), the only difference is the 'o' which has two sounds [o] in 'jogo' (game) and [O] in 'jogo' (I play), they are phonemes for they change the meaning of the word.

If I take [t] and [tS] for example: 'tia', can be pronounced both [tia] and [tSia], SO, [t] is a phoneme and [tS] is an allophone of [t] because I can exchange one another in this context (before [i]) and the meaning of the word will not be changed...

So, if we are talking about the language with the most vowels, ad if we want something concrete, we will talk about phonemes, and not possible realizations of these phonemes...

[Pittsboy]

Of course. But some might say a Portuguese [o~] is just a nasalized variant of the vowel [o].

It is indeed!

Oh, wow.
But isn't there a minimal pair that distinguishes [o~] and [o], so they are two different phonemes?

Just take a look at the examples I show on page 1. And also on the answer to Luis...

It is indeed possible when you say: "in Standard (Hoch?) German there are such and such number of phonemes" or "in this other variety of German, the phonemes are such and such"... We just need to keep in mind that phonemes are segments that distinguish meaning ONLY... the rest are allophones of a phoneme!

The standard is not 100% homogenous. And allophones should at least be mentioned. [C] and [x] are allophones in German, but they are pronounced at two entirely different positions, and it is important to distinguish them. Pronouncing [ix] instead of [iC] just marks a foreign accent, but [aUC] instead of [aUx] wouldn't be understood at all.

Allophones are mentioned when you describe the language's full Phonetic chart... and then you explain when those sounds appear in the language (the allophones)... [ç] and [x] are indeed allophones in German, but this doesn't mean they can be interchangeable anywhere you want to... there are allophones which we say they are in complementary distribution: this means that when allophone A occurs, allophone B cannot occur and vice-versa... and as I am concerned, German has, like, three allophones for written 'ich', they can (as I am told): [iç], [ix] and [ik] and all those occur in distinct areas inside Germany.

[Saaropean]

Allophones are mentioned when you describe the language's full Phonetic chart... and then you explain when those sounds appear in the language (the allophones)... [ç] and [x] are indeed allophones in German, but this doesn't mean they can be interchangeable anywhere you want to... there are allophones which we say they are in complementary distribution: this means that when allophone A occurs, allophone B cannot occur and vice-versa...

I take this as a nice explanation for anyone reading this, not for me in particular.

and as I am concerned, German has, like, three allophones for written 'ich', they can (as I am told): [iç], [ix] and [ik] and all those occur in distinct areas inside Germany.

That's a bad example. In northern dialects, the 1st person singular pronoun is "ik", just like in Dutch (which is actually closer related than High German). In central dialects, the [ç] phoneme becomes [ʃ], so you say [ɪʃ]. In southern dialects, the 1st person singular pronoun is [i:]. But some Swiss say [ɪx] when they try to speak High German.

In High German (the standard variant), the complementary distribution of CH is as follows:
- [x] after A, AU, O or U (auch [aʊx], Buch [bu:x])
- [k] at the beginning of a word when the following letter is not E or I (Chaos ['ka:ɔs], Chrom [kro:m])
- [ç] in all other cases (China ['çi:na], euch [ɔɪç])

As for regional variation (apart from the fact that some dialects pronounce G like CH in some positions):
- In central dialects, [ç] is always pronounced [ʃ]. (China ['ʃi:na])
- In southern dialects, CH and K are always pronounced [k] at the beginning of a word. (China ['ki:na])
- In some Swiss dialects, CH is always pronounced [qχ] or so.

[elgrande]

Actually English is a 12-(phonemic)vowel system language, which has:

[i:], [I], [3:], [e], [@], [u:], [U], [O:], [ae], [^], [A:] and [&] (& represents the inverted cursive a, and A, the cursive a)

Diphthongs are not taken into account!! Nor are English 'nasalized' vowels.

Why don't diphthongs count? Theys are phonemic in English, in contrast to Spanish for example. If diphthongs doun't count then a lot of people don't have [u:] and [i:] vowels either.

If length counts, you must also distinguish vowels in "bad" and "bat" for example, which makes the whole thing complicated.

I would simply consider vowel phonemes.

[elgrande]

"Pittsboy" wrote:
"a) diphthongs are the combination of two other phonemes in the language; so they are not phonemes themselves; (it all depends on the criteria and author you use to count them)"

No, this is not true. Diphthongs are vowels where the tongue does not remain in the same place during the articulation, but changes its position. The word "diphthong" doesn't say anything about the phonemic status.
In some languages vowels are phonemic (English, German), in others they aren't (Spanish), i.e. in Spanish a diphthong regularly comes into being when an unstressed [i] or [u] precedes or follows a vowel.
Thus, Spanish can't have phonemically different words [mai] (ai diphthong) and ['mai] (pronounced as two syllables) , whereas German can.




"Pittsboy" wrote:
"b) vocalic (?) /r/: you mean what is usually referred to as 'coloured' as American English word-final [r], which combines with the previous vowel (as in 'betteR'), the vowel still is the SAME vowel, with an extra /r/ phoneme. Then, not a big deal out of it! You are combining a vowel sound + a consonant sound, and it is not one single sound, but two."

No, German vocalic r is rather similar to "r" in British non-rhotic dialects, such as the word "beer". So, the whole thing is really diphthong ending in either [@] or upside-down a. I find it hard to decide whether those diphthongs can be considered phonemes in their own right in German, perhaps rather not.



"Pittsboy" wrote:
"c) Are there words in German with different meaning but spoken the same way except for the fact that one of the words have a long vowelthe other one a half-long? Is the answer is YES, then you probably have different phonemes! On the contrary, NO!"

On the other hand, it is not predictable where those vowels are used.
Anyway, the number of those words is small and most people don't make the distinction anyway and the distinction seems a bit artificial to me, so I'd say it's not phenomic too.




someone else wrote:
"I think the diphthongs should count, because vowel clusters are not permitted in German (unless there's a non-orthographic glottal stop in between)."

In this case, I'd say they are not phonemic, because then a diphthong is really just a predictable combination of two vowel phonemes with no [?] in between.

[el|grande]

In High German (the standard variant), the complementary distribution of CH is as follows:
- [x] after A, AU, O or U (auch [aʊx], Buch [bu:x])
- [k] at the beginning of a word when the following letter is not E or I (Chaos ['ka:ɔs], Chrom [kro:m])
- [ç] in all other cases (China ['çi:na], euch [ɔɪç])

Mostly true, but the syllable -chen is always pronounced with [C]. Usually this suffix causes preceding vowels to be umlauted, so none of the rules mentioned above is violated. However, there is the word "Frauchen", which violates these rules. Only almost minimal pairs like "fauchen" - "Frauchen", but IMHO that's "bad enough" to call them different phonemes. Some people try to get rid of this problem by condering "morpheme onset" as something separate, but this is usually not something that can be taken into a phonemic analysis AFAIK.

Besides, I don't think [k] spelt as "ch" can be considered an allophone of /x/ or /C/, because there's really nothing that gives any relation to those sounds other than the spelling.


(To make it all a bit more complicated, one could also consider words like Chanukka, which start with [x], but I think words starting with [C] are already unnatural enough.)

[Strigo]

And I just wanted to make a nice and friendly ranking....


Hey! This is interesting!

[Pittsboy]

Why don't diphthongs count? Theys are phonemic in English, in contrast to Spanish for example. If diphthongs doun't count then a lot of people don't have [u:] and [i:] vowels either.

I really don't get it (what's in bold)... of course diphthongs are phonemic in any language, if they are at the morphophonemic level... but diphthongs are two vowels combined, where one is the target and the other one is the glide vowel... they are not ONE sound, they are TWO sounds!!

If length counts, you must also distinguish vowels in "bad" and "bat" for example, which makes the whole thing complicated.
I would simply consider vowel phonemes.

It is known that vowels preceding voiceless stops, for instance, are 'slightly' longer than the ones preceding voiced ones... but no language makes it distinctive, languages rather change the quality of the vowel...
Please, do not confuse it with long/short vowels... whenvowel are long or short independently of the following segment!!

Vowels and Consonants can be phonemes in a language ONLY if they change meaning, for God's sake!

[Pittsboy]

No, this is not true. Diphthongs are vowels where the tongue does not remain in the same place during the articulation, but changes its position. The word "diphthong" doesn't say anything about the phonemic status.

That's what I said, diphthongs are two vowels combined, one is the target and the other the glide(=where the tongue changes its position to)... so, what's the deal???? They are phonemes in the sense that they are at the morphophonemic level, but they are not ONE sound to be considered a distinctive phonemes...

I must appologize, for, phoneme may be used in two situations, and I have probably not made myself clear:
a) phonemes are sounds which are distinctive in a language
b) phonemes are sounds that are at the morphophonemic level (mind level) of languages)

In some languages vowels are phonemic (English, German), in others they aren't (Spanish), i.e. in Spanish a diphthong regularly comes into being when an unstressed [i] or [u] precedes or follows a vowel.
Thus, Spanish can't have phonemically different words [mai] (ai diphthong) and ['mai] (pronounced as two syllables) , whereas German can.

EVERY language has vowels which are phonemic (this includes Spanish)! No language exist without vowels (please, do not say Hebrew has no vowels, it DOES have, do not confuse writting with phonology)... in every language diphthongs (there might be exceptions!) have a target vowel (the one which is the base of the diphthong) and another one which is the glide (the one which is the weakier)... if this does not happen, yo have what's called 'hiatus' (as in ['ma.i]... so what????

GUYS! ATTENTION here: we divide language into two distinc levels:

1) morphophonemic level: meaning making sounds (phonemes, or mental sounds) which are inside the speakers minds...
2) phonetic level: after rules of transformation applied (phonetic rules), what's represented in the morphophonemic level (inside one's mind) comes into being as spoken sound (as phones)... at this level, there appear allophones, context predictable sounds, all those are PHONES (sounds spoken)...

When I say that Portuguese doesn't have nasal vowels, it makes reference to the morphophonemic level... after I apply context phonetic rules, nasal vowels appear in the language, but they are nasalised at the phonetic level, this means that vowels in Portuguese get nasalization according to a specific context whn the word is spoken... example:

'assento' - word (meaning: seat)
/a.'sEN.to/ - morphophonemic level (what's inside the speakers mind)
/A.'sEN.tU/ - Rule 1 (reduction of unstressed vowels)
/A.'s~EN.tU/ - Rule 2 (nasalization os vowel preceding nasal consonant)
/A.'s~E.tU/ - Rule 3 (deletion of nasal consonant)
/A's~EJ.tU/ - Rule 4 (appearance of nasal ressonance - J is a nasalized superscript [j])
[A's~EJ.tU] - Phonetic level (spoken word)

After this simple example, I hope it is clear what's considered distinctive phonemes in a language and what are phones....
Any question, just ask me

[elgrande]

Pittsboy wrote:
"I must appologize, for, phoneme may be used in two situations, and I have probably not made myself clear:
a) phonemes are sounds which are distinctive in a language
b) phonemes are sounds that are at the morphophonemic level (mind level) of languages)"

This is already clear to me.

Pittsboy wrote:
"EVERY language has vowels which are phonemic (this includes Spanish)!"


Sorry, my mistake: I meant to write that diphthongs are not phonemic in every language. I totally agree that vowels are phonemic in every language (at least those I have heard of).

Pittsboy wrote:
"No language exist without vowels (please, do not say Hebrew has no vowels, it DOES have, do not confuse writting with phonology)... in every language diphthongs (there might be exceptions!) have a target vowel (the one which is the base of the diphthong) and another one which is the glide (the one which is the weakier)... if this does not happen, yo have what's called 'hiatus' (as in ['ma.i]... so what????"

As I can't think of a real example, let's just make one up for German. Let's say there are all the words ['baUm@] (diphthong), ['baUm@] (separate vowels, [ba'Um@]. None of those words exists, but as a native speaker I can say that they are phonemically/phonetically possible and could all have different meanings and they would all be distinguished in speech.
Now you say that a diphthong can be a phoneme in it's own right, so you have the phonemes [b] [a] [U] [m] [@] and stress. Try to build all the three words with only those phonemes. You will need to take a diphthong-phoneme into consideration additionally.

You yourself gave the example of [ts] affricate and that the affricate is two phonemes in German, but only one in different languages (I think you said you were referring to Polish). It's the same with German [aU]: It's one phoneme in German, it isn't in other languages like Spanish.

Again: Spanish is different. Spanish diphthongs can easily be considered as two following vowel phonemes, (most) German diphthongs can't.


Pittsboy wrote:

"When I say that Portuguese doesn't have nasal vowels, it makes reference to the morphophonemic level... after I apply context phonetic rules, nasal vowels appear in the language, but they are nasalised at the phonetic level, this means that vowels in Portuguese get nasalization according to a specific context whn the word is spoken... example:

'assento' - word (meaning: seat)
/a.'sEN.to/ - morphophonemic level (what's inside the speakers mind)"


In German, the diphthongs [aI] [OY] etc. don't consist of two different vowels in the speaker's mind.


Pittsboy wrote:
"/A.'sEN.tU/ - Rule 1 (reduction of unstressed vowels)
/A.'s~EN.tU/ - Rule 2 (nasalization os vowel preceding nasal consonant)
/A.'s~E.tU/ - Rule 3 (deletion of nasal consonant)
/A's~EJ.tU/ - Rule 4 (appearance of nasal ressonance - J is a nasalized superscript [j])
[A's~EJ.tU] - Phonetic level (spoken word)"

I understand what you are talking about. In the same way, we could most likely say that German doesn't have phonemic nasal vowels, although some speakers like me have nasal vowels (well, slightly nasal), perhaps that diphthongs that just come from a following r are not phonemic, that colloquial [eI] (diphthong) in some words like "welche" is just /El/ phonemically, that vowel length is not phonemic, that the different vowels in "Ding" and "Zinn" are just allophones, that aspiration after voiceless stops is not phonemic, that [E] and [@] are allophones, that affricates are not phonemic etc. etc.

But for the diphthongs in question such rules don't work.

[Pittsboy]

You really don't get it... I am almost crying here

Let's take your example:

['baU.m@] (diphthong)
['ba.U.m@]

Diphthongs are TWO sounds!! Is that clear or not?... it doesn't matter whether I can have the same two sounds that make a diphthong in different syllables or in the same syllable, they will always be two sounds... they are not an entity as a whole, they are TWO sounds!

Phonemically [aU] are the phonemes /a/ + /U/ which result as an [aU] phone (spoken)... they are two separate sounds that when spoken one of these, due to syllabification, is stronger than the other, and then you have a diphthong... there is not such a thing as a diphthong phoneme...

[ts] (affricate, with a tie bar ontop) are not two phonemes, it is a single phoneme!!!!! It is an [s] which starts as a [t] sound and soon becomes an [s], there're not two articulations... on the other hand, Polish can distinguish this affricate (tie barred) [ts] (which is one phoneme) from a common junction of two phonemes [ts]; here you have two articulations, really fast, but they are two, what differs enormously from the first one, which happens to be an [s] that starts as a [t]...

Is it clear now?

Pay real attention to whether I use [ ] or / /, so that you won't commit such a mistake as: "that colloquial [eI] (diphthong) in some words like "welche" is just /El/ phonemic[ally]"

That 'colloquial' [eI] in 'welch' is NO WAY phonemic, it is tottaly Phonetic... it only happens to exist in speech!! Be careful when you say such a thing...

I didn't describe all the rules that exist, I used rules 1 to 5, I think, to show you how the think works, there are rules on diphthonging for any language... don't worry, the theory accounts for it too...

[Saaropean]

Le'me get to Pittsboy first. I don't wanna see anyone crying here.

Diphthongs are TWO sounds!! Is that clear or not?... it doesn't matter whether I can have the same two sounds that make a diphthong in different syllables or in the same syllable, they will always be two sounds... they are not an entity as a whole, they are TWO sounds!

Yes, they are two sounds, but they are not just two random vowels that appear beside each other. In fact I'd rather say the [U] in German [aU] is a semi-vowel, but not a real consonant [w] either as the example "hauen" ["haU@n] shows.

I think you're contradicting yourself, Pittsboy. First you said only phonemic vowels (no allophones) should count, then you say diphthongs must be separated into the two vowels they consist of. The diphtongs [aI], [aU] and [OI] (plus [EI] from English loans) are phonemic in German. They developped from long vowels, and in some dialects they are still pronounced as long vowels.

So, for the original list, I would say German has 20 "vowels":
a, a:, E, E:, e:, @, I, i:, O, o:, 9, 2:, U, u:, Y, y:, 6, aI, aU, OI

And in words of foreign origin, you can also have French nasals and English [EI], but other vowels would be substituted with their German equivalents, e.g. "Batman" is pronounced ["bEtmEn].

In German, the diphthongs [aI] [OY] etc. don't consist of two different vowels in the speaker's mind.

Yes, and I guess that's why German-speakers don't think pronouncing "EU" as [OI] (Who says [OY]?) is strange. Many German speakers would even pronounce "dos euros" [dOs "OIRo:s] when trying to speak Spanish.

"/A.'sEN.tU/ - Rule 1 (reduction of unstressed vowels)
/A.'s~EN.tU/ - Rule 2 (nasalization os vowel preceding nasal consonant)
/A.'s~E.tU/ - Rule 3 (deletion of nasal consonant)
/A's~EJ.tU/ - Rule 4 (appearance of nasal ressonance - J is a nasalized superscript [j])
[A's~EJ.tU] - Phonetic level (spoken word)

My question again: Are their minimal pairs in Portuguese, where a nasal vowel (not followed by a nasal consonant) contrasts with a non-nasal one? If there is, I would say nasalization is phonemic in Portuguese, because the nasal vowel is merely orthographical/etymological.

colloquial [eI] (diphthong) in some words like "welche" is just /El/ phonemically

You pronounce "welche" [v"eIC@]? I've never heard that, but I'm not so familiar with the dialects of the north.

that vowel length is not phonemic

Well, you could say it's rather quality than length, because [Y] and [y:] do not only differ in length.

that the different vowels in "Ding" and "Zinn" are just allophones

Of course they are. The one in "Ding" [dIN] is slightly nasal, the one in "Zinn" [tsIn] is not.

that aspiration after voiceless stops is not phonemic

Yep, but some people would transcribe the SP or ST they pronounce as "schb" or "schd" rather than "schp" or "scht", because they think a voiceless unaspirated stop sounds like a voiced one.

that [E] and [@] are allophones

They are at least partly. Sometimes [E] sounds "more correct", sometimes it sounds "awkward"...

[Psi-Lord]

We're so off topic anyway... <sigh>

[A's~EJ.tU] - Phonetic level (spoken word)"

Thi, does EN in assento sound like EN in quente for you? What about you guys (the other Portuguese speakers here)?

And should it make any differences the fact that e.g. EI may sometimes be pronounced either [ej] ('educated') or [e] ('colloquial'), but not always?

e.g. beira = ['bej4a] / ['be4a]
but
leite = ['lejtSi] (never *['letSi])

Kind of talking to myself here, since I must confess I've long lost myself in this thread.

[ekalin]

My question again: Are their minimal pairs in Portuguese, where a nasal vowel (not followed by a nasal consonant) contrasts with a non-nasal one? If there is, I would say nasalization is phonemic in Portuguese, because the nasal vowel is merely orthographical/etymological.

There are some candidates for a, but there is a change in stressed syllable: 'manha vs. ma'nhã, or 'Ana (a name) vs. a'nã.

[elgrande]

You really don't get it... I am almost crying here

So am I. :/ Let's take your example:

['baU.m@] (diphthong)
['ba.U.m@]

Diphthongs are TWO sounds!! Is that clear or not?... it doesn't matter whether I can have the same two sounds that make a diphthong in different syllables or in the same syllable, they will always be two sounds... they are not an entity as a whole, they are TWO sounds!

"two sounds" is not the same as "two phonemes".
About: ['baU.m@] (diphthong) and ['ba.U.m@]:
What is the phonemic feature that distinguishes them in your opinion? They are not phonemically the same in German: ['ba.Um@] would be a completely different word than ['baU.m@]. So, what phonemic feature?

Phonemically [aU] are the phonemes /a/ + /U/ which result as an [aU] phone (spoken)... they are two separate sounds that when spoken one of these, due to syllabification, is stronger than the other, and then you have a diphthong... there is not such a thing as a diphthong phoneme...

No, if you don't use a diphthong, people will consider what you speak about as something very different.
Before I learnt that in Spanish diphthongs aren't phonemic, I was always very confused how I could see if I had to pronounce for example the ái in "llamáis" as a diphthong or not. Two Germans, it really makes a big difference and a phonemic difference. It's not like a diphthong is just a variant of the two sounds following one another. If you asked a German to pronounce a word very slowly, he would never get the idea to pronounce the diphthong as two separate vowels.

[ts] (affricate, with a tie bar ontop) are not two phonemes, it is a single phoneme!!!!! It is an [s] which starts as a [t] sound and soon becomes an [s], there're not two articulations... on the other hand, Polish can distinguish this affricate (tie barred) [ts] (which is one phoneme) from a common junction of two phonemes [ts]; here you have two articulations, really fast, but they are two, what differs enormously from the first one, which happens to be an [s] that starts as a [t]...

I don't think in German affricates are really separate phonemes, at least not [pf]. Here it is really as you said: The affricate just comes into being because the two phonemes /p/ and /f/ can easily be joined together into an affricate. No German would consider [pf] to be one phoneme.

Pay real attention to whether I use [ ] or / /, so that you won't commit such a mistake as: "that colloquial [eI] (diphthong) in some words like "welche" is just /El/ phonemic[ally]"

That 'colloquial' [eI] in 'welch' is NO WAY phonemic, it is tottaly Phonetic... it only happens to exist in speech!! Be careful when you say such a thing...

Be careful when you read such a thing: Why did you change my "phonemically" into "phonemic[ally]"? I never meant to say that [eI] is phonemic. I just meant: On a phonemic level (= Phonemically), the sound [eI] is just a variant of /El/ (the symbol after the E is a lower-case L).

(I admit that I sometimes confuse [] and //, but this time I didn't. )

I didn't describe all the rules that exist, I used rules 1 to 5, I think, to show you how the think works, there are rules on diphthonging for any language... don't worry, the theory accounts for it too...

I've never heard that diphthongs can't ever be phonemic and I don't know why it should be like that. I would be thankful if you could quote some source.

[Pittsboy]

Yes, they are two sounds, but they are not just two random vowels that appear beside each other. In fact I'd rather say the [U] in German [aU] is a semi-vowel, but not a real consonant [w] either as the example "hauen" ["haU@n] shows.

I've never said they were random, each language has its own constraints of syllable structure... some vowels are allowed to make diphthongs and others do not...

I think you're contradicting yourself, Pittsboy. First you said only phonemic vowels (no allophones) should count, then you say diphthongs must be separated into the two vowels they consist of. The diphtongs [aI], [aU] and [OI] (plus [EI] from English loans) are phonemic in German. They developped from long vowels, and in some dialects they are still pronounced as long vowels.

No contradiction here! Only phonemic vowles must count... vowels that are at the morphophonemic level!
Diphthongs are two sounds, not an entity, you do not produce two vowels as a single unit, but rather two! Diphthongs are phonemic in the sense that they appear in the morphophonemic level but they are not one whole... that's what I am trying to say... you can say that the phone which is a diphthong are two vowels and not one...

In German, the diphthongs [aI] [OY] etc. don't consist of two different vowels in the speaker's mind.

Yes, and I guess that's why German-speakers don't think pronouncing "EU" as [OI] (Who says [OY]?) is strange. Many German speakers would even pronounce "dos euros" [dOs "OIRo:s] when trying to speak Spanish. [/quote]

What is at the morphophonemic level is not something native speakers are aware of, they simply do that.. that's what a linguist is for... :LOL: samething in any language, you are a native speaker, and the linguistic process involved is mechanic, no need for thinking about it...

My question again: Are their minimal pairs in Portuguese, where a nasal vowel (not followed by a nasal consonant) contrasts with a non-nasal one? If there is, I would say nasalization is phonemic in Portuguese, because the nasal vowel is merely orthographical/etymological.

There 'might' be the nasal written 'a', but I would assure you of that... and it is not my opinion about whether Portuguese has nasal vowels or not, it is a concensus among linguists... the nasal vowel is not etymological, for in Latin there were no nasal vowels either, they were nasalized as well!!

[Pittsboy]

Thi, does EN in assento sound like EN in quente for you? What about you guys (the other Portuguese speakers here)?

I would say I speak both alike, and I tend to hear them alike too, athough some people claim that Portuguese has no nasal [E]...

And should it make any differences the fact that e.g. EI may sometimes be pronounced either [ej] ('educated') or [e] ('colloquial'), but not always?

e.g. beira = ['bej4a] / ['be4a]
but
leite = ['lejtSi] (never *['letSi])

This is more due to restrictions in Portuguese for the occurence of this sort of alternation, than anything else.

[Pittsboy]

So am I. :/ Let's take your example:

['baU.m@] (diphthong)
['ba.U.m@]

They are not my examples...

"two sounds" is not the same as "two phonemes".
About: ['baU.m@] (diphthong) and ['ba.U.m@]:
What is the phonemic feature that distinguishes them in your opinion? They are not phonemically the same in German: ['ba.Um@] would be a completely different word than ['baU.m@]. So, what phonemic feature?

If these two sounds are represented at the morphophonemic level they are phonemes... in the sense that they are part of the language's inventory of vowels... t is a completely different case when you have a diphthong spoken by a native who speaks the standard variation but he does speak a diphthong where there is a single vowel.. that's called variation and has no relation with this matter I am talking about...
Of course both words are different, but they are phonetically different... phonemically they are alike, the difference is how you syllabify the word when it is spoken... a syllable nucleus can only bear one target vowel and other vowels that appear are glides... there's a syllable difference but both strings of phonemes are alike...

Phonemically [aU] are the phonemes /a/ + /U/ which result as an [aU] phone (spoken)... they are two separate sounds that when spoken one of these, due to syllabification, is stronger than the other, and then you have a diphthong... there is not such a thing as a diphthong phoneme...

No, if you don't use a diphthong, people will consider what you speak about as something very different.
Before I learnt that in Spanish diphthongs aren't phonemic, I was always very confused how I could see if I had to pronounce for example the ái in "llamáis" as a diphthong or not. Two Germans, it really makes a big difference and a phonemic difference. It's not like a diphthong is just a variant of the two sounds following one another. If you asked a German to pronounce a word very slowly, he would never get the idea to pronounce the diphthong as two separate vowels.

It is completely different, it is not a matter of phonemes it is a matter of you you syllabify the word... each language has its rules on syllabification... how you syllabify depends on the language you speak... You are mixing things up!!!!!

[ts] (affricate, with a tie bar ontop) are not two phonemes, it is a single phoneme!!!!! It is an [s] which starts as a [t] sound and soon becomes an [s], there're not two articulations... on the other hand, Polish can distinguish this affricate (tie barred) [ts] (which is one phoneme) from a common junction of two phonemes [ts]; here you have two articulations, really fast, but they are two, what differs enormously from the first one, which happens to be an [s] that starts as a [t]...

I don't think in German affricates are really separate phonemes, at least not [pf]. Here it is really as you said: The affricate just comes into being because the two phonemes /p/ and /f/ can easily be joined together into an affricate. No German would consider [pf] to be one phoneme.

Affricates are NOT two phonemes, they are one single sound because you are not adjoining two sounds, you are starting one sound as another one for a brief period of time... it is acoustics, purely acoustics perception,... and it is very different of adjoining two sounds... that's like so that Poles can distinguish them...

I've never heard that diphthongs can't ever be phonemic and I don't know why it should be like that. I would be thankful if you could quote some source.

For God's sake, the diphthong can exist in the phonemic level, bacause it is two sounds combined, but they are still two sounds, they are not one single sound... if that was the case, you would have a tie bar linking both sounds to represent them as one, just as German [ts] which is a single sound.... that's what I have been trying to say... do not confuse thsi anymore...

[elgrande]

a syllable nucleus can only bear one target vowel and other vowels that appear are glides... there's a syllable difference but both strings of phonemes are alike...

It is completely different, it is not a matter of phonemes it is a matter of you you syllabify the word... each language has its rules on syllabification... how you syllabify depends on the language you speak... You are mixing things up!!!!!

So, the two words are distinguished by something that is not phonemic? How weird is that?

Affricates are NOT two phonemes, they are one single sound because you are not adjoining two sounds, you are starting one sound as another one for a brief period of time... it is acoustics, purely acoustics perception,... and it is very different of adjoining two sounds... that's like so that Poles can distinguish them...

I know what an affricate is.
Why can't one sound represent a combination of two phonemes? Some speakers of German use [n^] (like Portuguese nh) as an allophone of /nj/.

For God's sake, the diphthong can exist in the phonemic level, bacause it is two sounds combined, but they are still two sounds, they are not one single sound... if that was the case, you would have a tie bar linking both sounds to represent them as one

Oh, we are just talking about this tie bar? Yes, then give the two vowels a tie bar, they deserve one! So can we finally agree that German has the phonemes /OI/ (with tie bar, don't know the ascii symbol!), /aU/ (with tie bar!) and /aI/ (with tie bar!)?