Jump to content

Help talk:IPA/Slovene

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Dialectal vowels and vowel length

[edit]

Slovene alphabet says that the graphemes ö and ü (and by implication, front rounded vowels) occur in some dialects. Danilo Türk is mentioned. On his article, an IPA pronunciation is given which links here: [tyrk]. The dialectal sound [y] is not covered by this guide. Something needs to be done. (suoı̣ʇnqı̣ɹʇuoɔ · ʞlɐʇ) nɯnuı̣ɥԀ 18:02, 7 July 2014 (UTC)[reply]

I've just added the short ü. Give me an example of ö (preferably short) and I'll add it as well. Peter238 (talk) 12:36, 29 August 2014 (UTC)[reply]
Whatever, I've added it anyway. It still needs an example word though... — Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 08:17, 30 November 2014 (UTC)[reply]
You can use the surname Šömen as an example for [ø]. However, I think it would be better described as an [ø:] (the same goes for the [y:] in Türk), because stressed non-final vowels are always long (except for ə). It may also be better to move these two sounds in a separate, dialectal-only table. 78.153.35.53 (talk) 14:29, 26 November 2015 (UTC)[reply]
We should remove the artificial short-long distinction altogether. It's absent from modern Standard Slovene. The only issue is that the old /á-áː/ distinction has become /ɐ́-á/ distinction, and I know of only two sources that mention it. Also, which tone does the Šömen surname use? Peter238 (talk) 06:45, 14 December 2015 (UTC)[reply]
Standard reference works (e.g., Herrity, Toporišič, SP2001) support the short-long distinction, and so it should be retained. Doremo (talk) 07:03, 14 December 2015 (UTC)[reply]
Short-long distinction is not absent, it is just not arbitrary. I think it should be retained. Since "ö" and "ü" are used in the North-East only, there is no tone distinction (i.e. all vowels have a falling tone).78.153.35.53 (talk) 23:06, 20 December 2015 (UTC)[reply]
Herrity and SP2001 seem to be prescriptive rather than descriptive in this regard, and apparently so is Toporišič. Both of you should read the sources for the note number 5, especially the last couple of sentences from the last page of "On the vowel system in present-day Slovene". Regarding the front rounded vowels - so it means that they're kept falling in Ljubljana? AFAIK, some speakers from Ljubljana make conscious effort to correctly pronounce the front rounded vowels. Peter238 (talk) 23:25, 20 December 2015 (UTC)[reply]
78.153.35.53, I don't think that you mean to say that vowel length in Standard Slovene is 'not arbitrary' but rather that it's optional. Mr KEBAB (talk) 09:45, 1 December 2017 (UTC)[reply]
@Doremo: Just because they distinguish long vowels from the short ones in transcription it doesn't mean that they state that this distincition is in any way important or common in Standard Slovene. There's plenty of evidence to support the fact that vowel length is allophonic in Slovene (see Slovene phonology) and as such it shouldn't be transcribed on WP. It's a phonetic detail. Mr KEBAB (talk) 09:09, 28 November 2017 (UTC)[reply]
If standard references for the language (Herrity, SP2001, Toporišič, etc.) mark vowel quantity, then WP should do likewise. Doremo (talk) 09:14, 28 November 2017 (UTC)[reply]
@Doremo: But why when there's evidence that it's not phonemic (except for /á/ and /áː/, which we can transcribe [ɐ́, á])? It wouldn't be OR to omit vowel length as there are sources that state it explicitly that it's not phonemic. One of them is the Handbook of the IPA. (Redacted) Mr KEBAB (talk) 09:21, 28 November 2017 (UTC)[reply]
If you'd like to promote something that differs from the standard sources (you could add Greenberg, Derbyshire, and many others to the aforementioned ones), you're welcome to publish a paper on it. In time, if original research published elsewhere becomes the standard in the field, then those new sources become the standard sources. It's not the purpose of WP to be at the forefront of research and promote new paradigms that haven't yet become established in standard reference works. Of course, competing approaches/theories should be mentioned. Doremo (talk) 09:29, 28 November 2017 (UTC)[reply]
@Doremo: What makes you call Srebot-Rejec (1988) and especially Šuštaršič et al. (1999) non-standard references or original research?! What you're writing is extremely puzzling. Can you provide a source that proves (as opposed to states) that vowel length is phonemic in Slovene? Mr KEBAB (talk) 09:34, 28 November 2017 (UTC)[reply]
You're welcome to argue that Srebot-Rejec (1988) and Šuštaršič et al. (1999) are more authoritative than SP2001 for the language. Recall that Standard Slovene is an artificial language and is regulated by SP2001. Doremo (talk) 09:40, 28 November 2017 (UTC)[reply]
@Doremo: That's not what I'm arguing. I'm arguing that just because you have a list of references that transcribe length in Slovene it doesn't mean that they argue that it's phonemic (they may be just doing it out of tradition) nor does it mean that you get to dismiss other sources, especially a chapter in the Handbook of the International Phonetic Association. It and Srebot-Rejec (1988) are just as important and authoritative (whathever that means to you) as SP2001.
And yet people speak it. What is the prestigious dialect of Ljubljana if not spoken Standard Slovene?
Does SP2001 state (with proof) that Slovene has phonemic vowel length? Mr KEBAB (talk) 09:48, 28 November 2017 (UTC)[reply]
The residents of Ljubljana do not generally speak Standard Slovene; some will approximate it more or less closely as the situation becomes increasingly formal. SP2001 is not a linguistic treatise. It is a normative guide for an artificially regulated language. The phonetic description (§ 618) states that the language has long and short vowels. Doremo (talk) 09:56, 28 November 2017 (UTC)[reply]
@Doremo: Ok, but we're talking only about Standard Slovene. When they do speak it, and it's not mixed with the local dialect, the pronunciation they use closely approximates (or is identical with) what is described as Standard Slovene in the literature, probably minus the phonemic vowel lenght. Does that sound right to you?
I think that proves that we shouldn't consider SP2001 as a reliable source as far as the phonemicity of vowel length is concerned. Srebot-Rejec (1988) and Šuštaršič et al. (1999) make it clear that the traditional approach is to treat Slovene as having phonemic vowel length. What we should be concerned with is the actual pronunciation of Standard Slovene. If you have any sources that disprove the fact/assertion (however you want to treat it) that vowel length isn't phonemic in Standard Slovene, I'm all ears.
I'll try to check the sources you've listed, the ones listed in Slovene phonology and I'll also try to find more sources that discuss the phonemicity of vowel length in Standard Slovene. I also apologize for my tone, which was a bit combative. Mr KEBAB (talk) 10:27, 28 November 2017 (UTC)[reply]
You're welcome to ask here about the reliability of SP2001 (or any of its content) for Slovene. Doremo (talk) 10:46, 28 November 2017 (UTC)[reply]
@Doremo: I can judge it myself based on what you've written.
Another thing to consider is: who is this guide for? The answer is: laymen. If the vowel length isn't phonemic (and there's evidence that it isn't), we shouldn't transcribe it for the sake of laymen. There aren't even that many minimal pairs based on length to begin with!
By the way, why do you not include tone in your transcriptions of Slovene if you consider SP2001 to be the most reliable source? They do transcribe tone. You include the non-phonemic sound [i̯] which is the same as [j], you include vowel length but not tone. Why? Mr KEBAB (talk) 11:02, 28 November 2017 (UTC)[reply]
Wikipedia presumably presents information for the sake of laymen (i.e., it is not restricted to specialist readers). Most of the transcriptions that I have added come from sources that do not include tone in the transcriptions (generalist works generally view tone as less significant than vowel quality and quantity). I included [i̯] as parallel to [u̯] in offglides. Doremo (talk) 11:30, 28 November 2017 (UTC)[reply]
@Doremo: It's true in the case of this guide, but not necessarily in the case of other articles. For instance, on English phonology we can't help but use the IPA symbols, which are alien to the majority of readers. Here, we could use tonal orthography instead. Then again, there are millions of people that don't know it just as they don't know the IPA.
Either way, I'm dropping the subject. Jurgec (2005), cited in the Slovene phonology article says that length is contrastive, so we have conflicting reports. If someone wants to do a more thorough research on this issue, be my guest. For now, I see no reason to remove the length marks.
I don't know whether we want to include [i̯] just to parallel [u̯]. The latter is an allophone of /ʋ/ and it's distinct phonetically, but the former isn't phonetically distinct from [j], which is the main allophone of /j/. I think it's an unnecessary complication. Mr KEBAB (talk) 13:04, 6 December 2017 (UTC)[reply]
Why does it matter whether the distinction is phonemic? There is plenty of information in these language-specific IPA help guides that is not phonemic. For example, Help:IPA/Russian lists 19 vowels when the language only has five or six vowel phonemes. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 22:17, 8 December 2017 (UTC)[reply]
@Aeusoes1: In this case, it matters a lot because sources that state it's not phonemic say that stressed syllables are long, whereas the unstressed ones are short. As the traditional short-long distinction applies only to stressed syllables, I think you can imagine the confusion that transcribing all stressed syllables as long would cause. The anon above seems to be confused about that as well.
As the Handbook of the IPA says, transcribing length in Slovene is completely redundant (well, as long as you don't consider it to be phonemic - more on that below). In Greek, stressed syllables are also longer than the unstressed ones, yet nobody is arguing for transcribing this (completely predictable) allophonic length, as the presence of the stress mark is enough. In tonal transcriptions we also have tone marks, so stressed syllables are immediately obvious to the reader.
Plus, it's not quite as simple as saying that stressed syllables are long, whereas the unstressed ones are short. I did say that above, but it's an oversimplification. I can't find the exact page now, but somewhere in Srebot-Rejec (1988) there's a rather long list of things that influence length (open syllable vs. closed syllable, the type of the following consonant, etc.) She even states that there are instances of unstressed vowels being longer than the stressed ones.
At this point I'm against removing the length marks, as we have conflicting reports. Jurgec (2005), cited in Slovene phonology states that length is phonemic.
Or are you actually talking about [i̯]? The problem I have with it is that it's not distinct from [j]. Would you want to transcribe the same sound with two different symbols? Mr KEBAB (talk) 08:10, 9 December 2017 (UTC)[reply]
I was referring to the issue of vowel length. The [j]-[i̯] thing seems unnecessarily narrow. If the distribution of long vowels is not as simple as "only stressed syllables are long" then it's not completely redundant.
Just because something is fully predictable doesn't mean it isn't worth transcribing here. If it were just a matter of stress, I could see that. Is it the case that it's a distinction now lost? — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 14:25, 9 December 2017 (UTC)[reply]
@Aeusoes1: I don't feel comfortable discussing Srebot-Rejec's findings as I don't fully understand the study. If you have time and energy, you should read it yourself and tell us what exactly she's saying. I know that she's saying that vowel length is no longer distinctive, and objects of her study (judged to be speakers of 'pure' Standard Slovene) either can't distinguish length at all, or do so inconsistently.
There's another problem: it's the only study I'm aware of that discusses this. The Handbook of the IPA actually directly references her as far as the issue of vowel length is concerned, which might be a bit of an echo chamber. Another modern study, Jurgec (2005) actually says that vowel length is phonemic (and it's actually much shorter than Srebot-Rejec's paper). I don't think I can reliably judge which study is more accurate, or whether we should do that at all. Maybe there are some published in Slovene or Serbo-Croatian? @Doremo and No such user: are you aware of any?
When I was learning basic Slovene, the first thing that struck me was the discrepancy between vowel length as transcribed in dictionaries and its actual realization in words. So there must be something to Srebot-Rejec's findings, I'm sure of that. Mr KEBAB (talk) 15:03, 9 December 2017 (UTC)[reply]
In my experience (standard Slovenian reference works, current transcriptions by linguists working at SAZU, and North American / British standard works on Slavic languages), vowel length is always indicated in representation of Slovenian. It may indeed be predictable based on other features being marked, although non-specialists will not be able to make that prediction. In any case, whether phonemic or phonetic, there is a strong tradition (and continued practice) of vowel length being a highly salient feature in transcriptions, and this is a strong argument in favor of retaining such marking in the transcriptions. Doremo (talk) 15:16, 9 December 2017 (UTC)[reply]
Thanks. What we need though are studies that prove or disprove that vowel length has a phonemic function in Slovene. Mr KEBAB (talk) 15:22, 9 December 2017 (UTC)[reply]
But we don't need that. Phonemicity is not important here. If, as Doremo states, the feature is salient and unpredictable for nonspecialists, it's worthy of inclusion in our transcriptions. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 17:09, 9 December 2017 (UTC)[reply]
@Aeusoes1: That is precisely the problem: there's conflicting information from different sources - read above. I'm not sure why we keep coming back to discussing the same arguments over and over again. Especially given the fact that I've already dropped the subject of vowel length a few days ago due to exactly that: conflicting data from sources. As nobody else wants to remove the length marks (not even me now), what's the problem exactly? Mr KEBAB (talk) 17:23, 9 December 2017 (UTC)[reply]
All right! Consensus through exhaustion! — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 23:30, 9 December 2017 (UTC)[reply]
@Aeusoes1: If you want to put it that way :P Mr KEBAB (talk) 06:29, 10 December 2017 (UTC)[reply]

@Doremo and Aeusoes1: I still think that we should drop the length marks (and change a to ɐ wherever appropriate). I hope I won't sound like a broken record, I just want to get this off my chest. This is the last time I'm bringing this up and I'll accept any outcome of this discussion.

The fact that I don't understand Srebot-Rejec's study is irrelevant to the fact that in this paper she basically summarizes her study in a way that leaves no doubt as to what its conclusion is: that the majority of speakers of Standard Slovene (SS) are incapable of consistently producing correct vowel length wherever the prescriptive standard of pronunciation requires them to do so. This is backed up by at least three sources on Slovene phonology (the chapter on Slovene in the Handbook of the IPA and Peter Jurgec's papers), so this isn't a WP:FRINGE viewpoint. Guides like this are meant to help non-native speakers sound as close to the natives of the language with as little effort as reasonably possible. Dropping the artificial distinction between short and long vowels will surely help them with that.

AFAICS, Jurgec (2011:244) states clearly that Slovene doesn't have phonemic vowel length and quotes multiple scholars that back up that claim. I think we can consider that paper to be a bit more relevant than Jurgec (2005).

First of all, diphthongs. In the older prescriptive standard of Slovene, there's a supposed distinction between [ɛi̯, ii̯, iu̯, ɔi̯, ɔu̯] (=[ɛj, ij, iw, ɔj, ɔw]) on one hand and [ɛːi̯, iːi̯, iːu̯, ɔːi̯, ɔːu̯] (=[ɛːj, iːj, iːw, ɔːj, ɔːw]) on the other (maybe there are a few more contrasts - that I haven't been able to figure out). Since vowel length clearly isn't phonemic in Slovene, what sense does it make to force (well, kind of) our readers to distinguish between diphthongs that are pronounced the same by the vast majority of native speakers? Proper production of vowel length is hard to learn for an average native speaker of English and there's no reason he should do that when it's actually no longer phonemic. Not to mention that [ɛi̯] also contrasts with [eːi̯] and [ɔu̯] also contrasts with [oːu̯] (at least according to the prescriptive pronunciation - according to the Handbook of the IPA, both of these pairs are phonetically [ei̯] and [ou̯] with somewhat lowered starting points). The latter also contrast with [əu̯].

These contrasts are very hard to learn for an average native speaker of English (let's not forget that speakers of other languages also read Wikipedia) and we shouldn't add to the difficulty without a good reason. A mere note under the list of symbols in this guide is, IMO, not good enough. I think that there's plenty of evidence that an average reader of Wikipedia (if he bothers to learn the IPA at all) will treat our transcriptions in a rather literal manner, attempting to produce vowel length in Slovene in a prescriptive manner. But he won't hear it in recordings of actual Slovene speech, and the transcription system that we use here may actually prevent him from hearing the [a–ɐ] contrast - see below.

Speakers who've mastered the [ɛ–e] and [ɔ–o] contrasts and apply them consistently in, for example, German and Portuguese pronunciations can safely transfer that pronunciation to Slovene, with their exact phonetic length being irrelevant. There's no need to differentiate between those and two additional vowels [ɛː] and [ɔː]. There are many challenging aspects to Slovene pronunciations and our readers should focus on those which will actually make them sound more native (like stressing their schwa and not let it lower to [ɐ], a phone which is perceived as /a/ by native speakers of Slovene).

The short /a/ isn't always raised to [ɐ]. That phone appears only in closed word-final syllables. In other syllables an open [a] appears and so that distribution isn't the same as that of traditional /a/. But because speakers are no longer able to consistently realize vowel length where they should, because [ɐ] is consistently used by speakers of SS and because [a] can appear in the same positions as [ɐ] (compare čas [ˈtʃɐs] vs. pas [ˈpas]) there are multiple reasons to consider the [ɐ–a] contrast to be phonemic (and a remnant of the former length contrast). If you use a short central [a] where [ɐ] is expected (or vice versa) you actually risk sounding non-native because the traditional /aː/ may or may not be longer than the raised [ɐ] (so this isn't very similar to the /a–aː/ contrast in Standard German). [a] is also used in all unstressed positions which violates traditional phonotactics (specifically the rule that says that long vowels can appear only in stressed syllables). See Jurgec (2011:260) for more info (if you can read Slovene - I have some troubles with that). Transcribing the central vowels in words like čas [ˈtʃɐs] and časa [ˈtʃasa] with the same a symbol is, IMO, a rather clear mistake.

AFAICS, the only legitimate reason to show the vowel length in Slovene transcriptions would be to indicate which vowels are reduced to schwa in regional accents (that is, if the distribution of short and long vowels in prescriptive SS and dialects is more or less the same). But that's not good enough to me. Again, the purpose of this guide is to teach non-native speakers of Slovene to pronounce the language in a fairly native way with as little effort as possible and there's lots of evidence that vowel length isn't phonemic in SS. In the discussion above some of you also mixed up phonemic vowel length as described in papers on the prescriptive standard and phonetic vowel length as described in e.g. the Handbook of the IPA (the authors of the chapter on Slovene simply state that stressed vowels are longer than the unstressed ones. But they also say that transcribing both stress and vowel length is redundant and I agree with them).

The Handbook of the IPA also says that Slovene diphthongs occur without salient lenghtening of the first element. The fact that stressed vowels are longer than the unstressed ones isn't anything surprising and to me is just another reason to drop the length marks (I don't think that the schwa in Slovene pes is long enough to warrant transcribing it [əː], as the Handbook does. That word doesn't sound to me like RP English purse but more like puss as pronounced by a Welshman). I think we should be interested primarily in the way SS is spoken, not how some sources state it should be spoken (even if those sources are describing the traditional prescriptive pronunciation).

Vowel length isn't a salient feature of Slovene if it's non-phonemic (it might be a salient feature of transcriptions, but that's an entirely different thing). It's also predictable - stressed syllables are longer than the unstressed ones, as in pretty much every language. The supposed minimal pairs such as [ˈsit–ˈsiːt] aren't systematically distinguished by speakers of SS. Dropping the length mark will also help native speakers of SS transcribe their language more easily, without having to check the dictionary that often.

This isn't akin to the (now deprecated) distinction between /ɔər/ and /ɔːr/ in Help:IPA/English. The percentage of native speakers of English capable of making that distinction is much higher than the percentage of native speakers of SS that produce vowel length in a natural and consistent manner. We don't need to repeat mistakes of some of the sources and teach an artifical distinction to our readers without covering an actual distinction in height between the open central vowels.

Pinging @Kwamikagami, LiliCharlie, Nardog, and Phinumu: so that more editors join the discussion. Also pinging @Hominilupoid:. Kbb2 (ex. Mr KEBAB) (talk) 04:16, 24 January 2019 (UTC)[reply]

I strongly disagree. The usual Slovenian references works (dictionaries, the normative guide, and encyclopedias) and most grammars written in English and Slovenian all mark vowel length and do not use ɐ. Both suggestions would result in less clarity for nearly every WP user. Doremo (talk) 04:21, 24 January 2019 (UTC)[reply]
@Doremo: But what are you basing your disagreement on? What I'm talking about is the actual spoken standard Slovene. I think you're using the phrase "less clarity" to mean "lack of adherence to prescriptive sources". The rules for transcribing [ɐ] are very simple - you use it in closed word-final syllables for traditional /a/. Elsewhere you just use [a].
I find it unfortunate that you've disregarded most of my post. Kbb2 (ex. Mr KEBAB) (talk) 06:03, 24 January 2019 (UTC)[reply]
My disagreement is based on the notations provided in the usual Slovenian references works (dictionaries, the normative guide, and encyclopedias) and most grammars written in English and Slovenian. If [ɐ] can simply be predictably "calculated" from /a/ as a positional variant, then /a/ satisfies both those seeking phonetic detail as well as the majority of WP readers, for whom /a/ is clear. Doremo (talk) 07:22, 24 January 2019 (UTC)[reply]
@Doremo: Yes, but that assumes that the prescriptive sources are correct in stating that /i, u, ɛ, ɔ/ contrast short and long forms. Multiple other sources have proven otherwise. The short open central (or open front, whatever - this is probably variable) [a] can stand for both traditional /a/ and traditional /aː/; in fact, it's probably more prone to be heard as the latter and the one using [a] where [ɐ] is called for will sound non-native (not terribly non-native, but still). That's my conclusion anyway from what I've read. Kbb2 (ex. Mr KEBAB) (talk) 10:54, 24 January 2019 (UTC)[reply]

Doremo, we know you disagree. But you keep saying the same thing over and over, as if doing so makes you right. It's rather annoying.

I don't know Slovene, but if there are no minimal pairs in actual speech (not the prescriptive grammar), and the phonetic length of actual speech can be predicted fairly easily (stressed open syllables, or before voiced consonants, or whatever), then IMO it would be best not to specify length in the transcription. We could make a note about allophonic length in a footnote of the IPA key.

I'm generally in favor of retaining distinctions even if people don't always make them, unless they're predictable from orthography. I was opposed to conflating /ɔər/ and /ɔːr/ in English, and am still annoyed that we replaced reduced /ᵻ/ with /ɪ/, which to me seems as silly as replacing /ə/ with /ʌ/. But I'm not seeing good reason to transcribe Slovene words with length. — kwami (talk) 08:46, 24 January 2019 (UTC)[reply]

This raises the question of whether the Slovenian transcriptions are intended to reflect the proscribed standard pronunciation of the language (I believe they are), which stipulates length distinctions, or to reflect actual spoken production in one of any number of dialects (which one?). If the choice is to reflect actual speech, then a host of other changes should also be made, including very predictable syncope, apocope, /lj/ and /nj/ → /l/ and /n/, etc. Doremo (talk) 09:49, 24 January 2019 (UTC)[reply]
@Doremo: In my understanding they should reflect the actual speech of educated speakers from Ljubljana (whether we include tone in that or not is another story - see the discussion at the very bottom of this page). Kbb2 (ex. Mr KEBAB) (talk) 10:57, 24 January 2019 (UTC)[reply]
The actual speech of educated speakers from Ljubljana does not correspond to standard Slovenian. It is rich in non-standard features. Doremo (talk) 11:04, 24 January 2019 (UTC)[reply]
@Doremo: Can you name a few of them? Also, does that think that the Handbook of the IPA doesn't describe Standard Slovene (even though it claims that it does)?
Also, lack of vowel length isn't just a regional feature of the Ljubljana speech. Srebot-Rejec states clearly that the majority of native speakers of Slovene (so no matter where they come from) are incapable of producing it with any consistency. Kbb2 (ex. Mr KEBAB) (talk) 11:25, 24 January 2019 (UTC)[reply]
Educated speakers from Ljubljana regularly use predictable syncope (palicapalca), apocope (menimen), /lj/ and /nj/ → /l/ and /n/ (LjubljanaLublana), and selected adjective palatalization (takegatazga) in their actual speech, among other things. These are features regularly heard in the speech of university professors. They are not features of standard Slovenian. Doremo (talk) 11:29, 24 January 2019 (UTC)[reply]
@Doremo: Fair enough. I should've said something more like Standard Slovene as heard on national television and radio. Kbb2 (ex. Mr KEBAB) (talk) 13:59, 25 January 2019 (UTC)[reply]
I've heard all of these nonstandard features produced by educated speakers on Slovenian television and radio. Doremo (talk) 14:05, 25 January 2019 (UTC)[reply]
@Doremo: You can check the sources I've mentioned to see how they describe the variety they analyze. Pointing to them is the best I can do at the moment. Kbb2 (ex. Mr KEBAB) (talk) 14:11, 25 January 2019 (UTC)[reply]
As I've mentioned earlier, you're also welcome to ask here about the reliability of SP2001 (or any of its content) for standard Slovenian, including whether vowel length (as stipulated in SP2001) is relevant in the language. Doremo (talk) 09:53, 24 January 2019 (UTC)[reply]
@Doremo: I have multiple other sources that say that it's not. Shouldn't that be enough? I think it's fairly common knowledge that dictionaries occupy themselves with how their authors think X language should sound. Das Aussprachewörterbuch by Duden is a great example of that - in 2015, they dropped a distinction between [ɛ̃] and [œ̃] in their transcriptions even though native speakers of German very rarely use an unrounded vowel for [œ̃].
The fact that they still mark vowel length even though multiple papers have been published that disprove its existence in Standard Slovene might have to do with two things: first, dialects. The former short vowels tend to reduce to a schwa in regional Slovene. This is where they may be getting their idea that the vowel is short or long in any given word from. The second reason is mere tradition: they use a set of symbols they'd have to modify and they're too lazy to do so, or they think that it'd be too confusing and that the system is good enough as it is and native speakers will figure out that vowel length is irrelevant anyway. Also, the current contrast between [a] and [ɐ] serves as an additional cue as to whether the vowel is phonemically /aː/ or /a/. But that doesn't work in minimal pairs like čas [ˈtʃɐs] vs. časa [ˈtʃasa]. According to this logic, the latter word would have to be analyzed as /ˈtʃaːsaː/ and that violates traditional phonotactics (specifically, the rule that says that no long vowel can occur in an unstressed syllable). That rule, in fact, would have to be violated thousands of times. Kbb2 (ex. Mr KEBAB) (talk) 10:37, 24 January 2019 (UTC)[reply]
@Doremo: I guess časa could as plausibly be analyzed as /ˈtʃaːsA/, with the /A/ diaphoneme that represents neutralization between /a/ and /ɐ/. But I've never seen such an analysis in reputable sources, including those who write the unstressed mid vowels with /E/ and /O/. As far as I can see, the traditional transcription of časa would be /ˈtʃaːsa/. Kbb2 (ex. Mr KEBAB) (talk) 13:56, 25 January 2019 (UTC)[reply]
This also reminds me of Stankiewicz's proposal for revised accentuation of Serbo-Croatian transcriptions by eliminating (predictable) pitch notation after stress is marked on what are realized as non-stressed syllables. It made good sense as something that a linguist could easily "calculate" from the transcription, but the proposed transcription system was unintuitive for normal mortals (both native speakers and learners) and, as far as I'm aware, it was never adopted in any standard description of the language. Not marking Slovenian vowel length because it can be calculated by a linguist seems equally unintuitive to me. Doremo (talk) 10:26, 24 January 2019 (UTC)[reply]
@Doremo: I want not to mark it because it can be convincingly produced by anyone just by looking at the transcription without length marks, not because it can only be calculated by linguists (which isn't true). You seem simply not to believe multiple sources which state that native speakers are unable to produce this supposed vowel length in a way prescriptive dictionaries tell them to. Kbb2 (ex. Mr KEBAB) (talk) 10:37, 24 January 2019 (UTC)[reply]
In that case, it seems that the standard reference works (with vowel length) and various other reanalyses (without vowel length) disagree. Give the lack of agreement among sources, matters should be left alone, using the system already in place. A commentary or note mentioning alternative analyses is of course very welcome. Doremo (talk) 11:10, 24 January 2019 (UTC)[reply]
@Doremo: I think you're conflating the lack of agreement of the prescriptive sources with the descriptive ones with the lack of agreement (which, AFAICS, there's not) just among the descriptive ones. All of the descriptive sources that I've cited have come to the same conclusion: that vowel length is no longer phonemic in SS. The only disagreement among these sources is regarding the existence of [ɐ]; some of the sources just don't mention it. But that's no surprise, given the fact that only vowel length was the focus of these studies, not the exact phonetic realization of Slovene vowels.
I've listed some of the most likely reasons for which prescriptive sources keep describing SS as if it featured phonemic vowel length. Perhaps there are other reasons, I'm not sure of that. Kbb2 (ex. Mr KEBAB) (talk) 11:25, 24 January 2019 (UTC)[reply]
@Doremo: Of course there's another reason: consistency with the way Serbo-Croatian is transcribed in Serbo-Croatian dictionaries. The so-called high-tone vowels correspond to the falling vowels in SC and the low-tone ones to the rising ones in SC. The adherence to the traditional transcription of SC is so strong that the rising schwa is transcribed with ⟨ə̀⟩, with a diacritic corresponding to the short rising vowels in SC (used only with the schwa in Slovene), when in fact it could just as plausibly be transcribed with ⟨ə́⟩ and its falling counterpart with ⟨ə̑⟩ as /ə/ behaves like a long vowel as far as tones are concerned (in other aspects it seems to behave like a short vowel). Jurgec (2011) says that (if I'm not mistaken, my Slovene isn't good) phonemic vowel length started to disappear about 100 years ago. Slovene language says that Standard Slovene was formed two centuries before that.
AFAICS there's a strong need to reform that system. Slovene language#Diacritics lists some of the problems that could be easily fixed if dictionaries ceased to transcribe vowel length. The difference between /ɛ/ and /ə/ is phonemic and yet it's not always the case that they're differentiated. This is a bad practice, especially for language learners. Kbb2 (ex. Mr KEBAB) (talk) 13:52, 25 January 2019 (UTC)[reply]

I believe our transcriptions should reflect contemporary "Standard Slovene as spoken by educated speakers in Slovenia" which is the variety described by Šuštaršič et al. on page 135ff. of the Handbook of the IPA. Love —LiliCharlie (talk) 11:39, 24 January 2019 (UTC)[reply]

I'm unwilling to consider the normative guide (SP), the academy dictionary (SSKJ), and the work of Jože Toporišič, Marko Snoj, Metka Furlan, Mark Greenberg, Peter Herrity, and many others collectively wrong and unreliable at this point. If the analysis without vowel length is correct, eventually it will make its way into SP and SSKJ, and then there will be no controversy with WP adopting that approach. Doremo (talk) 14:03, 25 January 2019 (UTC)[reply]

@Doremo: I'd say that their analysis is rooted in prescriptivism and is simply outdated, nothing more. The fact that spoken SS lacks phonemic vowel length is not a WP:FRINGE viewpoint. Rather, it's supported by multiple sources published throughout the 20th century and later, one of which is the Handbook of the IPA. If anything, I'd expect the Handbook to get the phonemicity of vowel length right. It's a more serious issue than whether /t/ in e.g. French is laminal denti-alveolar or laminal alveolar, or whether the typical realization of /ə/ in e.g. German is mid or close-mid.
I see no controversy with adopting it. We can mention both analyses on Slovene phonology, Slovene language and Slovene alphabet. There's no problem with that. What I do (and at least two other editors) have a problem with is prescribing an actually outdated and much more complicated system to our readers. They have no reason to differentiate /i, u, ɛ, ɔ/ from /iː, uː, ɛː, ɔː/ if it's a distinction they won't hear from the natives themselves (and worse - the natives are likely not to appreciate the effort anyway since they're incapable of producing vowel length with consistency). Kbb2 (ex. Mr KEBAB) (talk) 14:41, 25 January 2019 (UTC)[reply]
I disagree and really have no more to contribute to the topic. Doremo (talk) 15:43, 25 January 2019 (UTC)[reply]
@Doremo: That's fair enough. Given that I have multiple descriptive sources to back up the removal and support of two more editors, if nobody else's gonna object I'm going to remove the length marks soon (probably along with the tones, as we discussed below) and add ɐ to the list of symbols. We can tweak Slovene phonology, Slovene language and perhaps Slovene alphabet so that they cover both viewpoints - descriptive and prescriptive. But here, only the descriptive viewpoint should be taken into account. Vowel length can be mentioned in a footnote. Kbb2 (ex. Mr KEBAB) (talk) 16:13, 25 January 2019 (UTC)[reply]
If so, I recommend that you add a clear note to the relevant pages that your changes are based on deeming the normative guide (SP), the academy dictionary (SSKJ), and most contemporary scholars of Slovenian as illegitimate or inaccurate. Doremo (talk) 17:19, 25 January 2019 (UTC)[reply]
@Doremo: The fact that these sources inaccurately describe contemporary spoken SS (at least regarding vowel length, I'm sure they get most of other stuff right) will be clear from the context (also, I don't know about most). If you have any doubts regarding what I've said on this page you can simply re-read my messages (at least those from this year). You can still answer any of them and we'll continue the discussion. It could as well be the case that I wasn't clear enough and some of what I've said requires further explanation. Kbb2 (ex. Mr KEBAB) (talk) 17:25, 25 January 2019 (UTC)[reply]
So own your position and clearly state that you deem the normative guide (SP), the academy dictionary (SSKJ), and most contemporary scholars of Slovenian as illegitimate or inaccurate. Or take it up here if you are confident. Doremo (talk) 18:37, 25 January 2019 (UTC)[reply]
@Doremo: I am owning my position. I also don't like where this is going. I'd be happy to answer any questions regarding my previous messages if you have any. Kbb2 (ex. Mr KEBAB) (talk) 18:43, 25 January 2019 (UTC)[reply]
@Doremo: Greenberg (2006:21) states that All words in Slovene that are not clitics have at least one prominent syllable. Prominence islexically determined, i.e., it is not predictable, so it is a feature of the word which must be learned along with the segmental sequence and the meaning. Prominence is marked in at least two broadly different ways depending on which dialect area a Slovene speaker comes from. Speakers from Ljubljana, the Upper and Lower Carniolan dialects, or the Carinthian dialect may have a pitch-accent (“tonemic”) system. Speakers from other areas generally have a dynamic stress (“stress”) system. In both the tonemic and stress systems, vestiges of a quantity opposition in stressed syllables may still be realized in many speakers’ systems. With the exception of some peripheral dialects, distinctions between long and short stressed syllables occur in the final or only syllable. Notice the words vestiges (he doesn't specify its meaning) and may. He marks vowel length out of tradition, that's it. He also marks the artificial distinction between open-mid and close-mid front vowels before /j/ and between open-mid and close-mid back vowels before /v/, which is another archaism that keeps being prescribed as standard by dictionaries (see p. 24 - he calls this neutralization phonetic but it's clearly phonemic if it occurs prevocalically).
Also, I'm not sure why you gave me a link to a website written in Slovene (and presumably expected me to ask a question in that language) when I've already stated that I don't speak it with any fluency (far from it). Kbb2 (ex. Mr KEBAB) (talk) 05:39, 26 January 2019 (UTC)[reply]
I.e., Greenberg has deemed it appropriate to mark vowel length for whatever reason. The link is the authoritative site for Slovenian language questions and should be consulted. They all speak English too. Doremo (talk) 06:07, 26 January 2019 (UTC)[reply]
@Doremo: The most likely reasons for that are tradition and the fact that the "short" vowels are prone to being reduced to schwa by many speakers. However, since the traditional short vowels occur mostly in final stressed syllables speakers can just expect those to be reduced to schwa. Not all speakers will do it, just as not every single word-final stressed syllable contains a short vowel in the traditional pronunciation. If you don't want to imitate this pronunciation (you shouldn't be taught it if you want to speak with a neutral accent - that's my understanding at least) this is a good enough explanation. You know it best that colloquial Standard Slovene is rich in non-standard features and this is no exception.
I can ask them a few questions, sure. They're as good of a source as any other. But if we were to apply your approach to articles such as Croatian language we could never state that it's a variety of Serbo-Croatian unless the Institute of Croatian Language and Linguistics said so. Just because one type of sources say one thing it doesn't mean that others are wrong. This is a rather extreme example of that (and perhaps I'm exaggerating a bit here) but I hope that you get what I mean. Kbb2 (ex. Mr KEBAB) (talk) 06:31, 26 January 2019 (UTC)[reply]
I fail to understand the meaning of "authoritative" when it comes to pronunciation. It is easy for authorities to control adherence to the prescribed orthography and to punish/dismiss state officials and school teachers who refuse to use it. Does the same thing happen to Slovene officials and teachers whose pronunciation differs from "authority" standards? Could you cite precedents/court decisions, or at least the law that authorises an institution to arbitrate pronunciation disputes? Love —LiliCharlie (talk) 06:44, 26 January 2019 (UTC)[reply]

@Doremo: The "authoritative" website https://svetovalnica.zrc-sazu.si/ that you recommend has the page Kateri znak IPA ustreza slovenskemu a? ("Which IPA symbol corresponds to Slovenian a?"). The anwer ends with the following sentence: Več o zapisovanju slovenskega knjižnega jezika v mednarodni fonetični abecedi lahko najdete v članku, ki je vzet iz priročnika Mednarodnega fonetičnega združenja in je dosegljiv na https://archive.org/details/rosettaproject_slv_phon-2. (Google translation: "For more information about writing the Slovenian book language in the international phonetic alphabet, see the article that is taken from the manual of the International Phonetic Association and is available at https://archive.org/details/rosettaproject_lv_phon-2.") This means that they recommend to consult the Handbook of the IPA example for Slovenian. Love —LiliCharlie (talk) 07:36, 26 January 2019 (UTC)[reply]

The recommendation for additional reading at the site follows the statement "Slovenskemu knjižnemu fonemu /a/ v transkripciji mednarodne fonetične abecede IPA ustreza znak [a]" (The standard Slovenian phoneme /a/ corresponds to the character [a] in IPA). Doremo (talk) 07:43, 26 January 2019 (UTC)[reply]
Note also that the linked Handbook of the IPA has, for whatever reason, transcribed Slovenian /a/ as [a] and has marked vowel length. Doremo (talk) 07:47, 26 January 2019 (UTC)[reply]
Thank you Doremo, I have the Handbook in my hands. On page 136 it reads: "Vowel length, however, can no longer be regarded as distinctive for most speakers, and it is generally accepted that long vowels occur in stressed, and short vowels in unstressed position..." And all transcriptions given on pp. 135–139 follow this simple rule. (Diphthongs are treated separately and are not lengthened when stressed; isolated monosyllables are treated as stressed syllables.) Love —LiliCharlie (talk) 08:07, 26 January 2019 (UTC)[reply]
Perhaps the linked version and your version (same pagination) are different editions of the Handbook. The linked version consistently marks vowel length. Doremo (talk) 09:10, 26 January 2019 (UTC)[reply]
@Doremo: Please re-read his post. Also, please re-read my messages. I said that I want to base the transcription of centralized [ɐ] (actually it's [ʌ] now that I checked it again, it's just that on the vowel chart it's in the near-open central position) on the research of Peter Jurgec, not the Handbook nor your website. I'm sorry but this is getting very annoying. We say something and then you pretend we didn't, you comment on something else or you start a fight. This is no way to discuss the issue. Kbb2 (ex. Mr KEBAB) (talk) 10:11, 26 January 2019 (UTC)[reply]
Sorry, I don't know what you mean. I don't have a website. Please don't accuse other editors of "pretending." Doremo (talk) 10:18, 26 January 2019 (UTC)[reply]
@Doremo: I mean this website. Unless this behavior ends I don't want to discuss anymore. I have a feeling that we've started going around in circles. If this doesn't sound civil to you then I apologize, but my position stands. Kbb2 (ex. Mr KEBAB) (talk) 10:23, 26 January 2019 (UTC)[reply]
@Doremo: I apologize, I misread LiliCharlie's quote. I thought that he quoted page 138 which reads In the transcription below, the different accents have not been marked, although both vowel length and stress are indicated despite considerable redundancy involved. His quote says something else. Kbb2 (ex. Mr KEBAB) (talk) 10:51, 26 January 2019 (UTC)[reply]

Rhotic consonant

[edit]

The chart clearly shows an alveolar tap/flap in the far-left column but then goes on to cite the nearest English equivalent as being a "rolled" r. If it's a tap, the English equivalent would be the intervocalic ⟨tt⟩ in American English "butter" (obviously there are other examples but that one seems to be one of the most commonly cited ones) and the chart should reflect something along those lines. If it's a trill, then the IPA symbol needs to be changed. Which one is it? — Preceding unsigned comment added by AnonymousMusician (talkcontribs) 15:38, 28 November 2018 (UTC)[reply]

It's definitely [ɾ] (tap/flap), in my experience. Native Slovenian speakers generally reject the [r] (trill) pronunciation when listening to the contrast between [ɾ] and [r]. I'll update the chart based on the same example for Spanish [ɾ]. Doremo (talk) 16:12, 28 November 2018 (UTC)[reply]

Allophones of /m/ and /n/

[edit]

Hello. I've added the labiodental [ɱ] (represented by m for simplicity) to the guide per the Handbook of the IPA, which describes the sound as an allophone of both /m/ and /n/ before labiodental consonants. However, there seems to be an additional type of assimilation, namely /n/ that becomes bilabial [m] before bilabial consonants (see [1]). Does that mean that Slovene nasals undergo neutralization before labial consonants (as it is the case in Spanish and Italian)?

Also, what do other sources say about the issue? I think I'm going to revert my changes for now.

Pinging @Doremo:. Kbb2 (ex. Mr KEBAB) (talk) 15:15, 16 January 2019 (UTC)[reply]

And @Rua:, who seems to be the one who added IPA to that Wiktionary entry (hope I'm not annoying you). Kbb2 (ex. Mr KEBAB) (talk) 15:46, 16 January 2019 (UTC)[reply]

Yes, /n/ assimilates to /m/ before labial stops. This assimilation is generally also spelled in native or nativized words (e.g., razčlemba 'analysis, etc.' < raz + člen +ba; Rihemberk < Reifenberg ; etc.). In the case of Istanbul, the more consciously borrowed character or more recent nature of the word (cf. older Slovenian Carigrad 'Istanbul') probably preserves the ‑nb- spelling, as in modern borrowings (e.g., inbriding 'inbreeding'; nanbudo 'Nanbudō'), while some lexemes show orthographic variation (e.g., bonbon ~ bombon 'candy'; inbus ~ imbus 'Allen wrench' < German Inbus) but would have an /mb/ pronunciation in all but very slow, deliberate speech regardless of spelling. Regarding informacija (i.e., before a fricative), I find plenty of people writing imformacija online, implying that assimilation may occur, but maybe it varies by speaker because my native informant denies doing it (but admits nb > mb). Doremo (talk) 16:07, 16 January 2019 (UTC)[reply]
@Doremo: I see, thanks.
Your informant may be denying it because (s)he produces a simultaneous alveolar-labiodental nasal, rather than a pure labiodental one. It's usual to transcribe such not-quite-fully assimilated forms as if they were fully assimilated. I think we can safely use m in all cases mentioned here. Kbb2 (ex. Mr KEBAB) (talk) 11:21, 17 January 2019 (UTC)[reply]
I agree. I'm not a native speaker of Slovenian, so my own pronunciation is not reliable. Doremo (talk) 11:28, 17 January 2019 (UTC)[reply]
@Kbb2:, thank you for the various IPA updates. However, the specific change [n ˈʋ] → [m ˈʋ] is dubious (my informant strongly denies it and I don't hear it in her speech); more likely, the /ʋ/ takes on the character of an onglide before the following vowel. I suggest that the changes at Martinj Vrh, Jerman Vrh‎, Kristan Vrh‎, Križan Vrh, Zgornji Dražen Vrh‎ be reverted to [n ˈʋ]. Doremo (talk) 12:34, 17 January 2019 (UTC)[reply]
@Doremo: No problem.
The Handbook of the IPA states that both /m/ and /n/ are labiodental before /f/ and /ʋ/ (what it doesn't discuss is the [n] -> [m] change before bilabial stops). I don't think we have a strong reason not to believe it, remember that [ɱ] and [m] are different sounds that we transcribe the same (so that [maɾˈtiːm ˈʋəɾx] stands for [maɾˈtiːɱ ˈʋəɾx] rather than [maɾˈtiːm ˈʋəɾx] in truly narrow transcription). Because the labiodental nasal isn't a phoneme in Slovene (in fact it's not a phoneme in almost any language but one) and because there may be additional contact of the tongue with the alveolar ridge in addition to the contact of the upper teeth with the lower lip that is the reason why your informant might be confused by the transcription.
Look, I actually support transcribing the labiodental nasal with a separate symbol, and I'm not alone in it. You should check this discussion to see what I have to say about the issue and possibly (if you want) voice your concern that such a transcription is actually confusing to (some) of the native speakers themselves. The reason I chose to represent it with m (and to actually go against the Handbook of the IPA) is because Aeusoes1 and some other users have continuously opposed using a separate ɱ symbol in our guides. I just don't want to see silly edit wars over this, that's all. Kbb2 (ex. Mr KEBAB) (talk) 12:48, 17 January 2019 (UTC)[reply]
Thanks, I'll try to gather some additional information from some other speakers over the coming days. Doremo (talk) 13:04, 17 January 2019 (UTC)[reply]

Remove tonal diacritics and change the symbol for /a/?

[edit]

Hello. Currently, we differentiate between rising and falling vowels. It's true that Standard Slovene can differentiate betwen them - but it doesn't have to. The variety of SS that lacks tones is somewhat less prestigious, but it's still Standard Slovene. As far as I can see, most of our transcriptions do not include tones, even the transcriptions of words that can be found in Slovenski pravopis 2001.

We say that Wherever possible, one should transcribe Slovene with both tonic and stress marks. If the correct tones are unknown, it is acceptable to put only a stress-based transcription. This clearly isn't our current practice. Therefore, I propose that we remove the tonal variety from this guide altogether. The non-tonal variety isn't marked and you'll sound just as native if you speak it, so the situation isn't very similar to Serbo-Croatian in which non-tonal pronunciations are perceived to be non-standard, or at least regional (Zagrebian, etc.).

Again, this proposal isn't controversial. I just want to adjust the guide to the way we already transcribe Slovene on Wikipedia.

EDIT: We shouldn't forget that there's a number of words which can be pronounced with either tone in SS and that the phonological short vowels (save for the schwa, but maybe it's phonologically long) can only be falling. These are additional arguments for getting rid of the tonal diacritics. Kbb2 (ex. Mr KEBAB) (talk) 10:37, 20 January 2019 (UTC)[reply]

Also, do you think that we should change a to ɐ? While the latter symbol isn't very commonly used in IPA transcriptions of Slovene, maybe it'd help some readers to properly differentiate between /ə/, /a/ and /aː/ (well, the main issue is not pronouncing /ə/ as too open). Kbb2 (ex. Mr KEBAB) (talk) 13:13, 20 January 2019 (UTC)[reply]

I do not think changing a to ɐ would improve clarity for the vast majority of WP readers. Just the opposite. Doremo (talk) 13:29, 20 January 2019 (UTC)[reply]
@Doremo: Fair enough. What about the tonal diacritics? You never use them (which, I guess, is understandable as not all sources use them as well). Kbb2 (ex. Mr KEBAB) (talk) 09:08, 21 January 2019 (UTC)[reply]
Most of the sources I've used for entering pronunciations of toponyms do not mark tonal diacritics, and so for the most part I have not been able to mark them. As you point out, non-tonemic transcriptions are also fully correct, and so it's a level of (optional) detail that I have not tried to represent. I have no objection to others adding (or removing) tonal diacritics, but I would not bother to do so myself. Doremo (talk) 09:32, 21 January 2019 (UTC)[reply]
@Doremo: I thought so. Well, if nobody objects, I'm going to begin simplifying relevant transcriptions as well as the guide itself in the next couple of days. Kbb2 (ex. Mr KEBAB) (talk) 10:26, 21 January 2019 (UTC)[reply]
I have no objection because most sources (encyclopedias, general dictionaries, technical dictionaries) that mark stress and vowel quantity/quality also do not consider tone important enough to mark, and a simpler transcription is more accessible/informative to most WP readers. However, it may be wise to wait and see if other editors disagree. Doremo (talk) 10:37, 21 January 2019 (UTC)[reply]
@Doremo: Ok, let's wait a couple of weeks (unless more editors come to support the proposal, then we can speed this up). Kbb2 (ex. Mr KEBAB) (talk) 10:39, 21 January 2019 (UTC)[reply]

Updated phonology article

[edit]

Article about Slovene phonology has recently been updated and expanded and it is self-evident that the help page also has to be updated. Most additions are allophones of minor importance, so I think that these (e. g. kʰ, p̪, tˡ) do not need to be seperately written, especially because they are fully predictable. The only consonant additions that seem big enough to be included here are omission that /dz/ can only be an allophone of /ts/, addition of /ʔ/ (even though it is fully predictable, but so are /v/ and /ɣ/) and addition of syllabic consonants (see below). I would also recommend addition of /lʲ/ and /nʲ/, because these two sounds are in contrast to the aforementioned allophones taught in grammar schools and pronunciation as a non-palatalized consonant is considered simplification. A footnote can be added that they can also be pronunced as geminated non-palatalized consonants or as non-geminated non-palatalized consonants. However, I do not know what to do with pre-labialized consonants. Should we add them or leave pronunciation with /w/ and /ʍ/ or change to pre-labialization?

I am also considering addition of /ɪ/ because this change does not always appear before /ɾ/ and I consider it to be a big enough change. I also strongly support addition of front rounded vowels. However, I am not sure whether front and back mid vowels should be added. From the same perspective as consonants, the change is not big enough to be included here. However, some dictionaries write these as open-mid vowels and some as close-mid (see Slovenski pravopis §628) and mid vowels also do not have neither retracted nor advanced tongue root. I would also change how unstressed e and o are transcribed. It is true that traditionally, they have been transcribed as always being open-mid, but that is now outdated and has received many critics. A lot more appropriate way of writing them would be as close-mid before the stress and when secondly stressed and as open-mid elsewhere.

Secondary stress should also be added. Even though modern Slovene is thought not to differentiate between length in the traditional way, I have not seen a dictionary that would not differentiate between length, so it probably has to be kept. There is also the problem of tonal vs. non-tonal Slovene. The article now explains the tone in more detail and this has now become a bit more complex. Apart from that, there are also other changes, such as /a/ vs. /ʌ/ and syllabic sonorants, so keeping both pronunciations and letting editors decide would be confusing for the readers. In fact, those reading the pronunciation are probably unfamiliar with Slovene and in that case non-tonal transcription will do just fine. In addition, tonal Slovene is spoken by the minority of Slovenes and the tone can be hard to find, especially for proper nouns. In my opinion, tonal transcription would just unnecessarily confuse readers and cause problems to editors trying to find the correct tone. Even in Slovenski pravopis and SSKJ, the non-tonal transcription is favoured. Not to mention that tone can change when declining and conjugating, nominalizing, with different prepositions and the way it is used or what it represents.

I am also thinking of addition of dialectal sounds. Should I add them here, create a new help page or not add them at all? What are your opinions? Garygo golob (talk) 10:09, 27 June 2022 (UTC)[reply]

@Sol505000: and anyone else who might be intersted: I have updated the Slovene phonology article again, and apart from what I have said above, I would also like to add that I recommend that /j/ is still transcribed as it is and not as /i̯/, because a very small portion of all articles transcribes it like so, even if it is a more accurate representation. Affricates /pf/ and /kx/ as well as their voiced counterparts should also be added despite their rare occurrence. What is your stance on the additions? Garygo golob (talk) 19:20, 2 October 2022 (UTC)[reply]
Not sure being "fully predictable" is a reason to exclude an allophone from an IPA help page - the point of the help pages is to help people who don't speak the language (thus don't have that intuition about what's predictable and what's not). Great work on the phonology article btw. – filelakeshoe (t / c) 🐱 14:52, 4 October 2022 (UTC)[reply]
First of all, thank you. Second of all, the mentioned allophones are are really some minor changes (e. g, [k] → [kʰ]), which do not really seem that important to have and also are not present in other languages' help pages (e. .g,, [p] → [pʰ] in English in worlds like pie). About the [ʔ], I have great trouble distinguishing this sound, so I really have no idea if it is a big enough change to add. Would you please listen to how word 'otrok' is pronounced (it should start with [ʔ]) and then post your comment about the inclusion? Thank you. Garygo golob (talk) 05:09, 5 October 2022 (UTC)[reply]
I don't think the English IPA help article is a good one to compare to, since it seems waaay more developed than any other on the wiki. I'd only include the core phonology of the language, plus a few positional variants speakers should be aware about (Help IPA English seems to show the positional variants of the long vowels, like /uː/ in "goose", but that's about it). I wouldn't include any aspirated stops like /kh/. Essentially, I would just include the non-positional phonology of Slovene, keeping everything in broad transcriptions and the long vowels, even if they aren't contrastive in modern Slovene.
For 'otrok', I personally don't hear a [ʔ]. Wiktionary transcribes it with a /t/ on the entry for Otrok, but it is just a broad transcription. I wouldn't add it since it seems to just be positional. JungleEntity (talk) 00:48, 6 October 2022 (UTC)[reply]
Ok, so no [ʔ] then? And about positional variants, voiced/devoiced allophones are already included so it would be sensible to keep those allophones. But in comparison, [ɪ] is also a somewhat big change n pronunciation and is represented by its separate letter, therefore, it might also good to add it. But what about the pre-labialization, palatalized consonants and, above all, which standard (tonal or non-tonal) is to be used. Apparently, most transcriptions are currently in non-tonal, which is also easier to comprehend, so I suggest we keep this version. Garygo golob (talk) 15:37, 6 October 2022 (UTC)[reply]
Yea, I would go ahead and not include [ʔ]. If it is indeed a positional variant, it doesn't seem to be known enough to place here. I think [ɪ]'s positional variation might be worth an add, but I think we should wait for a 3rd editor's opinion. I don't know if you can access this full journal (I can but I think that's because im on University wifi), but this journal from the International Phonetic Association on Slovene might help. The preview shows the consonants they recognize (as of 1995, so may be outdated), and the page after shows they don't use the tonal transcription. I don't think I can post images of the article on this talk page, but if you'd like the look at their vowel charts and examples, I'd be happy to send you some screenshots. You know more about this subject than I do! JungleEntity (talk) 23:57, 6 October 2022 (UTC)[reply]
The source you found was one of the main sources in the article and I don't have access to that version, but I already got the screenshots, so thank you. That article also does not differentiate between the palatalized and non-palatal /n/ and /l/. Choosing to not distinguish those would be easier to comprehend, however would eliminate distinction between two phonemes. The thing that worries me the most, however, is how to transcribe pre-labialization. It certainly cannot be eliminated as then we would have fake homophones (e. g., vstati – stati) and old transcription as [w, ʍ] is really inaccurate, especially because other uses of [w] tend to be pronounced more like /u/ (e. g., [təɾw]) The pre-labialization, however, occurs before a variety of consonants, so it would be inefficient to list it like [ᵂb, ᵂc, ᵂd, etc.]. Can we list it like [ᵂ] in a special section or something? Garygo golob (talk) 19:31, 7 October 2022 (UTC)[reply]