Help talk:IPA for German

From Wikipedia, the free encyclopedia
Jump to: navigation, search

English approximations are misleading[edit]

Can we have English examples like in the other languages, where there is no effort to make the English example any closer to the language example than necessary? It's just confusing.

Example: ball / ball. Why? The "b" is pronounced similarly/same, but are the two words also pronounced the same? Do they have the same meaning? Which column is which again? Et cetera. 118.7.219.80 (talk) 12:54, 13 July 2010 (UTC)

I'm confused. What's the problem? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 23:38, 13 July 2010 (UTC)
I think the more general problem (which should probably be discussed in WP:PHONETICS) is, shall we have a homogeneous set of English examples that is used across all "IPA for" pages? Of course there can be exception, but at the moment it's pretty much Wild West.
More in particular, the English examples for German are too close to the German word (as opposed to sound), which is unnecessary and in my opinion confusing. 205.228.108.58 (talk) 07:37, 14 July 2010 (UTC)
The English examples are only supposed to illustrate the sound in question. When German Ball and English ball are used for /b/, the /b/ is all that's relevant. Whether the rest of the word sounds similar in German and English is irrelevant. We could just as easily have used the English word bit to illustrate /b/. As for German short /a/, English doesn't really have an equivalent sound, but the closest we have is the vowel of bra, but shorter. The vowel of bat is much further away both articulatorily and acoustically, and the vowel of but has too much dialectal variation within English to be a good guide. Some dialects' STRUT vowel is quite similar to German /a/ (e.g. Australian English), other dialects' STRUT vowel is wildly different (e.g. Northern England English). +Angr 09:05, 14 July 2010 (UTC)
"Whether the rest of the word sounds similar in German and English is irrelevant." - Precisely my point. So why strive to use similar words? It's confusing, especially (but not only) for people who are trying to make sense of the table for the first time.
For similar reasons, we should try and avoid homographs in either column, i.e. don't use "Beet" as a German example and don't use "see" as an English example.
As I mentioned above, it would be nice to solve the problem, "What English word is representative of [b]?" only once, centrally, and use that word (with due exceptions) for all languages. 205.228.108.58 (talk) 09:56, 14 July 2010 (UTC)
I don't really see that the current state of affairs is confusing, but it if it bothers you, change it. I also don't see that it's necessary to have one single English representative of /b/ (etc.) used on this whole family of pages, and I rather doubt enough people would care about it to get a discussion at WP:PHONETICS off the ground. +Angr 10:08, 14 July 2010 (UTC)
(edit conflict) Well, I tried to do change the [a] earlier, but I guess I'm not competent enough. I still think that we can do better than "bra (but shorter)", and indeed other languages have better examples IMO. See Japanese, French, Czech, Danish, Greek, etc (I think Italian, Spanish etc, with "father" or even worse "Barack Obama" are wrong). Interestingly, the person who came up with the example for French and Arpitan used exactly the same example I was proposing, and you reverted (incidentally, in violation of 3RR). My main point being that this kind of discussions should take place only once and in a centralised place, else we keep wasting our time. 123.225.213.73 (talk) 14:59, 14 July 2010 (UTC)
I think the English examples were changed fairly recently to be similar to the German ones, so it shouldn't be too difficult to change them back.
The problem with having one set of examples for each language is that the sounds vary from language to language, even when they use the same IPA symbol. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 14:56, 14 July 2010 (UTC)
"one set of examples for each language" - do you mean "for all languages"? This is true only when the IPA column means /x/, but in most cases it means [x], otherwise the links to sound articles would not make sense. As I said, exceptions will be catered for. 123.225.213.73 (talk) 15:00, 14 July 2010 (UTC)

What variation of English is the "English approximation" using?[edit]

RP (British English) and GA (American English) are quite different in some aspects. Judging from the examples (ɔ Post caught (but shorter), for instance), I guess the article is using American English? I am not a native speaker of neither of the two languages, so I hope someone with better knowledge can write out the answer clearly. —Preceding unsigned comment added by BettyJJ (talkcontribs) 01:42, 27 July 2010 (UTC)

The example works for quite a few dialects, including RP, Irish English, and Scottish English. It works less aptly for American, Australian, Welsh, South African and Kiwi accents because said dialects have a vowel that is more close or more open than the German variant. It's worst for my variety (Californian) and Canadian, which have a lower unrounded vowel that's more like German a. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 02:10, 27 July 2010 (UTC)

Also, "says" is poor re "wähle", since "says" rhymes with "Pez", not "pays". — Preceding unsigned comment added by 184.147.122.14 (talk) 13:34, 3 March 2014 (UTC)

"Says" has a longish [ɛ] vowel, which is closer to the [ɛː] of wähle than the [eɪ] of pays is. Aɴɢʀ (talk) 18:36, 3 March 2014 (UTC)

Pluricentric languages should use phonemic transcriptions[edit]

Compare Wikipedia:IPA for German with Wikipedia:IPA for English: IPA for English uses phonemic transcriptions (/.../) while IPA for German uses phonetic transcriptions ([...]). I think that is not justified. Both languages are pluricentric languages. Of course, more people live in Germany than in Switzerland or Austria, but the same is true for the English language: More people live in the States than in England or Australia etc. Yet still, IPA for German defaults to Germany phonetic pronunciation, while IPA for English does not default to US phonetic transcription but instead provides a phonemic transcription that includes other varieties of English.

In the case of German, this mainly concerns the allophones of /r/: While in many places of Germany, the allophones are [ɐ̯] after long vowels and possibly after short vowels, [ɐ] for /ər/ and [ʁ] elsewhere, in other regions it is [r] elsewhere or even [r] everywhere. -- machᵗᵃˡᵏ 08:21, 5 August 2010 (UTC)

The difference is not whether the language is pluricentric, but whether the reader can be expected to know it, and therefore interpret its phonology. They can be expected to know English, but not German. We also don't default to a US pronunciation because much of our audience does not speak US English. — kwami (talk) 09:55, 5 August 2010 (UTC)
OK, I see. However, this page needs much updating. Other pages from Category:Wikipedia IPA that describe pluricentric languages, for instance Wikipedia:IPA for Spanish, describe regional allophones. And it would be certainly easier not to include three different transcriptions for the allophones [ʁ] and [ɐ̯] and then describe that both may be proncounced as [r] (and [ɐ] as [ər]), but instead one transcription for the allophone [r] and then describe that it may be pronounced as [ʁ] or [ɐ̯] (and [ər] as [ɐ]). I think the mere idea of prescribing certain allophones for all instances of Template:IPA-de is quite ridiculous. A more pragmatic approach would be allowing the allophones in articles with regional topics. -- machᵗᵃˡᵏ 10:32, 5 August 2010 (UTC)

Inappropriate English approximations[edit]

I'm a native German speaker and I have to say that two English approximations doesn't match the German pronunciation at all. Those are face for [eː] bone for [oː]. I'm not an expert in phonology but as I know the second letters of those words are pronounced /feɪs/ and /boʊn/ respectively. Since it's a combination of two vowels (I don't know if there's a technical term for that) where just one is supposed to be explained, I don't consider it an appropriate example. I'd suggest a better word if I was able to. Best wishes -- Rebell0209 (talk) 12:31, 8 August 2010 (UTC)

The term you're looking for is diphthong. And the English approximations are just that: approximations, not exact correspondences. Of course the way an English speaker pronounces bait and boat is different from the way a German speaker pronounces Beet and Boot, but as approximations, they're not too bad. If an English speaker uses the English sounds when speaking German, he'll still be understood. And some English speakers (e.g. those from Scotland or Minnesota) do have monophthongal [eː] and [oː] that are quite close to the German sounds. +Angr 16:25, 8 August 2010 (UTC)
Thanks for the information. I was irritated since the sounds are very different for me and the approximations don't represent the sound equally the way I've learnt to pronounce them. (I don't know anything about the dialects.) However you're right, I would probably understand the word. Consequently, the question is whether the approximations should give an example of the actual sound, or someone using those should be understood. Maybe I was thinking to fussily. -- Rebell0209 (talk) 22:15, 9 August 2010 (UTC)
In the traditional transcription of General American, the signs [e] and [o] are used for the vowels of bait and boat, even though many speakers may diphthongize them. BTW, we are not talking “dialects” here, but just regional varieties of a pluricentric language – similar to the different regional varieties of standard German. -- machᵗᵃˡᵏ 10:08, 10 August 2010 (UTC)

I (German native, English student) would suggest "bed (but longer)" as the English approximation to German [e:] ("Beet") and "cat" as the approximation to German [ɛ] ("hätte")! The vowel [e] in engl. "bed" has at least the same vowel quality as [e:] in "Beet" and is only the short version of it. The near-open [æ]-Sound in "cat" is closer to mid-open [ɛ] in "hätte" than [e] in "bed"! I'm still thinking about a better approximation to "Boot" than the really poor diphthong in "bone"...--139.133.7.11 (talk) 00:01, 11 October 2010 (UTC)

No, the vowel of German Beet has a very different quality from that of English bed in most accents. Probably only New Zealand English has a high enough DRESS vowel to compare it with the German /e:/. German /ɛ/ is much closer to English /ɛ/ than to English /æ/; Ger. Bett and Eng. bet are essentially homophones, while bat is quite different (at least to English-speaking ears; I know Germans generally pronounce bet and bat the same). —Angr (talk) 21:18, 11 October 2010 (UTC)

The R sound[edit]

The sound labeled as an Italian R should be also labeled as a Spanish R, I think english speakers are more used to the R in spanish and therefore more able to identify it. —Preceding unsigned comment added by 190.148.212.65 (talk) 17:05, 17 September 2010 (UTC)

The trouble is that Spanish has two R-sounds; it wouldn't be clear which is meant. —Angr (talk) 13:55, 19 September 2010 (UTC)

Possible error in ɐ̯[edit]

What's the deal with ɐ̯? I can't find it in the official IPA. The English approximation is shown as fun, which I find rather strange. I came here to sort out the proper pronunciation of Worms, Germany, which is shown as ˈvɔɐ̯ms. I know it has an R sound, but I've never heard fun pronounced with an R, so I'm very confused. —UncleDouggie (talk) 03:07, 23 September 2010 (UTC)

Obviously, combinations of letters with diacritics are not listed individually in the IPA chart, you have to look up ɐ and  ̯ separately. It denotes a nonsyllabic ɐ, the latter being a central not-quite-open vowel (indeed, quite similar to the vowel of fun in some English dialects). It is not really an R sound, but an approximant.—Emil J. 14:18, 23 September 2010 (UTC)
The pronunciation varies. Pronunciation dictionaries often follow the convention of using the [ɐ] allophone only after long vowels, not after short vowels, so according to these, Worms would be transcribed as [ˈvɔʁms]. Wikipedia seems to have a tendency to generalize [ɐ]. Yet other pronunciation dictionaries use [r] throughout, for instance, [ˈvɔrms]. -- machᵗᵃˡᵏ 06:34, 24 September 2010 (UTC)
I rarely heard someone say [ˈvɔʁms]. The 'r' is barely hearable. —Preceding unsigned comment added by 79.219.62.117 (talk) 12:54, 3 November 2010 (UTC)
But Wikipedia should stick to knowledge published in reputable sources such as pronunciaton dicitionaries rather than establish norms on the base of "I (rarely) heard someone say."--ペーター (talk) 09:36, 17 May 2011 (UTC)
[ʁ] in Worms does not denote a fricative, but an approximant or a uvular-r-coloured vowel. 89.79.133.41 (talk) 14:50, 1 July 2013 (UTC)

Fastelavend / Fastelabend: n or m?[edit]

The IPA transcription for "Fastelavend" / "Fastelabend" in Carnival in Germany, Switzerland and Austria is given as: "Fastelavend" = [ˈfastl̩.ˌɒːvm̩t], "Fastelabend" = [ˈfastl̩.ˌɒːbm̩t] using a syllabic where I expect a plain n. Can someone confirm my expectation? -- Michael Bednarek (talk) 05:50, 4 October 2010 (UTC)

This is an assimilation typical of the casual language (Umgangssprache, similar cases would be [ˈɡlaʊbm̩] instead of [ˈɡlaʊbən] for glauben ‘to believe’ or [ˈkʁaːɡŋ̩] instead of [ˈkraːɡən] for Kragen ‘collar’). A more dictionary-like pronunciation of the word in standard German would be [ˈfastəl.ˌaːbənt]. The article you are referring to, however, seems to indicate a special regional pronunciation, so I cannot tell whether it is accurate or not. By the way, the most remarkable thing about that regional pronunciation seems to be the rounded pronunciation of [ɒː]. -- machᵗᵃˡᵏ 07:47, 4 October 2010 (UTC)
It looks precise enough to suppose that the writer knew what they were doing. Should I delete the tags? — kwami (talk) 08:18, 4 October 2010 (UTC)

Marginal sounds from English[edit]

Do we really need the English phonemes /w/, /ɔː/ etc? I'd vote to remove them, as they clutter the table. Do any articles use them? Lfh (talk) 10:11, 13 May 2011 (UTC)

I wouldn't expect to need them. We can always put them back in on a case-by-case basis if we do need them. — kwami (talk) 17:19, 13 May 2011 (UTC)
OK. I'll take them out. Lfh (talk) 10:16, 14 May 2011 (UTC)
But I've decided to keep /ʒ/ and /dʒ/, I think they're necessary, due to words like Genie and Dschungel. Lfh (talk) 10:25, 14 May 2011 (UTC)

Non-native vowels[edit]

The following quote from the article is a little peculiar:

Non-native vowels
e Methan (short [eː])
i vital city (short [iː])
o Moral (short [oː])
ø Ökonom (short [øː])
u kulant (short [uː])
y Psychologie (short [yː])

I do see that the examples given are loanwords in the German language, hence non-native. The vowels, however, are native German, plain and simple. So I really do not see the point of them being on this list. I would like to entirely remove these lines. TobyDZ (talk) 22:33, 30 August 2011 (UTC)

These vowels occur short only in non-native words. In native German words, they're always long. Angr (talk) 23:04, 30 August 2011 (UTC)
The question for me is whether we would ever need them for our purposes. Why in the world would we want to transcribe non-native German words in German? Though I suppose there may be occasion. — kwami (talk) 23:33, 30 August 2011 (UTC)
The words may not be of German origin, but to Germans the words don't appear to be non-native. English has also many Latin and Greek words, but you still transcribe them, don't you? --2.245.242.131 (talk) 20:06, 30 June 2014 (UTC)
In names of political parties, such as Christian Democratic Union (Germany), or perhaps in names of books. — Eru·tuon 23:45, 30 August 2011 (UTC)
"These vowels occur short only in non-native words. In native German words, they're always long." That's not exactly true. Long vowels are shortened to half long [eˑ iˑ oˑ øˑ uˑ yˑ] when they're unstressed, for example the word Auto is phonemically /ˈaʊ̯toː/ and phonetically [ˈʔaʊ̯tʰoˑ]. The same applies to Dutch, at least upper class ("Posh") Netherlandic and standard Belgian accents. --89.79.133.41 (talk) 06:00, 1 July 2013 (UTC)
But Auto is a nonnative word. Angr (talk) 21:06, 3 July 2013 (UTC)
Doesn't matter. It happens in native and non-native words. --89.79.133.41 (talk) 19:41, 11 July 2013 (UTC)

R is not like t or ch[edit]

Two of the current entries are plain wrong:

ʁ roughly like loch (Scottish English)

r roughly like water (American English)

As there is obviously no English sound that comes near close enough to the R-sounds in question, I think it would be nice to link the sound recordings for Voiced uvular fricative and Alveolar trill, respectively. Unfortunately, I don't know how to do that. Help, please?--Senfteiler (talk) 11:23, 18 July 2013 (UTC)

The page already links to those articles, where the sound files can be heard. Angr (talk) 14:11, 18 July 2013 (UTC)

p, t, k[edit]

These consonants are aspirated. Shouldn't we add an "h"?--77.0.253.160 (talk) 22:23, 28 August 2013 (UTC)

No, in German /t/ and /th/ are not contrasting. It is customary to choose the least modified IPA symbol in case no phonemic distinction is made. −Woodstone (talk) 07:31, 29 August 2013 (UTC)
I fully agree! There is an unfortunate tendency in (the) German (wikipedia) towards using unnecessarily complex transcriptions. --mach 🙈🙉🙊 16:14, 29 August 2013 (UTC)
But it's absolutely wrong if e.g. French p, t, k and German p, t, k share the same transcription. Aspiration should be mentioned.--2.246.33.101 (talk) 20:18, 13 December 2013 (UTC)
Transcriptions can't be completely exact. /p/ represents a wide variety of sounds depending on the language and its phonetics and phonology. The sound it represents could be aspirated ([pʰ]), unaspirated ([p˭]), or unreleased ([p̚ ]), palatalized ([pʲ]), labialized ([pʷ]), velarized ([pˠ]), pharyngealized ([pˤ]), or glottalized ([pˀ]), and other things, to varying degrees. The transcription might include these features or might not, depending on whether it's necessary in the context where the transcription is being used. In transcribing German p, aspiration doesn't have to be mentioned, because German doesn't have aspirated and unaspirated ps that contrast. Maybe it would be good to mention aspiration in the context of comparing French and German, but that's a different situation. — Eru·tuon 00:55, 14 December 2013 (UTC)
This issue is very complicated, and it is best not to transcribe the aspiration. Here's why:
1. /p, t, k/ are usually aspirated in Northern Standard German (however, Krech et al. (2009) state that aspiration is merely "possible"), variably aspirated in Austrian Standard German and usually not aspirated in Swiss Standard German.
2. In varieties with aspirated /p, t, k/ they are not aspirated when /s/ precedes within the same syllable. This means that we must either list [pʰ, p, tʰ, t, kʰ, k] separately, or transcribe syllable breaks whenever /s/ ends the syllable preceding the one starting with /p, t, k/. The latter would lead to transcriptions like [spʰɔʁtʰ] for Sport, which is plain ridiculous. (The second possible pronunciation of Sport, [ʃpʰɔʁtʰ], would be transcribed correctly, as e.g. Mangold (2005) reports that /p, t, k/ are aspirated after /ʃ/ within the same syllable)
3. In many colloquial accents in the north of Germany unstressed /p, t, k/, as well as /p, t, k/ that follow /s/ within the same syllable are realised as voiceless lenis [b̥, d̥, ɡ̊]. This is probably also true for some colloquial Austrian accents. — Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 13:11, 13 February 2015 (UTC)

Which English word is most like ü?[edit]

I added the word cute as an approximation for ü, and Meemo16 reverted me, saying that shoe was a closer approximation. It looks like we need to have a discussion on this question.

Which is a better approximation for /yː/ and /ʏ/ – English /uː/ in shoe or English /juː/ in cute? My opinion is that cute is a better. Based on the International Phonetic Alphabet chart for English dialects, shoe has a back vowel [uː] in conservative dialects, a near-back vowel [u̟(ː)] in American and Canadian, and a central vowel [ʉː] in Australian, sometimes RP, and some other dialects. Therefore, shoe is most similar to /yː ʏ/ in dialects like Australian, not in North American and more conservative dialects. The sequence [ju], [ju̟], or [jʉː] in cute is a better approximation for /yː ʏ/, if we intend to have an example that applies to all dialects, since it contains a high front element plus a rounded element.

What do others think? And Meemo16, what's your argument for shoe? — Eru·tuon 05:27, 27 February 2015 (UTC)

Lenis obstruent symbols[edit]

I'm curious, are the symbols for voiced obstruents with the devoicing diacritic, like , actually used on Wikipedia to transcribe German lenis consonants? If not, the note on them should be removed. On the other hand, final devoicing is probably transcribed, so we need a note on it. — Eru·tuon 19:22, 29 March 2015 (UTC)

Actually the note should be retained, as Standard German spoken in Switzerland and Austria doesn't feature final devoicing. That note explains why. Peter238 (talk) 09:44, 24 June 2015 (UTC)

Where is the German "w"?[edit]

The German "W" is missing completely! --80.147.136.188 (talk) 08:49, 18 May 2015‎ (UTC)

It's not, it is transcribed /v/, and placed under /tʃ/ and above /x/. Also, don't forget to sign your messages. Peter238 (talk) 08:53, 18 May 2015 (UTC)

Austrian, Germany-German, Swiss[edit]

The possibility of having seperate table columns for the different varieties of the standard may be an improvement. But why lump Austrian and Germany-German together? At least with regards to the realization of /r/, they should be different, since in Austria, the pronunciation [r] prevails. There may be other similar cases (I am no expert for Austrian standard German).

Also, who says that cells should contain only one symbol? When there are variant pronunciations, they should be indicated. Especially when people take a rigid prescriptivist point of view and will only allow the pronunciations that are listed on this help page. (I am an inclusionist also with regard to pronunciation.) --mach 🙈🙉🙊 19:02, 24 June 2015 (UTC)

The only reason they are in the same cell is that I was not aware of any important differences between GSG and ASG. Apparently, the [ʁ] - [r] thing is one of them, so I'll create a separate Austrian cell.
Because this guide is not meant to be super-accurate, and because the more symbols we use, the more confusing this guide will get for laymen. It's enough to point out the difference in rhotics and /ɔʏ/ and have most common realizations here. By all means, if you have sources, add as much as you can (and want) to Standard German phonology. Peter238 (talk) 19:11, 24 June 2015 (UTC)
How is this guide not meant to be super-accurate? It super-accurately indicates allophonic variation. By this I mean that it is more accurate than what is necessary, indicating sub-phonemic allophones. This mainly concerns syllabic sonorants – we might as well have /ən/, /əm/, /əl/ and /ər/ – and the allophones of /r/ – we might as well have /r/. That is indeed the transcription we find in dictionaries, such as the Rechtschreibung Duden or even the Langenscheidt bilingual dictionaries. The sub-phonemic allophone distinctions are mainly found in specialist pronunciation dictionaries. Also, a broader dictionary-style transcription could be used regardless of the different varieties of the standard.
Yet Wikipedia has a strong tradition of super-accurately indicating allophonic variation in transcription of German (not only here, but also on the German wikipedia), which gives it a strong bias towards the Northern Germany-German (Prussian) pronunciation. Since I cannot change that, the only way to counter that Prussian bias is by equally including allophones from other standard varieties of German. --mach 🙈🙉🙊 05:08, 25 June 2015 (UTC)
I certainly would not describe this guide as super-accurate. If it were super-accurate, it'd include:
- aspirated and unaspirated fortis plosives
- voiced and voiceless lenes
- unreleased plosives
Or more, why stop there?
- denti-alveolar, laminal alveolar and apical alveolar coronals
- Both fricative and approximant variants of /j/. What's even funnier: scholars don't agree on its nature.
- Every possible variant of [ɐ̯], which not always is phonetically [ɐ̯]
- Why transcribe the diphthongs /aɪ, aʊ, ɔʏ/? According to Mangold (2005) and Wiese (1996), these are phonetically accurate. But according to Krech et al. (2009), these are really [aɛ, aɔ, ɔœ] (but they also list other variants for ASG and SSG), and according to the diphthong chart in Kohler (1990), they are [ae, ao, ɔe].
- Older sources disagree on the backness of /aː, a/. Do we list all [aː, äː, ɑː, a, ä, ɑ], or without the 'centralized' diacritic (if so, why?) [aː, ɑː, a, ɑ]? Note that none of the recent sources (at least as I am aware of) describe northern /a/ as back. If it's an obsolete realization, well, why stop there, and not add other obsolete realizations?
- The list can go on and on and on.
Can you enforce consistent two transcriptions (or more) in every article that would need that? Like this: [ˈtsyːrɪç], [ˈtsyːʁɪç]. I suggest that you can't, as it'd be widely considered a complete waste of space. It's another reason why we can't have more than one symbol per cell, as they all need to be used consistently. I want to remind that this article is a guide for non-native speakers, not Standard German phonology.
The biggest Polish-German German-Polish dictionary made by PWN treats [m̩, n̩, l̩, ɐ, ɐ̯] as phonemes (maybe [ŋ̍] too, I don't remember). So that's at least one relevant dictionary that does use these symbols in phonemic transcription. They certainly are relevant, as this assimilation is obligatory to sound 'normal' in the north (probably also the case in ASG and SSG), and they sound distinct enough to non-native speakers of German to include them here.
The only reason I picked /ʁ/ to represent the northern rhotic is that in The Sounds of the World's Languages, the authors state that the uvular trill "occurs in some conservative varieties" and that "most speakers with a uvular /r/ realize it as a fricative or an approximant.", both of which are correctly represented by [ʁ] in IPA. Peter238 (talk) 08:59, 25 June 2015 (UTC)
It is important that Zürich can be transcribed as [ˈtsyːrɪç] with an alveolar [r] and that Karl Barth, native of Basel, can be transcribed as [ˈbaʀt] with an uvular [ʀ]. That is in accordance with the natural local standard pronunciation. I would not want to take a rigid prescriptivist POV and enforce some unnatural artificial “standard” pronunciation in either article. That would be as inadequte as prescribing RP transcriptions such as [ˌnjuː ˈjɔːk] or [dʒɔːdʒ wɔːkə bʊʃ]. It infringes against WP:NPOV.
I think we should avoid a bias towards any one standard pronunciation. Instead, we should be flexible and allow for different standard pronunciations. Anyway, we cannot enforce a prescriptive standard transcription in the thousands of articles that use IPA-de. Therefore, I think the most sensible solution is mentioning the more common standard variants such as [ʝ ae ɔø ao ɑ ʀ] in this help page – aspiration and voiclessness are already mentioned. I think that this maximizes the helpfulness of this help page. Other solutions are less helpful to Wikipedia users. When we fail to explain valid variant pronunciations of standard German on this help page (for instance transcriptions with [ɑ] like on Oberaargletscher), users who come here do not get adequate help. When we change valid variant pronunciations of standard German in an article only to enforce our own prescriptivist POV (as you did in [1]), valuable local information from the article is lost to the users. When we redirect valid variant pronunciations of standard German that do not agree with our prescriptivist POV to a generic IPA template (as you did in [2]), users are deprived of the specific help this help page can provide.
Of course, we should not infringe against WP:OR. If there is any doubt, we should carefully provide WP:SOURCES for the variants of the standard. --mach 🙈🙉🙊 13:12, 25 June 2015 (UTC)
Krech et al. (2009:266) say that "As in Germany, any of [ʀ], [r] and [ʁ] are possible realizations of /r/ in Switzerland. Because most dialects of Swiss German use [r], it is often also used in the standard language." (My loose translation). It's best to use [r] when transcribing Swiss Standard German. If it's necessary, make a footnote (not here, in the article) which says "In Basel, [r] is realized as an uvular trill [ʀ].", or something similar. That way, if you use the IPAblink template, the user who clicks the ʀ symbol will be redirected to the uvular trill page, where he'll be able to learn about that sound.
The aim of the 'IPA for X' guides is to represent the spoken standard language as described by reliable sources. If there's some free variation in it, we simply pick one symbol (which either represents the most common realization, is easier to type, or both) and stick with it. I'd argue you're still trying to make Standard German phonology out of this page, which is not its purpose. Peter238 (talk) 13:49, 25 June 2015 (UTC)
Of course it’s not. It is a help page that should be helpful to users. But I insist on WP:NPOV, and I think I am right in doing so. I see no basis whatsoever for your insistence on a prescriptivist POV. --mach 🙈🙉🙊 15:24, 25 June 2015 (UTC)