Help talk:IPA/Italian

/a/ approximation

Hi. I saw this discussion Help talk:IPA/Italian/Archive 1#Southern American 'time' as an approximation of Italian /a/ involving Peter238, Aeusoes1 and IvanScrooge98 where it looks like father was chosen for the /a/ English approximation. However, I was messaging IvanScrooge98, and I think we agree that this is a tough one to approximate. However, I don't think father represents most cases. The Italian /a/ seems sharper and quicker. The "a" in father sounds more extended like f(awe)ther, whereas something like pasta is more sharp, not p(awe)sta. I thought something like apple may be more appropriate, but it also doesn't always fit with words that begin with "a". I think the examples in the table sound more like father than apple, but if we were to pick words like pasta or macchina, they sound more like apple than father. Vaselineeeeeeee^★★★ 16:06, 20 July 2019 (UTC)

Had forgot I had been involved myself. :D Anyway, I was telling Vaselineeeeeeee I saw pasta being used as an approximation for this sound in some helps, possibly because in most accents where the word is read with /ɑː/ the sound approaches the Italian and the ones that use /æ/ instead usually do not raise it. イヴァンスクルージ九十八_（会話） 16:14, 20 July 2019 (UTC)

Yeah, that makes a lot of sense. Pasta might be a better approximation for a lot of these help guides than father. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 17:39, 20 July 2019 (UTC)

Oh, I see, you want to use pasta as the English approximation. I suppose that could work, but I've actually heard pasta being said as p(awe)sta by English speakers far too often. Vaselineeeeeeee^★★★ 18:25, 20 July 2019 (UTC)

I'd like to remind everyone that millions of Italian speakers from Lombardy, Switzerland, and Piedmont realize Italian /ɛ/ as [æ] while New Zealand English speakers typically realize English /æ/ as [ɛ] or even closer. — As to pasta: I think that American English speakers usually pronounce the word as /ˈpɑːstə/, and speakers from most of the Commonwealth as /ˈpæstə/. Is that correct? Love —LiliCharlie (talk) 18:47, 20 July 2019 (UTC)

Yes, that’s correct: I thought of pasta because we would avoid creating confusion since /æ/-tensing is prevalent in North America, where /ɑː/ is used in that word instead, a phoneme which is often slightly advanced (closer to [ä]) in those accents. Conversely, many speakers who use /æ/ in pasta utter it as slightly retracted (is New Zealand an exception or is [ɛ] also used elsewhere?) Regarding Italian speakers realizing /ɛ/ as [æ], that should not make a difference, since this help focuses on standard pronunciation. イヴァンスクルージ九十八_（会話） 20:54, 20 July 2019 (UTC)

Non-Kiwi native English accents with /æ/ [ɛ] are used on several continents and include Broad Australian, Broad South African, and American accents featuring the Northern Cities Vowel Shift (in which /æ/ may be much closer than [ɛ]). Love —LiliCharlie (talk) 23:01, 20 July 2019 (UTC)

Found this where it looks like Australian English pronounces [ˈpɐːstə] (phonemically /ˈpɑːstə/). Is this realization close enough to what we want to approximate? イヴァンスクルージ九十八_（会話） 08:18, 21 July 2019 (UTC)

Pronunciation of English ⟨a⟩#In foreign borrowings says that pasta has /æ/ in Canada, Northern England and (unlike other words of the taco-llama-drama group) in RP, but /ɑː/ in American, Australian and New Zealand English, though "the pronunciation of certain words can vary even in regions which either usually assign the trap vowel or usually assign the palm vowel to such words". — I wonder if we should really use a word whose pronunciation in global Englishes is highly variable and depends on intra-regional factors that are unknown to us editors, let alone our readers. We cannot, and should not, conduct a dialect survey. Love —LiliCharlie (talk) 12:26, 21 July 2019 (UTC)

I guess father is fine then. :) イヴァンスクルージ九十八_（会話） 13:59, 21 July 2019 (UTC)

Yeah, I didn't really want to use a non-English word like pasta in the English approximation because of the possible regional differences, but I still don't think father represents most cases - surely there has to be a better word - I think an English word with /æ/ (North American) sounds a lot closer than one with /ɑː/. Vaselineeeeeeee^★★★ 15:07, 21 July 2019 (UTC)

I think people have explained why a word with /æ/ would be a bad idea. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 16:01, 21 July 2019 (UTC)

I get it, apple might sound like epple in New Zealand and parts of Australia, so a simple addition: we write apple (American English) or something of the sort, as we have done for future (Scottish English). This is better than having father which sounds off in the majority of cases. Vaselineeeeeeee^★★★ 16:10, 21 July 2019 (UTC)

American English apple is not a good approximation, as it's closer to [ɛə] or [ɪə] for a number of speakers. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 16:33, 21 July 2019 (UTC)

Never really heard that before, but that was just one word I thought of, something else can be chosen, like trap. Vaselineeeeeeee^★★★ 16:45, 21 July 2019 (UTC)

We shouldn't explain the value of Standard Italian /a/ with a phoneme realization used in an English accent that is less well known, or less unambiguously identifiable, than [ä] of a Standard Italian accent itself. — I'm in favour of the palm vowel rather than the trap vowel, though I think the strut vowel might be debatable. Love —LiliCharlie (talk) 17:16, 21 July 2019 (UTC)

So why do we do it with future (Scottish English)? American English is probably more well known than Scottish English. That is besides the point though. Between strut and palm, I think the latter would be better, however, Italian words like asso, attore, macchina, alzato, pastore, etc. all do not sound like "f(awe)ther", they sound more like the trap vowel in North American English (although not exact, sound better than palm) - not (awe)sso, p(awe)store. Vaselineeeeeeee^★★★ 22:08, 21 July 2019 (UTC)

That's unlikely to be the case, since, as I said, the trap vowel in NAE is often [ɛə] or [ɪə]. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 00:45, 22 July 2019 (UTC)

So specify a variant of English where it doesn't. We already use this solution in the article with Scottish English. The current version with father is misleading and not the way the vowel is said. Vaselineeeeeeee^★★★ 01:07, 22 July 2019 (UTC)

Sorry, I think it's the closest we're going to get. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 02:03, 22 July 2019 (UTC)

Father is the closest we can get? Really? It's not remotely close to any of the first few words that came to mind: alto, asso, attore, macchina, alzato, pastore. F(awe)ther does not fit - so instead of thinking of better words, we're going to leave an inaccuracy? The only ones that father may remotely fit are a's that follow r's, like soprano. Vaselineeeeeeee^★★★ 02:16, 22 July 2019 (UTC)

Italian /a/ doesn't "sound more like the trap vowel in North American English" because American English /æ/ can be pronounced in several ways, and some of them are more like Italian /ɛ/ or /e/, or even closer. Please read North American English regional phonology, Northern Cities Vowel Shift, and Flat A, and watch The Blues Brothers (film). Love —LiliCharlie (talk) 02:36, 22 July 2019 (UTC)

LiliCharlie, I am clearly not as well-versed as you in the seemingly dozens of ways /æ/ is pronounced in the United States, so please tell me, do you know of any areas where /æ/ is pronounced as /æ/ because father (if the palm vowel is pronounced the same everywhere, and if not how is this different than trap?) is not good enough and we should not settle for something that is not well represented by majority of the cases. If none exists, then tensing can be specified as we currently do with murder RP. Vaselineeeeeeee^★★★ 02:53, 22 July 2019 (UTC)

We can't expect readers from all continents to be acquainted with North American dialect geography (or North American geography in general), so we won't have them depend on any specific American English regional accent.

Your remark "if the palm vowel is pronounced the same everywhere" is interesting; John C. Wells chose the keywords of his lexical sets carefully, but later commented: "The least satisfactory keyword is PALM, and its set is also fairly incoherent. Amy says she prefers to replace it with FATHER, which is fine up to a point: but not if we are discussing Hiberno-English, where father often has not the expected aː of Armagh, Karachi, Java etc but the ɔː of THOUGHT." Which probably means we'd better replace father with palm, for the benefit of the Irish. Love —LiliCharlie (talk) 03:49, 22 July 2019 (UTC)

We won't expect them to know it, that's why we'd link it like we do RP, something that may not be well known. I think palm sounds slightly better than father, even though I still think a different word could be better suited. Vaselineeeeeeee^★★★ 04:26, 22 July 2019 (UTC)

Since Scottish English is already used in the approximations, why not have either trap or father or palm marked with the Scottish accent tag? Scottish /æ/ and /ɑː/ actually merge into a sound that is close or identical to the Italian A. That might be a good compromise, I think. イヴァンスクルージ九十八_（会話） 08:30, 22 July 2019 (UTC)

Alright, changed to a closer approximation as we seem to have reached consensus. イヴァンスクルージ九十八_（会話） 12:47, 23 July 2019 (UTC)

Didn't see this discussion, sorry for changing before reading. Arguments are convincing enough, and everyone else agrees with the current version. ITskandros (talk) 19:00, 5 October 2019 (UTC)

Why not audio?

English approximations of Italian speech sounds are inevitably rough and unreliable. This problem can, however, be overcome if a Standard Italian speaker makes audio recordings of the Italian examples, uploads them to the Commons, and finally links to them on this help page. Love —LiliCharlie (talk) 07:36, 22 July 2019 (UTC)

I may try to see if I can, but that may take me long. イヴァンスクルージ九十八_（会話） 08:30, 22 July 2019 (UTC)

Either of these options sound like a step forward, thanks. Vaselineeeeeeee^★★★ 15:50, 22 July 2019 (UTC)

Intervocalic ⟨s⟩

It has been brought to my attention that edits like this (by @IvanScrooge98:) may be problematic by showing the less common variant of a variable pronunciation first. Italian phonology seems to back this up. What do we think is the best way to deal with this issue? — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 16:00, 24 September 2019 (UTC)

@Aeusoes1: I felt it made more sense to list the traditional pronunciations first (as they have been the only standard for longer and still carry some distinctive value) and the ones that have become accepted standard only in recent times (and now maybe more common in some educated speech, even though I personally hear both in TV ads etc.) after. At least for the sake of having a definite order. 〜イヴァンスクルージ九十八［IvanScrooge98］（会話） 16:43, 24 September 2019 (UTC)

Bertinetto & Loporcaro (2005: 133) say:

Intervocalic /s/ is voiceless in [Rome Italian], so that [Standard Italian] minimal pairs, such as [ˈkjɛːse] 'asked.3sg' vs. [ˈkjɛːze] 'churches' (both spelled chiese), are neutralised as [ˈkjɛːse] in RI. On the other hand, the growing influence of the Northern pronunciation on [Florence Italian] has levelled out some traditional contrasts, such as the one cited, so that [ˈkjɛːze] is now increasingly accepted in both meanings, although normative pronunciation treatises still record the contrast. In [Milan Italian], the same pair is neutralised into [ˈkje̞ːze] ... for another reason, since in this variety (as opposed to RI) voiceless intervocalic [s] does not occur morpheme-internally, but only after vowel-final prefixes (e.g. risollevare [ˌrisolːeˈaːre] 'to raise again'), unless the prefix is no longer synchronically analysed as such (e.g. risaltare [rizalˈtaːre] 'to stand out').

Krämer (2009: 41) says:

Canepàri (DiPI 1999) criticizes contemporary dictionaries for giving forms which are not in use anymore even by the most trained professional speakers from Tuscany or Florence. Accordingly he gives almost all forms that have an intervocalic voiceless coronal fricative in the other dictionaries with a voiced fricative as the preferred form ("la piú consigliabile" 'the most advisable') and an 'outdated' form with the voiceless fricative ("la piú consigliata un tempo" 'the most advisable some time ago"). Those forms that are recommended with a voiced fricative by the other dictionaries have no alternative formerly recommended or non-recommended rendition. This leaves Canepàri's recommended pronunciation of Italian with an imbalanced contrast (if it is a contrast at all) leaning towards the other side: only very few forms, such as presidente 'president', or preservativo 'preservative, condom' are recommended with a voiceless fricative. These are those forms which, according to my experience, educated speakers from northern Italy, for whom intervocalic s-voicing can be regarded as operative, produce with a voiceless fricative. It is not quite clear if these speakers treat these words as loanwords or as morphologically complex.

So Canepari, which I assume you were referring to when you said Italian phonology seems to back this up, seems to be weighing heavily on Northern varieties, and the fact it is neutralized to [s], i.e. in the opposite direction, in Rome according to B&L is hard to dismiss. So the present question can hinge on what varietie(s) we choose to represent in our transcriptions. According to B&L (pp. 131–2):

Today's SI is based on the Tuscan (more precisely Florentine) dialect ... It is nowadays spoken with distinct local accents ... Among these, the Roman and the Milanese varieties are especially prominent owing to their use in the media, alongside the Florentine variety, which is very close to SI. Over the past few decades especially the Milanese accent seems to be increasingly gaining prestige ... SI is nowadays part of the active verbal repertoire of just a minority of educated people from Central Italy (especially Tuscany), besides being used by professional speakers or trained stage actors (the single idiolects spoken by these groups of people may, however, include sporadic features typical of RI). Due to its cultural, social and political relevance, the definition of (the status of) SI has been the topic of a long-lasting debate.

In B&L's samples, intervocalic s appear seven times, and while RI and MI consistently produced [s] and [z] respectively, SI and FI had [z] in sposo, mise (twice) and sposa, and [s] in riposarsi, riposatasi and rispose. The authors remark, "lexically-conditioned change seems to be in process under northern influence".

In Rogers & d'Arcangeli's (2004) sample of "a woman in her thirties who was brought up in a middle-class household in Rome, and has travelled extensively in Italy and abroad" and who is representative of "contemporary 'mainstream Italian' ... a variety under construction by speakers wishing to give themselves a national appeal, a process driven by the media and by workplace mobility", she pronounced all intervocalic s with [z] (in decisero, cosí and ammise).

Given these, I think I would give [z] precedence. The change is apparently undergoing and may be "lexically conditioned" in Tuscany, but for consistency's sake I agree with Ivan in that we should stick to one order. Nardog (talk) 22:49, 24 September 2019 (UTC)

I would like to point out that the Sicilian-derived word Puglisi which Ƶ§œš¹ linked to has only /z/ in Canepari's DIPI. By contrast, pugliese and other adjectives with the Tuscan-derived ending -ese show variation and are pronounced with /z~s/ in Standard Italian, and with /s~z/ in Tuscany, according to Canepari. (However, Canepari reports /z/ to be preferred in the proper name Pugliese even in Tuscany.) Love —LiliCharlie (talk) 06:37, 25 September 2019 (UTC)

This one case is peculiar, as the original Sicilian pugghisi is pronounced with the voiceless variant (as is actually usual for all instances of intervocalic ⟨s⟩ in southern Italy, but in this and several other cases the original local pronunciation matches the traditional Italian, which may be a reason to give it prominence; especially since the name has not been “fully Italianized”). I was about to publish this reply, but just noticed DOP, by far more conservative, only reports /z/ as well, meaning I made a mistake when providing the IPA—I apologize, I usually check the websites before, unless I am in some hurry.

In any case, we should not forget that Canepari is from northern Italy himself, so maybe that is one more reason why he tends to recommend voiced fricatives. As a northern Italian speaker who uses them natively, I still feel it is a little premature; I have heard Florentine speakers use /z/, but generally not in widespread suffixes etc. such as -oso, -ese (though this my personal experience and reality may be very different). And of course, I think we should be careful about not lying too much towards northern varieties, as we are supposed to be as neutral as possible. 〜イヴァンスクルージ九十八［IvanScrooge98］（会話） 08:38, 25 September 2019 (UTC)

My argument is based on Bertinetto & Loporcaro and Rogers & d'Arcangeli, not Canepari. Nardog (talk) 10:23, 25 September 2019 (UTC)

Yes, but considered the change is still in progress, and that both the voiceless and voiced seem to be fine now, I would keep the voiceless first until there is agreement that the main standard is /z/ (I had even encountered transcriptions that only used that on Wikipedia, which is a huge no-no if you ask me). And we should not forget that one thing are common pronunciations, and another is the standard (e.g. the sample from Rogers & d’Arcangeli lacks many instances of syntactic gemination, I suppose hypercorrection if the speaker is from Rome). 〜イヴァンスクルージ九十八［IvanScrooge98］（会話） 12:53, 25 September 2019 (UTC)

This is a spontaneous idea, and probably not a good one: We could show variation between /s/ and /z/ not through sequentially ordered symbols, but by using ⟨s̬⟩ and ⟨z̥⟩, in which the base symbol indicates the preferred pronunciation in the traditional standard.

Should we continue to use ⟨s⟩ and ⟨z⟩ for intervocalic intramorphemic orthographic ⟨s⟩ (and we probably will) and want to stick to a fixed order everywhere, I think the first pronunciation to indicate should be the progressive "contemporary 'mainstream Italian'" one that the model speaker of Rogers & d'Arcangeli (2004) uses, i.e., ⟨z⟩ first, ⟨s⟩ second. This seems more descriptive and less presciptive: it reflects what "pan-Italian" speakers do, not what they are traditionally expected to do. Love —LiliCharlie (talk) 17:30, 25 September 2019 (UTC)

Using ⟨s̬⟩ and ⟨z̥⟩ would be a nice solution, but yes, it’s probably overly complex for a help that we (hopefully) seek to maintain as simple as possible (four symbols for two sounds, and we would have to extend ⟨s̬⟩ to cases of preconsonantic ⟨s⟩, as in sbaglio etc.). Also, what I’m saying is, as there is a restricted number of (educated) speakers of Standard Italian, so the descriptive aspect still wouldn’t be “pan-Italian” anyways. And I just don’t want to make it look like most natural speakers from central-southern Italy should prefer the voiced realization over theirs even when that’s the traditional one in Standard Italian too. But as long as we find an agreement once and for all on which goes first, I’m fine. 〜イヴァンスクルージ九十八［IvanScrooge98］（会話） 19:07, 25 September 2019 (UTC)

I'd fully support ⟨s̬⟩, using it seems to be perfectly logical. Kbb2 (ex. Mr KEBAB) (talk) 06:44, 26 September 2019 (UTC)

Seriously? I would never support adding an archiphoneme to phonetic transcriptions. Nardog (talk) 09:08, 26 September 2019 (UTC)

I don't have a problem with archiphonenemes/diaphones in principle if it can reduce the burden of transcriptions, but this is not an instance where I think it's justified or transparent. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 15:18, 26 September 2019 (UTC)

@Kbb2: Aside from the question of whether to introduce such a symbol, do you have an opinion on which one to give precedence if we were to transcribe both? Nardog (talk) 18:55, 28 September 2019 (UTC)

Interesting topic. I'd like to say my piece too, I'm Italian and that could be useful. I think there're 2 options:

It's possible to strick to what Italian dictionaries generally prescribe, i.e. a single pronunciation, the most traditional, considered the "purest". This should be not only for s/z but also for e/ɛ and similar cases and would give readers the same information they can find in an orthoepy handbook.

It's possible to adopt a less narrow approach as suggested by the sources brought here, starting from the dictionary using I.P.A. transcriptions. The pronunciation of ⟨s⟩ between vowels, when it isn't the first letter of a suffix as in "sta⟨s⟩era", should be indicated first voiced and then unvoiced, to give reader the actual 3rd millennium pronunciation, recommended by the cited sorces.

They're both valid. Your opinions are slightly leaning toward the second as far as I can tell, I'm not saying my opinion for the moment because I'm not sure. ITskandros (talk) 19:09, 5 October 2019 (UTC)

I recommend, too, sticking to what Italian dictionaries prescribe, bearing in mind that people from Northern Italy have a more consistent attitude than Florentine speakers (in between "s" is always pronounced [z] like in rosa, casa, peso, posa, sposi, etc.), while speakers from the South cannot generally reproduce the sound [z], not even in their family name (like Puglisi or the inhabitants of Apulia). Although Italian film-makers from Cinecittà tend to have people from Rome starring in their movies, the Milanese [z] has gotten the upper hand, and that is an advantage for foreign learners as "s" is pronounced in a consistent way.--Luensu1959 (talk) 15:33, 29 October 2019 (UTC)

• I got a note on my German Wikisource talk page asking me to help wrap this discussion up. Have you all come to a conclusion on what the recommended order of transcriptions should be? Wug·a·po·des​ 21:50, 4 November 2019 (UTC)

Syllabification of /s/ + consonant

Do we have consensus on how to syllabify /sC/ clusters? In existing transcriptions, word-initial ones are transcribed as tautosyllabic, as in [ˈskaːla], and word-internal ones are split between tautosyllabic, as in [toˈskaːna], and heterosyllabic, as in [esˈprɛsso], though the former prevails.

For what it's worth, theoretical accounts mostly seem to regard /s/ in /sC/ clusters, including word-initial ones, as "extrasyllabic" (e.g. Krämer 2009 and Hermes et al. 2013)—which I assume is something akin to e.g. English /d/ in begged, where the otherwise illegal coda /ɡd/ arises due to the morphological complexity (cf. beloved, learned)—or "underdetermined" (e.g. Bertinetto 2004). (See Bertinetto & Loporcaro 2005:140–1 for a brief summary.)

However, I'm not so sure this means we should write [sˈkaːla], [tosˈkaːna], etc. (which is what DiPI does). I think we can agree [sˈkaːla] looks weird and confusing to most readers, since such a construction (word-initial [Cˈ...]) is seldom encountered. This notation is also problematic, as Bertinetto & Loporcaro point out, in that, if the previous word ends in a consonant, it would lead to a complex coda, which is highly marked in Italian. Rogers & d'Arcangeli (2004) have [ˈstaːvəɾ̃o]. Payne (2005) outright placed the stress mark before the affected vowel, as in [skˈaːla], to avoid the syllabification problem entirely, but I'm sure this also would be unnecessarily confusing. So, where stress falls on the first syllable of a word, placing the stress mark before the whole word even if it begins with /s/ seems uncontroversial.

What remains potentially controversial is the syllabification of word-internal /sC/. We could write [tosˈkaːna] etc., but that would create an asymmetry with word-initial /sC/, so my inclination is to write [toˈskaːna] etc. Monolingual dictionaries that indicate pronunciation also overwhelmingly treat /s/ in clusters as tautosyllabic: [1][2][3][4][5] vs [6] (although the non-IPA ones could be, in part or whole, about hyphenation in orthography). Either way, we should establish a consensus and stick to whichever way it favors (the same applies to [zC], I assume). Nardog (talk) 21:51, 24 June 2020 (UTC)

My preference is to maximize the onset, but I might be off base if there's some other consideration I'm missing. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 03:59, 25 June 2020 (UTC)

My 2¢. In Italy students learn from teachers that the so called "impure S" always makes one syllable with the following consonant, in fact a lot of Italian words begin with S followed by another consonant, unlike in most of the others Neo-Latin languages. All Italian dictionaries report this type of syllabication for Italian words and names, and almost all the Italian users who added or edited Italian phonetic transcriptions have followed this method. In a perspective of simplification for the readers, I think there's no point in distinguish between word-initial and word-middle "impure S", it's O.K. maintaining the /'sC/ transcription instead of changing all of them into /s'C/. Consider also that the way of simplification has already been chosen for other cases: no diacritics over symbols (/a/ in place of /ä/, /t/ in place of /t̪/, /ts/ in place of /t͡s/, /ai/ in place of /ai̯/), always /r/ even when "R" may be pronounced /ɾ/, fortis indicated with double consonant instead of /ː/, etc...--151.64.166.41 (talk) 09:01, 26 June 2020 (UTC)

Thanks to Nardog for the heads up re the discussion on this. I'm totally swamped today, but for the moment, the bit in Italian phonology under Onset CC summarizes, including brief explanation of phonetic (not necessarily phonemic/structural) [ˈsk] not [sˈk] word initially in isolation but [sˈk] both internally and in longer utterances when vowel precedes word-initial (super-obvious case not /sC/, but similar principle: psi-co-lo-go but lop-si-co-lo-go). There was quite a bit of discussion of /sC/ not terribly long ago. Some is archived (no idea why banished to archive) in the Italian phonology talk page. Just three points for now. First, the Maximum Onset Principle can work for traditional (prescriptive) orthographic practice (where to hyphenate), but there's no reason to assume that it necessarily churns out an accurate representation of either phonological structure or phonetics. Second, in English-language Wikipedia the transcriptions of Italian toponyms, names, etc. are phonetic; while quite rightly very broad with scads of detail not present, they shouldn't be inaccurate at a basic level (the article Pescara provides a good example; the transcription syllabifies pe-scara while "listen" triggers a rather clear pes-cara (if your Italian phonology is native or nearly so, try saying Trastevere in relaxed normal speech syllabifying tra- rather than tras-; weird at best). Third, dictionaries, especially but not only online versions, should be examined with care, always starting with the Guida all'uso of each one. Sabatini & Coletti, for example, make it almost clear in the print version that they're not reporting actual pronunciation, notwithstanding the use of [ ]. Sorry, must run; more later. Barefoot through the chollas (talk) 16:07, 15 July 2020 (UTC)

A quick follow-up to the bit about dictionaries, in this case Sabatini & Coletti, print version. On p. v of the Guida all'uso there's one point at which their intentions re transcriptions are made almost overtly clear. They exemplify the almost-completely-Italianized freudiano: [freu-dia-no, pr. /froiˈdjano/]. Square brackets don't enclose phonetics, but whatever information is being given beyond the basic spelling of the lemma itself. The form freu-dia-no in this case seems to be a writing guide, i.e. spelled with eu, syllabified in writing as freu-dia-no. Then they give the pronunciation (pr.) in IPA, phonemic rendition: /froiˈdjano/, showing that eu is [oi] and the i is [j], not [i]. Their second example helps with the s+C question. gestaltico [ge-stal-ti-co, pr. /gesˈtaltiko/]. Lots of Italians have at least basic German, so their purpose may have been to convey that the first consonant is [g], not [d͡ʒ], and the grapheme s represents [s] in the Italianized version, not [ʃ], but they also show stress, so a decision had to be made for the st cluster. Knowing what they're doing, they honestly (IMO) reported /sˈt/ notwithstanding their approval of the hyphenation ge-stal... in traditional orthography. Barefoot through the chollas (talk) 22:06, 16 July 2020 (UTC)

@Barefoot through the chollas: Thanks for the comment. I assume Talk:Italian phonology/Archive 1#syllabification of s+C clusters is the discussion you're referring to (archiving doesn't mean much, it's just done customarily to keep talk pages navigable).

I'm afraid I don't understand how syllabification in a transcription has any bearing on the phonetic realization it implies. How do [peˈskaːra] and [pesˈkaːra] differ phonetically? And aren't all sequences of /s/ + consonant heterosyllabic, even word-initially? So aren't Italian speakers theoretically incapable of pronouncing Trastevere as Tra-... no? And above all, how does any of this pertain to how we should represent the pronunciation of Italian words in our articles? @Aeusoes1: Do you understand Barefoot's argument? Do you agree?

As for syllabification in dictionaries, I only brought them up as potential references. If they make no phonetic or phonological claim like you say, we can dismiss them from this conversation. Nardog (talk) 14:48, 17 July 2020 (UTC)

I don't think it's such a big deal. The /ˈ/ is conventionally placed at the beginning of the stressed syllable, and in Italian an "impure S" makes one syllable with the following consonant, there's no real difference between /sˈkaldo/ and /ˈskaldo/ as there's no between /risˈkaldo/ and /riˈskaldo/ (both meaning 'I heat'). Maintaining the /ˈsC/ sequence would also make the reader know how the actual hyphenation of the Italian word is, so why preferring a /sˈC/ sequence which would add nothing to the previous?--151.64.164.98 (talk) 18:01, 17 July 2020 (UTC)

Maintaining the /ˈsC/ sequence would also make the reader know how the actual hyphenation of the Italian word is...

No, it wouldn't. Phonetic transcriptions are representations of spoken language and don't "make the reader know" any orthographical conventions. Love —LiliCharlie (talk) 18:57, 17 July 2020 (UTC)

Right. (I think I may have tried to post concurrently with you, Charlie. I'll put mine here, anyway; apologies for the repeat of what you say.)

I'm not sure I understand all the questions and comments, but I'll proceed as though I do. First this: I don't understand how syllabification in a transcription has any bearing on the phonetic realization it implies. If the transcription is phonetic, as the transcriptions of Italian such as Toscana are in English Wikipedia, any syllabification shown should be phonetically accurate. How do [peˈskaːra] and [pesˈkaːra] differ phonetically? Most noticeably by the presence or absence of [s] in the coda of the first syllable. Listen to the Pescara recording carefully; repeat what he says until you're sure you're saying pes-ca-ra; then produce versions with pe- as first syllable. And aren't all sequences of /s/ + consonant heterosyllabic, even word-initially? The majoritarian view amongst phonologists seems to be yes, phonemically. But not phonetically: scala uttered in isolation (post-pause) is [ˈsk] phonetically, la scala is usually [sˈk] (although there can certainly be lots of variability across word boundaries). So aren't Italian speakers theoretically incapable of pronouncing Trastevere as Tra- In principle no, depending on the linguistic agility of the speaker. It's just unnatural, odd, etc. No Italian would say [ˈbanka] naturally, but it can be done. (In a reverse of this, I can force myself to syllabify Cathleen phonetically in English as Cath-leen; not impossible, just unnatural.) how does any of this pertain to how we should represent the pronunciation of Italian words in our articles? You lost me there. If we're purporting to represent the pronunciation, we should do so.

in Italian an "impure S" makes one syllable with the following consonant In traditional orthography, yes (Maximum Onset Principle). That's why Sabatini and Coletti give ge-stal... in the script version of their transcription; it's misleading but does no real harm. there's no real difference between /sˈkaldo/ and /ˈskaldo/ as there's no between /risˈkaldo/ and /riˈskaldo/. There's very real difference in that different structures are described (assuming you're using phonemic slashes as in /k/ → [h] intervocalically in stereotypical Tuscan). It's most noticeable in /sˈkaldo/, which suggests to the attentive phonology student that something interesting is going on, perhaps to reveal itself in connected speech (then the student hears lo studente, /lo/ + /s.tu.../ realized as [los.tu...] -- Bingo!). Maintaining the /ˈsC/ sequence would also make the reader know how the actual hyphenation of the Italian word is I suspect this is mixing, if not apples and oranges, at least oranges and tangerines. Hyphenation is used in traditional script; the principles don't always follow phonological principles. If the speaker clearly says pes-ca-ra, it's irresponsible (at best) to give an IPA phonetic rendition of the syllabification pe-sca-ra. Barefoot through the chollas (talk) 19:56, 17 July 2020 (UTC)

Stressed vowels preceding /sC/ (as in maschera) require no lengthening, but vowels in non-final stressed open syllables do, so assuming a syllable break in front of /sC/ violates phonological rules. I am not sure what this means for our phonetic transcriptions, though. Love —LiliCharlie (talk) 20:43, 17 July 2020 (UTC)

That's one of the clear clues that the syllabification is VC.CV: as you say, short stressed vowel, closed syllable. Also the historical outcomes of stressed Ĕ and Ŏ (It. festa, Sp. fiesta; It. vostro, Sp. vuestro -- Spanish not constrained by syllable structure: puente, fuente but It. ponte, fonte) -- although in principle that syllabification can change over time. Barefoot through the chollas (talk) 21:53, 17 July 2020 (UTC)

Most noticeably by the presence or absence of [s] in the coda of the first syllable. That's not a phonetic difference. The syllable is a concept in the domain of phonology, not phonetics. Phonetic phenomena are indeed mere effects of speakers' cognitive processes, but since neuroscience isn't so advanced that we can directly observe the processes, we derive the phonological structure of a language from phonetic evidence. But that doesn't mean the derived phonological structure causes the phonetic phenomena; it's simply our best guess at how the speakers' brains are wired.

For example, English frustration is often syllabified as /frʌˈstreɪ-/ because the realization of the cluster /str/ resembles that of straight more than that of trait. But since /ʌ/ is otherwise a checked vowel, some syllabify frustration as /frʌsˈtreɪ-/, or regard /s/ as ambisyllabic. All three are valid positions to take, each taking different aspects of pronunciation to be more important than others. Some might be more convincing, but none are objectively "right". Syllabification is what we ascribe to it, not an observable fact (until science catches up).

scala uttered in isolation (post-pause) is [ˈsk] phonetically, la scala is usually [sˈk] Again, how does [sk] in scala uttered in isolation differ from that in la scala in phonetic terms? Does it differ in duration? In VOT? In intensity? Does it have a different spectral distribution of energy? Nardog (talk) 09:24, 18 July 2020 (UTC)

I think I should underline a point that maybe someone is missing: this discussion is about how writing phonetic transcriptions of Italian names or words in Wikipedia; we aren't talking about strict Italian phonetics. This Help:IPA page wasn't made for expert in phonetics who want to study Italian but for the average reader of this encyclopedy (in fact Italian phonology is another page, a dedicated article), I've made a few examples above. Italian /ä/ is written /a/ for simplicity, /t̪/ is written /t/ for simplicity, /t͡s/ is written /ts/ for simplicity, /ai̯/ is written /ai/ for simplicity... Letter "R" may be pronounced /ɾ/ in certain cases but here it's always written /r/ and there's a note explaining it. The aim is simplicity and clarity. And here we're "making a state matter" (as we say in Italian) about placing the stress symbol before or after /s/ followed by consonant depending on its initial or middle position in the word? I'm Italian and, even if in a fast speech I might pronounce /sˈC/, if in this moment I try pronouncing a sentence containing that sequence I have no difficulties at all at pronouncing /ˈsC/, it sounds neither unnatural nor odd to me. Are you Italian, Barefoot through the chollas? You talked as if you were, but I'm not sure about it. The /nk/ sequence would be unnatural and odd in every language not only in Italian, the example doesn't fit at all. I'll make another example to explain better what I was saying about simplicity and about the aim of a page like this. In French phonetic transcriptions the symbol /ˈ/ is missing, completely. That's because in French all the words end with the stress on the final syllable so it wouldn't allow to distinguish words on the base of its position, but transcribing "été" as /ete/ instead of /eˈte/ is phonetically "wrong", there's no other word to call it. Nevertheless it's the best solution for the aim it has here, because it's the simplest for readers. I really can't understand why somebody would want to complicate things for a language with such a simpler phonetic system compared to French or English. If we wanted to use a transcription so strict that it includes syllabification (if the "impure S" is really pronounced in another syllable in fast speech), then why don't we write [i.ta.ˈljaː.no] too? Simplification and clarity, not complication and irregularity. Writing (leaving written) /ˈsC/ both word-initially and word-inside, and adding a note in the page explaining that, when inside the word, the syllables may be separated in a different way (as it was done for the case of /ɾ/~/r/), would be a balanced solution, and much less incorrect than not writing the stress in any word of a language like French.--151.64.164.234 (talk) 08:36, 18 July 2020 (UTC)

As much as we seem to take the same position in regard to the issue of syllabification under discussion, it sounds like you should (re)learn the difference between phones and phonemes, and what the IPA purports to be, FWIW. /a, t/ are represented by ⟨a, t⟩ rather than ⟨ä, t̪⟩ not just for simplicity but because of the lack of phonemic contrast between [æ], [ä], [ɑ], etc. and between [t̪], [t̺], [t̠], [ʈ], etc. in the Italian language specifically. As the second clause of the Principles of the International Phonetic Association puts it, [p] is a shorthand way of designating the intersection of the categories voiceless, bilabial, and plosive; [m] is the intersection of the categories voiced, bilabial, and nasal; and so on. The sounds that are represented by the symbols are primarily those that serve to distinguish one word from another in a language. ⟨t⟩ sans diacritics doesn't canonically represent a voiceless apical alveolar plosive; it represents a voiceless dental, alveolar or postalveolar plosive with no regard to the tongue shape. [t͡s] is represented by ⟨ts⟩ not for simplicity but because the tie bar is merely optional, which is obvious if you look at the IPA chart. [ai̯] is represented by ⟨ai⟩ not for simplicity but because [i̯] in this position is an allophone of /i/ (though one could debate this).

The /nk/ sequence would be unnatural and odd in every language is simply not true. English is one of the few languages where the coda nasal assimilation is not mandatory. E.g. incapable may be pronounced with [ŋ] or [n], depending on speaker and speech style (see e.g. Roach et al. 2011, pp. xvi–xvii).

[T]ranscribing "été" as /ete/ instead of /eˈte/ is not wrong, phonetically or otherwise, because French doesn't have contrastive stress. First, stress is a completely language-dependent property that has no cross-linguistic set of phonetic correlates. It generally corresponds to the pitch, duration, and loudness of the affected vowel, but usually only two of these three are truly responsible for making the stress distinction and there's no telling which ones without external information about the given language (and they manifest in different ways; stressed vowels have rising pitch in Danish whereas they have falling pitch in English). In French stress falls on the last full (= non-/ə/) syllable in a prosodic phrase (there are exceptions), so there's no use in using ⟨ˈ⟩ when transcribing French unless one is transcribing a specific utterance of a longer phrase or sentence by a specific speaker. In languages like English, stress is indeed important since it has the power of distinguishing words, as in insight vs incite.

And there's no /ɾ/ in Italian, as slashes enclose phonemes, not phones. But even phonetically, the use of [ɾ] in transcribing Italian is debatable, as intervocalic singleton /r/ can have up to two contacts and, even when it has only one contact, the mechanism is usually aerodynamic, which is characteristic of a trill, not ballistic like taps. See the discussion on this page from 2018.

Again, we don't seem to disagree in terms of the conclusion, but the path you're taking to arrive at it is, IMHO, ill-advised. That we don't need to aim at as accurate a description of physical utterances as possible in our transcription is true, but it's not simply because we value simplicity but because our transcriptions are broad phonetic transcriptions. Just because a transcription is not phonemic doesn't mean it should therefore be narrow. Narrowness is not a binary but a continuum, and most of our IPA transcriptions in articles linked to keys under Help:IPA/ are broad phonetic transcriptions informed by our knowledge of the phonological structures of the respective languages. See the Handbook of the International Phonetic Association, particularly page 29 and thereabouts. And I don't find the fact our transcriptions are broad really pertinent to this converstaion in particular, because, as I said above, syllabification is not by itself a phonetic phenomenon. Nardog (talk) 10:43, 18 July 2020 (UTC)

Well, thank you Nardog for correcting my imprecisions and allowing me to learn new information about phonetics (the "check the lens well" example is very interesting to me)! I'm just an Italian guy who knows something about lingiustics in general because of my studies but I can't affirm I'm an expert in this area, I joined the discussion because I'd have liked to add a native speaker's considerations, I hope they can help.--151.64.169.20 (talk) 11:06, 18 July 2020 (UTC)

I'm sensing some danger of going down an endless rabbit hole here, so I'll respond with just a few observations. We can be grateful that word-initial clusters are not problematic in this: the Wikipedia entries have phonetic transcription, not phonemic, so a form like scala is honestly reportable as [ˈskaːla]. The issue is word-internal clusters. The question is whether to follow established phonological principles known to be derived from empirical research or to appeal to authority (whose ultimate sources may or may not be based on the results of empirical evidence). The majoritarian position amongst phonologists who work on Italian (most of whom are Italian) is that internal /sC/ syllabifies [s.C]; Bertinetto is the major holdout, and even he reports that [s.C] is frequent.

Dictionaries of various sorts presumably constitute the authority contingent. The first check is whether they're presenting actual pronunciation, and the test is items that are not subject to confusion/conflation between orthography and phonetics or phonemics. Entries like aglio or giudice make it very clear what they're up to: e.g. an entry like a-glio without [ˈaʎʎo] tells the tale -- useless for actual pronunciation (and phonemic or phonetic syllabification). Once those are discarded, then test to see, for example, if length is shown where it's not obvious from spelling: is azione shown to have [tts]? Etc. etc. Various tests, including whether the transcription is phonetic or phonemic. Among those Nardog provided above, 5 (SAPERE) looks rather good at first glance: [ŋ] indicates phonetic (basic allophonic assimilaton /banka/ → [ˈbaŋka])... but the more you search, the more you realize that their principles incohere. What is /ˈkà.sa/ supposed to represent? An idiosyncratic way of showing vowel length? Maybe, but azione shows oː and vita is given as simply /ˈvita/. Scroll down and you see that their source appears to be Wikizionario, which, sure enough, is enthusiastically inconsistent (euphemism for a mess).

It turns out that of the online dictionaries listed, 7 is the most coherent and accurate. A weakness for the purposes here is that, as presumably noted honestly by the slashes / /, the transcriptions are phonemic. banca is /ˈbanka/ with /n/ that surfaces phonetically as [ŋ], azione is /atˈtsjone/ without vowel length (yet slitta is /ˈzlitta/, either a strong theoretical claim that [z] is not /s/ → [z] before voiced consonant, or a cheat just to make sure that readers get the actual pronunciation).

In sum, of the possibilities easily available, 7 in Nardog's list is the most trustworthy source, although for phonetic transcriptions standard phones will have to be shown when the allophone is not obvious to non-native speakers. 7 has a weakness is that not many toponyms are shown, but that can be overcome by consulting Canepari's DiPi, which adopts the same principles and does contain lots of toponyms. Pescara, for example, shows up in 7 only as "variante arcaica di peschiera"; no reason to assume different syllabification of the toponym, but it's good to check, and yes, Canepari confirms that Pescara is pesˈkara, so it can be reported as phonetic [pesˈkaːra].

(Response to 151.64.169.20. I'm neither Italian nor a native speaker of Italian. Born (and raised) in the USA, to semi-quote Bruce. Since I was 19 (decades ago) I've been back and forth between the U.S. and Italy, and specialized in Italian phonology at the graduate level in Italy and the U.S. Your native speaker intuitions are very valuable, especially, but not only, the more you dig into phonetics and phonology. One little caveat to think about, which I'll illustrate with North American English: ask Americans and Canadians if they ever "drop their aitches" (h). Give them Southern Brit 'enry 'iggins for Henry Higgins as an example. If they give you a straight answer without thinking about it at length, they'll probably say no. Then ask them to say "He has a hat on his head" a few times in quick succession. It turns out there's a rather clear hierarchy of h-deletion, increasing through the gradations as speed or "carelessness" of speech increases, only one likely to never delete. And for Italian, just a further clarification. /nk/ (phonemic structure) is common in Italian -- banca and fango are presumably structured /banka/ and /fango/, a simple assimilation rule applies, et voilà, pronunciation with anything other than [ŋ] would be weird, indeed. But what about con Carlo in an utterance like È uscita con Carlo!? In normal speech you almost certainly produce [ŋ]. But what if you slow down just a little bit -- not enough for a pause between con and Carlo, but not fast speech. Does [n] ever emerge? And -- this one's more interesting -- what about a very deliberate, but not unnatural, pronunciation of incomprensible? A judge speaking, a politician, or anyone in a formal situation?) Barefoot through the chollas (talk) 17:31, 20 July 2020 (UTC)

It looks like opinions about this matter aren't univocal (neither users' nor specialists' opinions). For example, the transcription for the word "pipistrello" ('bat') can be found both as /pipiˈstrɛllo/ and as /pipisˈtrɛllo/ depending on the source. We'd better leaving things as they are currently, no need to make a modification which wouldn't have an actual meaning but could confuse a reader's ideas, at most a note in the help page might be added as I said before.--151.64.154.209 (talk) 08:48, 22 July 2020 (UTC)

The slashes you use for /pipiˈstrɛllo/ vs. /pipisˈtrɛllo/ seem to indicate you are talking about phonological/phonemic transcriptions, but the transcriptions on this help page are supposed to be broad phonetic. Love —LiliCharlie (talk) 09:27, 22 July 2020 (UTC)

Thank you LiliCharlie for correcting me. I thought that, about the current case (that is writing either /sˈC/ or /ˈsC/ for Italian words), there was no difference if we use phonemic or phonetic transcription, in fact when I made the example [i.ta.ˈljaː.no] I used on purpose the brackets. Sorry for my imprecision.--151.64.154.127 (talk) 17:03, 22 July 2020 (UTC)

You know, our notation includes things like [ŋ] and [ɱ] which are probably /n/ or /N/ in a phonemic transcription, and [ˈpriːmo] where /ˈprimo/ would certainly be more adequete on a purely phonemic level. That is to say, our transcription style is not phonemic, it is phonetic; Wikipedias in other languages may not see the need of indicating those allophones, other allophones might be indicated instead. — The question that remains is: Is there something phonetic (as well as phonemic) that distinguishes /V.sC/ from /Vs.C/ for at least some Standard Italian speakers? And: Which is the appropriate phonetic representation of /VsC/ for English-speaking Wikipedia users? Love —LiliCharlie (talk) 17:53, 22 July 2020 (UTC)

/pipiˈstrɛllo/ and /pipisˈtrɛllo/ in online dictionaries 5 and 7 are written between phonemic slashes, but it's clear from other examples, such as banca, that 5 intends the transcription to be phonetic, while 7 is genuinely phonemic. Unfortunately, however, 5 is precluded from being a source, as its source is Wikizionario, not at all reliable.

Charlie, I don't understand your last question, Which is the appropriate phonetic representation of /VsC/ for English-speaking Wikipedia users? If we're reporting Italian phonetics, e.g. for Toscana, that's what we report. If we're reporting English, such as Tuscany, we report the phonetics of English. As for Is there something phonetic (as well as phonemic) that distinguishes /V.sC/ from /Vs.C/ for at least some Standard Italian speakers?, yes. You identified it with the maschera example. If the structure is VC.CV and the first vowel is stressed, it won't be lengthened in a normal pronunciation. In something like Pescara or Città di Castello, the syllabifications [sˈC] and [ˈsC] are audibly different, with the first representing normal citation-form pronunciation. Not reporting it would be as inaccurate as not reporting the raddoppiamento triggered by Città. Hope this helps. Barefoot through the chollas (talk) 19:43, 22 July 2020 (UTC)

The representational choice is not only between [sˈC(r)V] and [ˈsC(r)V]. Some authors (Payne (2005) and Krech et al. (2009)) who obviously do not want their transcriptions to indicate/imply the position of syllable boundaries use [sC(r)ˈV] with the stress mark immediately preceding the nucleus, cf. Talk:International Phonetic Alphabet#Placement of stress marks. Maybe English-speaking Wikipedia users should not be given transcriptions that imply, or seem to imply, something that experts and native speakers don't agree on, or where there is variation between deliberate and, errh, careless speech. (Don't foreigners usually make a deliberate effort to pronounce Italian words?) Love —LiliCharlie (talk) 06:45, 23 July 2020 (UTC)

Yes, I know the difference between /ˈpala/ and [ˈpaːla], the first is, we may say, "how an Italian perceives and identifies that Italian word", while the second is exactly how that word is pronounced by mouth and tongue. The "raddoppiamento fonosintattico" is both a phonemic and a phonetic feature: for an Italian /pala/ is different from /palla/ and /la si/ (as in "la si trova qua") is different from /la ssi/ (as in "là sì che c'è"). Don't mix apples and oranges, Barefoot through the chollas. Writing /sˈC/ (or [sˈC]) and /ˈsC/ (or [ˈsC]) is perferctly identical, it doesn't give different information. Indeed, consider that the aim of the symbol /ˈ/ is just indicating the vowel or diphthong where the stress of the voice falls, period. You're spending each time thousand of words to move it from its current place before /s/, which would change absolutely nothing: the stress of [pipiˈstrɛllo] falls on the /ɛ/ (I'll say more: paradoxically, even writing it as [pipistˈrɛllo] would give the same information, that the stress falls on the /ɛ/!). One thing I can't understand is why you're pushing it so much, you're the only one here who seems to be so willing to change its position, and I don't think at all that you salary would increase if you put the stress symbol after the /s/, so I can't really get the purpose, the reason for your urgency.--151.64.153.120 (talk) 07:16, 23 July 2020 (UTC)

Stress affects voiceless segments as well as voiced ones, and not just "the vowel or diphthong where the stress of the voice falls." It is a property of larger units such as syllables and words. Love —LiliCharlie (talk) 08:18, 23 July 2020 (UTC)

Again, thanks for your precisions, maybe speaking about syllable instead of vowel or diphthong would have been more proper. But in this case I don't agree totally with you. If it was as you say, I wonder why different authors user different transcriptions (/sˈC/ and /ˈsC/), both Italian and English authors, in first place. And second, why that sequence transcriptions should differ depending on being in initial or middle position... Nobody, instead, would doubt that "città" triggers the following consonant doubling. I'm not expert in linguistics, I've already said, so I think I can make mistakes in my considerations, but I think we're making a much bigger deal than it is (not to say that it isn't even a deal to discuss, in my opinion).--151.64.156.9 (talk) 09:55, 23 July 2020 (UTC)

I'm not expert in linguistics. It doesn't take much expertise at all to realize that this Writing /sˈC/ (or [sˈC]) and /ˈsC/ (or [ˈsC]) is perferctly identical, it doesn't give different information is untrue, given that the information about syllable structure (and thus, in the detail, stress placement) is quite different. On the other hand, if the goal is merely to show which vowel is stressed, and the quality of the vowel, a form such as pipistrèllo would do. But that's not what English Wikipedia uses for Italian. Until that practice changes, the charge is to present accurate yet very broad phonetic transcriptions according to IPA norms. Barefoot through the chollas (talk) 15:09, 23 July 2020 (UTC)

I suppose that your final subtle innuendo in parentheses of your last reply to Nardog is referred to me. This being so, let me remind you (and allow me to let the other users know) what you were told here about your personal idea of Wikipedia. (All I've done in the previous lines is just linking a discussion that the user named "Barefoot through the chollas" joined containing other users' comments which weren't redacted, nothing more than that, in response to his comments about me.)--151.64.152.49 (talk) 20:39, 23 July 2020 (UTC)

As I wrote in the opening post of this thread, that /s/ in a cluster, word-initially or internally, belongs to a different syllable seems well established from all I can tell. I'm not disputing that. What I have yet to see evidence of is how writing [toˈskaːna] etc. is inaccurate in terms of the realization it represents or how [ˈskaːla] is acceptable but [toˈskaːna] is not. That vowel lengthening is prevented before /sC/ is a non-issue in our transcription, because not only is /sC/ never found between two stressed vowels (or is it?), we mark the allophonic length anyway.

Writing [sˈkaːla] or [skˈaːla] is confusing to readers. Writing [ˈskaːla] yet [tosˈkaːna], not [toˈskaːna], puts an unnecessary burden on the editors who instate transcriptions. Nardog (talk) 15:53, 23 July 2020 (UTC)

Nardog, real quick... Your thoughts are appreciated, and to the point -- for which, many thanks. Writing [sˈkaːla] or [skˈaːla] is not only confusing, but the first is arguably phonetically false and surely puzzling, and the second is either false or a deliberate mis-use of the stress mark. Agreed that accuracy and convention require [ˈskaːla]. [ˈskaːla] yet [tosˈkaːna] may be a bit of a burden, but it seems to me that it's like anything else on Wikipedia, i.e. one contributes what one knows, not what one doesn't know. (My guess is that many/most demanding the syllabification [toˈskaːna] are at least heavily influenced by orthographic norms, perhaps even believe that they're "true", and have little or no knowledge of phonetics/phonology, and almost certainly no knowledge of the state of the controversy among the experts.) Barefoot through the chollas (talk) 16:21, 23 July 2020 (UTC)

I didn't make this "personal", I've never made comments about you, not even masquerading them inside replies aimed to other persons, but you did it to me. Now it seems clear to me that you have a personal interest in changing these Italian phonetic sequences, whatever is the nature of this interest, but so far there isn't a blatant proof that /'sC/ is wrong (instead, the existence of sources which explicitly use such a transcription is a proof that it "isn't" wrong), so don't turn this thing into personal toward who doesn't agree with you.--151.64.152.49 (talk) 21:41, 23 July 2020 (UTC)

Nardog, sorry -- I missed one of your points due to being in a rush. Blocked lengthening of stressed vowel before sC actually is an issue in phonetic transcription, such as [ˈaskoli piˈtʃɛːno] versus that stressed first vowel being long. Just like the case of maschera, if the syllabification were a.sko.li rather than as.ko.li, the first vowel would be long, as in [ˈaːzolo]. Or did I misconstrue what you were saying? Barefoot through the chollas (talk) 21:23, 23 July 2020 (UTC)

Again, since we only (have to) mark syllabification when there is stress, whether we write [VˈsC] or [VsˈC] has no impact on the vowel length each transcription indicates because it's not stressed anyway (unless there are words with e.g. /ˌVsˈCV/). And even if there are, since we explicitly mark the allophonic length in our transcriptions, only the transcriber, not the reader, needs to know /s/ in clusters is heterosyllabic. I wouldn't call that an issue in phonetic transcription. Nardog (talk) 21:36, 23 July 2020 (UTC)

I think we may be saying the same thing, or nearly so, in different ways. I.e. the question of marking syllabic stress for sC clusters arises only if the vowel of C's syllable is stressed. Thus for [ˈaskoli], nothing controversial with regard to placement of [ˈ]. Items like [ˈaskoli] or the family name [ˈpaskolo] (etc. etc.) in turn inform re structure and phonetics of forms like Pescara, Toscana. etc. The stressed vowel of [ˈaskoli],[ˈpaskolo] is not lengthened phoneticaliy, thus the syllable structure is /ˈas.kol.li/, /ˈpas.ko.lo/, implying the same structure for other items: /pes.ˈka.ra/ → [pesˈkaːra], /tos.ˈka.na/ → [tosˈkaːna]. In fact, more than implies, I'd say; an as yet undiscovered principle that licenses as.ko.li but rejects tos.ka.na in favor of to.ska.na would have to be identified. Barefoot through the chollas (talk) 22:27, 23 July 2020 (UTC)

Again, how does that pose a problem to writing [toˈskaːna] etc. in phonetic transcriptions? Nardog (talk) 22:32, 23 July 2020 (UTC)

The structure determines the phonetics. If the structure is /tos.ka.na/, which the evidence of [a] not [aː] in Ascoli argues that it is, the phonetic projection is [tosˈkaːna]. Phonetic [toˈskaːna] implies structure /to.ˈska.na/, which in turn implies /a.sko.li/, shown by the phonetics of [ˈaskoli] to be false. I don't know why this isn't clear. I have to go for a bit. I'll see if I can think of a clearer illustration of the principles involved. Barefoot through the chollas (talk) 23:33, 23 July 2020 (UTC)

Yes, the structure determines the phonetics, but what our transcriptions show is that resultant phonetics, not the structure that determines it. As evident in the inclusion of the allophonic length. Sure, [toˈskaːna] would imply an inaccurate structure if someone reverse-engineered it to the phonemic state, but so would [ˈskaːla]. It shouldn't be any of our business anyone reverse-engineers it because what we show is phonetic, not phonemic. I just don't understand why you insist on [tosˈkaːna] while accpeting [ˈskaːla]. [toˈskaːna] and [ˈskaːla] are equally inaccurate; adopting both at the same time is at least consistent. Nardog (talk) 00:04, 24 July 2020 (UTC),

Ah, so that's the problem. First, one thing, though: nothing is being reverse engineered. Also, /ˈskala/ is inaccurate structurally (phonemically); the pronunciation in isolation [ˈskaːla] is accurate, however. Word-initial syllabic behavior of sC in Italian is revealed in [ˈstu.den.te] but [los.ˈtu.den.te], or, more obviously for /ps/ in the sense that it's very clearly audible, as I think I already mentioned, [psi.ˈko.lo.go] but [lop.si.ˈko.lo.go]. The scala-type thing was explained more fully elsewhere, though. I'll see if I can find the explanation and copy it to here for you. I may not be able to do it immediately. (Do you know Spanish well? If so, there's at least some insight available in the treatment of /sC/ in that language, in the sense that it's an exceptional case word-initially -- as it once was in French, too, but that died out long ago when the /s/ in the cluster bit the dust.) Barefoot through the chollas (talk) 01:19, 24 July 2020 (UTC)

Back in topic, and hoping for no further comments about users (such as me), I'd like to add a piece of information which you may find useful. The /ps/ case is different: "ps" in Italian is rarely encountered, mainly in scientific words ("sepsi", but the adjective is "settico"), and, to my Italian ears, it sounds like /pps/ between vowels in fast speech, more or less as for /tts/ and its phonosyntactic doubling "upside down". I must make an effort to pronounce [au.top.ˈsiː.a], while both [au.top.ˈpsiː.a] and [au.to.ˈpsiː.a] (the latter in a more careful speech) sound normal to me. Again, just a native speaker's impression about a phonetic matter.--151.64.157.228 (talk) 08:49, 24 July 2020 (UTC)

Nice contribution. Thanks. Yes, /ps/ syllable onset differs from /sC/ in that /ps/ is an introduction from high register, not developed "organically" within Italian historically as /sC/ is, and unlikely to be acquired in natural language acquisition at a very early age. It's such a "bad" onset in the Italian phonology machine that it will be broken up if possible in actual pronunciation, thus [au.top.ˈpsiː.a] is not difficult, but [au.to.ˈpsiː.a] requires some concentration. A bit of delay of the onset of the following syllable beginning with [s], et voilà, the sensation of gemination of /p/. Nice.

What's going on is clearer if /ps/ is word-initial. One indication that it is troublesome is the tendency to insert a ghost vowel if the word is said by itself: post-pause psicologo with [pⁱsi...]. Another is lo psicologo [lop.si...], in which /p/ escapes being trapped in the same syllable as the following /s/.

The technical description of /p/ in these instances is that it's extra-syllabic, i.e. in terms of structure, it's really outside of the syllable with the /s/. psicologo begins structurally as /pˈsi/. Thus /p/ escapes to a preceding syllable if possible in pronunciation ([lop.si]) and onset /ps/ is "bad" enough that it can be optionally broken up if the word is said utterance-initial ([pⁱsi]). (The fun bit in that is that those who insert the vowel automatically are unconsciously "obeying" Italian phonology.)

The principles are essentially the same for /sC/, the main difference being that /sC/ is well-integrated historically in Italian and widespread lexically, i.e. not infrequent. It's not at all "difficult to say" -- stavo male, sportello, scavi, no problem, nor fusto, rospo, vasca. It has had a hard life historically in most Romance languages, though, and Italian is no exception. It's settling down now (in [i]Svizzera seems to be just about gone; per scrivere presents no difficulty, so that per iscritto is now a fossil), but there's one peculiarity remaining: the evidence argues that it's still extra-syllabic. maschera is rather clearly syllabified [mas.ke.ra] (ditto mes.tolo, fis.tola, etc.).

I see that Nardog identified extrasyllabicity in his first post in this thread, so there's no need to copy and repeat from elsewhere. As mentioned repeatedly, the guy clearly says [pesˈkaːra] in the recording for Pescara. [tosˈkaːna] is a normal pronunciation, and Trastevere with first syllable [tra] actually sounds a bit strange. That leaves /sˈkala/ pronounced phonetically tautosyllabic [ˈskaːla]. It's part of the nature of extrasyllabicity that the extrasyllabic segment escapes the phonetic syllable if it can. If the cluster is especially "bad", like /ps/, it can escape even in absolute anlaut by inserting a ghost vowel. /sC/ isn't nearly that "bad", so [sk] is acceptable in that position. It's not problematic, nor is it inconsistent with the [tosˈkaːna] type that has the preceding vowel present. It just is. That's the way the extrasyllabicity of /sC/ works in Italian. Barefoot through the chollas (talk) 14:24, 24 July 2020 (UTC)

Good, your reply wasn't off topic and wasn't about users. Thanks for your contribution. All I can do as an Italian native speaker with some knowledge about phonetics is reiterating what I said before about [sˈC] versus [ˈsC]. There's no meaningful difference between them; Italian language hyphenation includes the "impure S" in the following syllable; it's neither unnatural nor strange for an Italian pronouncing [.sC] instead of [s.C]; the purpose of symbol /ˈ/ in phonetic transcriptions is indicating where the stress falls; even linguists don't use a univocal method to transcribe this sequence because some write [sˈC] and some write [ˈsC] (and some others even [ˈsC]) but this fact makes the second absolutely not wrong; in this talk it's you the one pushing to change the current convention; and most important the solution must be with a view to its aim, in this case being simple and clear to the readers as for other solutions adopted for phonetic transcriptions, and using [ˈsC] both at the beginning and in the middle of Italian words is the less confusing (which, inter alia, also stricks to Italian hyphenation). To sum up my considerations.--151.64.158.63 (talk) 18:27, 24 July 2020 (UTC)

Well... I can't think of a way to make this stuff any clearer. In the face of very distinct structures being peremptorily declared to have no meaningful difference between them in spite of the fact that they produce quite different phonetic results (/s.C/ maschera, mestolo, fistula... festa, vasto, rospo, vespa, bosco, casco ... all with the short stressed vowel of a closed syllable -- in every case closed by [s], thus syllable CVs.CV...), ignoring audible evidence such as the "listen" of Pescara (Pistoia is another one, even clearer), dismissing for unknown reason the information kindly provided free by Italy's foremost phonetician (Canepari)... etc. etc. and even the dead obvious seeming to just evaporate into the wind (e.g. the fact that conventional orthographic hyphenation is a completely different issue), there seems to be no way to try, and no point in trying, to have a rational discussion leading to an informed conclusion. Barefoot through the chollas (talk) 19:19, 24 July 2020 (UTC)

Sorry but: "you" say that writing [sˈC] and [ˈsC] is different in terms of correctness; the guy who pronounced Pescara (anyway, see the immediately previous point) also pronounced "Rieti" as [riˈeːti] (which isn't standard Italian) so he may be taken as an example just to a certain point; Canepari may have written /sˈC/ but here we're talking about [sˈC] if I've understood well what you all told me last week (he also writes /*dzinˈkini/ but according to our conventions we'd write it as [dziŋˈkiːni] which is phonetic and not phonemic); etc.--151.64.152.249 (talk) 10:24, 25 July 2020 (UTC)

See above, copied here (emphasized * *):

of the online dictionaries listed, 7 is the most coherent and accurate. *A weakness for the purposes here is that, as presumably noted honestly by the slashes / /, the transcriptions are phonemic.* banca is /ˈbanka/ with /n/ that surfaces phonetically as [ŋ] [...] 7 in Nardog's list is the most trustworthy source, although *for phonetic transcriptions standard phones will have to be shown* when the allophone is not obvious to non-native speakers. 7 has a weakness is that not many toponyms are shown, but that can be overcome by consulting Canepari's DiPi, which adopts the same principles and does contain lots of toponyms.

In simpler terms, online dictionary 7 and Canepari's DiPi use phonemic transcriptions. Whoever does phonetic transcriptions for Italian needs to adjust in accordance with basic Standard Italian allophony.

There's actually an interesting case for Castiglion Fiorentino. Canepari, quite rightly for phonemic representation, gives /kastiʎˈʎon fjorenˈtino/. Whoever did the transcription in the CF article applied two basic allophonic rules -- vowel length in stressed open syllables, partial assimilation of /n/ to adjacent consonant -- to arrive at phonetic [kastiʎˈʎoɱ fjorenˈtiːno]. The point of interest is [ɱ]. It's quite genuine and normal, of course, but it's also true that [n] can appear in deliberate speech, and a non-native speaker would not be "wrong" in using it. Whoever did the transcription took the job seriously and chose to inform readers accurately of what they may not know -- presumably the central purpose of an encyclopedia (rather than worry that [kastiʎˈʎoɱ] but, e.g., [benetˈtɔn] might be confusing to unititated readers). Barefoot through the chollas (talk) 14:36, 25 July 2020 (UTC)

I've said in the previous comment: "he also writes /*dzinˈkini/ but according to our conventions we'd write it as [dziŋˈkiːni] which is phonetic and not phonemic". Your example is out of place, moreover no word in Italian begins with [ɱf], while unlike most of the other Romance languages it's full of words beginning with [sC] in Italian. I repeat that there's no problem in leaving the stress symbol before the "impure S", I'm not the only one here and there're linguists too who claim that. Perhaps there might be, and so there'd be something to discuss, if we were talking about /ˈsC/, but being the talk about [ˈsC] (and about modifying it just in the middle of a word) there's really no issue here even if you don't think so.--151.64.159.26 (talk) 18:33, 25 July 2020 (UTC)

Long but read. At least partially. I agree with the argumentations and the doubts by User:Nardog. I do not see reasons sufficiently strong to make a change to the present rule. Scavvlo (talk) 14:59, 25 July 2020 (UTC)

I'm not aware of a present rule, but assume for the nonce that there is, and that it prescribes that phonetic transcription of /sC/ clusters should be represented as syllable onsets when that's relevant to the transcription. Thus, for example, [toˈskaːna] (with the actual usual pronunciation thus deemed irrelevant, whether it turns out to be [toˈskaːna] or [tosˈkaːna]). No big deal. Except that... phonological systems are exactly that: systems. One thing can impinge upon another.

In this case, prescribing /sC/ as necessarily tautosyllabic onset opens the gates to another transcription quandary, mentioned many times above.

Stressed vowels in open syllables are phonetically long. Wikipedia reports [ˈroːma], [miˈlaːno], [peˈruːdʒa], [kreˈmoːna] and so on, rightly indicating phonetic length in those stressed open syllables. [veˈnɛttsja], [ˈmantova], [ˈdʒiʎːo] (or [ˈdʒiʎʎo]) of Isola del Giglio do not have long stressed vowels because those vowels are in closed syllables. All standard stuff, presumably no controversies.

What to do with Sesto Fiorentino (quite often just Sesto amongst those who frequent the place)? The Wikipedia article gives [ˈsɛsto fjorenˈtiːno]. If syllabification is sɛs.to, the phonetic rendition [ˈsɛsto] is exactly what's expected, as reported in the article. But by the rule prescribing /sC/ as forming tautosyllabic onsets, [ˈsɛsto] is wrong. Syllabification has to be sɛ.sto, and the stressed vowel should be predictably long, just as it is in [miˈlaːno], etc. By the /sC/ onset prescription, the pronunciation is supposed to be [ˈsɛːsto]. But it isn't. It's [ˈsɛsto]. The stressed syllable is closed by [s], the subsequent syllable begins with [t]. Barefoot through the chollas (talk) 18:17, 25 July 2020 (UTC)

Hi Barefoot,

You seem convinced

1. that phonemic representations can unequivocally be derived from (surfacing) phonetic ones; and

2. that the phonotactics of stressed and unstressed syllables are necessarily the same, despite evidence to the contrarary from English.

Why? Love —LiliCharlie (talk) 22:39, 25 July 2020 (UTC)

Hi Charlie,

phonemic representations can unequivocally be derived from (surfacing) phonetic ones

In the case of a language never studied before, the discovery procedure is the other way around: given the phonetics of various utterances (variations in different positions, various speeds of speech, all the while testing for same/different meanings, etc.), what must the structures be? And then, in the other direction, what are the pronunciation rules? What are the principles to churn out the phonetics from the phonemic structures? Once those are known, you can analyze either way. Uniequivocally? Maybe, maybe not. But systematically, yes. If you run into an apparent exception or anomaly (equivocation of what you seem to have established), by all means it needs to be examined. (Crucial point: it has to be a genuine exception/anomaly, not just something the analyst doesn't prefer for extraneous reasons, and an objective analysis. "The data" determine; the analyst is just a conduit.) A plausible exception in the present case would be finding that items like festa, vasto, bosco have a long stressed vowel rather than short. If no principle can be found to account for that, the phonological system is telling us that /sC/ clusters are inherently tautosyllabic. Thus syllabifications pon.te, fat.to, al.to but fe.sta, va.sto, bo.sco -- the first three with phonetically short stressed vowels, the last three with phonetically long stressed vowels. (Short vowels in Sesto, maschera, festa, etc. etc. tell us that there's no such special class for /sC/ or its phonetic projections [sC])

the phonotactics of stressed and unstressed syllables are necessarily the same

I'm going to try to be brief: not necessarily at all. Again, "the data" decide. If there's good (hopefully systematic and coherently describable) reason to analyze them as different, then that's the way to analyze them. I've been through it too many times to have to repeat it much (Goundhog Day), but while the behavior of initial /s/ of /sC/ when an unstressed vowel precedes might seem odd given the phonetically unremarkable banality of [stuˈdɛn.te], a form like [los.tuˈdɛn.te] actually tells the tale; it fits perfectly with [ˈsɛs.to] (and [ˈsɛs.to] "explains" [los.tuˈdɛn.te], while *[ˈsɛ.sto] and [los.tuˈdɛn.te] examined together are at best puzzling). (I've been biting my tongue not to say this, but I will: Be careful with the work you linked. It's very good in some ways, but notice at least the strange syllabifications as a sort of tip of the strangeness iceberg.)

Barefoot through the chollas (talk) 02:07, 26 July 2020 (UTC)

@Aeusoes1, Barefoot through the chollas, and LiliCharlie: Since I don't see any consensus emerging from this discussion, I'm starting an RfC. Nardog (talk) 12:16, 6 August 2020 (UTC)

Requests for comments

In transcriptions using the Help:IPA/Italian key, should word-internal preconsonantal [s] or [z] preceding a stressed vowel be shown as tautosyllabic with the following consonant(s), as in [toˈskaːna, eˈsprɛsso], or heterosyllabic, as in [tosˈkaːna, esˈprɛsso]? 12:16, 6 August 2020 (UTC)

• Tautosyllabic. Our transcription is phonetic, not phonemic, as it shows the allophonic length of stressed vowels and the place assimilation of coda nasals. Although scholarship suggests that /s/ in /sC/ does not belong to the same syllable on the phonemic level, that is the case not only word-internally but word-initially as well, and writing [sˈkaːla] or [skˈaːla] is confusing to readers, while writing [ˈskaːla] yet [tosˈkaːna], not [toˈskaːna], puts an unnecessary burden on the editors who instate transcriptions. Writing [ˈskaːla, toˈskaːna, eˈsprɛsso] is not only consistent but least confusing for both readers and editors (it also happens to align with the orthographic conventions). Nardog (talk) 12:16, 6 August 2020 (UTC)
  • Neutral. I'm persuaded by the point made by LiliCharlie below. To reiterate, my position from the beginning was that the correct syllabification is no doubt [sˈkaːla, tosˈkaːna], but since that would be confusing to most readers it should be [ˈskaːla, toˈskaːna] for consistency's sake. But now I realize that would be taking consistency with what is already a compromise in exchange for theoretical rigor, and as an encyclopedia we shouldn't jettison rigor for editors' convenience. I'm not entirely persuaded to switch sides altogether, but I'd be fine with either way so long as we have a consensus. Nardog (talk) 19:37, 25 August 2020 (UTC)
• I don't understand the reason for opening this can of worms again. Yes, /sˈkala/ is a perceptive and informative deascription of the structure. But as already discussed ad infinitum, since we're doing phonetics, the transcription *[sˈkaːla] is spurious and does not arise in natural apeech; if no vowel precedes, it's phonetically [ˈskaːla]. As for "[tosˈkaːna], not [toˈskaːna], puts an unnecessary burden on the editors who instate transcriptions", I disagree. This isn't Pee-wee's Playhouse. It's an encyclopedia. The necessary burden on the editor is to know the phonological facts for the language in question. I wouldn't dream of doing phonetic transcriptions for, e.g., German, because my knowledge of German phonology, while not non-existent, is not up to snuff. That reluctance should apply to any editor doing transcriptions for any language. An editor who doesn't know how Italian phonology works should simply avoid trying to do transcriptions. Writing [ˈskaːla, toˈskaːna, eˈsprɛsso] is inconsistent with Italian phonology, and often contrasts with recordings available on the article page. (Orthographic hyphenation conventions are entirely irrelevant. French neuf ans does not syllabify [nœv.ɑ̃] in any sort of normal speech, in spite of the two distinct lexemes.) This is actually very simple: initial when not post-vocalic: [sC]; word-internally when post-vocalic: [s.C]. If we're not going to give the phonetics accurately, may as well not give it at all. Barefoot through the chollas (talk) 20:22, 6 August 2020 (UTC)
• I've gotta be honest, this whole conversation seems like a whole lot of work for very little payoff. I don't see any risk of confusion between e.g. [tosˈkaːna] and [toˈskaːna]. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 23:11, 6 August 2020 (UTC)

  Agreed, sort of. I.e. no reason why [ˈskaːla] and [tosˈkaːna] should bother readers. If they notice at all, they'll probably just think "it's different inside words" or some such, which is true. A report of [toˈskaːna] might lead to some confusion eventually, though, the first time a non-native with a good ear hears an Italian say [tosˈkaːna]. To see what I mean, go to the Wikipedia article for Pescara, see the transcription [peˈskaːra], then click on listen to hear the fellow's very clear [pesˈkaːra]. Barefoot through the chollas (talk) 02:25, 7 August 2020 (UTC)

  The reason we're doing this is that there have been edit wars similar to the one described below (see e.g. Special:PageHistory/Città di Castello, Special:PageHistory/Montespertoli). I don't really care if my position prevails; the important thing is to have an outcome (that is not "no consensus") at all so we can stick to it. Nardog (talk) 03:00, 7 August 2020 (UTC)

  Sure. It would be good to have this settled. This has been discussed at great length and the facts are pretty clear (bold added):

  Per quanto concerne la ➔ sillaba, /s/ e /ʃ/ mostrano un comportamento peculiare, che induce una palese contraddizione tra i principi di sillabazione fonetica e quelli convenzionalmente in uso nell’ortografia. [...] la sequenza /s/+ C (consonante) è eterosillabica, i due segmenti appartengono cioè a sillabe diverse. La sibilante chiude la sillaba, poiché rende breve la vocale che la precede e pertanto funge da coda, laddove l’altra consonante costituisce l’attacco della sillaba seguente. L’eterosillabicità di /s/ preconsonantica si apprezza sia all’interno di parola (rospo → ros.po, astro → as.tro, esca → es.ca), che in posizione iniziale (lo sposo > los.po.so); in ambedue i casi la sillabazione diverge da quella ortografica che assegna a una stessa sillaba l’intera sequenza /s/+ C (ro.spo). [7]

  Nevertheless, it seems that at least some of those who want syllabification V-sC word internally are trusting dictionaries that report orthographic hyphenation conventions to provide them with phonology, which those dictionaries have no intention or pretense of doing. As long as that position is held, consensus doesn't seem possible (alas). Barefoot through the chollas (talk) 15:47, 7 August 2020 (UTC)

  "Consensus" on Wikipedia doesn't necessarily mean a situation where all participants have the same opinion. I'm willing to accept whatever the uninvolved closer assesses the outcome of this RfC to be and enforce it, as is expected of every Wikipedia editor in good faith. Nardog (talk) 17:41, 7 August 2020 (UTC)

• Tautosyllabic. No need to complicate our lives, nor editors' and readers', by using different transcriptions for: words starting with [sC]; words starting with [sC] and preceded by vowel; words containing [sC] inside. For what can be seen in articles with such cases, almost always editors had rather using [ˈsC] when they added a phonetic transcription, I wonder why a discussion like this even started. Junghiano (talk) 18:06, 7 August 2020 (UTC)
• Heterosyllabic See above. It started because following a vowel, the cluster /sC/ syllabifies /s.C/ (and is pronounced thus: [Vs.CV]). Explained clearly here:

  la sequenza /s/+ C (consonante) è eterosillabica, i due segmenti appartengono cioè a sillabe diverse. La sibilante chiude la sillaba, poiché rende breve la vocale che la precede e pertanto funge da coda, laddove l’altra consonante costituisce l’attacco della sillaba seguente. L’eterosillabicità di /s/ preconsonantica si apprezza sia all’interno di parola (rospo → ros.po, astro → as.tro, esca → es.ca), che in posizione iniziale (lo sposo > los.po.so); in ambedue i casi la sillabazione diverge da quella ortografica che assegna a una stessa sillaba l’intera sequenza /s/+ C (ro.spo).

For purposes of Wikipedia's phonetic transcriptions, syllabification is shown if stress falls on the syllable following /s/: [tosˈkaːna]. For reasons never really explained coherently in phonetic or phonological terms, articles often have transcriptions like [traˈsteːvere], which comes close to downright weird in actual phonetic terms. Doing it accurately isn't complicating anyone's life. It's reporting the genuine phonetics. (The syllabifications of the sort [toˈskaːna] are mis-sourced, out of dictionaries that show traditional hyphenation slots for writing; they're not intended to be phonemic or phonetic.) Barefoot through the chollas (talk) 18:50, 7 August 2020 (UTC)

• Heterosyllabic. This is the syllabification that phoneticians agree on. And if that puts "burden on the editors who instate transcriptions" it's okay and nothing unusual; encyclopaedic accuracy always requires specialized knowledge and the willingness to expand it, and that's regularly burdensome. Love —LiliCharlie (talk) 13:34, 8 August 2020 (UTC)
• Heterosyllabic I was asked to partecipate as an Italian linguist who studied a a lot of phonetics some years ago; all reliable phoneticians and phonologists who studied Italian (such as Canepari, Fiorelli, Muljačić...) state it's /pes.ka/ and not */pe.ska/, although orthography prescribes pe-sca for analogy with co-scia and similar cases involving s.--Carnby (talk) 17:08, 8 August 2020 (UTC)

Off-topic accusations of bad-faith and disruptive non-sequitors to the discussion

@Carnby: "I was asked to partecipate as an Italian linguist who studied a a lot of phonetics some years ago": this means that, as everyone can verify, Barefoot through the chollas wrote a message in your talk page explicitly because you changed [toˈskaːna] to [tosˈkaːna] and because he knew your opinion about this matter, so he asked you to participate in order to be supported. This is exactly what must not be done to achieve consensus, as written in the section "Pitfalls and errors" of the project page Wikipedia:Consensus. That being so, he committed an infringement of the rules of Wikipedia, and since your vote in this survey is consequent to his request of your intervention aimed to alter the consensus I'm afraid it can't be considered legitimate. @Barefoot through the chollas: "Any effort to gather participants to a community discussion that has the effect of biasing that discussion is unacceptable [...] it is not acceptable to invite only people favorable to a particular point of view [...] but actions that could reasonably be interpreted as an attempt to "stuff the ballot box" or otherwise compromise the consensus-building process are considered disruptive": on this one, I'm not reporting you to sysops, on this one. I invite everyone to consider Barefoot through the chollas's conduct during this discussion, a conduct that confirms a something I said last week: what he was told here about his personal idea of Wikipedia. @Nardog: I promised myself I wouldn't join this survey you've launched and I'd just follow the discussion without intervening, but Barefoot through the chollas's conduct wasn't acceptable and I had to change my mind. Do you agree with what I've written in the lines above about this fact? And other than that, as far as I know a pair of users who disagree on a convention (in this case the spelling [ˈsC] used in almost every Italian phonetic transcription) don't constitute a new consensus, am I right? Maybe it's because we're in the middle of the summer and people have better things to do in this period (even I didn't want to come back here also because I preferred enjoying the summer), but it seems that almost nobody cares about this particular topic except Barefoot through the chollas. However, while I'm here, I'd like to add my vote too: tautosyllabic.--151.64.157.134 (talk) 18:31, 8 August 2020 (UTC) We aren't casting votes. Polling is not a substitute for discussion. Love —LiliCharlie (talk) 18:51, 8 August 2020 (UTC) @LiliCharlie: I'm not interested in fueling arguments. If you consider my contribution here unacceptable just ignore it. But please remember that the most reliable sources about Italian phonetics and phonology say what I've written.--Carnby (talk) 19:50, 8 August 2020 (UTC) Carnby, you didn't do anything wrong and your contribution was valuable. It was the IP who mentioned votes more than once. They don't seem to understand that voting is not the way to reach a consensus. Love —LiliCharlie (talk) 20:38, 8 August 2020 (UTC) Normally I wouldn't respond to spurious charges, but the rants above are too much. Here's what I wrote to Carnby: Carnby, I notice from your IPA corrections of Toscana and prosecco that you seem to have a firm grasp of Italian phonetics and phonology. There's ongoing discussion of syllabification of /sC/ clusters (e.g. [traˈsteːvere] vs. [trasˈteːvere]) and of reporting or not RS ([tʃitˈta ddi kaˈstɛllo] or [tʃitˈta di kaˈstɛllo]; [oltreˈpɔ ppaˈveːze], [oltreˈpɔ paˈveːze], etc.) that you might be interested in. Now at the survey or request for comments stage, looking for consensus to establish general English Wikipedia policies for Italian transcription. I had seen two of his corrections, for Tuscany and prosecco ([e], not [ɛ]). It seemed that he is one who knows his Italian phonetics and phonology, i.e. one in possession of informed knowledge, not just muddled impressions. (Wikipedia contents should not be based on "opinions", "points of view", "sides", etc., but, as much as possible, on knowledge.) I noted that there was discussion of syllabification of /sC/ clusters and of raddoppiamento that might be of interest to him. Obviously, I knew that he knows the syllabification of /sC/. I had and have no idea what his well-reasoned position may be regarding the use of RS for Italian transcriptions in English Wikipedia articles. I did not ask Carnby to contribute; I quite rightly left that decision entirely to him. He decided to do so, with admirable succinctness (and a lagniappe of good explanation of why orthographic hyphenation of /sC/ is not phonological). His contribution is perfectly legitimate and as welcome as (at least) any other. At this point someone owes him an apology. Perhaps I do, for unwittingly setting him up to be attacked for the sin of sharing his knowledge. If I do, there's another who owes him a much greater apology. Barefoot through the chollas (talk) 20:12, 8 August 2020 (UTC) @LiliCharlie: I've talked about "votes" improperly, it's because in Italian we talk about "votes" both for surveys and for polls. But the point is the same: Barefoot through the chollas wrote in a particular user's talk page an invitation, "particular" because he's a user whose opinion about this matter is the same he has. So much that here he wrote: "Carnby set this right last November.". He already knew Carnby's opinion about this, he already knew Carnby would support him. This is a fact, not an opinion. And that's exactly the conduct blamed in Wikipedia:Consensus. This is a fact too. Votes are another thing, all right, we're talking about consensus here: the sentences I wrote can be found in the project page Wikipedia:Consensus. As to your understanding of vote: I thought as much. That's why I provided the link to Wikipedia:Polling is not a substitute for discussion. BTW, in the languages I know best people use the same word for vote and voice (raised), too. However I grew up in Germany and Belgium, and both are highly consensual societies with a lot of respect for minority views because they also reflect large portions of the population, and where even oppositional parties have a great influence on legislation. I'm a bit like Nardog whose arguments I consider highly respectable. And I am also willing to support whatever consensus we reach. Still I'm in favour of transcribing heterosyllabicity, but none of the parties are insane. Love —LiliCharlie (talk) 22:29, 8 August 2020 (UTC) Wait, I'm not saying that the arguments expressed by users who prefer [sˈC] to [ˈsC] don't deserve respect or that they're insane. I didn't even talk about the syllabification of [sC] in my last comments, I've come back just to point out that Barefoot through the chollas's conduct in this case has been unfair, at the very least. I know I didn't speak highly about him in the previous weeks, but sincerely speaking I didn't expect him to do something like that, I couldn't believe it when I read that he asked to participate a user whose opinion about this matter he already knew.--151.64.156.130 (talk) 06:33, 9 August 2020 (UTC) @Barefoot through the chollas: I invite you to read my reply to LiliCharlie. You mentioned Pescara a lot of times: why didn't you invite to join this survey also the user who edited the phonetic transcription in the opposite way to Carnby? Why did you choose Carnby? It's because he knows his Italian phonetics and phonology? But you spoke about "correction" in this last comment of yours when you wrote about his edit, yet it isn't a "correction" but an edit that you would too, as opposed to his edit about Prosecco which is actually a correction... I could have asked for other editors' help too, particular editors whose opinion is the same I have, such as the ones who added phonetic transcriptions for Italian soccer players for example (Cristiano Biraghi, Cristiano Piccini). I didn't because it wouldn't have been fair, and against rules instead. But if I'd done would you have taken well to that? @Carnby: I want to tell you that you didn't do anything wrong too, you've been tricked but I know you were in good faith when you came here to tell us your opinion, an opinion that Barefoot through the chollas was already aware of when he invited you to join this survey.--151.64.156.130 (talk) 21:48, 8 August 2020 (UTC) Can we cut the assumptions of bad faith here? Accusing other editors of canvassing is a serious breach of WP:AGF. If you want to make the case that an editor has canvassed (and, to be blunt, there's not much of a case for it), do so in the appropriate noticeboard. Do not use it as a rhetorical tool to influence a discussion. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 23:18, 8 August 2020 (UTC) Excuse me, but how would you judge the conduct of a user who, finding himself apparently outnumbered in a survey like this, invites to join the discussion the only user he was sure that he'd support his opinion? If now I go and invite one of the users who'll support my opinion to join this discussion, will you assume good faith and defend me from whoever says that my conduct was unfair and that I broke the rules written in Wikipedia:Consensus? I didn't join this discussion again until I read that Carnby, who changed [toˈskaːna] to [tosˈkaːna], was invited here by Barefoot through the chollas, I didn't want to influence the discussion and didn't care so much about its outcome, but faced with this misconduct I couldn't just stand by and watch. If my attitude seems bad or too much for you please forgive me, I'll try refraining a little more.--151.64.156.130 (talk) 06:33, 9 August 2020 (UTC) You're assuming quite a bit and acting like your assumptions are established facts. They're not. The manner in which you're going about this is way out of line. Either drop it or report the user. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 16:33, 9 August 2020 (UTC) I'm sorry that you think that, I'm just putting two and two together based on what I read in Wikipedia:Consensus. Well, I have a pair of questions. Where could I report this misconduct (should it lead to an alteration of consensus because of the specific invitation to "Carnby" to join here)? If I do the same and ask another user to join this survey knowing that he'd support our position, will I be considered "not guilty" like him or will this constitute a breach of the rules? I'll decide what to do according to your answer. Probably I won't write any more here, currently there's no consensus to change the "status quo" which is writing [ˈsC], and a pair of users who disagree on that don't constitute a new consensus.--151.64.154.24 (talk) 17:36, 9 August 2020 (UTC) See WP:Canvassing#How to respond to canvassing. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 00:12, 10 August 2020 (UTC) I've already seen it. When Carnby joined this discussion after being invited by Barefoot through the chollas (because he DID). And you defended him (because you DID). But if I say I'd like to do the same I'm canvassing, unlike him. Double standards are such a great thing, aren't they?--151.64.154.24 (talk) 06:48, 10 August 2020 (UTC) Good idea. And then perhaps give some thought to whether informing one person who seemed to have "a firm grasp of Italian phonetics and phonology" that two topics dependent on knowledge of that field were under discussion and that he might be interested in them -- no "invitation", no "asked to participate" -- constitutes misconduct of any kind, much less disruptive canvassing. Barefoot through the chollas (talk) 02:11, 10 August 2020 (UTC) I have to repeat myself, and I hate it. Step 1: you wrote "Carnby set this right last November" when you changed again [toˈskaːna] to [tosˈkaːna] restoring Carnby's spelling. Step 2: this means that you were aware of Carnby's idea about which spelling was better for him. Step 3: you told Carnby about this survey hoping he would join. Step 4: you didn't do the same with any other editor including the ones who added or modified any phonetic transcription using the other spelling you don't like. Step 5: Carnby joined and with his intervention the consensus has undergone an alteration. Step 6: this wouldn't have happened if you didn't invite him and only him here. Now, don't take me for a naive simpleton, I don't care which words you used in Carnby's talk, you let him know about this survey, you already knew what he'd say if he joined, you obtained support to your position. This is exactly what's described in Wikipedia:Consensus in the section "Pitfalls and errors". Thank Aeusoes1 now I know that this conduct is described also in another project page, Wikipedia:Canvassing, where I read that "Canvassing is notification done with the intention of influencing the outcome of a discussion in a particular way", that it "is considered inappropriate", that "This is because it compromises the normal consensus decision-making process", and that "therefore is generally considered disruptive behavior". You notified about this discussion Carnby and only Carnby, and even if you say that you did because he's a great expert of Italian or something like that your good faith in this case is at the very least doubtful because of your edit summary about Tuscany. I'll go and ask some sysops for an opinion about your conduct, perhaps it's really legitimate but until a sysop tells me it is I'll keep thinking it is not. You're free to go on discussing, probably I won't interfere again here, after all currently there's no consensus to change Italian phonetic transcriptions in accordance to just a pair of users' preference. While I'm here, I'd like repeat myself one more time (just for you to be sure about mine): tautosyllabic.--151.64.154.24 (talk) 06:48, 10 August 2020 (UTC) Be sure to tell the sysop that there's an even more regrettable aspect of my misconduct: I looked it up in Treccani's Enciclopedia dell'italiano and reported the results here. By all means recount that therein, Patrizia Sorianello, co-author of Manuale professionale di dizione e pronuncia (Hoepli, 2011), and who holds the Dottorato di ricerca in Linguistica italiana, states very clearly that la sequenza /s/+ C (consonante) è eterosillabica, i due segmenti appartengono cioè a sillabe diverse. La sibilante chiude la sillaba, poiché rende breve la vocale che la precede e pertanto funge da coda, laddove l’altra consonante costituisce l’attacco della sillaba seguente. If announcing a discussion to a Wikipedia contributor who seems to have some expertise is punishable as misconduct, looking up the published statement of a guaranteed expert must be a far more grave offense. That, plus your explanation in phonological terms of why the stressed vowel in items like Sesto, maschera, etc, is not long, why festa and vostro, for example, are not fiesta, vuostro as would be expected of erstwhile Ĕ and Ŏ in stressed open syllables (e.g. piede, buono), and why the /s/ of an initial /sC/ resyllabifies phonetically when preceded by a vowel (stu.den.te but los.tu.den.te) should seal the case. Barefoot through the chollas (talk) 13:34, 10 August 2020 (UTC) Ok, this is enough, both of you. Anon, I've already asked you to stop and now you are actively being disruptive. Stop. Barefoot, this is not the place to defend yourself and responding is becoming akin to troll feeding. Stop. I'm collapsing this off-topic thread. If you two want to stick your tongues at each other, do so elsewhere. — Ƶ§œš¹ [lɛts b̥iː pʰəˈlaɪˀt] 13:54, 10 August 2020 (UTC) Fair enough, and thanks for shutting this down. My interest is in presenting accuracy to Wikipedia readers. The facts, as much as they can be ascertained, are presented clearly (and, ahimé, repeatedly at much too much length) in the portion of this section still visible. That should suffice. Barefoot through the chollas (talk) 14:18, 10 August 2020 (UTC)

RfC: When should we transcribe word-initial gemination?

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

---

Background: There appears to be a long-term, wide-ranging, low-key edit war going on in some articles such as Luca Zingaretti, Città di Castello, Life Is Beautiful, Dolce & Gabbana, and A cappella concerning the transcription of phrase-internal word-initial geminates. Without an explicit consensus any remedy could simply prolong the war like whack-a-mole, so I'd like to first gauge the opinion of the community. Nardog (talk) 12:17, 6 August 2020 (UTC)

Word-initial gemination in Standard Italian occurs in the following contexts (Payne 2005; Bertinetto & Loporcaro 2005; Maiden & Robustelli 2014):

• "Intrinsic" gemination: /ts, dz, ʃ, ɲ, ʎ/ following any vowel, as in la sciarpa
• Syntactic gemination (raddoppiamento sintattico):
  • Any consonant following a stressed vowel (and not followed by an occlusive), as in virtù diversa
  • Any consonant following a certain preposition or conjunction (and not followed by an occlusive), as in e Roma
  • /d/ in dio or its inflected form following any vowel, as in gli dei

Which of these should be reflected in transcriptions using the Help:IPA/Italian key? 12:17, 6 August 2020 (UTC)

Survey

• Transcribe all. I'm having a hard time understanding why some users have removed word-initial gemination in the first place. Any description of the sound system of Standard Italian mentions both types of gemination as far as I've looked. Although apparently northern varieties generally lack the syntactic type, they still exhibit the "intrinsic" gemination, especially /ts, dz/ (Bertinetto & Loporcaro 2005:134), and even syntactic gemination is spreading in Turin (Crocco 2017:111–2). Write [la ʃˈʃarpa, virˈtu ddiˈvɛrsa, e rˈroːma, ʎi dˈdɛi], or if a reasonable number of people find it confusing, write [la‿ʃˈʃarpa, virˈtu‿ddiˈvɛrsa, e‿rˈroːma, ʎi‿dˈdɛi], with the undertie often used for liaison in French, as done in the Syntactic gemination article. Nardog (talk) 12:17, 6 August 2020 (UTC)
• I agree we should transcribe all geminations that are obligatory in Standard Italian. — I'm not sure, but if we use underties we could also use spaces and write [la‿ʃ ˈʃarpa, virˈtu‿d diˈvɛrsa, e‿r ˈroːma, ʎi‿d ˈdɛi] which perhaps better preserves/shows the isolated "citation" forms of the constituent words. — I don't understand why the copula è in La vita è bella is transcribed with a stress mark. Love —LiliCharlie (talk) 14:24, 6 August 2020 (UTC)
• This topic has been discussed before, notably at [8] (December 2015).
  At that time, the consensus was that we should not show the syntactic gemination symbol (*) on individual words, except in articles that are specifically about Italian phonology, but that we should show it in the interior of phrases.
  On reflection, I'm not sure I even agree with the "interior of phrases" case, because many varieties of Italian spoken outside central Italy do not show syntactic gemination at all.
  Given the variability of Italian phonetics among speakers of standard Italian, it's not even clear that we should try for phonetic (as opposed to phonemic) transcriptions in general. --Macrakis (talk) 17:05, 6 August 2020 (UTC)
• Transcribe all. Standard non-regional Italian version. Must admit I don't understand why there's any question about this. As for the transcription norms, I vote for phonetic, in this version: [la ʃˈʃarpa, virˈtu ddiˈvɛrsa, e rˈroːma, ʎi dˈdɛi] (the ligatures are unnecessary, and might even be confusing). Barefoot through the chollas (talk) 19:34, 6 August 2020 (UTC)
• Transcribe all. The right transcription should be like [la ʃˈʃarpa] with a space between the two words. It's the same thing for words with reverted syntactic gemination, very rarely encountered in Italian and almost always for foreign words, when a word ending with [ts, dz, ʃ, ʎ, ɲ] precedes a word starting with a vowel and the preceding consonant doubles. Other solutions would be confounding: [la‿ʃˈʃarpa, laʃˈʃarpa, laʃ ˈʃarpa]. Junghiano (talk) 18:10, 7 August 2020 (UTC)
• Transcribe all. Per other users mentioned related points. Ali Ahwazi (talk) 14:13, 4 September 2020 (UTC)

Discussion

Let's do the Wikipedian thing and consult a Reliable Source:

Florentine features, such as... syntactic doubling as in [a kkása] are not used in actual fact in northern Italy and do not even constitute a theoretical model which people there try to imitate: they are felt to be either parochial and alien, or affected.

Note that the authors (university professors of Italian) are intimately familiar with this feature, and devote several pages to it in the section on Florentine phonology (p. 67-69), but are clear that it is not a feature of all prestige versions of standard Italian, and indeed even in central and southern versions which have it, the pattern is different (p. 75).

So I don't believe it should be notated. --Macrakis (talk) 23:45, 6 August 2020 (UTC)

Even the sources given above agree:

Raddoppiamento also applies in RI [Rome] and FI [Florence], ... Most Central and Southern varieties only possess the latter type of raddoppiamento, although with a lexical distribution that varies from place to place.... By contrast, raddoppiamento is unknown in MI [Milan], as in all Northern varieties. -- Bertinetto, p. 135

...the extent to which the rules of [raddoppiamento] can vary regionally... -- Maiden, p. 11

Type 4 is present also in other central and southern dialects and varieties of Italian, though trigger words vary regionally. Type 3 is present only in certain central dialects and varieties of Italian... -- Payne, p. 154, where Types 3 and 4 are varieties of raddoppiamento.

--Macrakis (talk) 00:21, 7 August 2020 (UTC)

1) The Lepschys are from the north (nevertheless, the first place I heard the striking unmistakable RS of (Italian pronunciation: [oltreˈpɔ ppaˈveːze] was in Pavia, from a native of the Oltrepò). 2) We're not doing either Florentine or Northern Italian. Sources should be describing region-neutral Standard Italian. Barefoot through the chollas (talk) 01:53, 7 August 2020 (UTC)

Yes, we should be transcribing neutral Standard Italian. Except that that doesn't really exist. All four sources make it clear that RS varies considerably by region. Which is why we should not try for a narrow phonetic transcription that commits us to a particular regional variant. Raddoppiamento is not consistent across regions, but is predictable given the region. So we should not be notating it. --Macrakis (talk) 02:33, 7 August 2020 (UTC)

Raddoppiamento ... is predictable given the region You just quoted Bertinetto & Loporcaro and Maiden & Robustelli saying the triggering words vary by region. Nardog (talk) 03:42, 7 August 2020 (UTC)

Exactly. What are you disagreeing with? --Macrakis (talk) 16:29, 7 August 2020 (UTC)

Don't northern varieties often lack word-internal gemination as well? But I don't think anyone would advocate doing away with that. And, in the other direction, some of them can have RS (see above). Also, Lepschys' proposal (or the idea of northern-influenced pronunciation as the new prestige replacing the Tuscan-based standard) is far from being accepted (Crocco 2017:110). I don't see any reason to deviate from describing the established standard in this case (unlike with #Intervocalic ⟨s⟩, where sources do suggest the northern influence is taking hold), the biggest and obvious benefit of which being it is easier to find reliable sources for individual words. Nardog (talk) 03:42, 7 August 2020 (UTC)

Agreed. Neutral Standard Italian does exist as a diastratic/diaphasic register, reported in just about any dictionary. Canepari, using his term co-gemination for RS, makes it clear, while also observing that it's creeping in in the north, in "chunks": "Co-gemination is part of neutral pronunciation, exactly as lexical gemination, which is marked in spelling [...] However, this is not the case in the north (natively, except in some common, set expressions, as è vero, ha detto, used by young people raised with high levels of exposure to the television)." Its status is also widely misunderstood by the general population: "Too often, it is erroneously considered as if it were a regional characteristic of the central-southern areas." p. 139 ([9]) It's true that RS varies diatopically in italiano/i regionale/i. But so do lots of other local features, all of which are irrelevant to providing Standard Italian citation forms for non-native speakers. Città [dd]i Castello, Oltrepò [pp]avese, etc. Barefoot through the chollas (talk) 14:54, 7 August 2020 (UTC)

All four sources we have are clear that RS varies significantly across regions in Standard Italian, not just between the North and the rest, but even between the Center and the South. The Lepschys' proposal about what variety to teach to foreigners is irrelevant here; the material of theirs I quoted above was descriptive.

Including RS is both pedantic and misleading. We should stick to a more diaphonemic notation. --Macrakis (talk) 16:29, 7 August 2020 (UTC)

We don't expect users to be familiar with the phonology of foreign languages, so our foreign-language transcriptions are not phonemic let alone diaphonemic. Love —LiliCharlie (talk) 17:11, 7 August 2020 (UTC)

Well said. Phonemic transcriptions would reduce the information offered, not enhance it (even less helpful, diaphonemic). Including RS is accurately descriptive of Standard Italian; not noting it would mislead non-native speakers. Again, yes, it's true that RS varies diatopically in italiano/i regionale/i -- good fodder for the Regional Italian article, but an obstacle to offering guidance to pronunciation of Standard Italian ("italiano senza aggettivi"). Barefoot through the chollas (talk) 17:47, 7 August 2020 (UTC)

There's not much to discuss, syntactic gemination is a phenomenon of standard Italian and occurs in a set number of cases. In a phonetic transcription it should be shown, unless there's a valid reason to propose not to show it here. It doesn't matter that in some regions it doesn't exist or it's different, we're talking about standard Italian. Maybe we could discuss about showing it or not in doubtful cases, for example the preposition da or the conjunction se which don't mandatorily double the following consonant, or some adverbs such as qui and là which south of Tuscany double the first consonant when preceded by vowel. Talking about the articles mentioned above, any of us could add the syntactic gemination right now, I can do it by myself if you agree, I've already fixed a few articles with wrong phonetic transcriptions and I'm looking for further errors to fix. Junghiano (talk) 18:14, 7 August 2020 (UTC)

This is an attempt to summarize the above discussion: 1. Everyone who accepts the existence of a "Standard Italian that requires no further qualification" (italiano senza aggettivi) is in favour of transcribing syntactic gemination and "intrinsic" gemination of /ts, dz, ʃ, ɲ, ʎ/, but one person argued that there are "regions in Standard Italian". So it seems a consensus can be reached if that person declares: "I believe syntactic and 'intrinsic' gemination are subject to regional variation in Standard Italian, but since Wikipedia's foreign-language transcriptions are phonetic, not diaphonemic, and we therefore have to decide on one form of speech anyway, we can choose the one that my fellow editors have in mind when they say Standard Italian, which is certainly within the limits of my notion of Standard Italian; it is the form used in Central Italy." — 2. Some said they favour [la ʃˈʃarpa]-style trancsriptions without underties, and no one said they prefer any other transcription style. Love —LiliCharlie (talk) 13:32, 9 August 2020 (UTC)

Standard Italian has "no regions": it is a sort of compromise between the central dialects (Tuscany and, to a lesser extent Rome) and Northern Italian. Basically it is a Florentine-based Italian without gorgia, scempiamento and other features perceived as dialectal (or vernacolari in Italian, since we don't use the term dialetti for Tuscan dialects). That "Standard Italian" is an abstraction: no one speaks it at home, but it's nevertheless very important. It's the form of Italian you'll find on most prestigious pronunciation dictionaries (such as DOP – Dizionario d'ortografia e di pronunzia by Migliorini, Tagliavini and Fiorelli and DiPI – Dizionario di pronuncia italiana by Canepari) and – most important of all – it's the language spoken in TV programs, by actors and by dubbers (remember that dubbing has a very important role in Italian culture since shows and films in original language are extremely rare in Italian).--Carnby (talk) 16:28, 9 August 2020 (UTC)

The original question was of a purely practical nature, and I believe we can reach a consensus whether or not we agree on the exact definition of Standard Italian (variation within Italy, Standard Italian of Switzerland, etc.). Love —LiliCharlie (talk) 17:42, 9 August 2020 (UTC)

References

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