Help talk:IPA/Alemannic German

Notation for length/fortis–lenis

I think we should not take a prescriptivist stance on the notation of length, but allow for different possible notations, especially since the choice of notation implies a certain analysis (it always does, but in this case, this is more obvious). In the current proposal, we are using the length symbol ⟨ː⟩ for sonorants (vowels and consonants), but not for obstruents. This implies that we are analyzing the sonorants as having a length contrast, whereas the obstruents have a fortis–lenis contrast. Other notations use the length symbol for all of these oppositions, thus implying that there is a length contrast throughout the system (for vowels and all kinds of consonants). Yet other notations use one kind of symbolization for vowel length, and another for the opposition in consonants (e.g., doubling), thus implying that the opposition in consonants is not of the same nature as the length distinction in the vowels. There is at least one analysis that distinguishes three degrees of obstruents (Ham, William Hallett (2001). Phonetic and Phonological Aspects of Geminate Timing. New York and London: Routledge.) – lenis, fortis and geminate, though I do not remember what symbols were used (probably /d t tt/ etc.).

What to do? I think it would be best if we strike the examples with ⟨ː⟩ from the table and add a section about length or fortis–lenis contrasts. That section should mention different ways of symbolizing the contrast and point out the signs that are potentially ambiguous (the simple voiceless obstruent signs ⟨p t k f s ʃ x⟩, used for lenes/short consonants in a few analyses, but more commonly for fortes). --mach 🙈🙉🙊 14:19, 2 January 2016 (UTC)

We can remove the long sonorants from the table, I don't see a much of a problem with that. The problem with allowing for different possible notations in this case is what you said - that [p t k f s ʃ x] can denote either fortes or lenes, which almost certainly will be really confusing to laymen. I think this is the case where we should take a prescriptivist stance for the sake of clarity and simplicity. When it comes to spelling, AFAIK, hardly any writing system of Swiss German distinguishes fortis ⟨sch, ch⟩ from lenis ⟨sch, ch⟩, so we can't completely rely on the spelling anyway.

You can, by the way, already transcribe the velars/uvulars, /r/ and the open vowels any way you want or need, which is anything but a prescriptivism.

I don't have much sources that talk about the fortis-lenis issue in Alemannic German besides the Fleischer & Schmid (2006) one, so I'm afraid I won't be of much help with creating a fortis-lenis section. Peter238 (talk) 14:47, 2 January 2016 (UTC)

You have chosen an analysis where sonorants (vowels and consonants) have a length contrast, whereas obstruents have a fortis–lenis contrast. How do you justify this POV?

Also, considering the options, it seems like you have made a poor choice because you have not avoided the ambiguous symbols ⟨p t k f s ʃ x⟩.

Also, in the transcription of Swabian or other dialects that only have lenis consonants, the transcription normally uses ⟨b d ɡ f s ʃ x⟩. Trying to prescribe ⟨b d ɡ v z ʒ ɣ⟩ would be as strange as, say, prescribing a non-rhotic notation for all varieties of English. (Incidently, the vowel symbols will need some expansions for Swabian).

I do not see any reason for taking a prescriptivist stance. You are saying it is for the sake of clarity and simplicity. I think your system does not provide much clarity, since it contains the ambiguous symbols ⟨p t k f s ʃ x⟩. And I think simplicity is not our aim. Our aim should be describing the possible ways of transcribing Alemannic that can reasonably be encountered. We do not require a one-size-fits-it-all system. We can have different approaches for different circumstances. I think pointing out the cases where ambiguity could arise (namely, ⟨p t k f s ʃ x⟩) is much more helpful than inventing a prescriptivist standard. --mach 🙈🙉🙊 15:52, 2 January 2016 (UTC)

I object allowing contradictory transcriptions of the obstruents. The primary goal of this kind of guides is to be of help to laymen. In such situation, allowing contradictory transcriptions is a very bad idea. Even after looking at the spelling nobody would know whether the consonant is fortis or lenis in case of ⟨sch, ch⟩, unless you happen to be a speaker of such dialect and/or a scholar. That's an unacceptable requirement.

We can mention the contradictory transcriptions somewhere in the guide (as I did), but we should use one analysis in our tables, and therefore our transcriptions (which have to match the system used in this guide). And yes, the dialects that use only the lenes should be transcribed with the corresponding lenis symbols (your comparison with non-rhoticity in English is not accurate). I'm not saying these absolutely have to be ⟨b̥, d̥, ɡ̊, ɣ̊, v̥, z̥, ʒ̊⟩, but it's a very logical (and non-OR - see Fleischer & Schmid (2006)) transcription. Why? Because in transcriptions of English, Dutch and Standard German ⟨p t k f s ʃ x⟩ are used for fortes in majority of cases, whereas the corresponding lenes are transcribed ⟨b d ɡ v z ʒ ɣ⟩. I think we should stick to it (although a devoicing diacritic on the lenes is fully appropriate, and I vote for keeping it), especially because it's the normal transcription in case of Standard German (the literary language of the vast majority of Alemannic speakers). Also, you yourself have admitted that ⟨p t k f s ʃ x⟩ is a less common transcription of the lenes in Alemannic.

About the sonorants... how else could you transcribe them if not as I did? Like this [mm, nn, aa]? The meaning is the same in the IPA, it's just looks different.

Let's also not forget that this guide, as well as the corresponding transcriptions are phonetic, rather than phonemic. Peter238 (talk) 16:57, 2 January 2016 (UTC)

Citing Fleischer & Schmid is no good when it comes to transcribing Schwabian. It is as inadequate as if you were citing a British English non-rhotic dictionary for prescribing a non-rhotic transcription of rhotic American English – it is theoretically possible, but it is highly counterintuitive and nobody out there ever does it.

If we are to invent a new prescriptive standard for transcribing Alemannic, then I strongly vote against using any ambiguous symbols. That is, I would recommend using ⟨pː b̥, tː d̥, kː ɡ̊, fː v̥, sː z̥, ʃː ʒ̊, xː ɣ̊, mː m, nː n, …, iː i, ɪː ɪ, …⟩.

I also don’t understand why you are so obsessed about contradictions when you have yourself proposed the contradictory use of ⟨ʁ̥⟩.

The more I think about it, the more it seems to me that inventing a prescriptivist standard for all varieties of Alemannic is a terrible bad idea. It cannot possibly work. For English, it only works because we are assuming that our made-up standard only needs to account for the most common varieties of English – more precisely, American English and British English – and because we assume that a phonemic transcription is good enough, typical readers having enough knowledge of English for filling in their actual pronunciations. For Alemannic, neither assumption works. There is no way for telling what the most common varieties are, and we need phonetic transcriptions since typical readers do not speak any Alemannic variety at all. At best, prescriptivism might work on a dialect-per-dialect basis. An overarching page can only describe the plethora of different conventions that are being used for the different dialects. Linking to the general IPA page may be more appropriate than creating a dedicated page with an artificial made-up standard that blatantly violates NPOV. --mach 🙈🙉🙊 17:26, 2 January 2016 (UTC)

It isn't fair to say that I'm "obsessed" about the contradictory transcriptions of the obstruents, because:

- If I counted that correctly, the symbols for fortis/lenis obstruents make for about 25% of the symbols we're using. That makes this issue even more important than I thought.

- I don't want to repeat myself, but my previous arguments about ⟨sch, ch⟩ and confusing the laymen still apply.

About the use of [ʁ̥]:

- Correct me if I'm wrong, but in the dialect of Basel, /r/ and /ɣ̊/ merge into the latter.

- Fleischer & Schmid (2006) do not say whether the /r/-/ɣ̊/ merger takes place or not when both of these are phonetically lenis voiceless uvular fricatives.

We are not making up any "artificial standard" of pronunciation (nor transcription). All we are doing is listing all of the symbols needed to transcribe Alemannic dialects on Wikipedia, and we're choosing one way of transcribing the obstruents out of necessity, rather than prescriptivism or wanting to violate the WP:NPOV policy.

If I were to assume good faith (which I do), I'd say you're confusing the IPA symbols and what they represent. As far as I remember, that's something that you've already been accused of in the past (see Help:IPA for English and the corresponding archives).

It is a big exagerration for you to say that because some dialects may have somewhat different conventions of transcribing them into the IPA (I'd like to see a proof of that, that is apart from the fortis-lenis issue of course), it is a terribly bad idea for us to have this help page. Even Fleischer & Schmid (2006:245) list as many as 8 ways of transcribing the fortis-lenis distinction for Zurich German alone - yet they choose one of them, stick to it and there's no problem - that's exactly what you did on Bernese German phonology.

One thing we agree on is that we need phonetic, not phonemic transcriptions of Alemannic, but, AFAIK, all of our IPA guides are phonetic, so that's not really an issue.

If Swabian (rather than just IPA convention(s) used to transcribe it) is that distant phonetically from the rest of the Alemannic dialects (AFAIK it's not, but maybe I'm wrong), we could always remove it from this guide and create a separate one specifically for Swabian. I don't see a problem in doing that (we have separate guides for Czech and Slovak, even though these are very similar phonetically), although I'm not knowledgeable enough to create it. Peter238 (talk) 20:23, 2 January 2016 (UTC)

If you want a transcription system that is not confusing to laypeople, then I would strongly advise against transcribing the fricatives as ⟨v̥ f, z̥ s, ʒ̊ ʃ, ɣ̊ x⟩, but rather use something like ⟨f fː, s sː, ʃ ʃː, x xː⟩ instead. It seems that you have been deceived by this very peculiar use of ⟨v̥ z̥ ʒ̊ ɣ̊⟩ yourself, seeing as you have provided inaccurate English correspondents such as van, zip, vision when words like fan, sigh, shy would be more appropriate.

The different ways of representing pairs of sounds imply different POVs (length, gemination, voice, fortis–lenis). The NPOV issue does not arise because you intend to violate NPOV – you most certainly don’t have any such intention. It arises because every choice necessarily implies a POV. If you really want to choose one transcription over the others, then I would like to see a very detailed and careful justification first. Merely copying the choice Fleischer & Schmid have made is by far not good enough. Their choice depends on the Zurich dialect, where consonant sonorants typically do not come in pairs, so adopting a different analysis for obstruents and sonorants makes sense.

A transcription like ⟨tː t⟩ implies length, a transcription like ⟨tt t⟩ implies gemination, a transcription like ⟨t d⟩ implies voice, and a transcription like ⟨t d̥⟩ implies fortis–lenis. And if different transcriptions are used for different classes of sounds (sonorants, fricatives, stops), it implies that these sounds behave differently. Marti (1985), for instance, only analyzes the stops as fortis–lenis pairs, but the fricatives and sonorants as singleton sounds that can be lengthened. This makes sense in Bernese German where “long” consonant sonorants and fricatives only occur medially, while “fortis” stops occur initially as well.

I think it would be much more helpful to readers and editors if we just listed the different symbols that are being used for transcribing Alemannic dialects. You are saying that there is a necessity for choosing one way. I honestly fail to see any such necessity. Why do you assume there is a necessity?

Not all our IPA guides are phonetic – Help:IPA for English is not, and that is the one I have been explicitly referring to. The fact that it is phonemic justifies simplifications. Such simplifications are not appropriate in a phonetic IPA guide. From what I understand, the difference between the Czech and Slovak languages may be similar to the difference between Bernese and Zurich German. Other dialects diverge much more, e.g. Swabian, Alsatian, or Wallis German. Where would you draw the line?

I don’t understand your previous arguments about ⟨sch, ch⟩. Of course ⟨sch ch⟩ do not provide any information about length/fortis–lenis. That is because they are not phonetic transcriptions, but ordinary spellings. But with regards to phonetic transcriptions, ⟨ʃ x⟩ are ambiguous.

To my knowledge, Basel German does not merge /ʀ/ and /x/ (or, if you prefer, /ɣ̊/), except maybe at the ends of words (after voicess consonant?). --mach 🙈🙉🙊 08:40, 3 January 2016 (UTC)

Sorry for not answering earlier. My original message got deleted, and I've been rewriting it. Anyway, I now have to leave my hometown for a couple of days (or maybe a week), so I'll try to finish my response and will probably post it next week. Take care. Peter238 (talk) 07:18, 7 January 2016 (UTC)

/r/-vocalization

Does any of the Alemannic dialects vocalize the /r/ to [ɐ]/[ɐ̯]? If so, it'd be reasonable to include it. Peter238 (talk) 15:18, 2 January 2016 (UTC)

Swabian

• Swabian has two series of diphthongs, traditionally transcribed /ai, au/ and /ei, ou/, or something along those lines. Phonetically, they are [aɪ̯, aʊ̯] and [ʌɪ̯, ʌʊ̯], but I'm not sure about the latter. Is anyone aware of the actual phonetic quality of /ei, ou/?
• [1] mentions [ᵻ, ɜ] among the 'lax' Swabian vowels. We can transcribe the former simply as [ɪ], but what about the latter?
• We should write something about the nasalization in Swabian. Peter238 (talk) 15:23, 2 January 2016 (UTC)

Egge

I don't hear any difference between <Egge> in my own dialect (BS) and the dialect of Berne. The /kk/ is longer in both dialects imho; and some people use /k/ for the fortis sound, whereas /kk/ or /k:/ is certainly more exact phonetically. It really depends on how you transcribe the lenis sound: if you use /k/ then the length of the fortis sound (which is in fact long) has to be indicated of course... Wathiik (talk) 16:24, 7 July 2017 (UTC)

The same applies to /pp/ and /tt/ of course. I object to the transcription <huupe> [ˈhuːpə] 'to hoot' (BS) - the vowel is short here! Hence [hup...] would be more correct, or [hupp...], depending on how you transcribe the fortis/long consonants. The fact that /u:/ /i:/ etc. are usually shortened before fortis/long consonants is one of the most typical features of the dialect of Basel; especially if you compare it to the dialect of Zuerich which always has nlong /u:/ etc. in these cases. For example, the name of the country Switzerland has a short [i] in Basel (not to be confused with the open vowel of course) but always a long vowel in the dialect of Zuerich. Wathiik (talk) 16:31, 7 July 2017 (UTC)

Move discussion in progress

There is a move discussion in progress on Help talk:IPA which affects this page. Please participate on that page and not in this talk page section. Thank you. —RMCD bot 16:15, 15 July 2017 (UTC)

Ideas for improvement

...or mere simplification, which in many cases is an improvement when it comes to guides like this. Their purpose isn't to list every phonetic nuance but to help native speakers of English pronounce any given language/dialect. I've already removed ⟨ɒ, ɒː, ɾ⟩ and rather inappropriate phonetic notes from the guide. Further improvements could include:

• Removing the distinction between velar and uvular affricates and fricatives and simply using ⟨qχ, ʁ̥, χ, χː⟩ for both variants. There seems to be a free variation between them in Zürich and Bern and, according to Alemannic Wikipedia, pretty much everywhere in Switzerland. To my ears, uvulars are prevalent. Is there any dialect that uses the velars to the exclusion of the uvulars? If there isn't, let's remove ⟨kx, ɣ̊, x, xː⟩ from the guide. It's an unnecessary complication and a nuance that an average native speaker of English is incapable of reproducing in a reliable and consistent manner. If ⟨kx, ɣ̊, x, xː⟩ is a more acceptable broad transcription (for whatever reason), then using only this set would be fine too. Either set is fine if we don't have to use both, but I'd rather use the symbols for uvulars (unless, as I said, I'm mistaken and the uvulars are not the prevalent phonetic variant among the Alemannic dialects).
• Merging some instances of Bernese [w] with [u̯]. After all, it forms a diphthong, no?
• Transcribing both [a, aː] and [ɑ, ɑː] with ⟨ɑ, ɑː⟩, unless there are dialects with front [a, aː] (but then... can't we transcribe the front variants with ⟨æ, æː⟩? Are there dialects that contrast open and near-open front unrounded vowels?). This transcription is good enough. It's a bit broad, but it's good enough. On the whole, native speakers of English will prefer more back realizations anyway. Kbb2 (ex. Mr KEBAB) (talk) 15:03, 24 July 2018 (UTC)

Also, is there any difference between [ʁ̥] that varies with [ɣ̊] and [ʁ̥] the rhotic (one of the realizations of /r/)? Kbb2 (ex. Mr KEBAB) (talk) 17:23, 24 July 2018 (UTC)

I honestly never got the point what benefit we would get from reducing the subset of available IPA signs for transcribing the diverse Alemannic dialects. It seems to be a spectacularly futile enterprise. Not only the most central articles about Alemannic dialects like Swiss German adhere to the prescriptivism that has been invented here.

Yes there are dialects that have exclusively velar kx, x, ɣ̊, e.g. the dialects from Bernese Oberland.

No, Bernese German vowelized /l/ cannot simply be described as diphthongs: For one, it can occur after the diphthongs /iə, uə, yə/, and also, it can be geminated like other sonorant consonants.

Using /æ/ and /a/ is a well-attested tradition in the transcription of many dialects. I agree with you that there might be valid reasons for discontinuing some signs. But why would you discontinue precisely the sign that is the most common of them all? I think the opposite approach would be much more convenient: keeping the common signs and discontinuing the oddballs. My honest opinion is that the IPA would be better off without [æ ɪ ʏ ʊ] – I believe they were only introduced because the more appropriate signs [a e ø o] felt too French to the English. The most important point, though, is that this is Wikipedia where we must not invent our own original research but instead stick to what is attested out there. So I advise you publish lots of peer-reviewed articles about Alemannic phonology instead of trying to impose your prescriptivist point of view. --mach 🙈🙉🙊 21:39, 24 July 2018 (UTC)

@J. 'mach' wust: I'm not imposing my point of view, just WP:BOLDLY introducing changes that I deemed necessary and asking questions about further changes. I didn't edit war with anyone here and you didn't revert me even once. So I'm really not sure what is up with the last sentence of your reply. Is there anything wrong with the edits I made to this guide that my edit summaries didn't clarify? If so, what is it?

I think you're using the word prescriptivism and its derivatives too excessively (is it perhaps something similar to WP:IDONTLIKEIT, just worded differently?) You're also confusing narrow transcription with the broad one and treating IPA too literally, when it's perfectly possible for a transcription to be both allophonic and diaphonemic at the same time. I'm sure I'm not the first person to tell you that (there have been similar discussions on Help talk:IPA/English).

Thank you for clarifying my doubts. As I said, I had no intention of changing any of this without being sure that I'm correct. Kbb2 (ex. Mr KEBAB) (talk) 22:10, 24 July 2018 (UTC)

On second thought though, what would be the problem with only using ⟨ɑ, ɑː⟩? If the back variants are predominant and no dialect contrasts open and near-open front unrounded vowels, we could use ⟨ɑ, ɑː⟩ for the back and central variants and ⟨æ, æː⟩ for the front ones. Neither this guide nor transcriptions linking to it are phonemic. Kbb2 (ex. Mr KEBAB) (talk) 22:59, 24 July 2018 (UTC)

There are at least three problems with prescribing [ɑ ɑː] over [a aː]:

1. It goes against a whole body of dialect linguistics that uses [a aː].
2. It goes against the common sense principle of using the simplest sign available, and special signs only if there is a sound reason.
3. Even if the back variants really were predominant – which I doubt – there are dialects where [ɑ ɑː] would be totally inappropriate, e.g. the dialects around the city of St. Gallen. Even by non-linguists, they are known for their blazingly front pronunciation of the *a.*

Also, the reverse question is equally valid: What would be the problem with only using [a aː]? And the answer is the same: It goes against a whole body of dialect linguistics that uses [ɑ ɑː]. I really see no benefit in prescribing one over the other. Uniformity is no benefit. --mach 🙈🙉🙊 06:14, 25 July 2018 (UTC)

@J. 'mach' wust: I'd say using, not prescribing (which is a less neutral term). They'd still be diaphonemic symbols that cover open back rounded, open back unrounded and open central unrounded vowel qualities. We could use ⟨æ, æː⟩ for dialects that have truly front qualities.

I'm dropping the subject of deprecating ⟨a, aː⟩ because arguing over it is counterproductive. We both have our viewpoints which probably won't change. Unless more editors join the discussion, there's no point in continuing it. Kbb2 (ex. Mr KEBAB) (talk) 15:19, 25 July 2018 (UTC)

I have to agree with J. 'mach' wust for including [a aː]. Outside of Switzerland, many dialects that lack an [æ] phoneme also use a more front vowel. I also don't understand why [ɒ, ɒː] should be removed. These are the most common realization of this phoneme in Zurich German and one of the more marked traits of this dialect. We should have all symbols listed here that could reasonably be needed for transcribing words such as the ones in Commons:Category:Alemannic pronunciation (there are actually still more missing, such as the [ɵ] that occurs in many dialects of the Markgräflerland. The purpose of these guides is not to simplify transcriptions to make them pronounceable for English speakers, but to give an overview of the symbols that might be used for transcribing the language on Wikipedia. With a completely unstandardized variety with immense regional variation like Alemannic, this will of course be complicated, but uniformity is no solution. -Terfili (talk) 06:35, 25 July 2018 (UTC)

@Terfili: The JIPA article says that the rounding of Zurich open back vowels is variable. The reason why I removed [ɒ, ɒː] from the guide is in one of the edit summaries.

The purpose of these guides is to not have an overly detailed list of symbols but one that is reasonably narrow so that phonemes and their major allophones can be identified. Are you able to reliably distinguish between a very back, cardinal-like realizations of [ɑ, ɑː] and the rounded [ɒ, ɒː]? Because [ɑ] is not just one vowel, it's a continuum of vowels. It can be fully back, advanced back, open or near-open. See [2] for further explanation. I really don't think we need [ɒ, ɒː] in the guide.

Also, if there are any signs that are missing, feel free to add them. Kbb2 (ex. Mr KEBAB) (talk) 15:15, 25 July 2018 (UTC)

I have felt free to add again the missing signs you had removed. It is obvious from this discussion that you do not have much knowledge of Alemannic German dialects. --mach 🙈🙉🙊 18:33, 25 July 2018 (UTC)

@J. 'mach' wust: Your arrogance has no place here. If you can't discuss in a civil manner then leave. I've provided reasons for my edits in the edit summaries. Kbb2 (ex. Mr KEBAB) (talk) 20:01, 25 July 2018 (UTC)

I apologize. I honestly did not mean to be arrogant. It just seems that you do not have much knowledge about Alemannic German dialects. You have made your bold edits first (which is totally fine), and then you have asked here wether these edits fit with the dialects this guide is meant to cover (which is totally fine, too). There is nothing wrong with not being an expert on Alemannic German dialects. I will add the signs you have removed again. --mach 🙈🙉🙊 16:39, 26 July 2018 (UTC)

@J. 'mach' wust: Apologies accepted, but reverting me after leaving a response that basically has no substance to it (and, again, is patronizing) still isn't appropriate. I had to recheck the source myself to make sure that [ɒ, ɒː] is the predominant realization of the open back vowels in Zurich (which, admittedly, was stated by Terfili though without any sources). This still doesn't mean that we have to include these symbols in the guide. On Help:IPA/Swedish we use ⟨ɑː⟩ that has a variable rounding depending on the region (strong in Gothenburg, light in Stockholm and nonexistant further north). Why can't we use ⟨ɑ, ɑː⟩ the same way in this guide?

There are more problems with the guide:

- Bernese German phonology states that the non-front open vowels are central and uses ⟨a, aː⟩ for them. Yet, in this guide, they're shown as examples of open back vowels and chaotically transcribed with ⟨ɑ, ɑː⟩ and ⟨a, aː⟩ throughout the guide. We need to know their actual backness and then decide on one way of transcribing them.

- Are the dark [ɫ, ɫː] phonemic? If not, we could remove them.

- The additional dialectal vowels mentioned by Terfili still haven't been added to the guide.

It'd be good if other editors offered their insights. Kbb2 (ex. Mr KEBAB) (talk) 02:21, 7 October 2018 (UTC)

I still have no idea why you see any benefit at all in reducing the number of signs. On the other hand, while I do not see any benefits, I see two severe disadvantages:

1. Removing signs will make this help page less helpful to readers when the sign they want to know about is no longer mentioned here.
2. Removing signs might incite over-zelous editors to purge the removed signs from all instances of {{IPA-gsw}} in the article space, which will lead to other editors using {{IPA}} instead.

The reason why I have not added signs like Markgräfler [ɵ] is that my knowledge about Alemannic dialects is limited. I might do some research about the dialects in Germany and Italy and then add dozens of new signs to this guide (I know there are many sounds we do not cover yet), but there are many other things I would do much rather. --mach 🙈🙉🙊 09:48, 7 October 2018 (UTC)

Kbb, with a completely unstandardized language like Alemannic, we should include all types of variation even when it only occurs in some dialects or in free variation.

Let me give an example from the dialects near my home region: all Alemannic dialects have a diphthong phoneme in words such as "guet" ('good'), but the realization varies widely. For instance, in the easternmost part of the Hotzenwald it is [uə̯], and in the more western part [ʊə̯], which is also what is attested for the village of Karsau (which is part of my hometown). The next two villages in my hometown for which there is data from the Dialect Atlas of Southwestern Germany (Adelhausen and Herten) have either [ʉə̯] or [ɵə̯]. Finally, in the very west, towards the French border and in Alsace, the diphthong is completely palatalized to [yə̯]. Clearly this is because there is a continuum of back [uə̯] in the east to palatalized [yə̯] in the west, with all kinds of variation in between. Of course we could simplify this by settling on one or two of these, let's say [uə̯] and [yə̯], and disregarding the rest as similar, or as not occurring exclusively in any dialect. We would then assign certain dialects/speakers to either [uə̯] and [yə̯], with the result of suggesting a sharp border between non-palatalized and palatalized [uə̯, yə̯], when really there is a broad continuum. To me, this would not be very helpful.

This is basically what you are suggesting with the removal of [ɒ, ɒː], which (at least for the long vowel) is a transitional feature between unrounded [ɑː] and rounded and raised [ɔː]/[oː]. The same applies to your question about [ɫ, ɫː]. There are dialects where /l/ is never velarized, dialects where it is velarized allophonically, and then dialects where it is completely vocalized. I have two friends who grew up near each other in such a border area, and one of them vocalizes the /l/ completely, while the other uses a velarized version in syllable-final position some of the time. Including the symbol allows for showing this continuum, where excluding it would suggest a sharp division between dialects with complete vocalization and dialects with regular [l, lː]. --Terfili (talk) 13:29, 7 October 2018 (UTC)

Transcribing the fortis–lenis contrast

@J. 'mach' wust:, the point of this guide is not describing different transcriptions of Alemannic German (at least not in the main section, you can create a separate section below, cf. Help:IPA/Danish) but to help our readers understand what the symbols we use to transcribe Alemannic German mean. We cannot use the same set of symbols ⟨p t k f s ʃ x ~ χ⟩ to denote two different types of sounds. Standardizing your transcriptions is not "artificial", it is an absolute necessity in this case. You are mistaking an in-house convention for WP:OR, which it is not. We want our transcriptions to be intelligible. If the length transcription is better then let's choose that, I don't mind that at all. Sol505000 (talk) 08:15, 22 February 2024 (UTC)

This guide is meant to be useful for readers who click on a transcription in the main namespace and are directed here. And it is meant to be used for all Alemannic dialects, which makes sense because otherwise, we would end up with a multiplication of guides. There is no consensus on the transcription of the obstruent oppositions in Alemannic dialects. We must therefore expect the main namespace to use different systems. For this guide to be useful, it must accomodate different systems.

Some Alemannic varieties have been analyzed to have a three phonetic consonant quantities, lenes, fortes, and extrafortes (cf. William Ham 2001: Phonetic and phonological aspects of geminate timing, p. 61, Urban Zihlmann 2020: Vowel and consonant length in four Alemannic dialects and their influence on the respective varieties of Swiss Standard German, p. 33). We must not set up this guide in a way that rules out such differentiations. --mach 🙈🙉🙊 12:28, 23 February 2024 (UTC)

We must therefore expect the main namespace to use different systems. For this guide to be useful, it must accomodate different systems. is a violation of MOS:IPAINTEGRITY, which 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." "Use the conventions of that key" means that conventions must be established first, one set of symbols cannot have two opposite meanings. WP:OR cannot possibly be an issue here since we're choosing from the systems that are already in use in the literature and we must not confuse symbols not matching the transcription system in the source with OR. See Help:IPA/Danish, where ⟨p t k⟩ have the opposite meaning to what you can find in most sources.

Phonetic or phonemic? It's a big difference. In Zurich, the fortes can also be lengthened/geminated in some positions, but that's a mere allophonic effect. Fleischer & Schmid transcribe the long fortes simply as [tː] etc. We can copy that, assuming we need to transcribe the difference at all. Sol505000 (talk) 23:46, 26 February 2024 (UTC)

Thanks for helping my argumentation by citing MOS:IPAINTEGRITY: “If you wish to change those conventions, bring it up for discussion on the key's talk page.” Yes indeed. Don’t impose your unilateral changes without prior discussion and consensus. “[C]hanging the key so that it no longer conforms to existing transcriptions will confuse readers.” Exactly.

On the other hand, MOS:IPAINTEGRITY does not contain anything at all to back up your intention to eliminate variation. If there is a key, then its conventions should be used. It does not say anything at all about what those conventions should be. In my opinion, normal content policies apply. Over the years we have had a tendency to replace idiosyncratic transcription symbols by more mainstream choices.

Phonetic or phonemic? Obviously, phonetic, see Wikipedia:Manual of Style/Pronunciation: “For foreign-language pronunciations, a phonetic transcription is normally used, with a link to Help:IPA or to various language-specific IPA keys.” --mach 🙈🙉🙊 06:21, 27 February 2024 (UTC)