Help talk:IPA/Standard German

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
WikiProject iconWikipedia Help NA‑class
WikiProject iconThis page is within the scope of the Wikipedia Help Project, a collaborative effort to improve Wikipedia's help documentation for readers and contributors. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks. To browse help related resources see the Help Menu or Help Directory. Or ask for help on your talk page and a volunteer will visit you there.
NAThis page does not require a rating on the project's quality scale.
WikiProject iconLinguistics: Phonetics NA‑class
WikiProject iconThis page is within the scope of WikiProject Linguistics, a collaborative effort to improve the coverage of linguistics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
NAThis page does not require a rating on Wikipedia's content assessment scale.
Taskforce icon
This page is supported by Phonetics Task Force.

Overly detailed descriptions of vowels[edit]

Do we really need to describe e.g. [aː] as bad (RP, Irish but slightly more backed), mark (Northern England, Australian), bod (American)? We don't do that on any other page, and the title of the column clearly says English approximation. Mr KEBAB (talk) 18:01, 25 September 2017 (UTC)[reply]

Pinging the editor who introduced that: @Ranníocóir:. Mr KEBAB (talk) 18:03, 25 September 2017 (UTC)[reply]

Why not? Surely this is helpful for speakers of these varieties and for people with intricate knowledge of English varieties. Better than nothing. And there is no denying that these are still approximations. --mach 🙈🙉🙊 18:49, 25 September 2017 (UTC)[reply]
@J. 'mach' wust: Surely this is helpful for speakers of these varieties and for people with intricate knowledge of English varieties. Is that what we want though?
There is no denying that these are still approximations. The one I quoted is an exact description of the sound in question. It's quite clearly not an approximation. I'm not sure if we're talking about the same tables, because I can't see many approximations there.
By writing better than nothing you're suggesting that we have a different (or at least a more narrow) choice than we actually have. The choice we have is between giving general English approximations of German sounds as it's done on other Help:IPA/X pages and what we have now. I don't think that anyone would take the idea of removing English approximations altogether seriously (and I'm not even saying that that's what you're proposing, I'm sure that's not the case). It'd be much better to remove this page instead, because doing the former would render this guide pretty much useless. Mr KEBAB (talk) 22:50, 25 September 2017 (UTC)[reply]
I agree that the overly detailed description doesn't help anyone. IMO, the concise version from 27 January 2017 is less confusing and more helpful than the current one: diff. -- Michael Bednarek (talk) 04:36, 26 September 2017 (UTC)[reply]
You have still not given any reasons why you wish to remove the additional information other than keeping symmetry with other similar pages. I am just saying that this is not a valid reason for removing information. A valid reason would be the WP:CCPOL. Put a WP:V flag on the article and wait some time so others have a chance to find sources. If they do not, then you have a valid reason for removing the information. --mach 🙈🙉🙊 05:55, 26 September 2017 (UTC)[reply]
@J. 'mach' wust: Just because we can source something it doesn't automatically mean that we should keep it. We're not arguing that the information is wrong or unverifiable but that it's overly detailed and it's misleading to call these extremely elaborated descriptions 'approximations' because they're not that. This page is not designed to be a complete guide to pronouncing standard German. There are books for that, like Modern German pronunciation: An introduction for speakers of English cited in this guide. We also have Standard German phonology, which also covers phonetics.
I object to tagging the article. Nowhere else do we require English approximations to be sourced nor do we allow narrow phonetic descriptions of sounds to be called 'approximations' of them. I see no basis for treating this page as if it were different than other guides. German is not special, it's a language like others. Mr KEBAB (talk) 08:10, 26 September 2017 (UTC)[reply]
Just because the approximations on this help page are slightly elaborated it does not mean that we should simplify them again. And just because the approximation on this help page is different from others does not mean we have to change it back. Neither reason is valid.
And concerning the word “approximation”: Of course it is an approximation. With a mere dozen of words, it cannot be anything else. An exhaustive description – as opposed to a mere approximation – would demand entire paragraphs, if not chapters. --mach 🙈🙉🙊 16:45, 26 September 2017 (UTC)[reply]
@J. 'mach' wust: Ok, let's turn this around: do you have a valid reason for keeping elaborate descriptions of vowels on this page? We find them unhelpful, too specific and harder to relate to for people that don't know the IPA compared to what we had before.
So a detailed description of something can't have the length of one sentence and it must be called an approximation. Well, almost all descriptions of German vowels on this page prove you wrong. You can describe German vowels in a more detailed way, but the way we describe them crosses the line between approximations and fairly exact phonetic descriptions. Plus, explaining IPA with the IPA? Is that what we want? Mr KEBAB (talk) 17:33, 26 September 2017 (UTC)[reply]
@Mr KEBAB: I concede that my reasons for the longer approximations may be just as subjective and unconvincing as your reasons for the shorter approximations. However, my main point is the following: If there are two acceptable revisions, we should always prefer the more recent one because this is Wikipedia. It thrives on change and participation and diversity. WP:DONTREVERT. --mach 🙈🙉🙊 20:08, 26 September 2017 (UTC)[reply]
@J. 'mach' wust: Fair enough. The question is: are they equally acceptable? Judging by what this page actually is (a guide for laymen, often with little knowledge of IPA), that's questionable. Especially because we're trying to help people understand IPA by using more IPA. This is a bit silly and I think you can admit that we should at least remove any IPA symbols from that column. That'd be a good start. Mr KEBAB (talk) 20:17, 26 September 2017 (UTC)[reply]
@Mr KEBAB: Agreed, removing the IPA symbols from the descriptions is a good idea. Other than that, I think the descriptions are helpful. They all refer to well-known varieties of English. That covers people without any knowledge in phonetics. And then, they go on to give some hints about where the German sounds differ. This is additional information for people with sufficient understanding of phonetics. --mach 🙈🙉🙊 20:24, 26 September 2017 (UTC)[reply]
@J. 'mach' wust: Ok, will do. Mr KEBAB (talk) 20:28, 26 September 2017 (UTC)[reply]
@Mr KEBAB: I am reverting your edits. The only thing we have agreed upon is removing the IPA, but you have gone much further than that. You have pretty much reverted all the changes, and you have obscured your edits by making tons of subsequent edits. Please do not do that. --mach 🙈🙉🙊 06:06, 27 September 2017 (UTC)[reply]
@J. 'mach' wust: You've missed the change to lead. Mr KEBAB (talk) 06:08, 27 September 2017 (UTC)[reply]
@J. 'mach' wust: What did you reinstate the incorrect information for? I wrote the edit summaries for a reason. To call my multiple edits 'obscuring what I really do' or whatever you meant is completely incorrect. Mr KEBAB (talk) 06:11, 27 September 2017 (UTC)[reply]
@J. 'mach' wust: Sorry, I might've overreacted a bit but my point still stands. Maybe I went too far with removing certain varieties (e.g. Irish) from the table, but it was full of errors. It's as if the editor who put it there just took a quick look at Wikipedia articles on varieties of English without reading any sources that discuss their phonetics. We both completely missed that in our discussion above. Mr KEBAB (talk) 07:14, 27 September 2017 (UTC)[reply]
@Mr KEBAB: I am sorry for accusing you of obscuring. However, your tendency of making dozens of subsequent tiny edits makes it rather hard to follow. Of course, I can view all your edits at once, but tying the numerous edit summaries to the individual edits is a labyrinthical task.
I am now trying my version: Keeping the expanded descriptions, but – as we agreed upon – removing the IPA.
BTW, I agree that a few of the descriptions are questionable. However, I think that most are very accurate, including many of those you dismiss as completely wrong (or do you – it is kinda hard to tell), e.g. “bod (American)”, that is, [ɑː], for German /aː/, also (potentially) [ɑː], or “square (Modern RP)” for German /ɛː/, both [ɛː]. --mach 🙈🙉🙊 19:02, 28 September 2017 (UTC)[reply]
@J. 'mach' wust: Yeah, I should've listed the wrong examples. I'll do it soon. Mr KEBAB (talk) 19:15, 28 September 2017 (UTC)[reply]
Nardog cleaned up the guide for us a few months ago. Sorry for not responding back in September. Kbb2 (ex. Mr KEBAB) (talk) 14:14, 24 July 2018 (UTC)[reply]

a pronunciation[edit]

why does it say "The Austrian and Swiss pronunciation of /a/ and /aː/ is [ɑ] and [ɑː]" in a separate note instead of just having the column split like a | ɑ (like with r)? LICA98 (talk) 16:48, 12 February 2018 (UTC)[reply]

The symbols in the first column comprise the set of symbols allowed to use inside {{IPA-de}}. The column can be split, yes, but then we would be prescribing the use of ⟨ɑ⟩ and ⟨ɑː⟩ rather than ⟨a⟩ and ⟨aː⟩ as far as Austrian and Swiss notations are concerned. An argument against it may be that it would be too many instructions to editors and difficult to maintain, or that it would be too much precision when there's no phonemic contrast or (unlike [ɐ̯] vs. [r] etc.) phonotactic variation. Nardog (talk) 17:00, 28 February 2018 (UTC)[reply]
didn't understand what you mean by "we would be prescribing the use of ⟨ɑ⟩ and ⟨ɑː⟩ rather than ⟨a⟩ and ⟨aː⟩ as far as Austrian and Swiss notations are concerned" LICA98 (talk) 17:08, 6 March 2018 (UTC)[reply]
Currently, the key instructs editors, as designated by MOS:PRON, to transcribe within {{IPA-de}} the German rhotic as [ʁ] for German (or regionally unspecified) German and as [r] for Austrian and Swiss German. If we split the cells for [a] and [aː] into [a, ɑ] and [aː, ɑː], we would be demanding those editors use [ɑ, ɑː] in place of [a, aː] in Austrian or Swiss German notations the same way they (are supposed to) use [r] in place of [ʁ]. Nardog (talk) 18:13, 6 March 2018 (UTC)[reply]

RP?[edit]

The use of “RP” in “Scottish, or RP and Irish” (for example) is not, at least for me, self-explanatory. Received Pronunciation? Like English via BBC? or what? Thanks — Jo3sampl (talk) 03:35, 28 February 2018 (UTC)[reply]

Yes, Received Pronunciation is what it stands for. But IMO the current vowel table is utter overkill. Especially for vowels that sound practically the same in both RP and General American, those qualifiers are definitely extraneous. Also the quality of /uː/, /ɑː/, etc. varies quite a lot even within RP or GA. The "English approximation" column is for those who have no idea what "back" or "closed" or "cot–caught merger" means—were they so linguistically sophisticated, they wouldn't need those reference words in the first place. Others interested in the precise quality of a vowel can simply click on the symbol. Nardog (talk) 14:21, 28 February 2018 (UTC)[reply]

Variants of Standard German[edit]

As of now, we cover only Standard German as spoken in Germany, Switzerland and Austria. So what about countries like Belgium, Luxembourg, Lichtenstein or Namibia? Is it safe to assume that SG as spoken in Lichtenstein is basically Swiss SG, whereas the other 3 countries adhere to German SG? Mr KEBAB (talk) 15:41, 13 May 2018 (UTC)[reply]

Luxembourgish phonology says that there's generally no [m̩, n̩, l̩] in Luxembourg, where only [əm, ən, əl] are used. So that's at least one difference between Luxembourgish SG (if that's even a thing...) and German SG. Mr KEBAB (talk) 15:45, 13 May 2018 (UTC)[reply]

I doubt a distinct standard pronunciation which is taught in drama schools, broadcasting services, etc. has emerged in Belgium, Luxembourg, Liechtenstein or Namibia. And I know for sure that some of Luxembourg's and Namibia's news anchors, television presenters, radio announcers and/or show hosts are Germans or grew up in Germany. — Let me note two points: 1. Luxembourgish is a stanardized language of its own, so it is unsafe to reason by analogy. 2. As long as we don't have reliable sources it is a mute point to discuss this topic. Love —LiliCharlie (talk) 22:31, 13 May 2018 (UTC)[reply]
@LiliCharlie: I actually didn't reason by analogy. Luxembourgish phonology lists that example and explicitly mentions Standard German in the sentence (and so does the source). But I agree, there's really no point in discussing this without reliable sources. Kbb2 (ex. Mr KEBAB) (talk) 14:16, 24 July 2018 (UTC)[reply]

vowel like RP "heart"[edit]

Is the close-mid front rounded vowel denoted by ø (as in, e.g., ökonom) really "somewhat like RP heart"? I'm not a native speaker of Standard German or Standard British English, but this seems wrong. JGambolputty (talk) 09:12, 27 May 2018 (UTC)[reply]

Yeah, totally wrong. RP heart is a long back vowel. I replaced it with hurt. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 13:50, 27 May 2018 (UTC)[reply]
Isn't the vowel of New Zealand English heard/hurt much closer to Standard German [øː/ø]? It is rounded, much higher and a bit fronter than its RP equivalent. Love —LiliCharlie (talk) 04:44, 28 May 2018 (UTC)[reply]
If we can avoid dialect-specific pronunciation equivalents, it's more helpful IMHO. Pretty much no matter what dialect of English you speak, hurt will be the closest approximation. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 05:46, 28 May 2018 (UTC)[reply]

[o][edit]

@JGambolputty: Dubious how? Also RP THOUGHT is usually higher than the cardinal [ɔ], sometimes approaching [o], so not that "roughly". Nardog (talk) 05:14, 30 May 2018 (UTC)[reply]

I was just typing a comment in which I was about to say that more competent others should feel free to revert my edit if they think it made things worse. The reason I took out the example of American English low was that it seemed to me to be suggesting an equivalence between the diphthong [oʊ] and the sound [o]. Also, the (previously unqualified) example of RP thought was already present, and I added the qualifier "roughly" because, as you point out, the sound seemed a bit higher than [o]. However, as I say, if you think my logic is wrong, please feel free to change it back. JGambolputty (talk) 05:29, 30 May 2018 (UTC)[reply]
Looking more carefully, I see now that only the letter 'o' in low was bolded, which I suppose should not be ambiguous. I have added back low while retaining the qualifier "roughly" in connection to the RP example. JGambolputty (talk) 05:40, 30 May 2018 (UTC)[reply]
Thanks for the feedback. Yes, that was the intent as for the low example. Still not sure about what you mean about thought though. RP /ɔː/ is not higher than [o]. (I assume you meant to say "higher than [ɔ]"?) Standard German [o(ː)] seems to be exactly like the cardinal [o] according to Standard German phonology#Vowels, to which RP /ɔː/ is fairly close. In fact we're already giving law as the example for [oː]. Nardog (talk) 05:52, 30 May 2018 (UTC)[reply]
You might have been confusing phonemic notation and actual realization. When I say "RP THOUGHT is usually higher than the cardinal [ɔ]", I mean that the vowel in thought, conventionally referred to as the lexical set THOUGHT and traditionally transcribed phonemically as /ɔː/, is realized in Received Pronunciation with a quality between the Cardinal Vowels [ɔ] and [o]. Compare the vowel diagrams seen in IPA chart, Received Pronunciation, and Standard German phonology. Nardog (talk) 06:17, 30 May 2018 (UTC)[reply]
I'm sorry; I misread what you wrote *and* typed the wrong the character. What I meant to say is that the [o] in originell seemed higher to me than [ɔ] as in RP thought (/θɔːt/), hence "roughly." When you said that thought is "not that 'roughly'" close to originell, I took you to be saying that they aren't that similar, whereas I now see that you meant they aren't that different. If you think that RP thought is actually higher than [ɔ] and thus closer to [o] than I had thought, then by all means remove roughly. I apologize for the confusion: I should have just left well enough alone! JGambolputty (talk) 06:23, 30 May 2018 (UTC)[reply]
Ah, now I see what happened. No worries, sorry for the vague wording in the original post. Nardog (talk) 06:28, 30 May 2018 (UTC)[reply]
I've removed low as an example. As can be seen by this thread, it's confusing. In general, breaking up diphthongs in these approximations is often less than helpful.
I've also tweaked a few of the other English approximations. I don't like delving into more obscure dialects to find an exact phonetic replica. These are approximations and so we shouldn't be worried if the English approximations are imperfect. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 16:31, 30 May 2018 (UTC)[reply]

Remove Austrian and especially Swiss SG from the guide[edit]

There are multiple reasons to remove these from the guide.

First of all, Northern Standard German is clearly the most prestigious variety, and many (EDIT: maybe that's an exaggeration, some is probably a more accurate description) speakers of Standard German from Austria imitate some of the features of NSG when they speak Hochdeutsch. Moosmüller, Schmid and Brandstätter (2015) even had to select specific speakers for their study because the variety of Standard German some Austrians (e.g. news presenters) speak is too mixed with NSG.

Swiss Standard German is the least prestigious, because many speakers don't vocalize their /r/'s and when that's coupled with alveolar realizations thereof (not to mention using velar/uvular fricatives where [ç] is present in other varieties of SG), the result sounds strongly regional if not borderline non-native (though this might be an exaggeration) to other speakers of SG. The same goes for overly open realizations of /eː, øː, oː/. It's somewhat like speaking Dutch with a strong West Flemish accent. It just doesn't sound standard, even though it might be in that particular region of Belgium. My point is, it's highly probable that SSG is standard in Switzerland and pretty much nowhere else. It's not a widely acceptable variant of pronunciation, and because of that we shouldn't include it here. NSG is acceptable everywhere, SSG probably isn't. Let's not forget that for many southeners SG is a second native language, if not an actual second language (it's kind of like the situation of Standard Dutch in Limburg). We really shouldn't encourage speakers of English to pronounce German in the Swiss way. It should be their conscious choice to adopt such a marked accent (that is, if my reasoning is correct). Would you teach the New Zealand accent to someone who is learning English?

Also, let's not forget that this guide can be both allophonic and diaphonemic. When we transcribe besser as [ˈbɛsɐ] this also covers variants such as [ˈb̥ɛsɐ], [ˈb̥esəɾ], etc. Of course, this doesn't work in all cases: [ˈtɑːɡ̊] isn't recoverable from [ˈtaːk], though it kind of is when you also look at the spelling.

There also are reasons not to include ASG, or at very least to change our representation of it, which seems to be unsourced and simply untrue. The source of the following descriptions is Moosmüller, Schmid and Brandstätter (2015).

  • Consonants
    • [z] appears just as a possible intervocalic variant of /s/.
    • Lenis plosives (which normally are voiceless) may be phonetically voiced when intervocalic, either as stops or as fricatives.
    • Intervocalic fortis plosives can be realized as long voiceless lenis stops. This and the aforementioned voicing of the lenis plosives introduces a complication to the transcription. Word-final /s/ can also be phonetically long.
      • This also occurs in colloquial NSG, for example in the (former) Low Saxon-speaking area.
    • Furthermore, fortis plosives might or might not be realized as voiceless lenis also in some other positions.
    • Conversely, word-final lenis plosives might actually be fortis, as in NSG.
    • /n/ tends to assimilate to the place of articulation of the following obstruent.
      • Nothing special, this also occurs in actual spoken NSG and doesn't need to be transcribed.
    • The predominant pronunciation of /r/ isn't [r] but [ʁ], as in NSG.
    • The distribution of [x] and [ç] in ASG is a bit different to what we can hear in NSG, as in ASG [x] is also used after the centering /r/-diphthongs, as in Kirche [ˈkɪɐ̯xɛ].
    • The glottal stop is much rarer in ASG than in NSG.
    • The non-native [dʒ] and [ʒ] seem not to exist in ASG.
      • This kind of merger appears in many varieties of SG and ASG isn't special in this regard.
  • Vowels
    • /a, aː/ are back [ɑ, ɑː], not central.
    • /ɛː/ doesn't exist and it completely merges with /eː/.
      • Many speakers of SG across Europe merge the two, so it's nothing special. However, speakers of ASG might be a bit more consistent in merging them, kind of like Americans are in flapping their alveolar stops in comparison with Australians.
    • The schwa /ə/ doesn't exist and it's replaced by either /ɛ/ or a short [e].
      • This also happens in many regional variants of SG.
    • There can be a phonetic zero where a phonetic (and phonological) schwa would appear in NSG, as in gesagt [ɡ̊sɑːkt] (notice that it has one syllable) or [ɡ̊əˈsɑːkt].
    • /ɪ, ʏ, ʊ/ as well as /ɛ, œ, ɔ/ are variably tensed to [i, y, u, e, ø, o].
    • The centering diphthongs that begin with tense vowels tend to be laxed, which means that there's no actual difference between [iːɐ̯, yːɐ̯, uːɐ̯, eːɐ̯, øːɐ̯, oːɐ̯] on one hand and [ɪɐ̯, ʏɐ̯, ʊɐ̯, ɛɐ̯, œɐ̯, ɔɐ̯] on the other, with both sets being realized as [ɪɐ̯, ʏɐ̯, ʊɐ̯, ɛɐ̯, œɐ̯, ɔɐ̯]. However, this merger doesn't seem to be categorical.
      • ASG is identical to colloquial NSG in this regard - see our article on the phonology of SG.
    • These diphthongs also appear before intervocalic /r/, as in Lehrer [ˈlɛɐ̯ʁɐ] (NSG [ˈleːʁɐ]).
    • /r/ is completely absorbed by the preceding /ɑ, ɑː/, so that /ɑr, ɑːr/ and /ɑː/ all merge to [ɑː]
      • Nothing special, it also happens in practically all areas with /r/-vocalization.
    • Some speakers (a minority) make no strong distinction between the lax /ɛ/ and the tense /eː/, and the former can be realized as [e] and the latter as [ɛː]
      • This would probably be viewed as rather strongly non-standard by speakers from other regions (especially most of Germany).
    • The JIPA article transcribes our /aɪ, aʊ, ɔʏ/ with aɛ, ɑɔ, ɔɛ. This is probably uncalled for and aɪ, ɑʊ, ɔɪ would be a better set of symbols. Also, there are far more possible realizations of these than just [aɛ, ɑɔ, ɔɛ].
      • We could change the NSG transcription of these diphthongs to aɛ, aɔ, ɔœ (per Krech et al. (2009)), but probably no other source transcribes them as such. It's far better to keep transcribing the first two sounds with aɪ, in NSG and to simplify the transcription of /ɔʏ/ to ɔɪ per the Handbook of the IPA and the latest edition of Duden's Aussprachewörterbuch. These symbols are more familiar to native speakers of English that can read English IPA, and tell them that these sounds aren't that different from their English counterparts (in fact, some native speakers of English pronounce these diphthongs pretty much exactly like the Germans do, it's just that the first element is too long).
    • We don't know how the nasal vowels, the shortened tense vowels and the non-native [ɛɪ, ɔʊ, œːɐ̯] are pronounced in ASG.

I might've missed something, so you need to read the paper yourself. The link is here. Generally, a lot of these features remind me of regional SG.

What if these inconsistencies and variations also apply to the actual spoken Swiss Standard German? Do we have sources to confirm that that's not the case? If not, that's another reason to have only one (Northern) variant here. Also, SSG transcriptions aren't as local as the Alemannic ones, and we already have Help:IPA/Alemannic German. There's no problem with including both Alemannic and and quasi-pandialectal Standard German (using established symbols for NSG) transcriptions. There's also no problem with creating Help:IPA/Bavarian and linking truly local Austrian/Bavarian pronunciations there. Also, even if we choose to change our representation of ASG in this guide, are there sources that actually transcribe the Standard Austrian accent as described in the JIPA article? What about authentic Swiss Standard German? Because there are multiple (well, at least 2) pronunciation dictionaries which transcribe German words (and loanwords) into NSG - another reason to prefer this variant on WP.

Maybe I got something wrong in the first part of my post. If so, feel free to correct it. Also, my apologies if my message looks too chaotic. I had to cover multiple topics.

Also, do we have a method of checking how many Austrian and Swiss pronunciations there are on Wikipedia? If not, removing these varieties will be problematic, unless there's someone crazy enough (:P) who would go through all of our German transcriptions.

I think being of help to our readers and having uniform German transcriptions across WP is far more important than not being prescriptive (and editors who say they have a problem with prescriptivism seem to miss the fact that there are levels of narrowness to IPA transcriptions and that they can be both allophonic and diaphonemic, at least to some degree).

I suggest that we remove ASG and SSG from the guide. They might be complicating things too much and SSG isn't a suitable model for most learners of German - and we should also think about them. Kbb2 (ex. Mr KEBAB) (talk) 17:02, 24 July 2018 (UTC)[reply]

I do not have any issues with removing the pseudo-three-column layout. However, I strongly object to banning [r] or prescribing [ʔ]. The inescapable outcome would be people imposing a totally inappropriate northern German pronunciation onto Swiss or Austrian names and places. If we want to simplify, I would actually prefer the good old dictionary tradition of using only [r]. It is simpler to foreign speakers, and the Northern pronunciation with [ɐ] is easily recoverable from it, and we do the same thing in the English transcriptions already. --mach 🙈🙉🙊 21:04, 24 July 2018 (UTC)[reply]
@J. 'mach' wust: r is a very broad transcription, and we have multiple sources that use ʁ instead (Kohler (1999) and Krech et al. (2009) just to name two). ʁ is an objectively superior symbol because the vast majority of speakers of SG realize /r/ as a uvular continuant, and remember that we should prefer NSG here (which is the most prestigious variety and even more superior of a choice for this guide because NSG is what English-speaking learners of German learn all around the world). As I said, there are multiple reasons not to cover SSG and probably also ASG in this guide (and therefore in all of the transcriptions that link to this guide).
Transcribing [r] where a (non-syllabic) low central vowel mandatorily appears in NSG would be doing a massive disservice to learners and would be a bad practice for multiple reasons:
- That's not how pronunciation dictionaries treat /r/, and those should be our primary sources (not in the Wikipedia sense, you know what I mean)
- It'd be sacrificing showing an authentic NSG pronunciation just for the simplicity of transcription and to accomodate speakers of SSG, a variety which we probably shouldn't even cover here. It's just a regional pronunciation standard, it doesn't have the same prestige as NSG.
- [ɐ] and [ɐ̯] do just as good of a job of being diaphonemic symbols that correspond to consonantal realizations of /r/ in Switzerland.
You seem to be confusing broad and narrow phonetic transcriptions. Just because there'd be a glottal stop in the transcription doesn't automatically mean that it needs to be pronounced. If, for whatever reason, someone chooses to sound more Swiss or Austrian, they can ignore the distinction between [z, dʒ, ʒ] on one hand and [s, tʃ, ʃ] on the other and merge the two sets into [s, tʃ, ʃ]. If they can do that, they can do the same with the glottal stop. I object to not transcribing it because it doesn't have a completely predictable distribution. It must be transcribed in order for readers who aim at NSG as a model to truly sound authentic when they speak German. Others can ignore it. We can drop phrase-initial glottal stops (as pronunciation dictionaries do) but not others.
I'd also like to point out that a POV of a native speaker of Swiss SG is just as much of a POV as a prescriptivist one (not that I really identify as such, those are your words, maybe not from this discussion in particular but from the other ones). Kbb2 (ex. Mr KEBAB) (talk) 22:36, 24 July 2018 (UTC)[reply]
We gain nothing by forcing Swiss or Austrian names or places to be transcribed in a Northern German way. Uniformity is no benefit. The vast majority of readers will never find out that this is not the standard pronunciation of such names or places. By providing the actual standard pronunciation, and not the one that might be used in Northern Germany, our readers get valuable information. --mach 🙈🙉🙊 06:04, 25 July 2018 (UTC)[reply]
@J. 'mach' wust: Pardon me for insisting, but I still haven't seen many of my concerns addressed. Perhaps that's my fault for writing such a long message. So, let me rewrite them here:
- Is what I wrote about Swiss SG correct? Is it perceived as noticeably less prestigious than Austrian and especially Northern SG? If so, why would we want to tell our readers that it's just as fine to use SSG when you speak German? It seems to me that you just don't want us to not cover SSG in the guide.
- What about the JIPA article on Austrian German? It makes transcribing it problematic and our representation of it (as it is now) is just plain wrong.
- Where can we find dictionaries, especially pronunciation dictionaries that transcribe Austrian SG as described by the JIPA article?
- What about Swiss SG? Are there descriptions of the actual spoken standard of Switzerland? Is transcribing it as complicated as transcribing ASG? Are there dictionaries that transcribe SSG words as they would actually be pronounced by educated Swiss people? Because the last thing we want is to allow WP:OR here, which we've already done in allowing this guide to describe an artificial standard of AG that Austrians don't use.
- What about the fact that NSG is actually an ideal that speakers from Germany (at least some of them), Belgium, Luxembourg, Namibia and some from Austria approximate their speech to? It just so happens that Germans from Northern Germany are usually the most successful in that, but that's only because their local accents tend to be closer to NSG than any other.
- Also, isn't it the case that SSG and ASG actually aren't truly local? I thought most southerners only treated Alemannic and Bavarian dialects as their mother tongue and SSG/ASG as something closer to a foreign language that has an overtly formal feel to it. People from Northern Germany often speak NSG (or something close to it) as their only native language. I know that doesn't change much, but it might influence the fact that NSG is perceived as the most prestigious variety.
- You're still ignoring the fact that IPA transcriptions can be allophonic and diaphonemic at the same time and that they don't have to be read completely literally. This makes many transcriptions that look like NSG not wrong if you know how to read them to get the ASG/SSG version (and it's dubious whether we should cover these here, so please address my original message).
Most of our transcriptions of SG on Wikipedia don't have regional labels and that should change if we want to retain SSG and ASG in the guide. Transcriptions such as [ˈbɛsər] must be labelled as Swiss. When you read German IPA as an Englishman, you almost always expect it to represent a NSG pronunciation, not just a "Standard German" one (again, it is dubious whether the three standards are equally prestigious and therefore useful to cover here).
It's also just wrong to say that NSG pronunciations are not standard in Switzerland and Austria, especially if the standards are truly equal (and I have doubts about that, especially when it comes to SSG). NSG is an international standard of pronunciation, also in Austria. Why would (some) Austrian journalists aspire to NSG as an ideal model of speech if that wasn't the case? Kbb2 (ex. Mr KEBAB) (talk) 12:07, 25 July 2018 (UTC)[reply]
As the current trend in German linguistics and in studies of Deutsch als Fremdsprache in particular is going more towards recognizing the variation within German as a pluricentric language (and also of introducing learners to the variation between Northern- , Southern- , Swiss-, and Austrian Standard German), I am a bit baffled by the suggestion here to go in the opposite directions by prescribing a northern German pronunciation as the sole standard. And that is about all I have to add to this discussion. --Terfili (talk) 08:08, 25 July 2018 (UTC)[reply]
@Terfili: But what (or rather who) we should be concerned about is laymen who generally expect German IPA to always represent NSG. It's not our place to teach various Standard German pronunciations (Wiktionary would be a fine place for that), especially if they can't be backed up by sources and when they can be recovered from the NSG transcriptions. Kbb2 (ex. Mr KEBAB) (talk) 12:07, 25 July 2018 (UTC)[reply]
I do not buy your presumption about what laypeople expect. I strongly doubt that they would have the necessary knowledge about the pluricentricity of standard German in order to expect northern Standard German. Instead, I believe laypeople simply expect the transcriptions to reflect the normal pronunciation. And when it comes to Swiss places or names, the normal or most prestigious Standard German pronunciation is certainly not the one from Northern Germany, but the one from Switzerland. Naturally, it is the Swiss Standard German pronunciation that is used in the commons:Category:German pronunciation of toponyms of Switzerland.
I think your arguments based on diaphonemicity are void. Diaphonemicity only works if we can reasonably assume that potential readers are native or near-native speakers. This being the English Wikipedia, we cannot. (And of course, nobody uses diaphonemicity outside of Wikipedia.)
And now your individual points:
  • Is what I wrote about Swiss SG correct? Is it perceived as noticeably less prestigious than Austrian and especially Northern SG? – This is certainly not correct when it comes to Swiss places or names.
  • What about Swiss SG? Are there descriptions of the actual spoken standard of Switzerland? – There are Ortsnamenbücher that indicate the pronunciation of place names. They do not use obscure overly detailed phonetic symbols no laypeople will understand like this guide does.
  • What about the fact that NSG is actually an ideal that speakers from Germany (at least some of them), Belgium, Luxembourg, Namibia and some from Austria approximate their speech to? – That’s a myth.
  • I thought most southerners only treated Alemannic and Bavarian dialects as their mother tongue and SSG/ASG as something closer to a foreign language that has an overtly formal feel to it. – Nevertheless, Swiss people expect Swiss Standard German when it really concerns them, e.g. in T.V. or radio news shows. If a news anchor uses a Standard German pronunciation from Northern Germany, people will reject it.
  • You're still ignoring the fact that IPA transcriptions can be allophonic and diaphonemic at the same time and that they don't have to be read completely literally. This makes many transcriptions that look like NSG not wrong if you know how to read them to get the ASG/SSG version. – And you are still ignoring the fact that needless allophonic detail makes a transcription hard to read. Everybody knows what an [r] is, but only very few specialists know what [ʁ ɐ ɐ̯] are. Also, by entering in such allophonic depths, you will create unsurmountable ambiguity: There is no telling whether /ˈbʊrɡ/ should be transcribed as [ˈbʊɐ̯ɡ] or [ˈbʊʁɡ], or whether /ˈfyːrər/ should be transcribed as [ˈfyːʁɐ] or [ˈfyɐ̯ʁɐ] – the choice is entirely arbitrary.
--mach 🙈🙉🙊 18:24, 25 July 2018 (UTC)[reply]
@J. 'mach' wust: English Wikipedia isn't written for Swiss people, it's written for native and non-native speakers of English. Swiss people make up a tiny, tiny minority of that.
Diaphonemicity would be important only for native speakers of German that speak something different than NSG and for a tiny minority of learners of German that consciously choose to speak with an accent that's different from SG (which would be going against Assimil recordings, pronunciation dictionaries, etc.). Most of these aren't laymen who are native speakers of English.
Swiss SG recordings might be better suited for Wiktionary.
Learning to pronounce a foreign language is a huge effort. If Swiss SG in popular perception is truly inferior to other varieties of SG (and that does seem to be the case, based on your reaction and my limited knowledge of the subject) then we have no reason to transcribe it. Do you want laymen to make an effort just to be ridiculed for it afterwards? Maybe the reaction won't be so extreme (in fact it almost certainly won't be), but why would we want to deliberately teach Swiss SG to people who aren't aware of what they're learning? Prescribing Swiss SG pronunciation of Swiss places is WP:POVPUSHING because of Swiss SG not being neutral enough, and we don't need to have more than one German IPA in WP articles because WP is WP:NOTADICTIONARY, let alone a pronunciation dictionary of German. The simplest option is to use NSG for all of our transcriptions and to use one set of symbols for SG across all of the Wikipedia.
We can pick one source that transcribes NSG and stick to that. It can be the latest edition of Das Aussprachewörterbuch, I don't see a problem with that. It uses r, which is your preference. Personally I don't like this symbol (it's overused in an inappropriate manner, there's no need to simplify transcriptions like this in 2018), but that's just my personal POV. Also, alveolar trills are way harder to produce for native speakers of English than uvular fricatives, which are by far the most common and therefore the de-facto most neutral realization of German /r/, both in NSG and Standard German as a whole. [r] is regionally marked and it's associated with Southern German and the speech of the elderly.
Also, who decides which allophonic detail is needless and why does it have to be the allophones of /r/? Learning to produce the low schwa and to reliably distinguish it from the high schwa is absolutely essential to anyone who wants to speak NSG (which is the majority of learners of German). Sorry, but your Swiss bias is showing here (and above). It's time for other people to join the discussion (if they so choose) so that we can read more perspectives on the issue. Kbb2 (ex. Mr KEBAB) (talk) 21:50, 25 July 2018 (UTC)[reply]
  • If Swiss SG in popular perception is truly inferior to other varieties of SG ...
So how do you know what "popular perception" is? I mean, this is an encyclopedia, not some populist mobocrat's press conference.
Incidentally, my perception is that Swiss SG is on a par with other varieties of SG. Love —LiliCharlie (talk) 00:19, 26 July 2018 (UTC)[reply]
@LiliCharlie: I don't know for sure, I asked a question. What I am aware of is that Swiss accents are often mocked by people from the north (and perhaps some Austrians). Unless I'm mistaken, we don't want our readers to make effort to pronounce SSG just to be mocked for it afterwards. We need to use a variety that's acceptable everywhere, and we're not a pronunciation dictionary of German. Kbb2 (ex. Mr KEBAB) (talk) 10:38, 26 July 2018 (UTC)[reply]
I don't believe that anyone with the slightest trace of a foreign accent runs the risk of getting mocked for an apical trill or more or less voiced syllable-final obstruents. People who speak like that are by no means rare in any part of the German sprachraum. Who are those mockers you have in mind? Love —LiliCharlie (talk) 11:59, 26 July 2018 (UTC)[reply]

This thread is giving me a bit of a tl;dr glaze, but I do want to interject to say that it's not a good idea to treat non-English transcriptions the same way we do English as far as phonetic imprecision. NSG uses uvular rhotics and it would be misleading to write it with ⟨r⟩ just because some lay readers might not be familiar with it. The point of this IPA guide is to familiarize such readers with IPA symbols they may not understand. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 04:11, 26 July 2018 (UTC)[reply]

@Aeusoes1: Mach's point is that ʁ is too exclusionary a symbol, but that's only because he chooses to read it too literally. Besides, it's not like any other realization is as common as [ʁ]. Kbb2 (ex. Mr KEBAB) (talk) 10:38, 26 July 2018 (UTC)[reply]
I am not advocating that we drop [ʁ ɐ ɐ̯], even though I we would be better off without them (for the reasons I have: complicated, not well-known, not normally used in dictionaries, ambiguities). All I am saying is we should not prescribe them (or [ʔ]) exclusively.
Do you want laymen to make an effort just to be ridiculed for it afterwards? – No, and that is precisely why I think it is very important we keep the freedom of choice between [r] or [ʁ ɐ ɐ̯]. Imagine we exclusively prescribe [ʁ ɐ ɐ̯], and an English native speaker layperson tries to impress someone from Bern by making an effort to correctly pronounce the city’s name. Having exclusively prescribed [ʁ ɐ ɐ̯], our IPA guide happens to say [bɛɐ̯n]. That layperson’s best effort – approximately /ˈbɛɑːn/ – will be ridiculed for being a totally inappropriate pronunciation from Northern Germany. – Or worse, if our IPA guide said [bɛʁn] (you can never know which one will be chosen because of the inherent ambiguity), the approximation /bɛxn/ would be ridiculed even more for sounding like eastern Swiss German, which has lower prestige still. --mach 🙈🙉🙊 17:56, 26 July 2018 (UTC)[reply]
@J. 'mach' wust: If you think that the SSG pronunciation of "Bern" is the only correct one then I really don't know how to respond to that. It's a completely unreasonable statement that has nothing to do with the reality. Kbb2 (ex. Mr KEBAB) (talk) 02:18, 7 June 2019 (UTC)[reply]
I have never said such a thing. It is certainly not the only pronunciation, but it is the most appropriate one. --mach 🙈🙉🙊 04:25, 7 June 2019 (UTC)[reply]

@J. 'mach' wust: I'm not gonna address your latest message yet (though I will do that eventually), but you should know that edit warring without addressing my edit summaries is dishonest and probably violates WP:GASLIGHTING as well. Kbb2 (ex. Mr KEBAB) (talk) 10:42, 3 November 2018 (UTC)[reply]

This guide is supposed to cover standard German in all its varieties. Your removal one of the major national varieties is a major disruption and cannot go without discussion. --mach 🙈🙉🙊 10:48, 3 November 2018 (UTC)[reply]
@J. 'mach' wust: I removed it because the way we transcribe it is very different to what the JIPA article says. To call it a disruption is, again, dishonest. We've already had a discussion and you haven't presented a good way of representing ASG in this guide that'd be based on reputable sources (including pronunciation dictionaries). Therefore, it needs to be removed because it's an OR transcription that doesn't match the source. All of my arguments are above.
We can use the prescriptive set of symbols for Northern Standard German for ASG. It's good enough. Kbb2 (ex. Mr KEBAB) (talk) 10:51, 3 November 2018 (UTC)[reply]
The basis for your argument is that the JIPA is the ultimate source for Austrian German. That is of course untenable.
My point, which you have not yet addressed, is that only showing the flags of Germany and Switzerland is de facto excluding Austria. I agree that you could largely use the same symbols for Germany German and for Austrian German, and I agree that the multi-column layout was a bad idea to start with – as I have said, I would much prefer just mentioning the variants and then explaining in the footnotes what they are about. However, making it appear as if this guide were only for Germany and Switzerland is wrong on a fundamental level.
And please try to abstain from name-calling. It is not helpful in any way. --mach 🙈🙉🙊 10:59, 3 November 2018 (UTC)[reply]
@J. 'mach' wust: I apologize for getting worked up.
Maybe I did sound like that, but at the same time we can't ignore it. JIPA is an important phonetic journal and that article is very recent.
I'll check Deutsches Aussprachewörterbuch and will get back to you. Kbb2 (ex. Mr KEBAB) (talk) 18:12, 3 November 2018 (UTC)[reply]
@J. 'mach' wust: SSG and ASG were added in this diff, AFAICS without any prior discussion. I think that I've already proven that the ASG column is wrong on multiple levels. "We can't just include German SG and Swiss SG and leave out the Austrian standard" is, I'm afraid, a non-argument. We have multiple sources that contradict each other on what ASG is specifically. The situation of including GSG and SSG and leaving out ASG is easily fixed by removing SSG as well. WP isn't a pronunciation dictionary of German and native speakers of German anywhere will be happy to hear their language spoken by a non-native speaker, no matter the accent they use (the NSG one is the safest bet in each case).
It takes deliberate misunderstanding of the IPA to assume that you can't convert those transcriptions to regional German, e.g. when you need to ignore the glottal stops, read [n] as [ŋ] etc.
Deutsches Aussprachewörterbuch backs up some of the statements made in the JIPA article. It's a pain to read it because my German isn't very good. But I think that we have enough reasons not to transcribe ASG on WP and just use the NSG transcription. It's correct everywhere in Germany and Austria and speakers from Switzerland will recognize it as being native. Kbb2 (ex. Mr KEBAB) (talk) 02:18, 7 June 2019 (UTC)[reply]
I have never liked the multi-column layout. But you should not use the layout switch to sneak in your prescriptivist point of view. I have therefore really restored what was there before the layout change: both [ʁ] and [r]. --mach 🙈🙉🙊 04:25, 7 June 2019 (UTC)[reply]
@J. 'mach' wust: Then let's just follow Duden and use r which our readers will be free to interpret however they wish. Plus, the alveolar trill (which, as you know, is the canonical value of the IPA symbol r) is certainly not wrong nor unused in Germany, it's perfectly standard (though not always local) in all regions of Germany and Austria, it's also common in Switzerland.
I can partially get behind your proposal of writing [ɪr, ʏr, ʊr, ɛr, œr, ɔr, ar] because /r/ in those sequences is either a uvular approximant (rather than a fricative) or forms a centering diphthong [Vɐ̯] with the preceding vowel, depending on the region and speaker (other realizations are also found, including the trilled/tapped ones). Both pronunciations are considered correct - Duden writes these [ɪr, ʏr, ʊr, ɛr, œr, ɔr, ar], whereas Deutsches Aussprachewörterbuch prefers [ɪʶ, ʏʶ, ʊʶ, ɛʶ, œʶ, ɔʶ, aʶ], which is a non-IPA way of writing [Vʁ̞] (a vowel followed by a uvular approximant).
[iːɐ̯, yːɐ̯, uːɐ̯, eːɐ̯, øːɐ̯, oːɐ̯, ɛːɐ̯] and [ɐ] should be transcribed just like they are right now, per Duden and DA. Their transcriptions cover much more than the prescriptive German German accent, you just have to read them in a non-literal way. /aːr/ can be safely written [aːr] because the diphthong [aːɐ̯] simply doesn't exist for many native speakers, and neither does [aɐ̯]. Rather, both /aːr/ and /ar/ can merge with /aː/, which is normally the case in Austria and probably also for many speakers of NSG. Writing [ar] and [aːr] allows the reader to correctly identify the underlying phonemes and then read the transcription however he wishes to (either [aʁ̞, aːʁ̞], with a uvular approximant or [aː] for both).
I think that we've agreed on the fact that ASG should be transcribed the same as NSG. Sources are too contradictory as to what exactly defines ASG and many of the features described in the JIPA paper apply to colloquial German German as well. These can be dealt with in notes, and I think that we can do the same with SSG. Roger Federer, Schweizerpsalm and Freiburg im Üechtland can be safely transcribed [ˈrɔdʒɐ ˈfeːdərɐ], [ʃvaɪ̯tsɐˈpsalm] and [ˌfraɪ̯bʊrk ʔɪm ˈʔyːɛçtlant]. [ɐ] can be readily mapped to [ər] by speakers of SSG, they can ignore the glottal stops (if "Freiburg im Üechtland" were transcribed in either of the main pronunciation dictionaries of German, the glottal stops would be there) and lack of final fortition is a feature that can be found in many if not all Southern German varieties, not just SSG (in ASG it's variable). I'm not sure why we'd need to transcribe the lack of final fortition and ignore e.g. two varieties of the low central vowels found in Bavaria or two varieties of /aɪ̯/ and /aʊ̯/ in Swabia, both of which are features of the corresponding regional standards of pronunciation (which are as regional as SSG, it's just that neither of those is a national standard - then again, we don't deal with Luxembourgish SG either, which is probably similar to the regional standard of Western Germany).
Again, much can be done by just expanding the notes. I don't know why the lack of distinction between [ɛː] and [eː] (observed in millions of speakers in Germany and Austria) should be treated as any different than the lack of distinction between [ɐ] and [ər] or between [ʔ] and a phonetic zero. Kbb2 (ex. Mr KEBAB) (talk) 15:05, 7 June 2019 (UTC)[reply]
The unspoken assumption behind all your proposals is that there should be a single standard transcription for German pronunciatin on the English Wikipedia. As I have said over and over again, I do not agree with that point of view. --mach 🙈🙉🙊 08:51, 8 June 2019 (UTC)[reply]
@J. 'mach' wust: There's a difference between enforcing a single standard transcription (or transcriptions, if we transcribed multiple varieties of Standard German - see Help:IPA/Portuguese and similar guides for comparison) for German pronunciation (per MOS:PRON, this isn't my POV) and transcribing multiple standards of German pronunciation, especially if you don't explicitly label the latter (at least when you're not transcribing NSG).
Das Aussprachewörterbuch, Deutsches Aussprachewörterbuch, German Wiktionary and other works on German phonetics and phonology (Wiese's Phonology of German is another example) all use one set of symbols, whatever it consists of. Whether we choose to religiously follow any of those sources or to mix those transcriptions up is irrelevant as long as we're being consistent and the symbols we use are found in contemporary sources. What would be the benefit of allowing our editors to transcribe German however they wish (MOS:PRON aside?)
Your problem seems to be just that: a selectively literal reading of our transcriptions. Whether we choose to write the consonantal variety of /r/ with r or ʁ does not mean that that symbol can only symbolize an alveolar trill or a voiced uvular fricative to the exclusion of everything else. That literal reading of those symbols goes against both the principles of the IPA and the pronunciation dictionaries themselves. In Das Aussprachewörterbuch the authors state clearly that when they write [ɡɛrn] the transcription stands not only for [ɡɛrn] (with an alveolar trill) but also [ɡɛɐ̯n] (the most common realization in Germany and Austria), [ɡɛʁn], [ɡɛʀn], etc. The same applies to [ɡɛʶn] - the corresponding transcription from Deutsches Aussprachewörterbuch.
I don't know what to do with SSG, but given the contradictory descriptions of ASG and the fact that many of its features are present in colloquial NSG as well (and therefore are already covered by our transcription) there seems to be little to no need to differentiate between ASG and NSG on WP. In SSG, we have a noticeable lack of glottal stops, r-vocalization and final fortition. We also have different vowel qualities, perhaps a lack of distinction between the short lax [ɪ] and the short tense [i] and other similar pairs, l-velarization (perhaps variable from region to region). [ɛː] and [eː] are separate phonemes, as in NSG but not necessarily ASG. The lack of glottal stops and r-vocalization is easily dealt with by not reading the NSG transcriptions in a literal manner. The only problematic feature is the lack of syllable-final fortition.
Transcriptions containing e.g. the Swiss diphthong [yə̯] can be simply enclosed within the IPA template, again per MOS:PRON. I think those are rare enough that they won't be a major issue for us. Kbb2 (ex. Mr KEBAB) (talk) 12:46, 9 June 2019 (UTC)[reply]
You are claiming that your point of view is backed by MOS:PRON. However, what MOS:PRON really (and in my opinion rightfully) says is that “[f]or foreign-language pronunciations, a phonetic transcription is normally used”. German is a pluricentric language with different standard pronunciations. Obviously, for a phonetic transcription there is no other way but to phonetically represent these differences. That is just how phonetic transcriptions work (and I am sure you know the difference between phonetic and phonemic transcriptions).
What would be the benefit of allowing our editors to transcribe German however they wish (MOS:PRON aside?) I have never said that our editors should transcribe the different varieties of standard German “however they wish”. Instead, I think we should allow for a reasonable amount of variation within a set of traditionally used symbols. What I have in mind are the following:
  • Different ways of transcribing /r/: [r ʁ ɐ]
  • Transcribing [ʔ] or not
  • Differentiating between /eː/ and /ɛː/ or not
  • Transcribing final fortition or not
There are at least three main benefits in allowing the different phonetic transcriptions for the different standard varieties:
  1. For place names (and other things that are tied to a locality) using anything other than the local standard German pronunciation will mislead non-locals.
  2. For place names (and other things that are tied to a locality) using anything other than the local standard German pronunciation will offend the locals.
  3. Allowing different transcriptions accurately reflects the current academic consensus: that standard German is a pluricentric language with differences, among other things, in pronunciation.
Whether we choose to write the consonantal variety of /r/ with ⟨r⟩ or ⟨ʁ⟩ does not mean that that symbol can only symbolize an alveolar trill or a voiced uvular fricative to the exclusion of everything else. Please have a look at Template:IPA-de and the pages that use it. You will notice that it uses square brackets to enclose the pronunciation. This signals that we are using phonetic transcriptions, which means that the signs are supposed to represent their IPA values. Now if we provide some hidden-away pronunciation guide with some hidden-away footnote saying that the signs really represent something else, then I think we are not being helpful to our readers at all. To the contrary, I think that we would be very misleading. Furthermore, it would be totally unnecessary, given that we can easily allow for different transcriptions to be used (as we are doing right now).
That literal reading of those symbols goes against both the principles of the IPA and the pronunciation dictionaries themselves. Regarding the IPA, I have no clue what principles you might possibly refer to. The IPA is, by its very nature, a phonetic alphabet, not a random collection of meaningless arbitrary signs. Regarding the dictionaries, there are dictionaries, unfortunately, that exclusively choose a single variety of standard German for all their transcriptions, or other dictionaries that really use phonemic transcriptions even though they employ square brackets. No matter. We do not need to repeat the shortcomings of these dictionaries.
Allow me to ask you back for once: What possible harm is there in continuing to allow a transcription of the different pronunciations of standard German? I fail to see any harm whatsoever – all I see are the various benefits I have mentioned. Why are you so resolutely advocating for uniformity? --mach 🙈🙉🙊 14:21, 9 June 2019 (UTC)[reply]
@J. 'mach' wust: Please read User:Nardog and tell me what you think about it. You're basing your messages on a faulty understanding of how the IPA works. Kbb2 (ex. Mr KEBAB) (talk) 15:05, 9 June 2019 (UTC)[reply]
I understand very well what the IPA is. Please answer my question: What possible harm is there in continuing to allow a transcription of the different pronunciations of standard German? --mach 🙈🙉🙊 15:12, 9 June 2019 (UTC)[reply]
@J. 'mach' wust: I'll answer that when you read his user page. Kbb2 (ex. Mr KEBAB) (talk) 15:19, 9 June 2019 (UTC)[reply]
Of course I have read it. It is not relevant to this discussion. Please do not embarass yourself by forcing me to demonstrate you point by point how it is not relevant. Instead, stop evading and answer my question. --mach 🙈🙉🙊 15:23, 9 June 2019 (UTC)[reply]
@J. 'mach' wust: If you can't see how it's relevant to this discussion then we have nothing to discuss. I don't want to go around in circles. Your messages seem to be deliberately misleading. Kbb2 (ex. Mr KEBAB) (talk) 15:54, 9 June 2019 (UTC)[reply]

If my skimming of the thread is correct, it sounds like Mach would like us to transcribe German in a more permissible way with regional variation in e.g. the rhotics while Kb would prefer having one pronunciation to either reflect this variation in pronunciations in a more abstract way or use one SG variant over others because of its prestige. There is merit to both of these approaches, but if we are going to go with Mach's approach, we want the guide here to be clear about when to use variant over another and how to transcribe each variant. Despite what (it seems like) Mach is saying, the IPA and our in-house IPA policy do permit the use of characters in a diaphonemic way, even for other languages, but readers familiar with the IPA are primed to interpreting brackets as phonetic transcription and so this nuance might get lost on them. I lean towards one pronunciation for the sake of simplicity and verifiability, so the question I guess is what benefit towards incorporating this variation would outweigh the burden that comes from this added complexity? — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 17:19, 9 June 2019 (UTC)[reply]

@Kbb2: While it is clearly the standard of some German transcription systems to use /r/ in slashes to represent [r ʁ ʀ ɐ], do any use [r] in square brackets with the same meaning? Some of these sounds diverge very widely from the narrowest interpretation of the symbol ⟨r⟩. Or are there other phonetic transcription systems that use symbols to represent sounds that are as different from the narrowest interpretation of the symbols as [r] is from ʀ ɐ]? Is the level of imprecision that is permitted in a phonemic transcription also permitted in a phonetic transcription?

The quote on User:Nardog's user page that you may have seen as bearing out your point gives the example of two phonetic transcriptions, [tʃɛkðəlɛnzwɛɫ] and [tʃe̞ʔ͡kð̞əlɛ̃nzwæ̠ɫ], of which the latter is very narrow. The differences between the two transcriptions involve whether the indication of precise details of tongue position, secondary articulation, glottal closure, nasalization. They don't involve something like the difference between a coronal or an uvular place of articulation, or a trill consonant and a vowel. So to justify the assertion that the level of imprecision you are arguing for is regular IPA practice, other examples would be helpful. — Eru·tuon 18:58, 9 June 2019 (UTC)[reply]

@Erutuon: It was just an example (read the discussion above for more examples of Mach's literal understanding of the IPA), but Das Aussprachewörterbuch has always used r that way. Really, what I meant was that whatever symbol you choose for the consonantal variety of /r/ (e.g. the variety of /r/ that is a consonant regardless of the variety of SG), the reader is never required to interpret it according to the canonical IPA value of that symbol.
With that being said, ʁ is a superior choice to me. The consonantal /r/ is uvular for the vast majority of native speakers of German, and it's still spreading (Munich is becoming a sort of a uvular enclave in the region due to the amount of immigrants from other regions, though even that description may be already somewhat outdated). The majority of speakers of NSG and ASG use the uvular fricative and it's a possible realization in Switzerland. The difference between uvular and alveolar rhotic is not and has not ever been phonemic in Modern German, and so I see little reason to use more than one symbol for the consonantal variety of /r/. I said that we could perhaps use r just to end one of the arguments with Mach, but now I prefer ʁ for the sake of clarity.
Now, the issue with /r/-vocalization is a bit complicated. The long vowels + /r/ are all vocalized, so they should be written [iːɐ̯], etc. But /aːr/ isn't [aːɐ̯], but either [aːʁ] (with an approximant rather than a fricative) or a bare [aː], in which case it falls together with /aː/. If /ar/ is vocalized, then it falls together with those two as well, at least in NSG and ASG. The supposed diphthongs [aːɐ̯] and [aɐ̯] are probably alien to many speakers of German, much more alien than a phonemic /ɛː/. If a speaker with a uvular /r/ really differentiates between /aː/ on one hand and /aːr/ and /ar/ on the other, the difference lies in the final consonant, which is a uvular approximant. When it comes to transcription, Duden writes those as [aː, aːɐ̯, ar], whereas Deutsches Aussprachewörterbuch uses a more reasonable transcription [aː, aːʶ, aʶ] (the superscript voiced uvular fricative stands for an approximant).
The short vowels + /r/ traditionally weren't vocalized in SG, producing [ɪʁ] etc. (again, with an approximant rather than a fricative). In contemporary SG, they're very often vocalized to [Vɐ̯] - except /ar/, which merges with /aː/ and /aːr/ to [aː]. The diphthong [aɐ̯] is rare, and my theory is that sources write [aɐ̯] and [aːɐ̯] instead of [aː, aː] in order to facilitate identification of the underlying phonemes. Both Duden and DeA write those with the consonantal symbols r and ʶ. I've seen many times e.g. [ʊʁ] replaced with ʊɐ̯ on the grounds that "it's vocalized" or "it's not pronounced with a consonant". I wonder whether those editors can tell a difference between a uvular approximant and a non-syllabic [ɐ̯] (which really is [ɐ̯] or a vowel very close to it, *not* [ə̯], [ɨ̯] or [ɤ̯]). Just because there's no friction typical for syllable-initial varieties of /r/ it doesn't mean that it's a vowel.
I propose that we write the consonantal /r/ with ʁ, the variably vocalized one (after all short vowels including /a/ and the long /aː/) with ʁ and the mandatorily vocalized one with ɐ̯. The /ər/ sequence should be written with ɐ. Speakers of regional varieties of German and the tiny minority (if not less than that) of learners of German that want to speak a specific regional accent can interpret those symbols however they wish. Kbb2 (ex. Mr KEBAB) (talk) 20:27, 9 June 2019 (UTC)[reply]
Okay, so Das Aussprachewörterbuch uses ⟨r⟩, but I was asking about how broad IPA phonetic transcriptions should be, and it's not clear to me whether their transcriptions are phonetic or phonemic or some weird hybrid because they don't seem to use any bracketing. — Eru·tuon 21:33, 9 June 2019 (UTC)[reply]
@Erutuon: They are, of course, phonetic. Any transcription of German that uses ɐ̯ (with the non-syllabic diacritic) or marks syllabic consonants is phonetic by definition. Hall (2003) also uses r like Duden. Kbb2 (ex. Mr KEBAB) (talk) 15:59, 11 June 2019 (UTC)[reply]
@Kbb2: I am sorry I got heated up. I do not like being accused out of the blue of ignoring a text I have read, especially when my own repeated questions keep being ignored.
Here is why I think Nardog’s quotes are not relevant:
  • 1.1: This quote is an interesting illustration of how a consensus about the adequate level of detailedness can change over time. Back then, it was apparently common both French _fil_ and English _fill_ be transcribed as [fil]. Nowadays, it is much more common to differntiate them as [fil] vs. [fɪl] – which is a fruit of Gimson’s compromise after decades of debates over the precise use of signs (which very much matter to the serious linguists). This does not tell us anything about how we should handle diatopic variation, though.
  • 1.2: There is a range of possible transcriptions between a phonemic transcription and a super-detailed narrow transcription. This does not tell us anything about how we should handle diatopic variation.
  • 1.3, first quote, 1: This is an argument for using a phonemic transcription. It is funny how Kbb2 instead argues for transcribing a surprising amount of allophonic variation that will even blur phonemes. Kbb2 proposes that the German phoneme /r/ should be transcribed in roughly three different ways: by itself, it can be either [ʁ] or [ɐ̯] (it is inherently difficult to draw a clear line), and in the sequence /ər/, it can be [ɐ] as long as it is not [əʁ]. This does not tell us anything about how we should handle diatopic variation.
  • 1.3, first quote, 2: This is an argument for using the same signs for similar sounds in different languages. It is probably not meant to apply to diatopic variation, but it might be read in such a way. By choosing to represent the allophonic variation of one variety, Kbb2 looses the opportunity of having one same sound that could be found in several varieties.
  • 1.3, second quote: This is nice and concise, but it does not tell us anything about how we should handle diatopic variation – unless we understand that “[t]he IPA is designed to be a set of symbols for representing all the possible sounds of the world’s languages” including the different sounds of the different national varieties of German.
  • 1.4: Again arguing for phonemic transcriptions, which is again at odds with Kbb2’s insistence on representing allophonic variation, and again this does not tell us anything about how we should handle diatopic variation.
  • 1.5: There are different ways to syllabify English. This does not tell us anything about how we should handle diatopic variation.
My initial question remains: Why? What can we possibly gain by changing our practice and prescribing a single standard? To reiterate my arguments once again, I see nothing but disadvantages: It will mislead readers, it will upset users, and it goes against linguistic consensus. What are the advantages? What do you hope to gain? I honestly fail to see anything. I also think it is very unbalanced that we should represent allophonic variation but not diatopic variation. --mach 🙈🙉🙊 21:21, 9 June 2019 (UTC)[reply]
@J. 'mach' wust: It's not "me" but "pronunciation dictionaries" and other reputable sources. I'm arguing for following them, in one way or another. Per WP:NPOV it's on you to provide arguments to the contrary.
Transcribing German SG, Austrian SG and Swiss SG in the appropriate (or would-be appropriate) articles isn't our current practice. Our current practice is to rarely and inconsistently change German SG to Swiss SG and that's it (or perhaps it's more often than "rarely" - either way, we're not being consistent and transcriptions that diverge from NSG aren't explicitly labelled). The only articles in which I saw ASG being actually ones are the ones I've already changed to German SG for the reasons I've mentioned above. There were about 20 of them. Many entries (e.g. Salzburg) are transcribed in NSG.
Why are you (again) ignoring the inconsistency with which ASG is described in various sources and the free variation in it (see the JIPA article and DeA)? We don't have a consensus to represent all 3 varieties at all. There are multiple issues with representing ASG.
Wikipedia:Manual of Style/Pronunciation#Other languages says that [I]f the language you're transcribing has such an IPA key, use the conventions of that key. If you wish to change those conventions, bring it up for discussion on the key's talk page. Creating transcriptions unsupported by the key or changing the key so that it no longer conforms to existing transcriptions will confuse readers. This means that we shouldn't represent other varieties than NSG if these aren't explicitly laid out in the key (good luck doing that with ASG) and probably also explicitly labelled in the articles. Linguisticly aware readers really do expect German IPA to represent Northern Standard German, as per most German dictionaries, which in my experience are more uniform than the English ones. I've never encountered any dictionary that would transcribe German in an accent different from NSG. English dictionaries transcribe RP, GA and sometimes also Canadian English and Australian English. To the best of my knowledge, there's no such variation in German.
The standard described in pronunciation dictionaries isn't really Northern Standard German but a non-regional Standard German, spoken or at least approximated by millions of people from Northern Germany, Southern Germany and Austria (probably also Luxembourg and Eastern Belgium) alike, just as it is the case with RP. It's a non-regional accent like RP in England, Parisian French in France, Central Standard Swedish in Sweden and Northern Standard Dutch in the Netherlands, that's why it's used in the news and it's taught to foreigners. This means (or seems to mean) that NSG, unlike other standards isn't regionally marked in any major way. Its features are seen as prestigious and non-regional. Speakers with pronounced southern features are sometimes/often regarded (or they regard themselves) as non-native speakers of SG [their native languages are regional, non-standard dialects of German], even though they can write, read and understand spoken SG like a native. So this may also contribute here: the fact that those who don't speak any dialects generally speak SG with an accent that is a close approximation of the non-regional variety of SG - but it doesn't change the fact that it's still non-regional. You have the exact same situation in Randstad in the Netherlands, where the concentration of speakers of Northern Standard Dutch is the highest, but it's still regarded as a non-regional form of Dutch. It's impossible to tell whether you're from Amsterdam, The Hague, Groningen or Maastricht (all have different accents of SD and, in the case of the last two, regional dialects) or if you're a flawless non-native speaker if you speak NSD. The same applies in the case of NSG: it's impossible to tell whether you're from Hannover, Berlin, Hamburg, Stuttgart, Vienna or Salzburg if you use it.
The JIPA article on Austrian Standard German states clearly that ASG has traditionally been geared towards German Standard German and that goes back to at least the year 1750. Austrian newsreaders are instructed according to the norms of Das Aussprachewörterbuch and Siebs's Deutsche Aussprache. Bühnenaussprache. This means that pronouncing Bern [ˈbɛʁn], [ˈbɛʀn], [ˈbɛrn] or [ˈbɛɐ̯n] says nothing about where you come from as all four variants are acceptable in the non-regional Standard German (which even a Swiss person can speak, if they choose to). There's nothing inappropriate in using this accent (even if it makes you sound formal) in any situation when you're a foreigner. It's misleading to suggest that someone will take offense if you say [ˈbɛʁn] or [ˈbɛɐ̯n]. I'd walk away if they did because it's an utterly unreasonable position to hold, no different than expecting a foreigner to understand or speak one of the Alemannic dialects of Switzerland when they just speak Standard German (this also contradicts the idea that all forms of SG are equal - if they're equal [and they aren't, NSG is more neutral and prestigious than any variety and it's non-regional], there's nothing to be offended by in this context).
Our goal is to represent German in its most neutral form that's acceptable everywhere. It's neither insulting nor inappropriate to speakers of other accents. WP isn't a pronunciation dictionary of German, and they all describe this non-regional accent anyway. I think it's WP:UNDUE to give such importance to SSG if SSG transcriptions can be easily derived from the NSG ones just by reading them in a non-literal way, final fortition aside (of course, first you have to be willing to do that, and to be willing to do that you need to be aware that it's possible. I don't buy that you aren't aware of that, not after all these years you've been dealing with IPA and the amount of explanations I've provided. If you still think that writing [ʔ] is 'prescribing a glottal stop' then I guess that's your problem. It's not true, never was, and never will be. Send an e-mail to the International Phonetic Association if you think otherwise, they should provide you with satisfactory citations [and since my position is common sense and it's shared among linguists, including the authors of modern material on SG, I won't bother discussing that particular aspect of IPA anymore]).
I've asked editors of German Wiktionary to come and join this discussion. We'll see what they have to say. If you ask me, there's no better option than to replace all SSG transcriptions with NSG on WP. Places such Üechtland are best dealt with with the IPA template: [ˈyə̯xtland̥]. Kbb2 (ex. Mr KEBAB) (talk) 20:28, 11 June 2019 (UTC)[reply]
@Kbb2: Let me try again. You wish to find a single standard transcription that works for all varieties of standard German. However, the proposals you have made so far are biased towards specific varieties of standard German because of the phonetic and allophonic features they use:
  • /ər/ is represented with the allophones [əʁ] or [ɐ]
  • /r/ is represented with the allophones [ʁ] or [ɐ̯]
  • [ʔ] is always used even though its status as a phoneme is being debated
One might argue that a transcription the other way round would be equally biased. However, there are two different things going on. When a single sign is supposed to be read in multiple ways, e.g. ⟨r⟩ as [ʁ] or [ɐ] (besides [r]), it means only accepting other readings. When different signs are supposed to be read in the same way, e.g. both ⟨ʁ⟩ and ⟨ɐ̯⟩ as [r] (besides [ʁ] and [ɐ̯]), it also means ignoring an explicit differentiation.
The number of speakers of the different varieties of standard German may differ noticeably. However, there is a clear consensus, at least among linguists, that this does not make any of them better or more standard than the others. There is no super-regional standard opposed to mere “regional varieties of German”. All varieties are regional.
Many people identify strongly with their place of origin. Prescribing that e.g. their hometown be transcribed in a way that uses phonetic and allophonic features of a foreign variety will be perceived as imposing and arrogant. By contrast, people who do not speak German have no indication that these phonetic and allophonic features are not from the standard German variety used in that town, but from a different variety of standard German used in another country. I see no reason why we should change our current practice. I think we should keep transcribing local German names in the respective local variety of standard German. --mach 🙈🙉🙊 20:31, 11 June 2019 (UTC)[reply]
Just so we're clear: this isn't a reply to my post but an addendum to one of Mach's previous posts. I've already addressed some if not most of the points he's raised in this message - see above. Kbb2 (ex. Mr KEBAB) (talk) 20:49, 11 June 2019 (UTC)[reply]
Your contribution does not answer my contribution at all. To the contrary, my contribution is a very valid response to yours, especially to your sad insistence that Northern standard German is superior to other varieties of standard German. --mach 🙈🙉🙊 21:00, 11 June 2019 (UTC)[reply]
@J. 'mach' wust: This reply reassures me that your only agenda here is preservation of SSG transcriptions. You have no interest in an honest discussion, which is too bad. If nobody else replies I'll start replacing SSG with NSG in a couple of weeks. There's no consensus nor any strong reason to differentiate the two on WP. Kbb2 (ex. Mr KEBAB) (talk) 21:03, 11 June 2019 (UTC)[reply]
I concur with mach. Why should we use a foreign standard if a well-defined national standard is available? We don't use the standard of Spain for the entire Spanish-speaking world, nor that of Portugal for the entire Portuguese-speaking world, nor that of the PR of China for the entire Mandarin-speaking world. And many of our users will have noticed this and will expect Wikipedia to follow this policy throughout. Love —LiliCharlie (talk) 21:22, 11 June 2019 (UTC)[reply]
@LiliCharlie: Because this "foreign standard" is a diaphonemic notation which aims to accomodate all three national standards in all aspects possible (when it's not possible, the "bias" is towards NSG - but as far as I can see this only applies to the fortis-lenis distinction).
The 7th edition of Das Aussprachewörterbuch clearly uses a diaphonemic notation. For example, on page 52 they admit that the transcription [aːɐ̯] is mostly wrong as the vowel is almost always the same as [aː], the main allophone of /aː/. The reason they write [aːɐ̯] is to accomodate those speakers which use a consonantal variety of /r/ in words such as Haar, which they write [haːɐ̯]. If they wrote [haː], the fact that speakers of Swiss Standard German (and some regional varieties in other countries) use [r], [ʀ] or [ʁ] in this position would have to be inferred from spelling. [haːɐ̯] makes it obvious, even though this transcription doesn't make sense when you read it literally (it's a difficult diphthong to produce and it's probably at least twice as difficult to use it consistently, rather than using it in free variation with [aː] and [aːʁ]).
Pronunciation dictionaries already use [ɪʁ (...)] (or a similar notation) to mean "either [ɪʁ (...)] or [ɪɐ̯ (...)]" in NSG as the vocalized pronunciation is very common (this also shows that not only there's no distinction between /aːr/ and /aː/ in non-prevocalic contexts, /ar/ often (if not most often) also joins this neutralization in NSG and ASG - so the transcription [aːɐ̯, aː, aʁ] is fully (or doubly, depending on how you want to put it) accomodating SSG and a rather unrealistic distinction when it comes to NSG and ASG). What's the problem of using [iːɐ̯ (...)] to mean "either [iːɐ̯ (...)] or [iːr (...)]"? Pronunciation dictionaries already use this convention.
[ʔ] also already means "either a glottal stop or nothing" in some positions as it's dropped in some (mostly word-initial) positions in NSG. What'd be the problem if we used [ʔ] to mean "a glottal stop that is much more often dropped in Southern German than in Northern German"?
We already use [ɛː] to mean "either [ɛː] or [eː]". The latter pronunciation is just somewhat less prestigious, but it's still standard in many if not most regions.
So all of this isn't "my idea". My idea is only to align the way we transcribe German on WP with pronunciation dictionaries. When you know the rules, you can derive the SSG form from any transcription (sometimes you have to look at the orthographic form). All it takes is the willingness not to read the IPA literally.
Also, it's better not to include SSG when ASG isn't included either. It's not a big loss, especially when you know how to read the IPA in a non-literal way in order to derive the SSG form from our transcriptions. Kbb2 (ex. Mr KEBAB) (talk) 10:51, 3 September 2019 (UTC)[reply]
A diaphonemic transcription only works under the condition that the readers are native speakers. Then one may claim that they will naturally pronounce the words in the right way according to the region where they live. This condition is not met on the English Wikipedia. We must assume that most readers are not native speakers of German and therefore have no way of knowing what the intended pronunciation is. Our readers do not “know how to read the IPA in a non-literal way in order to derive the SSG form from our transcriptions”.
Also, if you honestly wanted to provide a one-size-fits-it-all, over-regional, “diaphonemic” transcription, [r] would be a much better choice than delving into the needless details of allophonic variation which even fuses phonemes. This is what the most widespread German dictionary of them all does, the plain orthography Duden, which is not aimed at specialists, but at the general public – much like Wikipedia.
Ceterum censeo you are ignoring decades of research that disprove your prescriptivst point of view. Please stop ignoring the Varientenwörterbuch. --mach 🙈🙉🙊 04:40, 4 September 2019 (UTC)[reply]
Addendum: I forgot that the transcription you want to prescribe is not “diaphonemic” at all. It does not feature any diaphonemes whatsoever. Instead, it represents the allophones of certain varieties of standard German up to a degree where the actual phonemes are blurred: the phoneme sequence /ər/ is sometimes represented by [ɐ], sometimes by [əʁ]. --mach 🙈🙉🙊 04:30, 5 September 2019 (UTC)[reply]
Me, too, concur with mach and LiliCharlie. I strongly reject your idea, Kbb2. German is a pluricentric language, and that's acknowledged by numerous publications. Switzerland and Austria have their own rules about pronouncing Standard German; see e.g. Schweizerhochdeutsch (= Swiss Standard German) by Hans Bickel and Christoph Landolt, 2nd ed. Dudenverlag (!), Berlin 2018, p. 99–104 and the long bibliography there. Your proposal is like replacing the U.S. pronunciation of U.S. place names by the U.K. pronunciation – a no-go. --Freigut (talk) 09:25, 13 June 2019 (UTC)[reply]
@Freigut: That doesn't address my concerns regarding Austrian Standard German though. The fact that there are multiple national varieties of German is undisputable. Kbb2 (ex. Mr KEBAB) (talk) 09:43, 1 September 2019 (UTC)[reply]

--mach 🙈🙉🙊 17:02, 4 November 2018 (UTC)[reply]

French as English Examples?[edit]

Why is French the main go-to for sounds which are apparently not available in English? How would the average English speaker know how the 'an' in 'chansons' is pronounced? --IronMaidenRocks (talk) 04:52, 16 April 2019 (UTC)[reply]

Yes, it doesn't make sense to say a French sound is an English equivalent. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 15:53, 16 April 2019 (UTC)[reply]
Under English natives, French is probably the most spoken language that uses these nasals.--A11w1ss3nd (talk) 12:29, 30 December 2019 (UTC)[reply]

Cat vs father[edit]

Listen please:
cat
father
AVS (talk) 04:50, 16 July 2019 (UTC)[reply]

@Avernarius: Right, the trap vowel in English cat, accent is a perfect illustration of Standard German /ɛ/ [ɛ] in Ende, hätte. Please listen to Amy Walker.😉 Love —LiliCharlie (talk) 09:57, 16 July 2019 (UTC)[reply]
Sorry – deutsches Sprach schweres Sprach, dass die 'der' 'die' 'das' der Teufel hol ...
but there is also a difference between -1 Ende and -2 hätte: -1 rather like 'pension', -2 battle. Yours AVS (talk) 10:51, 16 July 2019 (UTC)[reply]
@Avernarius: Not unless you speak with a Western Swiss accent. There's a difference in spelling but not pronunciation (compare English 'bed' vs. 'bread'). The recordings of Ende and hätte have the same vowels, as expected (people on the recordings speak Standard German). Kbb2 (ex. Mr KEBAB) (talk) 12:09, 16 July 2019 (UTC)[reply]
Ouch, since 1966, I'm living in Kebabvillage Berlin. AVS (talk) 12:27, 16 July 2019 (UTC)[reply]
Wikipedia is an encyclopædia that depends on reliable sources, not on where users have chosen to settle. — My reply was intended to show that this help page is limited to a couple of very similar Standard German accents targetwise, but not to RP, General American or any other native English accent departurewise. Love —LiliCharlie (talk) 12:55, 16 July 2019 (UTC)[reply]

Wrong. trap vowel is [æ], German Ende or hätte is [ɛ] like in „end“ (mouth a bit more closed). A11w1ss3nd (talk) 09:16, 20 November 2020 (UTC)[reply]

Should we use Austrian (or Swiss) Standard German pronunciation for topics related to Austria (or Switzerland)?[edit]

There is a clear consensus to use Austrian (or Swiss) Standard German pronunciation exclusively for topics related to Austria or Switzerland.

Cunard (talk) 09:27, 15 December 2019 (UTC)

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

When we provide a German pronunciation for a topic related to Austria (or Switzerland), what variety of Standard German should we use? There are at least three options: use Austrian (or Swiss) Standard German exclusively, use German Standard German exclusively, or use both. --mach 🙈🙉🙊 14:12, 3 November 2019 (UTC)[reply]

Survey[edit]

  • Use Austrian (or Swiss) Standard German exclusively (for topics related to Austria or Switzerland). Numerous reasons: This is our current practice with German and other languages. Standard German is widely recognized to be a pluricentric language, so German Standard German has no precedence over Austrian (or Swiss) Standard German. Using Austrian (or Swiss) Standard German for topics relating to Austria (or Switzerland) is more helpful to readers. Imposing a different Standard German variety on topics that relate to Austria (or Switzerland) will alienate Austrian (or Swiss) editors. The differences are minute, so there is no point in additionally using German Standard German (e.g. allophones of /r/, absence of [ʔ], Final-obstruent devoicing). There are reliable sources we can refer to (e.g. Variantenwörterbuch des Deutschen or Österreichisches Aussprachewörterbuch). --mach 🙈🙉🙊 14:12, 3 November 2019 (UTC)[reply]
  • Use Austrian (or Swiss) Standard German exclusively. This just makes sense. We don't give British pronunciation/spelling for American topics.--Ermenrich (talk) 14:44, 3 November 2019 (UTC)[reply]
  • Use Austrian (or Swiss) Standard German exclusively. I'm Swiss and to me it would seem quite absurd to give the German Standard German pronunciation for a place name in the German-speaking part of Switzerland, for example. Gestumblindi (talk) 16:45, 3 November 2019 (UTC)[reply]
  • Use Austrian (or Swiss) Standard German exclusively. We should also make sure the guide is clear on the transcription conventions for each dialect. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 03:30, 4 November 2019 (UTC)[reply]
  • Use Austrian (or Swiss) Standard German exclusively. This logically follows from MOS:ENGVAR; if necessary, we should add a MOS:LANGVAR principle immediately under it, along these lines: "Similarly, use of a non-English language in its non-English context should be responsive to strong national ties. For example, in an article on an Austrian topic, the German pronunciation key for the subject's name should follow Austrian Standard German norms."  — SMcCandlish ¢ 😼  05:07, 4 November 2019 (UTC)[reply]
    I like the MOS:LANGVAR idea. --mach 🙈🙉🙊 21:14, 6 November 2019 (UTC)[reply]
  • Use Austrian (or Swiss) Standard German exclusively. Wikipedia:Manual of Style/Pronunciation#Foreign names says: "Transcriptions ... are normally given in the national or international standard of the language in question..." As there is no international standard of German we naturally use national ones. Love —LiliCharlie (talk) 07:48, 4 November 2019 (UTC)[reply]
  • Use Austrian / Swiss Standard German unless good reasons exist (can be discussed on the article talk page) to additionally use another variant. At Vorarlberg, we currently use the German pronunciation (stress on the first syllable), and I think we should use the Austrian one (stress on the second syllable), but mention the German pronunciation somewhere (it looks more natural for most German speakers, including those with German as an additional language; the stress on the second syllable after "Vor-" is rather unusual). This is not comparable to British vs American spelling, as Vorarlberg directly borders Germany. —Kusma (t·c) 10:10, 6 November 2019 (UTC)[reply]
    Agree. Duden Online does this. I lived in Germany for many years and always preferred Vor'arlberg because I found it easier. Jmar67 (talk) 11:10, 6 November 2019 (UTC)[reply]
    @Kusma: It's a digression, but I wonder... people in Vorarlberg speak a form of Alemannic German (like Swiss German), unlike the rest of Austria, where the dialects are Austro-Bavarian. I would assume that this also influences stress in Austrian Standard German usage in Vorarlberg. Myself, I'm Swiss (from the northwestern part of the country, so quite far away from Eastern Switzerland and Vorarlberg, I might be entirely wrong about this), and I would stress Vorarlberg on the first syllable. Are you sure that in Vorarlberg itself people use stress on the second syllable? Gestumblindi (talk) 21:16, 6 November 2019 (UTC)[reply]
    Gestumblindi, no, I am not sure how the locals pronounce it (I am Southwest/West German), but have heard non-Vorarlberg Austrians say VorARLberg. Googling gives me this, which points towards this being a local thing, and an exception also for most Austro-Bavarian speakers. (In this case, the debate may even be notable enough to mention in the article). Back to the point: There are going to be some exceptions to the rule, so "exclusively" should not read "other ways to pronounce the word must be removed". I fully agree that the minor differences as pointed out by User:J. 'mach' wust should usually not be included. —Kusma (t·c) 21:35, 6 November 2019 (UTC)[reply]
  • Use Austrian / Swiss Standard German per the arguments above. RockingGeo (talk) 00:24, 8 November 2019 (UTC)[reply]

Discussion[edit]

[Moved to the very top for the sake of visibility.] In this discussion above I bring up the issue of conflicting descriptions of Austrian Standard German in the literature. Please familiarize yourselves with it before you participate in the survey. Kbb2 (ex. Mr KEBAB) (talk) 14:50, 3 November 2019 (UTC)[reply]

In that section, you have referred to a detailed allophonetic description of Austrian German (Moosmüller et.al. 2015) as a reason to dismiss our broad dictionary-style Austrian German transcriptions. That reference is seemingly plausible, though ultimately irrelevant because nobody has ever considered using detailed allophonetic descriptions of Austrian German instead of broad dictionary-style transcriptions. --mach 🙈🙉🙊 15:31, 3 November 2019 (UTC)[reply]

I do not think that we should seek to impose "one rule fits all" for this. I can imagine situations where it would be appropriate to apply Vienna dialect, but if you start doing it in an article about an internationally famous figure such as Bruno Kreisky or for that matter Gustav Mahler, Sigmund Freud or the Emperor Joseph. I think that would look a little ... odd. And when you talk of Swiss German are you thinking of Chur, Zuerich, Bern or Basel? If you try and apply "one rule for all occasions" you will get into endless pissing contests over the exceptional cases where the rule does not make sense. We prefer wiki contributors to contribute content rather than spend time on digressive discussions of this nature, surely. Or...? Better, surely, to leave it to the individual writing / translating in respect of the individual entry in question. Regards Charles01 (talk) 14:41, 3 November 2019 (UTC)[reply]

He's referring to Swiss and Austrian Standard German, not Dialect.--Ermenrich (talk) 14:43, 3 November 2019 (UTC)[reply]
"Austrian standard German" is the official Language used in Austria. It's like British and American English. In de.WP we use Austrian Standard German for Autrian topics, and Swiss Standard German (for example no use of "ß"9 for Swiss related articles. --Kanisfluh (talk) 15:31, 3 November 2019 (UTC)[reply]
Legally speaking that's not quite right. Article 8 (1) of the Federal Constitutional Law merely stipulates that "German is the official language of the Republic without prejudice to the rights provided by federal law for linguistic minorities." There is no mention of either an Austrian or a standard variety of German. [Cited after Bundes-Verfassungsgesetz (B-VG)/​Federal Constitutional Law issued by the Austrian Rechtsinformationssystem des Bundes (RIS).] Love —LiliCharlie (talk) 08:51, 4 November 2019 (UTC)[reply]
And? The US has no official language, it doesn't prevent us from using American English on US articles.--Ermenrich (talk)
I don't think LiliCharlie is disagreeing on that. It's just a minor factual correction. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 15:23, 4 November 2019 (UTC)[reply]
That's right. "Official language" is an exclusively legal term, and lawyers disagree on which varieties are subsumed by the legal terms "German" and "German language." In Germany, for instance, it is even unclear whether the Low German language, which is not mutually intelligible with Standard German, legally belongs to the German language, see here. And in Austria foreign dialect speakers of German use the constitutional official language without having to change their linguistic habits. Similarly, 28 or so states in the US define their official languages without specifying particular national or stylistic varieties of English, Spanish, etc. Love —LiliCharlie (talk) 19:56, 4 November 2019 (UTC)[reply]
The constitution does not, however, say that the "German" it refers to is the German German. Regarding vocabulary, the official (!) norm is the ÖWB. Regarding speach, let me point to the inauguration of our current head of state. Both the president of the Federal Assembly and the president-elect used an Austrian interpretation of spoken standard German. If that were not considered as granted by the constitution, there for sure would have been yet another legal battle :) → «« Man77 »» 19:51, 6 November 2019 (UTC)[reply]
Of course not Germany German in particular! Still the entire German language is constitutionally official in Austria (i.e., Staatssprache not Amtssprache), so speakers of Germany or South Tyrolese German who demand an interpreter at tribunals will be told that they speak the official language of the court and that no translation is required or even possible. And when the Austrian authorities deal with their correspondence, the national or socioliguistic varieties used are not an issue either, and constitutionally they may not be. Having contributed to article de:Österreichisches Wörterbuch I am aware of the status of the dictionary. Note that Austrian German vocabulary is not limited to what it contains; neologisms not found in that (or any other) dictionary are frequent. Love —LiliCharlie (talk) 01:17, 7 November 2019 (UTC)[reply]
I was summoned by the bot, but do not speak German well enough to have an intelligent opinion. However, as an analogy to other languages, my thought is that local place names should reflect local dialect, whereas larger places and more famous people should perhaps show multiple pronunciations if need be, but should include whatever the German equivalent might be of Parisian French or Castillan Spanish, ie the German I would have learned in high school in France. That's my best answer to this question, going away now. Elinruby (talk) 02:36, 17 November 2019 (UTC)[reply]

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Implementation[edit]

Now that the RfC has been closed in favor of the regional varieties, are we to restore the previous form of the key? Nardog (talk) 12:25, 30 December 2019 (UTC)[reply]

I think so. That presentation helps users spot the differences that exist between the major national standard varieties of German. Love —LiliCharlie (talk) 13:14, 30 December 2019 (UTC)[reply]

what about x [ks] and qu [kv]?[edit]

Currently [ks] and [kv] are lacking. --A11w1ss3nd (talk) 12:21, 30 December 2019 (UTC)[reply]

It would be a bit cumbersome to add every possible consonant cluster. We usually don't include any in these guides. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 16:15, 30 December 2019 (UTC)[reply]
We could theoretically add a note that'd say that [ps] and [ks] are sometimes counted among the affricates of Standard German (following the logic that they aren't homorganic, like /pf/), though a better place for that is probably Standard German phonology. Kbb2 (ex. Mr KEBAB) (talk) 16:55, 30 December 2019 (UTC)[reply]
Yeah, I can't see that informing editors or readers in how to transcribe German or understand German transcriptions. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 17:11, 30 December 2019 (UTC)[reply]
A11w1ss3nd, please understand that this page is only about IPA representations of speech sounds. Phonotactics including possible consonant clusters should be dealt with in Standard German phonology, though that article seems to contain little information on that topic. — For current as well as historical orthographic representations using the the Roman alphabet go to article German orthography. Love —LiliCharlie (talk) 18:52, 30 December 2019 (UTC)[reply]

x/χ phoneme and uvular trill [ʀ] mention[edit]

Hi, I'm just wondering whether it would be worth mentioning as a note here that the velar fricative (x) in German can sometimes be realised as a uvular fricative (χ) by some speakers, especially after some vowels in German as the Standard German phonology article mentions, because even though I don't speak German, I have heard some German speakers pronouncing it closer to a uvular fricative rather than the listed velar fricative. I had attempted to do this last year but my change got reverted by User:Erutuon on the basis that the uvular fricative was not listed but even so, I see no reason to exclude this mention at least as a note (I do think it shouldn't be mentioned as a main phoneme when it is listed as x, but I think χ should be mentioned as a note) because articles like Help:IPA/Arabic and Help:IPA/Dutch both give mention as notes that the x phoneme can be phonetically realised as a uvular fricative even though they both mainly list it as a velar fricative (with χ being an allophone of x) so I don't see how IPA Standard German has to be any different in not mentioning this even as a note. I had attempted to raise this issue in the Talk:Standard German phonology page last year after my edit got reverted but since no one bothered to respond to my query there (honestly, whats the point in having a talk page if no one is keen on responding to a simple query?), I've raised the issue again in this talk page.

Another thing I'm wondering is if we should also mention in the ʁ/r phoneme note is that /r/ can also phonetically be realised as a uvular trill [ʀ] as well as the voiced uvular fricative (e.g. "In other regions, the uvular pronunciation prevails, mainly as a fricative/approximant [ʁ] or trill [ʀ]")? I would be very grateful if anyone here can please point me in the right direction for these queries, many thanks. Broman178 (talk) 08:12, 7 February 2020 (UTC)[reply]

A good way of approaching this is to consider whether this note would help editors transcribe German and/or if it would help readers understand the transcriptions. If we aren't indicating the allophony between [x] and [χ], I don't think it would be. But if this is a common feature, I don't see why we shouldn't indicate this in our transcriptions. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 15:30, 7 February 2020 (UTC)[reply]
Thanks for your reply, looking at the Standard German phonology article, it does seem to mention that depending on the preceding vowel, the /x/ phoneme may be pronounced either in the velar or uvular region (it mentions χ can occur after a, a, ʊ, aʊ̯ and to a lesser extent ɔ, while x occurs in other areas) and it doesn't seem to mention that this is merely dialectal feature or something restricted to a certain city/german-speaking region so in a way it does indicate that this could be a common feature. I can understand its exclusion if the uvular fricative is restricted to a certain German-speaking area or a city where the language is predominant (e.g. if it was restricted to Berlin or Munich, then I believe would not need to be mentioned here as that would just be dialectal rather than a common feature) but if the information indicates that this phoneme applies to most german speakers then I do believe it should be mentioned in a note. If it were to mentioned, I think a wording similar to what's written in Help:IPA/Arabic would be the best format as the note there says "/x/ is pronounced as [x] or [χ].", especially as that article lists it as an /x/ phoneme but doesn't exclude the χ fricative, through the note I just quoted. However, if most people support its exclusion here, then I will go along with that agreement/consensus even if I disagree. Broman178 (talk) 17:45, 7 February 2020 (UTC)[reply]
The sources for [ꭓ] we have so far are sketchy. The sources we have in Standard German phonology#Consonants are by Kohler or by Moosmüller, Schmid & Brandstätter. The former describes the speech “of many educated Germans in the North”, whereas the latter describes Standard Austrian German. Central and Southern Germany or Switzerland are thus not covered, so we cannot say whether or most German speakers have a [ꭓ] allophone with our current sources. --mach 🙈🙉🙊 20:44, 7 February 2020 (UTC)[reply]
In that case, it probably would be useful to add more sources on that matter in the Standard German phonology article covering Switzerland and the Central/Southern bits of Germany. However, I'm no big IPA expert as its only from last year I've started editing article in the IPA field so while I could attempt to add more sources if I am free, it may not be an easy task for me. That being said, while I agree the information is sketchy, it still does support the fact that the /x/ allophone could be pronounced either as [x] or [χ]. Broman178 (talk) 10:29, 9 February 2020 (UTC)[reply]
I reverted the edits because I didn't see the purpose of describing the distribution of [χ] when [χ] is not even used in transcriptions. The notes should describe allophones that are written in transcriptions; other information about allophones is not really needed, and in this case could confuse editors so that they use [χ] instead of [x], so that there would be inconsistency between transcriptions. However, as there's confusion about [χ] being mentioned in Standard German phonology but not used in transcriptions, perhaps there should be a note saying something like "we write [x] even though the actual pronunciation is sometimes [χ], depending on dialect and environment" just to make it clear what's going on. — Eru·tuon 21:29, 9 February 2020 (UTC)[reply]
I agree that we should avoid mentioning it if we aren't transcribing it. But if we are confident enough about [χ] occurring that we would put a footnote, we might as well transcribe with it. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 00:41, 10 February 2020 (UTC)[reply]
Thanks for explaining your reasoning further for reverting my edits last year Erutuon, I can understand your point about the transcription purpose of [χ] but I might also point out (as I have done above) that the IPA help pages for Dutch and Arabic both use the velar fricative [x] in transcriptions but still mention the uvular fricative in a footnote (unless you can explain further to me that there is a way the uvular fricative is transcribed for Dutch and Arabic despite being listed as [x]) so I personally don't see how that can be confusing (though I perhaps have a different perception on what is confusing and what isn't). However, I am nevertheless willing to not include mention of this in a note here if everyone agrees with it as I do believe any agreement/consensus should be followed here.
By the way, I would be interested in hearing your opinions on giving mention to the uvular trill here Aeusoes1, Erutuon and mach alongside the uvular fricative in the note here as that hasn't been mentioned yet in this discussion even though I have mentioned it in my query. Broman178 (talk) 19:27, 10 February 2020 (UTC)[reply]
I can't speak with confidence regarding Dutch (though it looks like there might be a bit too many footnotes), but I know that Arabic is variably described as having a uvular or velar articulation due to the free variation/dialectal differences, so I can see merit in including it there. That said, The Arabic IPA guides are such a muddled mess that I don't even know how to fix them without doing some heavy research on the topic. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 19:38, 10 February 2020 (UTC)[reply]
@Broman178: I think the note on the Dutch fricatives gives more information than is necessary, so it should be shortened (though I don't want to get involved in such a detailed page on a language that I know less about), but the note on Arabic /x/ seems to stay on task (simply stating that the phoneme has two allophones); it seems closer to what I think the note on this page should do, simply clarify that one symbol is used rather than another. Given the Dutch note, it seems that not everyone agrees about keeping the notes purely focused on helping editors write transcriptions, so I regret reverting your addition of the note rather than editing it or raising the issue here on the talk page. — Eru·tuon 22:23, 10 February 2020 (UTC)[reply]
Thank you for both your responses, I think later if I get time I might mention a note similar to what is in IPA Arabic in here for the velar fricative because even if a bit sketchy, the information for Standard German Phonology seems sufficient enough for me to support that [χ] or [x] can be allophones of the /x/ phoneme. Would any of you by chance agree with mentioning the uvular trill [ʀ] by any chance because even though it occurs in a specific German speaking area, it does seem to be common (I might point out that the alveolar trill is also a dialect thing because it mostly occurs in Switzerland, Austria and some parts of Southern Germany while elsewhere it is mostly uvular) or are you happy with just the uvular fricative alone being mentioned in the note? Broman178 (talk) 22:56, 11 February 2020 (UTC)[reply]

I have now reinstated the note I attempted to add last year but I have worded it differently and kept the description more simple based on what has been discussed here because in my opinion, it still is worth giving mention to [χ] in a note even though it is mostly listed/transcribed as /x/ because it helps people more in using the correct transcribed/listed symbol. If there are any issues with the way I have worded it, please feel free to change it. However, I won't give mention to the uvular trill [ʀ] in the /r/ note unless any of you are willing to comment on that here and agree to its mention. Broman178 (talk) 07:02, 13 February 2020 (UTC)[reply]

I don't like it. There's no indication that editors would be confused as to how to transcribe German without such a note and it doesn't help readers understand out transcriptions. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 15:38, 13 February 2020 (UTC)[reply]
Thats your opinion but I think Kbb2's edits have improved my addition a bit because the bottom of the note now says that /x/ is the phoneme used for transcriptions regardless of whether speakers realise it as either [x] or [χ]. Its not really about whether people get confused with the symbols or not, its about guiding them to use the correct transcribed symbol while also informing users about the allophones within that phoneme, thats why I personally think the note is useful. Anyway, I won't be making any more changes on this myself so I'll leave it here. Broman178 (talk) 17:46, 13 February 2020 (UTC)[reply]
We don't need to guide editors to not use ⟨χ⟩ with a footnote because we already guide them not to use it by not including it in the key. Again, there's no indication that editors are confused about whether to use ⟨χ⟩ or not so the footnote is unnecessary. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 18:27, 13 February 2020 (UTC)[reply]
Per The Principles of the IPA (1949) it is acceptable to transcribe the voiceless uvular fricative with x even if there's no variation in pronunciation (meaning: when the fricative can only be uvular, as in Northern Standard Dutch). Also, German isn't the only language in which x stands for more than one sound. In Polish, the sound can be velar or glottal, voiceless or voiced depending on the environment and speaker (before voiced consonants it can only be voiced and velar, though). In Serbo-Croatian /x/ is often also closer to glottal, and in Spanish it can signify a velar, uvular, glottal or even palatal fricative, depending on the variety. Kbb2 (ex. Mr KEBAB) (talk) 20:37, 13 February 2020 (UTC)[reply]
The Association's 1989 principles published in the Handbook (1999:159f.) still stipulate that “[o]rdinary roman letters should be used as far as practicable, but recourse must be had to other symbols when the roman alphabet is inadequate.” — In the Handbook (1999:88) Kohler writes that “[x] is used instead of [χ] after high and mid back tense vowels (e.g. [ˈbux] Buch ‘book’, [ˈhox] hoch ‘high’)”, but other authors have described other distributions. On the same page Kohler also writes that “[ç] occurs [...] morpheme initially (e.g. [çeˈmi] Chemie ‘chemistry’, [-çən] = diminutive suffix)” while current pronunciation dictionaries indicate word initial [x-] as the only option in certain well-established words (e.g. Junta, Chuzpe etc.). — In my opinion the situation is far too complex for a help page with the purpose of guiding users. Love —LiliCharlie (talk) 23:01, 13 February 2020 (UTC)[reply]
I have no issue myself with its mention however, if you all feel the mention of the uvular fricative is an issue in the footnote then feel free to remove its mention from the note, even though I disagree with that. But if you feel its mention is problematic here, I might as well point out that Help:IPA/English uses symbols that are pronounced differently by many English speakers even though they are listed/transcribed differently there e.g. [r], the symbol for the trill is used even though a footnote mentions that most speakers pronounce it as an approximant [ɹ] while [ʌ] in English is actually pronounced closer to [ɐ] by many English speakers (except some speakers in the USA, Canada, South Africa and some parts of Southern England), especially those speaking RP, and that is covered in the footnote there even though the actual [ɐ] symbol is not listed as a main transcribed phoneme of English (except in Australian English where it is listed), so with that logic, I don't see what the issue is with mentioning a simple thing like [χ] in a footnote here just out of concern of whether people will use the symbols for transcriptions or not (e.g. people don't often use [ɹ] for transcriptions even though the footnote mentions it). The reason why I believe mentioning that here was worthwhile was because it is mentioned in the footnotes in articles like the help pages for Dutch and Arabic even though the transcribed phoneme for them is [x] otherwise I wouldn't have attempted to mention it if that wasn't the case. So if its mention is to be removed here, it should be removed in those articles too in order to maintain consistency with the IPA help guides here because I believe in maintaining consistency here no matter what. I won't be making any more changes regarding this topic so I'll leave it for you all to decide whether to keep [χ] or remove it completely but if you choose to remove it, then it should be removed from the Dutch and Arabic help pages too (and other pages which transcribe it as [x] but mention [χ] in a footnote) otherwise its pure double standards to me on this matter if that doesn't happen. Broman178 (talk) 05:57, 14 February 2020 (UTC)[reply]
See WP:OTHERSTUFFEXISTS. Each of these languages are treated differently because each language is different. You may be right that they should be removed from the other languages, but the situation is different for the three languages and it's not necessarily inconsistent to have footnotes in one language and not another. In the case of English, many of those notes are a response to editors bringing up the issue in the talk page (the English rhotic has been brought up several times) and so it's guided by an indication that readers and editors have been confused in the past. We don't have such an indication with the German uvular allophone. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 15:11, 14 February 2020 (UTC)[reply]

Does /z/ exist in ASG/SSG?[edit]

Standard German phonology indicates the /s/–/z/ contrast is maintained in "southern varieties", saying, The obstruents /b, d, ɡ, z, ʒ, dʒ/ are voiceless lenis [b̥, d̥, ɡ̊, z̥, ʒ̊, d̥ʒ̊] in southern varieties, and they contrast with voiceless fortis [p, t, k, s, ʃ, tʃ] ... /b, d, ɡ, z, ʒ/ are voiceless in most southern varieties of German. But Moosmüller, Schmid & Brandstätter (2015) don't list /z/ (but do /b̥, d̥, ɡ̊/). It seems either the key or the phonology article needs to be amended. (cc @Austronesier, J. 'mach' wust, LiliCharlie, and Sol505000) Nardog (talk) 15:52, 18 February 2021 (UTC)[reply]

@Nardog: Not to my knowledge ([ʒ̊, d̥ʒ̊] also don't exist, at least on a phonemic level). Sol505000 (talk) 15:56, 18 February 2021 (UTC)[reply]
Krech & al. (2009:235) Deutsches Aussprachewörterbuch distinguish between three registers of Austrian Standardaussprache (standard pronunciation):
"I. Die gehobene Standardaussprache geschulter Sprecher. Sie folgt weitgehend Siebsschen Grundsätzen unter geringer Berücksichtigung der österreichischen Sprechkonvention.
II. Die gemäßigte Standardaussprache geschulter Sprecher. Zwar schließt auch sie sich den Siebsschen Grundsätzen an, bringt aber in wesentlich stärkerem Umfang die österreichische Sprechkonvention ein.
III. Die Standardaussprache der Laien. Als »regionales Hochdeutsch« folgt sie der österreichischen Sprechkonvention auf der Grundlage der verschiedenen großräumigen dialektbedingten Lautungen, Lautdistributionen, Lautkombinationen und Silbenverhältnisse, so dass sie entsprechend regional differenziert ist.
Während die Register I und II eng zusammengehören und in Bezug auf Einzelheiten individuell gegeneinander durchlässig sind, setzt sich Register III davon deutlich ab. Hier können individuell auch umgangssprachliche und teilweise sogar einzelne unmittelbare dialektale Lautungen einfließen, was aber im Folgenden unberücksichtigt bleibt. Nicht berücksichtigt wird ferner die den Schreibungen unmittelbar folgende Leseaussprache von Laien.
Was die Breitenwirkung der Register I und II und damit ihre Akzeptanz bei der Bevölkerung betrifft, so werden sie im Österreichischen Rundfunk und Fernsehen (ORF) von Ansager(inne)n, Nachrichtensprecher(inne)n und Moderator(inn)en verwendet. Dazu kommen einerseits in literarischen Sendungen und Hörspielproduktionen Rezitatoren und Schauspieler meist mit Register I und andererseits für spezifische Sendungen engagierte Moderator(inn)en mit oft nur geringer oder keiner Sprechausbildung, so dass sie im allgemeinen zwischen den Registern II und III liegen. Trotz der Unterschiede im Einzelnen garantiert die österreichische Herkunft der Sprecher(innen) und besonders ihre heimische Intonation Akzeptanz."
On page 242 they write about the Austrian coronal fricatives:
"In den Registern II und I werden die schriftlichen Verteilungen von ⟨s⟩ gegenüber ⟨ss⟩ und ⟨ß⟩ als stimmlose Lenis und Fortis beibehalten, also [v̥rˈeːz̥n̩] fräsen : [v̥rˈɛsn̩] fressen : [ɡ̊e̞v̥rˈeːsiç] gefräßig. Dabei werden in I nach den Siebsschen Empfehlungen teilweise auch stimmhaftes [z] und [ʒ] gesprochen."
That is to say, the fortis–lenis opposition is retained in registers I and II (trained speakers) and the phonemes may be realised as voiceless vs. voiced in register I (trained speakers' "elevated" standard pronunciation).
On page 265 they describe the distribution in Switzerland as follows:
"Ähnlich wie die Plosive verhalten sich die s-Laute. Steht am Wortende oder am Silbenende vor Vokal oder Sonorant ein einfaches ⟨s⟩, wird dieses oft als stimmlose Lenis realisiert, z.B. in den Wörtern Gras [ɡ̊ʁaːz̥] (neben [ɡʁaːs]), auseinander [aɔ̯z̥aɛ̯nˈandəʁ] (neben [aɔ̯saɛ̯nˈandɐ]) oder löslich [lˈøːz̥lɪç] (neben [lˈøːslɪç]). An allen Stellen, an denen für die Aussprache in Deutschland [z] kodifiziert wird, wird in der Schweiz häufig der stimmlose s-Laut [z̥] gesprochen, z.B. Sonne [z̥ˈɔnːə] oder Esel [ˈeːz̥əl̩]."
So what they say is that pronouncing the eqivalent of Germany German /z/ as a voiceless lenis sound [z̥] (that is identical with one Swiss realisation of G.G. /s/) is "frequent", but they don't expressly say that the voiced sound or the voiceless fortis sound also occur. Love —LiliCharlie (talk) 20:28, 18 February 2021 (UTC)[reply]
Right, so based on that what is your advice on how we should transcribe those fricatives? Sol505000 (talk) 21:14, 18 February 2021 (UTC)[reply]
and ʒ̊ look good to me, as this keeps the distinction that trained speakers maintain while it also shows that a voiceless pronunciation is the common one. (The IPA Handbook (1999:15) says: "It is a moot point whether [k̬] and [ɡ] refer to phonetically identical sounds, and likewise [s] and [z̥].") That'd be my solution — unless native speakers of those varieties protest. Love —LiliCharlie (talk) 15:14, 25 February 2021 (UTC)[reply]
a voiceless lenis sound [z̥] (that is identical with one Swiss realisation of G.G. /s/) – I do not think Krech & al. (2009: 265) are saying that [z̥] is a Swiss realization of Germany German /s/. What they are saying is that there is often no final fortification. The fortis [s] is not really discussed.
There is a solid contrast between /s/ and /z/ in German-speaking Switzerland. I do not have any literature about Swiss Standard German at hand, but cf. Fleischer/Schmid (2006) for a discussion of the fortis-lenis contrast and its transcription in a Swiss German dialect. Swiss Standard German phonetics uses the same sounds that occur in the dialects.
/z/ should be added. I do not think /dʒ/ should be included. There can be no fortis-lenis distinction in Southern German obstruent clusters, so [d̥ʒ̊] is the same as [tʃ] – the fortis-lenis contrast only exists between voiced sounds or pause. The affirmation that the distinction is based on “articulatory strength” is just one POV.
In addition to Austrian and Swiss Standard German, Southern Germany Standard German should be included as well. --mach 🙈🙉🙊 06:41, 26 February 2021 (UTC)[reply]
We cannot add phonemic /z/ to our key, it must be etic [z], [z̥], [s] or [s̬], or perhaps several of those. Love —LiliCharlie (talk) 22:28, 26 February 2021 (UTC)[reply]
I was talking about the footnote. I had not seen that the key merges /z/ and /s/ for Austrian and Swiss Standard German, which is clearly wrong, so I am going to fix it right away.
Regarding the footnote, here is my suggestion:
In Austrian Standard German and Swiss Standard German as well as in Southern Germany, the lenis obstruents /b, d, ɡ, z, ʒ/ are usually voiceless [b̥, d̥, ɡ̊, z̥, ʒ̊] (only /v/ is really voiced). In Northern Germany Standard German, they are often voiced [b, d, ɡ, z, ʒ] if surrounded by other voiced sounds or after a pause. There is no consensus about what distinguishes the voiceless lenis obstruents from the fortis obstruents /p, t, k, s, ʃ/. The distinction might include articulatory strength, duration, or fortis stop aspiration. In Northern Germany Standard German, the fortis and lenis obstruents are merged in final-obstruent devoicing.
--mach 🙈🙉🙊 06:50, 27 February 2021 (UTC)[reply]

After reading this page I'm quite convinced that edits like this one are not really covered by the sources and don't represent the outcome of this discussion. The sources say that in southern varieties of Standard German /z/ might be realised only in certain registers and even there only partially, right? --Mai-Sachme (talk) 06:28, 1 March 2021 (UTC)[reply]

The problem arises because our IPA key for German is not diaphonemic like the one for English. It is purely phonetic, so the symbol z here is supposed to mean voiced [z] rather than abstract /z/ which doesn't necessarily have to be realized voiced and can even mean the same thing as /s/ in certain varieties if the transcription is not only phonemic but diaphonemic. I am still in favour of the symbol for Austrian and Swiss Standard German. Love —LiliCharlie (talk) 13:32, 1 March 2021 (UTC)[reply]
I think I understand, but since the symbol z here is supposed to mean voiced [z] this edit should be addressed as factually incorrect, because that's not what the sources say and as a native South Tyrolean and proficient speaker of Standard German I can assure you that I have never heard the town name pronounced as [zaˈlʊrn] in my lifetime. Which is not to say, that I can't imagine anyone ever pronouncing it that way (tourists from Germany might do it, there might also be cases of hypercorrection), but it's surely not the way it is regularly done by regional speakers of Standard German.
As for the , I have carefully read your source, but... honestly, is there anyone else out there who postulates the pardoxical existence of such a thing as a voiceless voiced alveolar fricative? That seems to me like a rather... ephemeral opposition :-) --Mai-Sachme (talk) 14:04, 1 March 2021 (UTC)[reply]
[z̥] is usually taken to refer to a lenis alveolar fricative, a "voiceless-ized" lenis [z], or a "lenis-ized" [s]. [s] and [z̥] mean the same if voicing is the only difference between [s] and [z]. I don't want to go into details here, but let me remark the following: Air-pressure from the lungs being equal more air passes through the wide open glottis for "normal" voiceless sounds than through the extremely narrow and intermittently closed glottis for modally voiced ones, and this is seen as the basis of the fortis–lenis contrast by some authors such as Luciano Canepari, who describes lenis voiceless sounds as sounds with a narrow glottis as for (some forms of) whisper. However the exact nature of the fortis–lenis distinction remains unclear; other authors see a difference in duration, or one in "articulatory strength," as mach noted above. In any case, the transcriptions [s] and [z̥] may refer to different sounds, and one of the reasons for that is that voiceless is a cover term for quite different states of the glottis. Love —LiliCharlie (talk) 16:12, 1 March 2021 (UTC)[reply]
Thanks for the explanation, very interesting!
More in general, after skimming through this talk page I think that many issues (and probably also the current one) arise from fundamentally different answers users give to a few basic questions. What kind of readers do we cater to with our transcriptions (experts vs. laymen)? What is the message we want to convey (a prescriptive pronounciation guide favoring the allegedly most prestigious variant vs. a descriptive account of how the standard pronounciation is most likely/frequently realized)? And what are the expectations of our readers (do they prefer uniformity throughout all German language transcriptions or do they accept local variations)? No wonder this talk page is as long as it is :-) --Mai-Sachme (talk) 17:04, 1 March 2021 (UTC)[reply]

By the way, a few empirical datapoints: [1], [2], [3], [4]. Not that much [z] or [z̥] in sight in the south... --Mai-Sachme (talk) 21:56, 1 March 2021 (UTC)[reply]

Moved from User talk:Nardog
 – Nardog (talk) 21:58, 1 March 2021 (UTC)[reply]

Hey Nardog, in the link you provided I can read that /z/ is only realised in certain registers in the south and even in these registers only partially (teilweise). No need to change anything based on that. --Mai-Sachme (talk) 06:18, 1 March 2021 (UTC)[reply]

@Mai-Sachme: The participants concluded that /s, z/ do contrast initially (whatever their realization) and decided to represent the phoneme /z/ with z, which is now reflected in Help:IPA/Standard German. Sie in the table is represented by z for all three varieties, so your reversion doesn't make sense. Nardog (talk) 21:16, 1 March 2021 (UTC)[reply]
Sorry, but I can't find that part in the discussion, where the participants concluded /s, z/ do contrast initially in ASG or STSG. Can you point it out to me?
Apart from that, I tend to thing my reversion makes a lot of sense, because I'm a proficient speaker of Standard German and native South Tyrolean, and [zaˈlʊrn] sounds as alien as it gets in my ears. And if a proficient local native speaker finds a transcription strikingly odd, there might be a strong suspicion that something is wrong, I suppose.
If your are interested in a few empirical data points: [5], [6], [7], [8]. The picture is far from being complete, obviously, but /z/ really is a rarity in the south, no matter in which position. --Mai-Sachme (talk) 21:51, 1 March 2021 (UTC)[reply]
@Mai-Sachme: See Mach's first post. He restored z for Sie, diese and no one has contested. Saying all that to me is completely and utterly pointless, so I'm moving it here. Nardog (talk) 21:58, 1 March 2021 (UTC)[reply]
Ah, I see, this edit is based on this post and ASG isn't really discussed. I hereby contest. Is there really any kind of reliable non-prescriptive academic literature that alleges a widespread use of /z/ in ASG, especially in word-initial position? Moosmüller, Schmid & Brandstätter (2015) mention it exclusively in intervocalic position as a possibility. --Mai-Sachme (talk) 22:22, 1 March 2021 (UTC)[reply]
Peter Wiesinger who was responsible for the Austrian part in Krech et al. (2009) says that "Register I und II ... werden ... im Österreichischen Rundfunk und Fernsehen (ORF) von Ansager(inne)n, Nachrichtensprecher(inne)n und Moderator(inn)en verwendet" (for full wording and context, see above) whereas Moosmüller et al. (2015) assert that "pronunciation used in the Austrian broadcasting media is unsuitable for defining SAG", i.e. they exclude the higher-register accents that Wiesinger says have the contrast. Love —LiliCharlie (talk) 23:22, 1 March 2021 (UTC)[reply]
Yes, and even in the "highest" register [z] is only partially (teilweise, i.e. not consistently) realized. In other words, [z] isn't really a common feature of ASG. It occurs in high registers, but is barely heard in the vast majority of speech production. So Help:IPA/Standard German should be amended accordingly, correct? --Mai-Sachme (talk) 08:04, 2 March 2021 (UTC)[reply]
That reminds me of comedian Christoph Grissemann who frequently mocks his fellow Tyroleans (and Vorarlberg Austrians, who speak an Alemannic/Swiss variety of German) for not being able to speak intelligible German. Higher registers certainly enjoy more acceptance in eastern Austria that in the West. As to South-Tyrolean German, article de:Südtiroler Deutsch, which treats the "in Südtirol geschriebene und in förmlichen Situationen gesprochene Varietät des Standarddeutschen", expressly says: "Die deutsche Standardsprache wird oft als fremd empfunden, und Defizite im sprachlichen Ausdruck sogar der Gymnasiasten wurden im Auftrag der Südtiroler Landesregierung u. a. durch die Studie DESI nachgewiesen." [User Mai-Sachme removed that whole section along with its seemingly reliable source a few hours after I wrote this. Love —LiliCharlie (talk)] So the question seems to be whether or not to cater for those who have difficulty acquiring Standard German. Love —LiliCharlie (talk) 10:05, 2 March 2021 (UTC)[reply]
Wait, are we talking about [z] or /z/? Mach's point was that as long as the contrast is made, it should be reflected in our transcriptions, which I share. So in those high registers, is the contrast made (whatever its realizations may be)? Nardog (talk) 10:22, 2 March 2021 (UTC)[reply]
(after edit conflict) I don't think that we are talking here about catering for those who have difficulty acquiring Standard German, but about offering transcriptions to our readers that reliably reflect real-world usage. Therefore I'm sorry, but your remarks seem a bit off-topic... Nardog pointed out to me, that this edit was substantially and formally correct, because on this talk page participants [had] concluded that /s, z/ do contrast initially in ASG. If he was correct, then we should also change the transcription in Salzburg... But I contest that this is correct and I'm yet to see any evidence for this allegedly widespread usage of [z] in ASG.
For fun's sake I just listened to the first three minutes of the news show of the Austrian national public service broadcaster. One can hear three journalists and several politicians in a formal setting. Only the male journalist produces [z] (in sieben- and in Sarkozy), but not in other instances it would be expected. The female journalists and all politicians together don't produce, as far as I can hear, one single [z], not even in a word like sonnig. --Mai-Sachme (talk) 10:24, 2 March 2021 (UTC)[reply]
I'm yet to see any evidence for this allegedly widespread usage of [z] in ASG. Hence my question. Mach wasn't alleging widespread usage of [z], but of /z/. Nardog (talk) 10:33, 2 March 2021 (UTC)[reply]
Nobody here seems to contests that the (phonemic) contrast is observed by at least some Austrian speakers. Even Mai-Sachme reports in their anecdotal observation that one of the speakers in the show used /z/ [z] in sieben and Sarkozy. Love —LiliCharlie (talk) 10:48, 2 March 2021 (UTC)[reply]
Yes, and the key observation is that while [z] surely exists in ASG, it is rarely used, even in formal settings (according to literature and even more so according to my anecdotal evidence). This should be reflected in Help:IPA/Standard German and in the transcriptions we offer to our readers. --Mai-Sachme (talk) 10:56, 2 March 2021 (UTC)[reply]
So you do agree that the fricative in Sie, diese still needs to be transcribed with something other than s even for ASG/SSG, correct? If so, what do you suggest it be? ? Nardog (talk) 11:13, 2 March 2021 (UTC)[reply]
Uhm, sorry, but I fail to see the deeper meaning of your question and its real-life ramifications. What exactly prevents us from amending Note 1 with something like: [z] is rarely realized in Austrian Standard German, it occurs mostly in highly formal registers? That would then prevent an edit like this one, correct? --Mai-Sachme (talk) 12:05, 2 March 2021 (UTC)[reply]
There's no deeper meaning. The first consonant of Salurn should be transcribed with whatever this guide lists for Sie under the column "AT". LiliCharlie said Nobody here seems to contests that the (phonemic) contrast is observed by at least some Austrian speakers and you responded Yes. So I ask you again: What do you think the consonant transcribed with z for GSG should be transcribed with for ASG/SSG in our IPA-de transcriptions? Nardog (talk) 12:46, 2 March 2021 (UTC)[reply]
Hm, no, wait a second. There is a deeper meaning and I disagree with this part: Salurn should be transcribed with whatever this guide lists for Sie under the column "AT". What you are implying here is that users should follow in their article edits a guideline framed by the strict formal structure provided by another internal Wikipedia page and controlled by a bunch of Wikipedians. This is quite obviously not in line with several fundamental Wikipedia policies (please spare me to list them all, let's start with Wikipedia:Wikipedia is not a reliable source...). Academic sources cited above specify that in ASG [z] is almost exclusively and inconsistently found in special registers which represent only a tiny fraction of speech production and which by another academic source are even excluded as being non-representative for real ASG. This completely suffices to justify my revert and I'm quite sure you realize that as well. I don't know how to fit this datapoint within the formal structure of Help:IPA/Standard German, but, honestly, it's also irrelevant. If the formal structure doesn't allow a randomized occurance of a sound in specific registers, then the formal structure is wrong and should be changed. But I also feel that's not my burden. My best take is this one, if you don't like it, then change it accordingly with the cited sources in mind. --Mai-Sachme (talk) 13:26, 2 March 2021 (UTC)[reply]

An editor with a strong opinion is also "a bunch of Wikipedians"—a bunch of one. Krech et al. (2009), a descriptive source, gives zaˈlʊʶn (so much for the "guideline framed by the strict formal structure provided by another internal Wikipedia page"). This is a legitimate representation of the German pronunciation of the place name, as it represents a widely used and accepted pronunciation, and at the same time gives justice to all common regional variations by simple rules which are stated in the introduction to the book (equivalent to our key). z can be voiced or unvoiced, a can be back, mid or front, l can be "clear" or velarized, and so on. Transcribing the pronunciation in the local variant of Standard German IMO equals forcing on our readers a "mock" regional accent. –Austronesier (talk) 19:35, 2 March 2021 (UTC)[reply]

Austronesier, while it's fair to say that zaˈlʊʶn represents an "accepted pronunciation", it doesn't, by any standards, represent a "widely used" pronounciation. You could make that claim for ˈzaltsbʊɐk, but surely not for Salurn, a small alpine town with less than 4,000 inhabitants. A vast majority of all speech tokens mentioning Salurn originate in South Tyrol, where the realization of [z] is a rarity, even in high registers.
There are quite a few remarks one could say now about Krech et al. and their "descriptive" method, but let's leave it at the observation that in our specific case the "mock" "regional" "accent" (no less than three scare quotes) comprises half of the German sprachraum, how empirical data without prescriptive intentions show.
Regarding your scepticism about using different standard varieties, a few sensible things have already been said higher on this page, especially in this post, and then there was also a poll. --Mai-Sachme (talk) 21:04, 2 March 2021 (UTC)[reply]
Krech et al. "allow" me to say iˈbɪtsa, while you disenfranchise Northerners and non-native speakers aiming at a /z/~/s/-distinguishing variant to say zaˈlʊʶn. Where now lies the prescriptivism? –Austronesier (talk) 21:41, 2 March 2021 (UTC)[reply]
The main problem with the transcription of German lenis obstruents is that they are only allophonically voiced in Northern German, but the IPA defines the respective signs [b d ɡ z ʒ] as modally voiced (I am not including [v] because it is truly voiced and does not pattern as an obstruent in German). However, the German lenes do not have modal voice in any variety (at least not in standard pronunciation), because voiceless pronunciations occur in all varieties, including in the North, whereas voiced pronunciations are rare in the South.
Regarding the sibilants, the opposition between (voiceless) /z/ and /s/ is most prominent between vowels in minimal pairs like heiser – heisser. I know that all registers of Swiss Standard German have a very solid opposition between the lenis and fortis sibilants between vowels, and I thought the same would be true of Austrian Standard German, but I am not sure. At the ends of words in minimal pairs like Reis – reiss, the opposition is neutralized in Northern German due to final fortition. At the beginning of words, I have heard anecdotal evidence that some Northern German speakers may have a marginal distinction between a lenis sibilant in sechs and a fortis in Sex. I do not think that such an initial sibiliant opposition exists in Southern German. At least in Switzerland, the two words are homophones unless sechs is pronounced /zɛxz/.
There are two questions:
  1. What signs do we use for the fortis-lenis opposition of obstruents?
  2. What signs do we use when there is no fortis-lenis opposition of obstruents?
Regarding the first question, it seems to me that the use of [b d ɡ z ʒ] is well established even though the respective German sounds are not truly voiced. Currently, we have a footnote explaining that the respective sounds are not voiced. I see no harm in mentioning [b̥ d̥ ɡ̊ z̥ ʒ̊] in this guide as well. After all, I understand this guide is meant to be descriptive.
Regarding the second question, I do not think there is any harm in transcribing the initial sibilant as [z] or as [s] as long as we do not claim that this has any implications on the very existence of the fortis-lenis contrast in sibilants. In other words: [saˈlʊrn] is totally fine to me as long as we keep distinguishing [ˈhaɪzər] (or [ˈheɪz̥ər]) from [ˈheɪsər]. Also, we should probably add another footnote explaining that Austrian or Swiss Standard German do not have an opposition between initial fortis and lenis sibilants and that therefore, either [z] or [s] are sometimes used for transcribing the neutralized initial sibilant. --mach 🙈🙉🙊 06:33, 3 March 2021 (UTC)[reply]
Austronesier, my understanding is that we are catering here to non-native speakers interested in a pronounciation that reliably reflects the most common real-world usage, much rather than to language learners who set themselves the goal to aim at a word-initial /z/~/s/-distinguishing variant. Offering that has nothing to do with "prescriptivism". And you might want to think again about that "disenfranchisement" claim... --Mai-Sachme (talk) 07:13, 3 March 2021 (UTC)[reply]
FWIW, Rudolf Muhr's Österreichische Aussprachedatenbank has no entry for Salurn but the similar salopp is transcribed as [s̬aˈlo̞p] rather than [saˈlo̞p]. (As no audio is available for that word you have to search the Gesamtwörterbuch to find it.) Love —LiliCharlie (talk) 07:03, 3 March 2021 (UTC)[reply]
Thank you, I didn't know that one. You can hear the sound, when you search for Sonne. They seem to use [s̬] consistently. --Mai-Sachme (talk) 07:13, 3 March 2021 (UTC)[reply]
@J. 'mach' wust: How robust is the contrast between the initial sounds in Serbien and Servus in the Swiss samples of the Österreichische Aussprachedatenbank (there is no such contrast in the samples for sechs and Sex)? Is it an artifact of the elicitation setting (which tend to create acro-acrolectal pronunciations)? @Mai-Sachme: Dito for the Austrian samples (although here it only affects one speaker): would you expect the speaker to make the contrast in running speech too?
As for the initial /s/~/z/ contrast in sechs vs. Sex in Northern German: it is marginal, but robust; definitely more robust than the opposition between /s/ and /ts/ following /n/ and /l/ in absolute final position, or the opposition between /eː/ and /ɛː/. –Austronesier (talk) 10:06, 3 March 2021 (UTC)[reply]
Hm, very interesting. The Swiss sample for Sex seems quite odd in the light of this map, though.
As for the Austrian samples, all the speakers sound plausible to me, but on the street I'd expect [s] to be significantly more frequent (also check out Seriosität and Serie), maybe the occurance of [s̬] in Serbien is indeed an artifact of the elicitation setting or, simply, an idiosyncrasy. --Mai-Sachme (talk) 11:14, 3 March 2021 (UTC)[reply]

@J. 'mach' wust: All your suggestions make a lot of sense to me. --Mai-Sachme (talk) 12:26, 3 March 2021 (UTC)[reply]

I have not been able to get more than a brief crackle out of the adaba site, even after trying various browsers. Thus I have not been able to listen to the recordings. A Swiss distinction between the s in «Serbien» and «Servus» seem extremely unlikely to me, though. I faintly recall listening to adaba recordings a few years ago and thinking that the Swiss recordings sounded very Northern German to my ears.
Anyway, I can see the adaba transcriptions, and they confirm a fortis-lenis distinction for all three varieties:
  • When searching for the string "eise", the first word is Abreise. In Austrian German, the pronunciation is transcribed with [s̬], which is s with COMBINING CARON BELOW, the IPA sign for voicedness (so we have either a voiced voiceless [s̬] or a voiceless voiced [z̥] ☺). The corresponding Swiss pronunciation has simple and simplistic [z].
  • The first result for the string "eiße" is abreissen. It is transcribed with [s̟] in all three varieties, that is s with COMBINING PLUS SIGN BELOW, which is weird because this is really the IPA sign for an advanced pronunciation. It appears they have expanded the meaning of the sign in their own transcription to include «Fortiskonsonant», cf. X-SAMPA-AUSTRIA-Überblickstabelle.
So the adaba confirms the intervocalic fortis-lenis contrast for Austrian German, even though they are using weird transcriptions.
I am very disappointed of the Atlas zur Aussprache des deutschen Gebrauchsstandards. While they variously claim that this or that sound is a fortis or a lenis, they fail at describing the regular fortis-lenis contrast in German sibilants, concentrating entirely on marginal phenomena such as the Sex–sechs contrast in Northern Germany. They completely ignore the Southern fortis-lenis contrast in minimal pairs such as heiser–heisser or Reis–reiss.
Incidently, they confirm that the Sex–sechs contrast only exists in Northern Germany. They transcribe Sex as voiceless [s] all over the German-speaking area, but they are using the same voiceless [s] for transcribing the initial lenis in Austria, Southern Germany and Switzerland, cf. /z/ im Anlaut in Sirup und Saison.
So here is my suggestion for the new footnote, for changing the fortis-lenis footnote, and for new example words (dropping example words with initial s that are now explained in the footnote):
Consonants
DE AT CH Examples English approximation
s lassen, groß, HausFest[1] fast
z Sie, diese, unser[2][1] (in Austrian German sparsely used) DE: zebra
AT, CH: soup
  1. ^ a b Initial s is normally transcribed as [z]. This reflects the Northern German pronunciation, which can be voiced initially. The voiceless initial s from Southern German is sometimes transcribed with [s] instead, since Southern German does not have any distinction between initial /s/ and /z/.
  2. ^ In Austrian Standard German and Swiss Standard German as well as in Southern Germany, the lenis obstruents /b, d, ɡ, z, ʒ/ are usually voiceless [b̥, d̥, ɡ̊, d̥ʒ̊, ʒ̊] and are distinguished from /p, t, k, tʃ, ʃ/ only by(only /v/ is really voiced). In Northern Germany Standard German, they are often voiced [b, d, ɡ, z, ʒ] if surrounded by other voiced sounds or after a pause. There is no consensus about what distinguishes the voiceless lenis obstruents from the fortis obstruents /p, t, k, s, ʃ/. The distinction might include articulatory strength, duration, or fortis stop aspiration. (/v/ is really voiced, and /s/ is the only alveolar fricative). The distinction is also retained word-finally. In German Standard German, voiceless [b̥, d̥, ɡ̊, z̥, d̥ʒ̊, ʒ̊] as well as [v̥] occur allophonically after fortis obstruents and, for /b, d, ɡ/, often also word-initially. See fortis and lenisIn Northern Germany Standard German, the fortis and lenis obstruents are merged in final-obstruent devoicing.
What do you think about the suggestions? --mach 🙈🙉🙊 20:40, 3 March 2021 (UTC)[reply]
I don't see a recommendation how to transcribe word-initial ⟨s⟩, the "normally" and "sometimes" in footnote 1 don't seem to do the job. I also wonder where users might imagine the boundary between the north and the south is. Otherwise that looks good to me. Love —LiliCharlie (talk) 21:20, 3 March 2021 (UTC)[reply]
My point is we do not need a recommendation for initial s. When a reader comes to this help page after clicking on a pronunciation, they do not need a recommendation, but a description of the sounds, which is what we are providing. --mach 🙈🙉🙊 06:39, 4 March 2021 (UTC)[reply]
I am not sure what all that means. If speakers from Frankfurt and Wiesbaden which are the largest city and the capital of the German state of Hesse have no intervocalic /s~z/ contrast and typically pronounce that as [z], should we then also transcribe Hessen as [ˈhɛzn̩]? Or do we accept that there are national pronunciation standards? Love —LiliCharlie (talk) 07:36, 4 March 2021 (UTC)[reply]
@J. 'mach' wust: I also had problems with adaba initially, but then I found a method to hear the soundfiles. 1) Write a word in the "Orthographische Suche", 2) click on the magnifier, if a result is found, the word appears in the field "Ergebnis", 3) double-click on the word you want to listen to in the field "Ergebnis", strangely all soundfiles are played mutually, 4) only now the individual soundfiles can be started individually :-)
Talking about the AADG by the Institut für Deutsche Sprache: you shouldn't mistake the AADG for a lexicographical project or even a systematic, all-encompassing mapping of the German language. It is work in progress and only covers select features and, as I was told many years ago, they have found great joy in debunking long-held beliefs, eternally reiterated in literature, which is why they always report the takes of Duden and Siebs. Sirup, traditionally reported exclusivly with word-initial [z], but in fact realized by more than half of all German-speakers with [s], turns out t be a good case study...
As for your proposal, why not something like this in footnote 1: There is no consensus regarding the transcription of initial s, since there are regional differences between voiced and voiceless realizations. Therefore transcriptions like [z], [z̥], [s̬] and [s] occur, with voiced pronounciations more commen in northern parts of the German sprachraum and voiceless pronounciations in the southern parts. --Mai-Sachme (talk) 07:23, 4 March 2021 (UTC)[reply]
Chiefly based on mach's arguments in this and earlier threads, my final take is:
  1. Since we don't have a consensus for a diaphonemic notation, advocating for uniform initial [z] based on diaphonemic arguments (including my previous arguments) is obviously moot. In a consequent diaphonemic notation, we would also distinguish [s ~ z̥] before consonant and pause, with [z̥] as default in most lemmata. (Off-topic: @mach: Do you have example for a lemma with [s] in final position or before a consonant in Swiss Standard German? I am only familiar with inflected forms.)
  2. The broad phonetic notation in the key has to guide non-native speakers, and should not contain the potential for fatal mispronunciations. For stops, it is certainly better to use [t]/[d] than [tʰ]/[t], since many readers would then be prone to pronounce [t] in a way which Germans associate with a fortis articuation (e.g. with slight aspiration etc.). Fully voiced [d] is within the range of allophones of /d/, so no harm here, even if [d̥] actually is the default value.
  3. The same holds for intervocalic [s]/[z]. Most readers won't have an unvoiced fortis/lenis contrast for sibilants, so fully voiced [z] will help readers to make the oppostion where it matters. Does it matter elsewhere? Apparently not. Even if the default pronunciation in (Northern) German Standard German (in most native and nativized words) and Swiss Standard German (in all words) is lenis (regardless of voice), the optional transcription with [s] is less "fatal" as in the case of stops. Alternatively, we could introduce [z̥] to the key, but even then, we would have to approximate it with "soup". Plus, for Austrian German it would be spurious (per Moosmüller et al. in JIPA); having three potential transcriptions (Austrian [s], Swiss [z̥], "Northern" [z]) for the same thing (phonologically) would be too much.
Austronesier (talk) 08:15, 4 March 2021 (UTC)[reply]
That's all fine, but mach seems to imply that there is no national Standard German pronunciation for initial ⟨s⟩ in Germany, so Sigmaringen being a city in Germany's South may have initial [s] in the then no longer national standard. And I speculated that Hessen might analogously have a [z] in another regional standard. Love —LiliCharlie (talk) 08:41, 4 March 2021 (UTC)[reply]
@LiliCharlie: The AADG shows that the pronunciation of initial s varies within Germany, cf. /z/ im Anlaut in Sirup und Saison. It appears you want this help page to have examples with initial s. How would you suggest to include them? Only in the [z] entry, or in the [s] entry as well? With explanations in parenthesis or footnotes? What explanations? As I have already said, I personally think this help page works perfectly fine without examples of initial s, but I am not against including such examples, I just do not see any easy way for doing so.
@Austronesier: Examples of Swiss standard German words with final /s/ after a long could be heiss, gross, Spiess, and after short vowel Hass, Biss (the latter a minimal pair with bis). After long vowel and before consonant e.g. schliesslich, weisslich, after short vowel e.g. wässrig, hässlich (before obstruents, /z/ and /s/ are neutralized, cf. de:Heuslersches Gesetz). --mach 🙈🙉🙊 19:33, 6 March 2021 (UTC)[reply]

[i̯][edit]

Phonetically speaking, [i̯] and [j] are completely equivalent. The symbols may represent slightly different sounds depending on the conventions, but doesn't even the sound we represent by in transcriptions of Standard German still fall within "voiced palatal approximant"? What's wrong with linking it to Voiced palatal approximant? [o̯] would also fall within "voiced velar approximant" according to Catford (1977). Nardog (talk) 08:37, 25 February 2021 (UTC)[reply]

Maybe it's more a thing of native speaker intuition (thus phonemic rather than phonetic), but linking to the same thing as j appears odd to me. It evokes to me non-standard pronunciations like [ʃtuːt.jʊm]/[ʃtuːt.çʊm] instead of [ʃtuː.di̯ʊm], where /d/ is devoiced because it appears before a consonant other than in an allowed syllable-opening cluster (some speakers even turn Medien and Mädchen into homophones). Even more odd is the rendering [ʔjambʊs] Iambus instead of [ʔi̯ambʊs] (although in actual speech, most people would fluctuate between rapid [jam.bʊs] and careful [ʔi.am.bʊs]). For me, is more "vocalic" than j, but this is admittedly more a gut feeling than laboratory phonetics. –Austronesier (talk) 10:07, 25 February 2021 (UTC)[reply]
Yeah, saying "[ʔjambʊs] instead of [ʔi̯ambʊs]" means nothing to me. j indeed often represents an articulation with greater constriction than i, but AFAIK that still doesn't make [i̯] not a palatal approximant. Nardog (talk) 10:23, 25 February 2021 (UTC)[reply]
Krech et al. (2009) distinguish between [bi̯ˈaŋkaː] Bianca and [bjœʁn] Björn, and I hesitate to believe they introduced a spurious/meaningless contrast just guided by the spelling, or the transcription conventions for the source languages of these two fully nativized names. –Austronesier (talk) 10:39, 25 February 2021 (UTC)[reply]
I don't doubt they represent different sounds; I doubt either of them falls outside of what the article Voiced palatal approximant describes. Nardog (talk) 10:44, 25 February 2021 (UTC)[reply]
Austronesier is right, " is more 'vocalic' than j," and that's more than a gut feeling. Traditionally symbols for pre-vocalic non-syllabic vowels indicate variation with syllabic vowels. As Mangold (2005:62) puts it: "Die unsilbischen Vokale [i̯ o̯] der Standardlautung (vgl. S. 42) erscheinen in der Bühnenaussprache (bis 1957) als die silbischen Vokale [i y u o], doch wird [i̯] nicht völlig abgelehnt. Beispiele: Aktio̱n [akˈt͜si̯oːn] > [akt͜siˈoːn]; loya̱l [lo̯aˈja:l] > [loaˈja:l] sexuẹll; [zɛˈksu̯ɛl] > [zɛksuˈɛl]." This happens not only in traditional stage pronunciation but also in slow or extremely precise, and in northern-ish speech. I don't think non-syllabic vowels are imperative to speak native-like German, except in the diphthongs /aɪ ɔʏ/.
As to Krech's distinction between [bi̯-] and [bj-]: It's certainly possible to say [biˈaŋka(ː)] for Bianca, but *[biˈœʁn] for Björn sounds a bit strange to me. Love —LiliCharlie (talk) 12:02, 25 February 2021 (UTC)[reply]
So what is your recommendation as to where (or whether) to link the symbol in the key? Nardog (talk) 12:21, 25 February 2021 (UTC)[reply]
I suggest if we want link it anywhere, then rather to the vowel i. And FWIW, phonologically, there is a discrete gap between and j. A nice example is standard /ma:.ri̯oː/ vs. (predominantly Eastern) colloquial /ma:r.joː/ Mario. The first is realized as [ma:.ʁi̯oː], the second as [maː.joː] (for basilectal [mɔˤː.jɵː]). ʁ is not vocalized before (because the latter is still vowel-ish), but before j it is. –Austronesier (talk) 13:05, 25 February 2021 (UTC)[reply]
If we keep the symbols for non-syllabic vowels at all we should link to the articles on vowels. That leads to less confusion, and even if users pronounce a syllabic vowel instead of a non-syllabic one that does not make the pronunciation uneducated, let alone foreign. Love —LiliCharlie (talk) 13:38, 25 February 2021 (UTC)[reply]
OK, done. (@LiliCharlie: Speaking of, do you have an answer to Sol505000's question above?) Nardog (talk) 13:44, 25 February 2021 (UTC)[reply]
@LiliCharlie: Another small thing: should we restore then? Krech et al. don't use it, not sure about Duden. Anyway, there's hardly an article where we could make use of it (except maybe for Die Sexualität im Kulturkampf). –Austronesier (talk) 14:02, 25 February 2021 (UTC)[reply]
Krech et al. introduce the symbols , , , and ɔ̯ (=Mangold's ) in chapters 5.4.3 through 5.4.9 on pp. 57, 63, 66, and 68. (Both Krech et al. and Mangold use instead of /.) For consistency's sake we should either include them all, or none. I agree that use cases are extremely rare, but what names come up in tomorrow's news is unpredictable. Love —LiliCharlie (talk) 14:23, 25 February 2021 (UTC)[reply]
And then there's also the diphthong that I pronounce in pfui, hui and alternatively in ruhig. I think we can do without that one. Love —LiliCharlie (talk) 14:30, 25 February 2021 (UTC)[reply]

What is implied by this punctuation of /some symbol/ versus [some symbol]  ?[edit]

E.g. (one of many possible) from Note 2: "In German Standard German, voiced stops /b, d, ɡ/ are devoiced to [p, t, k] at the end of a syllable." Nothing states why // gets used and then [] gets used. Neither here, nor in https://en.wikipedia.org/wiki/Help:IPA/Introduction nor in https://en.wikipedia.org/wiki/Help:IPA/Conventions_for_English ... It would be helpful. Thank you for your time. — Preceding unsigned comment added by 71.176.129.25 (talk) 02:29, 1 March 2021 (UTC)[reply]

True. The explanation is here: International Phonetic Alphabet#Brackets and transcription delimiters. User:John Maynard Friedman just created the template {{Linguistics notation}} that links there, and I think we'll add it to this and similar pages before long. Love —LiliCharlie (talk) 02:58, 1 March 2021 (UTC)[reply]
{{Contains special characters|IPA}} and {{Linguistics notation}} added to this article. I have started adding them to the Help:IPA/xxxxxx articles, thus far just the Help:IPA/Axxxxx set to check for any adverse reactions. (For Armenian etc, I also added {{Contains special characters|Armenian)}}.) --John Maynard Friedman (talk) 15:26, 1 March 2021 (UTC)[reply]
That is _very_ nice. Thank you, very much.

Should this guide represent conversational pronunciation phenomena like progressive assimilation?[edit]

In the spirit of WP:BRD, I am reverting the recent changes by User:Sol505000. Those changes introduced the progressive assimilations of /ən/ to [m̩] in words like «graben» or to [ŋ̍] in words like «liegen» along with an explanatory footnote. The explanatory footnote cited a source (Hall 2003, pp. 55, 143), but it went beyond that source. According to the source, such progressive assimilations are optional: “the alveolar nasal /n/ can be assimilated to preceding labial and velar sounds” (Hall 2003: 143, my highlighing), and they are “very common in conversational pronunciation” (Hall 2003: 143, my highlighting). This means that these progressive assimilations can be expected in conversational situations like talks before small audiences, discussions or more formal dialogue, but not in the “moderate formal style” used by newsreaders on radio and television or before large audiences (cf. Hall 2003: 5).

Also, the explanatory footnote contained the following sentence: “In IPA transcriptions of German outside of English Wikipedia, all three are most often written with ⟨n̩⟩, so that graben and liegen are transcribed [ˈɡʁaːbn̩] and [ˈliːɡn̩], instead of the phonetically explicit [ˈɡʁaːbm̩] and [ˈliːɡŋ̍].” I think this is problematic for several reasons. It is never good when Wikipedia does things differently from what is common outside Wikipedia. Also, I think the statement “all three are most often written with ⟨n̩⟩” is dubious, because it seems to me that the transcription ⟨ən⟩ is at least as common. And, last but not least, I have shown above that according to the source, the transcriptions [ˈɡʁaːbm̩] and [ˈliːɡŋ̍] represent optional pronunciations common in certain styles, so we cannot simply call them “phonetically explicit”.

I would tend not to include the progressive assimilations in this guide. Wikipedia is an encyclopedia. As such, I would think we should stick to the more formal styles of German, and not include the numerous options of conversational pronunciation. But if we include them, we should certainly stick to the source and say that they are optional. --mach 🙈🙉🙊 20:15, 19 March 2022 (UTC)[reply]

We probably need a better source, but [ˈɡʁaːbn̩] and [ˈliːɡn̩] look like BS to me. This, for instance, is not [ˈfɔlksˌvaːɡn̩] but [ˈfɔlksˌvaːɡən] - both the lack of nasal plosion on /ɡ/ and the vowel between it and /n/ is very audible.
I can see that you are a speaker of Swiss Standard German which, I think, lacks syllabic consonants like Luxembourgish. Pedantic speech aside, [ˈfɔlksˌvaːɡən] (i.e. File:De-Volkswagen.ogg) is how the name would be said in Switzerland, or in Western Central Germany (due to the Limburgish/Ripuarian influences). The normal German and Austrian pronunciation is [ˈfɔlksˌvaːɡŋ̍], with a nasally released /ɡ/. Sol505000 (talk) 20:22, 19 March 2022 (UTC)[reply]
Regarding my edits, I tried to transcribe the corresponding audio clips (you can analyze them in Praat or other software yourself and tell me whether I'm correct):
File:Abenberg.ogg [ˈaːbənbɛʁk]
File:De-Alpen.ogg [ˈalpən]
File:Alpenrhein.ogg [ˈalpənʁaɪn]
File:De-Arthur_Schopenhauer2.ogg - this one MIGHT be [ˈʃoːpn̩haʊɐ] - it was pretty difficult to find that sequence in those recordings!
File:De-Bad_Liebenwerda.ogg [ˈbaːt liːbənˈvɛʁda]
File:De-Bad_Liebenzell.ogg [ˈbaːt liːbənˈtsɛl]
File:De-Benneckenstein.ogg [ˈbɛnəkənʃtaɪn]
File:Birkenkopf.ogg [ˈbɪʁkəŋkɔpf] - the nasal ends up velar anyway due to the assimilation to the following /k/
File:LL-Q188_(deu)-Sebastian_Wallroth-Berlin-Blankenburg.wav [ˈblaŋkənbʊʁk]
File:De-Blankenfelde.ogg [ˈblaŋkŋ̍fɛldə]
File:De-Brackenheim.ogg [ˈbʁakənhaɪm]
File:De-Bundesminister_für_besondere_Aufgaben.ogg [-ˈʔaʊfɡaːbən]
File:De-Cloppenburg.ogg [ˈklɔpənbʊʁk]
File:De-Das_Leben_der_Anderen.ogg [-ˈleːbən-]
File:Donauschwaben.ogg [ˈdoːnaʊʃvaːbm̩]
File:De-Erbendorf.ogg [ˈɛʁbəndɔʁf]
File:De-Eupen.ogg [ˈɔʏpən]
File:De-Falkenberg.ogg [ˈfalkənbɛʁk]
File:De-Falkenhagener_Feld.ogg [ˈfalkŋ̍haːɡənɐ-] (or [ˈfalkŋ̍haːɡnɐ-], which is how I hear it on the recording)
File:De-Franken.ogg [ˈfʁaŋkən]
File:Frankenbach.ogg [ˈfʁaŋkənbax]
File:De-Franz_Beckenbauer.ogg [-ˈbɛkənbaʊɐ]
File:Franz_Joseph_Hermann_Michael_Maria_von_Papen.ogg [-fɔn ˈpaːpən]
File:Gemeinnützige_Krankentransport_GmbH.ogg [-ˈkʁankŋ̍ˌtʁanspɔʁt-]
File:De-Grebenstein.ogg [ˈɡʁeːbənʃtaɪn]
File:Hadersleben.ogg [ˈhaːdɐsleːbm̩]
File:De-Hakenfelde.ogg [ˈhaːkŋ̍fɛldə]
File:De-Haldensleben.ogg [ˈhaldn̩sˌleːbm̩]
File:Heilbronner_Falken.ogg [-ˈfalkən]
File:De-Hockenheim.ogg [ˈhɔkənhaɪm]
File:Hockenheimring_Baden-Württemberg.ogg [ˈhɔkənhaɪmʁɪŋ-]
File:De-Iris_Berben.ogg [-ˈbɛʁbən]
File:De-Karawanken.ogg [kaʁaˈvaŋkən]
File:De-Kempen.ogg [ˈkɛmpən]
File:De-Kerpen.ogg [ˈkɛʁpən]
File:De-Leoben.ogg [leˈoːbən]
File:De-Meckenheim.ogg [ˈmɛkənhaɪm]
File:De-Meppen.ogg [ˈmɛpən]
File:De-Merkendorf.ogg [ˈmɛʁkəndɔʁf]
File:De-Mittelfranken.ogg [ˈmɪtl̩ˌfʁaŋkən]
File:De-Oschersleben.ogg [ɔʃɐsˈleːbən] (isn't this stressed on /ɔ/?)
File:De-Oppenheim.ogg [ˈɔpənhaɪ̯m]
File:De-PaulvHindenb-long.ogg [-fɔn ˈbɛnəkŋ̍dɔʁf-]
File:De-Reinickendorf.ogg [ˈʁaɪnɪkŋ̍dɔʁf]
File:De-Rodalben.ogg [ˈʁoːtʔalbən]
File:De-Rudolf_Eucken.ogg [-ˈʔɔʏkən]
File:De-Saarbrücken.ogg [zaːʁˈbʁʏkŋ̍]
File:De-samuel-von-bruckenthal.ogg [-fɔn ˈbʁʊkəntaːl]
File:De-Siebenhirten.ogg [ziːbənˈhɪʁtn̩]
File:De-Siebenbürgen.ogg [ziːbm̩ˈbʏʁɡŋ̍]
File:LL-Q188_(deu)-Sebastian_Wallroth-Steinstücken.wav [ˈʃtaɪnʃtʏkən]
File:De-Staaken.ogg [ˈʃtaːkŋ̍]
File:Untergruppenbach.ogg [ʊntɐˈɡʁʊpm̩bax]
File:de-Volkswagen.ogg [ˈfɔlksvaːɡən]
File:De-Zweibrücken.ogg [ˈtsvaɪbʁʏkən]
Also, [ziːbm̩ˈbʏʁɡŋ̍] is exactly how Krech et al transcribe Siebenbürgen. Can someone check their conventions? I might've mistaken Duden for "all dictionaries". Krech et al seem to use [m̩ ŋ̍] the way I do. Sol505000 (talk) 21:29, 19 March 2022 (UTC)[reply]
According to Krech et al., the syllabic assimilated nasals belong to the level of "Standardaussprache mit hoher bis mittlerer Artikulationspräzision" (p. 101). –Austronesier (talk) 11:50, 20 March 2022 (UTC)[reply]
That passage from Krech et al. applies to German standard German. --mach 🙈🙉🙊 12:14, 20 March 2022 (UTC)[reply]
Thanks for correcting my revert, that's what I meant (which is obvious from the edit summary).
Krech et al (p. 268) say that there's a free variation between [əm ən əl] and [m̩ l̩] in Switzerland. But they don't say whether the assimilated variants of [n̩] occur in Switzerland. The lack of assimilation would be notable and they'd state it explicitly. Because they don't say that, we should assume that the syllabic nasals work the same in Switzerland as they do in Germany (and in Eastern Netherlands, Denmark, etc.), so [ˈʃtaːkŋ̍] and [ˈalpm̩], not [ˈʃtaːkn̩] and [ˈalpn̩].
Again, maybe in your idiolect you only have [ən] which you think is [n̩] and hence your confusion. Sol505000 (talk) 13:22, 20 March 2022 (UTC)[reply]
There is certainly no confusion on my side. I am just skeptical what depth of phonetic detail we should represent. From the two big German reference pronunciation dictionaries, one writes [bn̩, ɡn̩], and the other writes [bm̩, ɡŋ̍]. If we really must prefer one over the other (why should we, though?), then I would opt for [bn̩, ɡn̩] for sake of simplicity and also because this allows transcribing the [bm̩ – bn̩] contrast. --mach 🙈🙉🙊 15:06, 20 March 2022 (UTC)[reply]
Duden (3rd edition, p. 33, chapter "Genormte Lautung"): "An Stelle von [pn̩], [bn̩], [kn̩], [ɡn̩] wird im allgemeinen häufiger [pm̩], [bm̩], [kŋ̍], [ɡŋ̍] gesprochen". Transcription is one thing, but the actual description is more relevant for us (similar as in the case of Help talk:IPA/Danish, where conventional and exact phonetic transcriptions quite diverge). So we have two sources describing [pm̩], [bm̩], [kŋ̍], [ɡŋ̍] as most the common realizations of /-pən#/ etc. Krech et al. are explicit about its geographical range (Germany and Austria).
I disagree that [pn̩], [bn̩], [kn̩], [ɡn̩] are "BS"; they occur in Swiss standard German (this what Krech et al. talk about in p. 268), and can also be heard in semi-pedantic speech in Germany. We should not make any OR-ish comments in a footnote as if these were just transcriptional artefacts when they're not.
We thus have a spectrum from [pm̩] to [pən], with the former as unmarked standard realizations in Germany and Austria, and the latter most common in Switzerland. The "intermediate" [pn̩] is fact the least common of all in actual standard German speech, as far as I can see. It deserves a mention, but shouldn't be in the primary key, being the pronunciation that is the least likely encountered by our readers. –Austronesier (talk) 19:39, 20 March 2022 (UTC)[reply]
I think introducing another difference between Swiss pronunciation and German or Austrian pronunciation would be unfortunate. Therefore, I would favour using one of the pronunciations that is common everywhere, [pn̩ bn̩ kn̩ ɡn̩] or [pən bən kən ɡən]. Such pronunciations occur everywhere in careful and explicit speech. On Wikipedia, we typically indicate pronunciations for isolated lemmata. For such a use case, a careful and explicit citation form pronunciation is not unnatural.
A problem of transcribing syllabic consonants are the exceptions. There are various cases when schwa pronunciations are more common. This becomes problematic when non-native speakers apply the key mechanically to hundreds of pronunciations.
A problem of transcribing the progressive assimilation is that it implies a nasal release of the stop, so [pm̩ bm̩ kŋ̍ ɡŋ̍] are really [pⁿm̩ bⁿm̩ kⁿŋ̍ ɡⁿŋ̍]. While this is natural to native speakers, the pronunciations we indicate here are meant for non-native speakers. Not knowing about the nasal release, they might misread ⟨pm̩ bm̩ kŋ̍ ɡŋ̍⟩ as if it were [pᵊm̩ bᵊm̩ kᵊŋ̍ ɡᵊŋ̍] or [pəm bəm kəŋ ɡəŋ], which would be completely wrong. --mach 🙈🙉🙊 06:45, 21 March 2022 (UTC)[reply]
I fully concur with the last point. Nothing would be more fatal than involuntarily inducing people to say something like [ˈbɛkʰŋ̍baʊɐ] with orally released [kʰ]. –Austronesier (talk) 18:48, 22 March 2022 (UTC)[reply]
If that was a problem we wouldn't use and ŋ̍ in Help:IPA/Danish. If [n̩] is the least common realization after bilabials and velars, we shouldn't transcribe it - it's simple as that. Especially given the fact that we already use for the syllabic nasal in großem. If [m̩ ŋ̍] vs. [n̩ n̩] is a legitimate dialectal difference it should be transcribed. Sol505000 (talk) 16:34, 23 March 2022 (UTC)[reply]
I do not see how Danish is relevant. It is a different language that has more consonantal weakenings and assimilations and less geographical variation compared to German. And nobody is talking about any “dialectal difference”. What we are talking about are differences between varieties of standard German. When talking about German, the differentiation between dialects and varieties of standard German is essential. And I do not get your point about the use of [m̩] in «grossem» – it is different from [n̩] in «grossen».
According to the literature we have cited so far, a pronunciation with progressive assimilation ([pm̩ bm̩ kŋ̍ ɡŋ̍]) is not used in all regions. Speakers in Luxemburg, Switzerland, and maybe some regions in Western Germany do not normally have it. On the other hand, a pronunciation without progressive assimilation ([pn̩ bn̩ kn̩ ɡn̩]) can be used in all regions in more explicit pronunciation styles. To me, this looks like a very strong argument for Wikipedia not to transcribe the progressive assimilation. This is also what one of the two reference pronunciation dictionaries does.
Another important point is our clear consensus to use Austrian (or Swiss) Standard German pronunciation exclusively for topics related to Austria or Switzerland. In other words, we should not prescribe progressive assimilation for Swiss topics (or, by extension, for Luxemburgish topics).
So we have two options: we can either continue not to transcribe progressive assimilation (which is acceptable everywhere), or we can allow for the variation between transcribing progressive assimilation or not. We cannot prescribe transcribing progressive assimilation everywhere because it is not used everywhere. --mach 🙈🙉🙊 22:21, 23 March 2022 (UTC)[reply]

[uə̯][edit]

According to Duden, the surname Ueli is [ˈuə̯li], but this key says [u] is only found in unstressed syllables and includes neither the diphthong [uə̯] nor the semivowel [ə̯]. How is it to be transcribed conforming to this key? Nardog (talk) 13:40, 9 July 2022 (UTC)[reply]

Krech et al. says (p. 263):

Die fallenden Diphthonge [iə̯], [uə̯], [yə̯] sollten bei der Aussprache von Ortsnamen wie Brienz, Fiesch, Spiez, Buochs, Muolen, Schlossrued, Flüelen oder Üetikon sowie Familiennamen wie Dieth, Lienert, Vieli, Buess, Fueter, Ruoff, Grüebler, Lüönd oder Rüegg verwendet werden. Auch das aus dem Alemannischen in die Standardsprache übernommene Wort Müesli sollte mit [yə̯] ausgesprochen werden.

Does this mean we should add ⟨iə, uə, yə⟩ under "Diphthongs"? Nardog (talk) 14:03, 9 July 2022 (UTC)[reply]

If we include them at all, they should be under "Non-native vowels" since they occur only in unassimilated loanwords from Alemannic. But it might be better to link such pronunciations to Help:IPA/Alemannic German instead of to this page. —Mahāgaja · talk 15:44, 9 July 2022 (UTC)[reply]
Yeah, they should probably be included with the non-native vowels. Sol505000 (talk) 05:41, 10 July 2022 (UTC)[reply]

Swiss pronunciations[edit]

All the existing discussion aside about whether to include Swiss SG pronunciations at all, I see a bigger problem: many of the pronunciations listed for Switzerland are just plain wrong. (For example, [k] in China.) Others omit essential information. (Like [z] being like soup, instead of zebra, which is incorrect as a blanket rule for CH, since the [s]-[z] distinction is situational.) In total, it seems like they were added by someone actually quite unfamiliar with Swiss SG (and Swiss German) pronunciation. — tooki (talk) 23:39, 24 November 2022 (UTC)[reply]

It seems like we should've never splitted the variants after all. Not to sound rude, but WP:COMPETENCE is required to transcribe variants not covered in pronunciation dictionaries. IPA transcriptions of Northern SG are easily verifiable.
The addition of Austrian and Swiss SG (or at least the latter, if we can establish that there are no problems with Austrian transcriptions) to the guide should be reverted if there are no editors who are willing to moderate the transcriptions.
One problem would be the extremely broad transcription [b d ɡ]. If there is no voice involved, it is wrong. Sol505000 (talk) 07:54, 16 February 2023 (UTC)[reply]

Post-vocalic rhotic question[edit]

Can anyone point to somewhere on the talk page that explains this decision: In many regions except for most parts of Switzerland, the /r/ in the syllable coda is vocalized to [ɐ̯] after long vowels or after all vowels (in this guide [ɐ̯] is used only after long vowels, following the pronunciation dictionaries), and /ər/ is pronounced as [ɐ]? Thanks! Wolfdog (talk) 23:20, 1 March 2024 (UTC)[reply]

I do not think it has been discussed, let alone decided. I guess it was introduced by our vociferous advocate against variants as a reaction to my pointing out that it is inherently difficult to draw a clear line between [ʁ] and [ɐ̯] (Special:Diff/913480762). However:
  1. There are numerous articles out there that use [ɐ̯] after short vowels. Therefore, the change goes against MOS:IPAINTEGRITY: “[C]hanging the key so that it no longer conforms to existing transcriptions will confuse readers.”
  2. It is not true that pronunciation dictionaries only use [ɐ̯] after long vowels since [ɛɐ̯] is widely used in prefixes.
Therefore, I believe we should restore the previous version. --mach 🙈🙉🙊 08:12, 2 March 2024 (UTC)[reply]
I concur that for many speakers in Germany, [ɐ̯] is always used after long vowels in closed syllables, whereas there is more variation after shorter vowels (except with final written -er(C) and in the prefixes er-, ver- and zer-). Many speakers consistently use [ʁ] after short vowels, whereas others (including myself) prefer [ɐ̯] but optionally use [ʁ] in "careful" speech; NB this option is not available after long vowels: both [vɪʁ] and [vɪɐ̯] wirr are not perceived as "regionally marked", whereas [viːʁ] wir definitely is.
Talking about dictionaries, both Duden and Krech et al. have this distribution of [ɐ̯] after long vs. [ʁ] after short vowels. And only Duden uses [ɛɐ̯] in prefixes, while Krech et al. use the syllabic monophthong [ɐ] (e.g. [ɐˈnɛːʁɐ] Ernährer).
So personally, I think we can keep it as a rough guideline. Whether this entails radically streamlining every entry from let's say [ˈɛɐ̯fʊɐ̯t] Erfurt to [ˈɛʁfʊʁt] is another story. I'm perfectly happy to see [ˈkɛʁpn̩] Kerpen next to [ˈɛɐ̯bax] Erbach (Rheingau and Odenwald), reflecting local preferences that yet sound "non-regional" to most German ears ([ˈkɛχpn̩] for Kerpen or [ˈhɛɹbɔɹn] Herborn would be a different story), and I guess J. 'mach' wust feels the same. –Austronesier (talk) 09:51, 2 March 2024 (UTC)[reply]
Wait: don't you believe in the status quo while mach believes in the status quo ante? To mach: was the previous situation all [ʁ] or all [ɐ̯]? With my limited knowledge of Germans speech, the one I hear consistently is [ɐ̯], but of course I'm probably almost always hearing (German) Standard German from worldly, upwardly mobile speakers. Wolfdog (talk) 13:47, 2 March 2024 (UTC)[reply]
Indeed, I think we should restore the status quo ante. The article space has both [ʁ] and [ɐ̯] after short vowels. Both make perfect sense phonetically. As per MOS:IPAINTEGRITY and as per Template:IPA key at the beginning of this guide, there must be integrity between this guide and the transcriptions that link to it, and the guide should not be changed without prior consensus. Since there was no prior consensus on the exclusive use of [ɐ̯] after long vowels, I am changing it back. --mach 🙈🙉🙊 08:33, 3 March 2024 (UTC)[reply]
So it was neither all [ʁ] nor all [ɐ̯], but all [ɐ̯] after long vowels, and [ʁ]/[ɐ̯] after short vowels, broadly speaking. To be clear: if we want to be unambiguous, I prefer "[ɐ̯] after long vowels and in unstressed prefixes; [ʁ] after short vowels". But like 'mach', I believe that both pronunciations are good (and generally considered supraregional), so unless we have a consensus to go into diaphomenic terrain (i.e. consistently transcribing /r/ as ⟨ʁ⟩ after short vowels, and mention the [ɐ̯]-variant in a note), we cannot give exclusive preference of one over the other. –Austronesier (talk) 11:27, 3 March 2024 (UTC)[reply]
I fully agree.
A side note: Transcribing r-vocalization has nothing to do with diaphonemes. Diaphonemes means a lowest-common-denominator analysis of different but overlapping phonemic oppositions. Within the German r sounds, there are no phonemic oppositions, let alone different overlapping phonemic oppositions. There is just one /r/ phoneme. Our guide offers different allophonic transcriptions of that phoneme, [ʁ] or [ɐ̯], and different allophonic transcriptions of the two-phoneme sequence /ər/ (or /ɛr/), [əʁ] or [ɐ]. A diaphonemic analysis is not necessary (or applicable) since all varieties have the same /r/ phoneme. --mach 🙈🙉🙊 13:37, 3 March 2024 (UTC)[reply]
Oh yes, you're right, I should have said "diaphonetic". It only becomes diaphonemic with the Schafscharf merger. :) –Austronesier (talk) 13:51, 3 March 2024 (UTC)[reply]
I'm a little bogged down in semantic satiation. So is the current pronunciation at Arthur Schopenhauer reflective of the consensus at this time: German: [ˈaʁtʊʁ ˈʃoːpn̩haʊɐ] , where /r/ after short ANY vowel is transcribed as [ʁ] but syllabic /ər/ specifically is transcribed as [ɐ]? — Preceding unsigned comment added by Wolfdog (talkcontribs) 22:00, 3 March 2024 (UTC)[reply]
Yes it is, though I would transcribe it with a long [uː] as [ˈartuːr], [ˈaʁtuːɐ̯] or [ˈaɐ̯tuːɐ̯], and that’s also what I hear in the recording. --mach 🙈🙉🙊 06:36, 4 March 2024 (UTC)[reply]