Talk:Italian phonology/Archive 3

flap vs trill
I take it from the wording that the diff tween /r/ & /rr/ is a long vs short trill, not flap vs trill as in Spanish, so I made this explicit. A one-vibration trill is not the same as a flap (the aerodynamics differ), but please correct me if I got it wrong. kwami (talk) 14:19, 6 May 2009 (UTC)

You're wrong, though not on here, Italian has flap R exactly like Spanish, it's a natural allophone in all languages having trill R.. sadly Wikipedia won't accept it —Preceding unsigned comment added by 80.182.77.236 (talk) 00:47, 9 July 2009 (UTC)


 * Where is your source for that? In Spanish, the flap is not at all an allophone of the trill. They are different phonemes, with a minimal pair such as "perro" / "pero". You can't use a trill in the latter. You can in Italian. --LjL (talk) 01:00, 9 July 2009 (UTC)


 * Not to undermine your authority, LjL, but according to some pretty prominant authorities regarding Spanish phonology, the flap and trill are basically the same phoneme with the trill surfacing always when geminated. — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  01:33, 9 July 2009 (UTC)


 * Eh, maybe, after all when it comes to details such as this it's mostly just a matter of how one decides to analyze it, isn't it? I find it's a peculiar analysis, though, because it would make it basically the only geminated consonant in Spanish - Occam's razor and all... --LjL (talk) 13:25, 9 July 2009 (UTC)
 * Only a decade late: Bingo. I'm reluctant to label it a true phonological geminate, but it's definitely long, the only phonologically (phonemically) long consonant left in Spanish. Length loss was hierarchical in the history of Hispano-Romance, in Castilian running through all of the inventory except /nn/, /ll/, /rr/ as degemination. /nn/ and /ll/ palatalized, leaving only rhotic length. Remnants of the late stages were evident a few decades ago when Badía Margarit reported on the speech of Bielsa, where long liquids and nasals were still possible (although his exemplification of /mm/ is a bit iffy). Hispanists are loathe to recognize this, thus do the world the disservice of transcribing the rhotic in, e.g., carro without overtly signaling length. Barefoot through the chollas (talk) 14:00, 27 April 2019 (UTC)

Here's my source : http://venus.unive.it/canipa/pdf/HPr_03_Italian.pdf .. Luciano Canepari canIPA, one of the best phoneticians all over the world.. flap R in Italian EXISTS, I'm Italian and I always use it, sometimes, even replacing the trill.. but it's mainly the big allophone of the /r/ sound in unstressed syllables.. Saying a /r/ in unstressed positions is almost impossible and considered ridicolous.. —Preceding unsigned comment added by 80.182.77.236 (talk) 01:17, 9 July 2009 (UTC)

And this so called famous "dictionary" of Canepari, no longer exists. please quote competent sources. use for example the website of the academia della crusca, or the dictionary of devoto-oli or the Treccani, not the invented ones. So lame. — Preceding unsigned comment added by 2.233.66.85 (talk) 21:24, 1 June 2019 (UTC)


 * So you don't use a /r/ in "arrivato"? Because, you know, that's an unstressed syllable. As for your source - yes, I see what it's saying; I don't think it matches my pronunciation, however, and anything that prescribes allophones for a language with as many varieties as Italian should be taken with a grain of salt. Feel free to add information about usage of flaps, for what I'm concerned, but please don't do it the way you did in your last edit, but turning a cautious "sometimes" statement into a blanket "always, everyone, everywhere" one. That will just earn you more reverts, you know. --LjL (talk) 01:23, 9 July 2009 (UTC)

Yes, I say arrivato with a double flap, and rarely with a trill... this is standard Italian... ever heard something using /r/ for caro, or diario?? Come on... —Preceding unsigned comment added by 80.182.77.236 (talk) 01:53, 9 July 2009 (UTC)


 * And how does one pronounce a double flap, exactly? And yes, in very careful pronunciation, I can very much see non-geminated "r" being pronounced as a trill (quite possibly with two touches). Anyway, the point is also that a one-touch trill is not the same thing as a flap - or at least, that's what Trill consonant and Flap consonant, and my sense of hearing agrees although that hardly matters. While a trill might possibly only have a single contact, though, I really can't think of something like a double flap. --LjL (talk) 02:00, 9 July 2009 (UTC)
 * I've never heard of a double flap. Since we're getting into phonetic particularities here, remember that we should find published sources.  Native speaker insight and OR will often steer us in the wrong direction when it comes to phonetics.  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  02:43, 9 July 2009 (UTC)
 * Of course, but I think I'll let my native speaker's common sense guide me in deciding which statements to require sources for, and the like. 80.182.77.236 does provide a source... I'm not sure it looks like a very peer-reviewed, paper-published reliable source, but at least it's not just hot air. It does feature "double" flaps in its transcription, however; as I said, I really doubt such a thing exists... maybe, by duplicating the phone, it means long (geminated), but does that exist, either? --LjL (talk) 13:15, 9 July 2009 (UTC)


 * This discussion belongs to years ago, so I hope that my comment will still be helpful. First of all, a necessary preamble: consonants in Italian can be double or single, as is well known; however, what is often overlooked in this description is that this contrast can only occur in certain positions, i.e. in between vowels or between a vowel and a "semivowel" (/w/ and /j/) or, for plosives, between a vowel and "r"/"l"; in all other position all consonants have an indeterminate length. So in the word "pasto" ("meal") neither the "s" nor the "t" can be described as being "single" or "double", they have an intermidiate, indeterminate length (pronouncing something like "passto" or "pastto" would be meaningless in Italian phonology, unlike other languages featuring geminate consonants like Finnish). So, applying this concept to the pronunciation of "r": single "r" is pronounced as a flap, double and indeterminate "r" are pronounced as a trill in standard pronounciation; for example "caro" (dear) contains a flap, while "terra" (earth) and "treno" (train) both contain a trill. You can check this listening to the entries of the "Dizionario di Ortografia e Pronuncia", published by the Italian state television RAI and which is (or used to be) available on this page. Using the trill in "caro", no matter how short, would inevitably turn it into "carro" (charriot).


 * I am a native speaker, I've taught Italian to foreigners and I have studied and compared Italian phonology in depth. Of course, this claims are not sufficient for wikipedia, but there's an inherent problem when referencing publications on the subject. For long time phonology has been considered in Italy just a lesser part of grammar; more importantly, most study focussed on internal controversies and variations, i.e. describing the language from within the language, a more objective and scientific study of the phonetics is quite limited. It suffices to say that even IPA has hardly ever been used in the description of Italian phonetics, many authors using custom phonetic conventions even in promimenent works (such as the one I linked before). This means that while the contrast between /e/ and /ɛ/ or /ɔ/ and /o/ is widely discussed in many pubblication because there's a lot of variation in distribution, little work can be found about other distinctive traits of Italian phonology that are uniform or nearly so among native speakers (e.g. the assimilation of place of articulation of "n" with the following consonant; the lengthening of the vowel in stressed, open syllables inside words; phonetic value of syllables like in /pas.to/ contrasting with the prescribed orthographic split "pa-sto", syllable-timed rythm, etc.). The difference of distribution of flap vs trill for "r", as well as an accurate description of consonants of incontrastive length in general, has been largely disreguarded. Geon79 (talk) 00:51, 10 October 2012 (UTC)

I found a sourcepp. 213-214 which reviews Canepari (1999:97-98), who says: "“[N]ella pronuncia neutra odierna effettiva abbiamo, normalmente [r] in sillaba accentata: [(C/V)ˈrV-,ˈCrV-, ˈVrːC(V),ˈV(ː)r#] (oppure, solo come variante occasionale, non sistematica, e non enfatica, [ɾ]). Mentre negli altri casi si ha [ɾ]: [ˈVːɾV, (V/C)(ˌ)ɾV-, Vɾ-, -ɾ(ˈ)C-] (oppure come variante possibile, specie per enfasi, [r]). Per /rr/ si ha: [ˈVrːɾV, VɾˈrV, (ˌ)VɾɾV, Vɾ(ˌ)ɾV] (oppure anche [rːr, rr], soprattutto per enfasi).”" However, the analysis concludes that [ɾ] "is restricted to the intervocalic unstressed position or to the ‘explosive’ phase of /rr/", and that the other supposed systematic realisations of it, i.e. in non-intervocalic unstressed syllables, are just shorter trills with fewer taps. As for the earlier discussion, my thinking is that since [r] is a particularly hard sound for non-native speakers to pronounce, it tends to be overdone (louder, longer, with more breath) compared with natural Italian, and so perhaps for someone learning the language [ɾ] is a better approximation in these cases.

In any case, if carro vs caro shows a phonemic contrast it should be added to the list of phonemes. ─ ReconditeRodent « talk · contribs » 15:03, 7 October 2018 (UTC)
 * Good point. Perhaps it's possible to do it without triggering too much quarreling, so that pala/palla, ano/anno, casa/cassa, fato/fatto, etc., can handled along with it.47.32.20.133 (talk) 16:14, 7 October 2018 (UTC)
 * wait, what is going on here? It’s true in certain positions short is realized as a flap, but wouldn’t it be much easier to simply add a note? I think it’s pointless to have a help that lists it separately when in Italian all consonants (except  which is always short) contrast by length and not by quality. [[File:Italy.png]]  イヴァン スクルージ 九十八 （会話）[[File:Italy.png]] 21:54, 7 October 2018 (UTC)
 * The help pages are organised around the IPA, not just letters or phonemes, so, for example, the Help:IPA/Standard German page lists both [ç] and [x] even though they are in complementary distribution, which is useful because it means we can a) give examples of the context(s) in which the sound occurs and b) be more precise about the best approximation for an English speaker, which is the ultimate goal of these pages after all. This second point is particularly relevant because quite a few English accents feature this sound, and so for people who speak or can imitate those accents an entry means pronouncing a sound you already know instead of trying to pronounce a new one which it seems here would sound less natural anyway. A footnote might be preferable if it were only found in a regional dialect, but my understanding is that this is a common to universal feature of Italian. ─ ReconditeRodent « talk · contribs » 23:41, 7 October 2018 (UTC)

I've realised that carro and similar words are usually transcribed two /r/s, as in, so the distinction with words like caro probably doesn't count as a phonemic contrast like in Spanish. ─ ReconditeRodent « talk · contribs » 23:41, 7 October 2018 (UTC)
 * If the difference weren't phonemic, caro and carro couldn't contrast, and they do. If you mean are "geminates" separate phonemes from singletons in Italian, the question is almost as fraught as it is for Spanish, but helped along by phonological behavior strongly suggesting that not just geminate plosives, but even long liquids and nasals are structurally distinct. Singleton V.CV vs. VC.CV. For Italian this is a a pseudo-problem (and quite possibly for Spanish as well). 47.32.20.133 (talk) 13:39, 8 October 2018 (UTC)


 * yes, but those allophones are listed in the helps because they are actually either a standard or a pretty much universal feature of the language, while in Italian short and  are in free variation, since the phonemic contrast stands rather between  and, as for all other consonants; for consistency reasons, I don’t think we need a whole row in the column at all for a possible (however common) realization of short . [[File:Italy.png]]  イヴァン スクルージ 九十八 （会話）[[File:Italy.png]] 06:34, 8 October 2018 (UTC)
 * The only way I can see [ɾ] can be added to Help:IPA/Italian is if /r/ was consistently realized as [ɾ] in intervocalic positions. ReconditeRodent edited the article to suggest this (by changing "may be" to "is realised as"), while you say it's in free variation. As far as I can tell you're the one who is correct, per Rogers & d'Arcangeli. Nardog (talk) 07:01, 8 October 2018 (UTC)
 * I admit they are in free variation but in the context we are discussing, [ɾ] is clearly much more usual, to the point of being standard. Romano writes: "Regional varieties of Italian follow the same distribution, with intervocalic single rhotics realised as single-strike sounds." A footnote reads: "Inouye (1995) demonstrated that intervocalic tapping of trills is widespread crosslinguistically (in this case only as realisations of a single consonant)."
 * Alternatively, in the phonetic transcription in Rogers & d'Arcangeli I count 7 intervocalic [ɾ]s vs 2 [r]s. I agree "is realised" was too definitive a statement and I'm happy to leave the phrasing as is on Italian phonology but help pages and the IPA given in articles should reflect the most common and most useful pronunciation. ─ ReconditeRodent « talk · contribs » 11:46, 8 October 2018 (UTC)
 * See this source, based on the 2002 Lo Zingarelli dictionary, which consistently transcribes [ɾ] as standard. — Preceding unsigned comment added by ReconditeRodent (talk • contribs) 14:01, 8 October 2018 (UTC)

The point here is: in Spanish, Catalan and a few other Romance languages, the phonemic contrast between long and short consonants only occurs for few sounds: Spanish contrast only vs  and  vs  besides  vs, and it isn’t even a common opposition, since it is only found in words such as those prefixed with in-; in Italian, on the contrary, all consonants except  are each opposed to its geminate counterpart, and replacing the – opposition with – just to align to other helps is imho simply inconsistent, especially given that  and  are both free variants of phonemic short. イヴァン スクルージ 九十八 （会話） 15:26, 8 October 2018 (UTC)
 * I appreciate that for a native Italian speaker the [ɾ]-[r] distinction feels the same as the distinction between other geminate and non-geminate pairs but for non-native speakers (who these pages are designed for) they seem like entirely different sounds, which is why they have separate IPA symbols. ─ ReconditeRodent « talk · contribs » 20:34, 8 October 2018 (UTC)
 * It doesn’t feel the same, it is: that’s precisely why I would avoid listing . [[File:Italy.png]] イヴァン スクルージ 九十八 （会話）[[File:Italy.png]] 21:29, 8 October 2018 (UTC)
 * I think I've found the issue: I'm interpreting anything that sounds like a tap as a tap and you're interpreting anything that you make the same motion as a trill to pronounce as a trill, and it turns out there's such a thing as "monovibrant trill" which is essentially both of those things, in that it's aurally indistinguishable from a tap but is made using the same tongue vibration as in a trill. I also found a source which transcribes "monovibrant trills" with [ɾ]. More detail at the RfC on Help:IPA/Italian. ─ ReconditeRodent « talk · contribs » 23:45, 9 October 2018 (UTC)
 * oh great! So basically single postvocalic may be uttered as pretty much all the realizations between actual  and, is that so? Do you concur now a note is more practical, considering the variation? [[File:Italy.png]]  イヴァン スクルージ 九十八 （会話）[[File:Italy.png]] 08:13, 10 October 2018 (UTC)
 * No, because it is rare for it to be realised as a "multivibrant" which is the only kind of trill which sounds like a trill to most people. I know it's a shame to mess up the system of double letters being distinguished only by duration, but this is a pronunciation guide and monovibrant trills and taps are indistinguishable in spoken language, while both are distinguishable from multivibrants. ─ ReconditeRodent « talk · contribs » 12:12, 10 October 2018 (UTC)

Gio' is not a real italian Word
Hello, and welcome to you unknow canceler and deleter.

Explanations has been given in this talk section and someone just deleted it. You have canceled the posed question, without giving explanations of the content merit in this section. So you did not address the subject in the talk and just deleted my observations and concerns without facing it, just deleting it from the talk. And being irritated is normal, since you didn't follow the rules and are pretending to be right by default. You are not responsible or authorized to delete comments from the talk section. You don't like it, answer it instead of deleting it. Stay in-topic.

The discussion session is not a "nice page", or a report you should like, but a place for discussion; and you haven't discussed. You just deleted the posed fact. This is bullying.

Also, you are not promoting consensus, nor following the NPOV, and you are too prone or to quick to just apply "rules" instead of entering in the merit and face the posed questions.

Wikipedia:Consensus#Consensus-building WP:What Wikipedia is not (WP:NOT) https://en.wikipedia.org/wiki/Criticism_of_Wikipedia#Excessive_rule-making

I have modified contributed and corrected wikipedia since 2007, so I am not a newcomer. And I am an actual scholar, and as you know it is not necessary to have an account to make changes or introduce corrections and talks. Like it or not. Instead, you need to follow the rules and not behave as if you were right in the first place because you don't like the correction inserted or the question posed in the talk session. Blatant fakes must be removed. It is true you are not Italian, and you don't know italian, and hence you need a competent source to defend a position, and the source you cite, is not recognized as enough competent.

Here is the question you might answer, that you deleted it.

''the word gio' for gioia is a false, the source cited is not competent, no one knows Luciano Canepari works, and the word gio' does not even exist in Italian.

"gio '(abbreviation of gioia)"'' '''

Now going into the merit.

Speaking of Italian as a language, we (scholars) have been dealing with linguistic and phonetic questions about italian for more than 800 years. Dante was the first to pose matters about the Italian language, namely in "de vulgari eloquentia", which you surely read it, even if it is written in Latin.

Being a recognized source is not a side issue if you face a linguistic question about the Italian. Canepari is just "a guy" (really who is he?) who wrote something online and pretends to be competent, which is not true, on phonology: this source is unknown to Italian scholars and is not sufficiently competent as it is not recognized by anyone, in terms of linguistic and phonological historiography. Publishing on a site is not a demonstration of competence and for a language discussed for more than 800 years is not a lateral fact to be recognized as a valid source.

So please, take a look at real competent sources such as the Treccani Institute Dictionary - Treccani is an eminent State-founded institution and is there exactly for this purpose - or the Academia della Crusca (and it website) institution founded 400 years ago by eminent cultural personalities, who is there exactly to verify and check Italian linguistics matters, and is universally recognized competent in italian  linguistics questions by more than 4 centuries, or the devoto-oli, the most recognized Italian dictionary. Or similar truly competent sources.

You can't just take a web source and say, it's valid because it's online and it's in Venice. This is not the way, either to control or to address a question.

"gio '(short for gioa)" is a fake. and the source is considered not competent in the scholars ambient.

bye. — Preceding unsigned comment added by 2.233.66.85 (talk) 14:47, 2 June 2019 (UTC)


 * Hmm. You could have written something like: "It is not really possible to show that a word does not exist in a given language, and though it might be true that some native speakers use the word gio’ I have been unable to find it in current dictionaries except Canepari's Dizionario di pronuncia italiana (print and online versions), so it seems too rare to include it here. (The dictionaries I checked only listed the bisyllabic gio, an outdated variant of geo, and gio. or gio, an abbreviation of giovedì.)"
 * Instead you wrote something that sounds like: "Galileo Galilei is a fake astronomer, not a scholar. We scholars studied the universe for centuries, we know the Earth is at its centre, and we are authoritative."
 * This makes me suspicious. Love —LiliCharlie (talk) 16:42, 2 June 2019 (UTC)
 * Actually I never heard gio'  for gioia, and this DIPI dictionary should be considered as a personal website (not one that should use as a reference) imho: at http://www.dipionline.it/guida/ you can read: "DiPI online is based on Dizionario di Pronuncia Italiana written by prof. Luciano Canepari, who generously provided its content". If you search trustable modern sources written by a committee of experts (e.g. Treccani), this word doesn't appear anywhere. Nonetheless, gio'  is a 14th century Italian word meaning "go!" (see Tesoro della Lingua Italiana delle Origini), abandoned few centuries ago. I then propose to replace it with another word, with the same phonological impact: paltò. -- Ruthven ( msg ) 15:50, 23 July 2019 (UTC)
 * Paltò is a good example. If, for some reason, a truncated form is desired for the text, frequent [tɔ] 'prendi' would seem to be an obvious choice. It raises the question, though (along with gio' and ciò) of what "word-final stressed" might mean, if anything, in the case of words of one syllable. -- As for Canepari and his DiPi Online... He's easily Italy's best-known phonetician, both at home and abroad. DiPi is not a lexicon or a "vocabolario", but what it claims to be, a dizionario di pronuncia. Given Canepari's reputation as Italy's best-known phonetician it's a bit disappointing that his transcriptions are phonemic rather than phonetic (e.g. /ˈbanko/, not [ˈbaŋko]), but he does supply useful information that others usually don't, such as major regional Italian variants, structure that doesn't correspond as expected to spelling (/aʦˈʦjone/ 'azione') and syllabification based on actual phonology (/kasˈtɛllo/ rather than the usual misleading orthographic convention /ka-stèllo/). DiPi is the standard for phonemic transcription, as any comparison with the usual dictionaries reveals very quickly. Barefoot through the chollas (talk) 18:35, 30 August 2019 (UTC)
 * In contrast to paltò and the other sample words, gio’ is an example of word-final stressed that came about through troncamento, but I don't think we need an example for that.  was okay, though we now have five (!) examples for a phenomenon that isn't exactly extraordinary, rare, or surprising: però, ciò, paltò, dormirò, and parlò. Love —LiliCharlie (talk) 22:02, 30 August 2019 (UTC)
 * It was as an example of troncamento that I suggested [tɔ]; very common, none of the controversy of gio'. As for the others, they're the same phenomenon in the sense of all being final [ɔ], but they arise from different conditions: paltò as a borrowing from French (paletot), però and ciò as native Italian lexemes, and the verb forms by way of morphology marking subject. Perhaps the trouble is that "Word-final stressed /ɔ/ is found in a small number of words" might be a bit misleading. It's true that the number is limited with regard to lexemes, but that it repeats morphologically as indicated, and also in monosyllables such as ciò, first-person presents such as ho, do, sto (and Tuscan  fo, vo), do of do-re-mi, etc. Barefoot through the chollas (talk) 01:23, 31 August 2019 (UTC)
 * Right. However we must be very careful with monosyllabic words, as many of them are often unstressed. (For example, ho and sto are regularly unstressed when used as auxiliary verbs.) What we are talking about here is word-final stressed, and there should be no risk of reduction to a closer -like vowel quality due to loss of stress in our sample words. Love —LiliCharlie (talk) 05:03, 31 August 2019 (UTC)
 * Lots of cans of worms there, but they needn't be troublesome. It would be enormously cumbersome to cover even just a few of the phonetic renditions of phonemic structure in connected speech at various speeds and in various registers. Canepari's stɔmˈmɛʎʎo 'sto meglio' suggests that main stress shift away from the monosyllable can leave /ɔ/ surfacing as [ɔ], but even if/when that fails to hold, it's not problematic for the word-level observation in the article, especially since the monosyllable "behaves" as stressed in triggering RS in stɔmˈmɛʎʎo (and sto [pˈp]arlando) (and, in any case, back to connected speech, "[A]ll (non-emphasised) primary stresses but the utterance-final one are considerably weakened in the speech chain" (Bertinetto 1981)).
 * There's a larger issue buried in all this, though (I'll add italics to the main point): "Final stressed /o/ only occurs in the pronunciation of foreign names such as Bordeaux; however, the corresponding loan word, indicating a nuance of red, has been adapted as [borˈdɔ], often spelled bordò. The same lowering historically applied to no [nɔ] 'no' (instead of the expected *[no] from Lat.NŌN)" (p. 137 of Bertinetto & Loporcaro in the article bibliography). In other words it's word-final stressed /o/ that's rare, while word-final stressed /ɔ/ is restricted lexically, but highly frequent due to ho, sto, do (and no) and due to productive morphology for the entirely open set of -are verbs. Upshot once again: the claim "Word-final stressed /ɔ/ is found in a small number of words" can be very misleading. Barefoot through the chollas (talk) 17:28, 31 August 2019 (UTC)

"pseudo" -> "sseudo"?
I do not doubt that the source given describes "sseudo" as a usual careless pronunciation for "pseudo" and the like, but I, an Italian living in Italy, never heard such a way of pronouncing this sets of consonants. If anything, somebody introduces a ghost vowel, or perhaps a schwa, sound, saying almost "peseudo". I suppose that, as the source is a 1944 book, something has changed in non-written Italian uses. But of course I am not an acceptable source, so I hope I or somebody else will find a more recent one. (Thanks for correcting the tag!) Goochelaar  (talk) 21:39, 2 January 2009 (UTC)


 * I agree, this assimilation may be true for consonants not in syllable-initial position. 84.223.133.56 (talk) 02:33, 15 January 2009 (UTC)
 * Never heard. At most there are people pronouncing pneumatico as /(il) neu'matiko/ (not /(lo) nneu'matiko/, such an effort would be unnecessary). --Erinaceus (talk) 17:05, 23 March 2009 (UTC)


 * Neither I, native speaker, have ever heard "ps->ss" or "pn->nn". In some southern variety (i.e. in Sicily) it can be added a vowel and a syllable (psychologist, "psi-'cɔ-lɔ-go" becomes "pis-si-'cɔ-lɔ-go"), but it's far from being standard language. — Preceding unsigned comment added by 78.134.19.249 (talk) 18:32, 11 July 2011 (UTC)

Same, and I’m a native speaker, too. I never heard no simplification whatsoever of initial consonant clusters. “Psiche”, “pneuma”, “pterosauro”, “ctonio”, “Cnosso”, “xeno-”, etc., are always pronounced entirely, that is: as a cluster of two different consonants, by everybody; except that some people put a schwa or an “e” in between the two; but that sounds quite clumsy. Mauro Maulon69 (talk) 00:04, 28 April 2014 (UTC)

According to Luciano Canepari’s spelling dictionary (http://venus.unive.it/canipa/pdf/DiPI_3_A-Z.pdf), /ps/ has the standard realisation of [ps], and in some common case also [*s] or [piss]. (The asterisk means the consonant is doubled after a vowel.) The latter two are both marked with a downward arrow, which seems to mean (going by their use in other entries) that these forms may occur but are incorrect and considered uneducated. — Preceding unsigned comment added by 87.64.90.68 (talk) 16:00, 25 June 2014 (UTC)

and this so called "dictionary" of Canepari, no longer exists. please quote competent sources. use for example the website of the academia della crusca, or the dictionary of devoto-oli or the Treccani, not the invented ones. — Preceding unsigned comment added by 2.233.66.85 (talk) 21:20, 1 June 2019 (UTC)
 * Canepari's Dizionario di pronuncia -- which is not intended as a spelling guide -- is now fully operable on line at http://www.dipionline.it/dizionario/ . It should not be consulted without reading the Guida al dizionario portion, but with the information there understood and a grasp of basic Italian allophony (the transcription is in principle phonemic), it's a more useful guide to phonological matters. Barefoot through the chollas (talk) 16:49, 7 September 2019 (UTC)

Two "a" values in Italian language, then EIGHT vowels
Check "a" in "banca" and "a" in "palco": the first one is open, the second not. P.s. I know, the second "a" is quite uncommon, but it nevertheless exists. Ardashir (talk) 08:01, 18 May 2020 (UTC)
 * What you're talking about is a contextual variant or allophone. If you can identify minimal pairs (and a source backing it up) then there would indeed be 8 vowel phonemes. — Æµ§œš¹  [lɛts b̥iː pʰəˈlaɪˀt] 16:15, 18 May 2020 (UTC)

Velar nasal not a phoneme
Dear ,

It's not clear why you keep writing that the velar nasal is not a phoneme in Italian language. Is it because it's not contrastive?--171.96.189.177 (talk) 06:17, 7 September 2019 (UTC) I think it is contrastive. Anyway, I'd love to have an answer from you.--171.96.189.177 (talk) 06:27, 7 September 2019 (UTC)


 * This is not a matter of what some user thinks. All Italian phonologists agree that is not a phoneme of Italian because it doesn't contrast with  as in mai (i.e., * is illegal),  as in noi, Don (*, *), or  as in gnor (*). Note that the phoneme  (or, if you like, the archiphoneme, that is, any nasal phoneme) immediately followed by a consonant other than  and  adopts that consonant's place of articulation, so the phoneme  is realized as a velar nasal phone  in front of the velar plosives  and . Love —LiliCharlie (talk) 08:00, 7 September 2019 (UTC)

What you write is somewhat incorrect; please provide reliable sources for your original research. Example: --223.24.148.122 (talk) 09:41, 7 September 2019 (UTC)
 * fango                 ['faŋo]
 * Fano (Italian city)   ['fano]


 * No, the ng in fango is pronounced as in English finger, not  as in singer. The two syllables are unambiguously stressed   and unstressed  . Love —LiliCharlie (talk) 10:15, 7 September 2019 (UTC)

Please, provide realiable sources for all of this and prove that it's not original research.--223.24.148.122 (talk) 10:45, 7 September 2019 (UTC)


 * Provided that it's not a phoneme, is the above reasoning enough to omit that consonant from the table? It's still a sound that occurs quite often in Italian and it's worth to mention it. — Preceding unsigned comment added by 58.8.154.116 (talk) 12:23, 7 September 2019 (UTC)
 * The label on the table answers your question: consonant phonemes (not sounds, i.e. phones). As LiliCharlie explained, [ŋ] appears only allophonically in native Italian speech: /kon karlo/ → [koŋˈkarlo], no contrast with [n] should someone produce [konˈkarlo] (ditto word-internally: fango with normal [ŋ] or exceptional [n] are both interpreted as /fango/). Barefoot through the chollas (talk) 13:35, 7 September 2019 (UTC)


 * Here's one of countless reliable sources for the non-phonemic status of the Italian velar nasal: It's Derek & Luciana d’ (2004:117f.) [=Illustrations of the IPA: Italian in: Journal of the International Phonetic Association (2004) 34/1]. — I didn't think we needed a source proving a fact that isn't contested among phonologists at all, but it seems you haven't even read a superficial introduction to the field of Italian phonology that you're making claims about. My source is less than five full pages. Carefully read at least this little before you post more. Love —LiliCharlie (talk) 15:13, 7 September 2019 (UTC)
 * Also, as noted in the article, Bertinetto & Loporcaro 2005, pp. 132, 134-135, with the PDF just a click away in the article Bibliography. Reliable sources were already cited and easily available before the poster requested them. Barefoot through the chollas (talk) 16:22, 7 September 2019 (UTC)

Omitting ŋ from the consonant chart because it's called a consonant phoneme chart and hiding this sound in the notes in a sentence that is incomprehensible to almost all users of Wikipedia (/n/ has a velar allophone [ŋ] before /k, ɡ/.) is very misleading to say the least. Leaving out clarification of whether or not Italian pronunciation pronounces words with the spelling ng in two different ways as English does is really sloppy. Absence of this info and any mention of the sound ŋ on Italian_orthography is even more sloppy. --Espoo (talk) 09:54, 28 October 2019 (UTC)
 * It's not "hiding" the information in notes. We put all the relevant information in article prose and avoid cluttering the table with allophones. We treat the consonant table at Spanish phonology the same way. — Æµ§œš¹  [lɛts b̥iː pʰəˈlaɪˀt] 14:18, 28 October 2019 (UTC)
 * This isn't Italian orthography. I think that most readers assume that $⟨ng⟩$ isn't pronounced because there's no * phoneme in Italian. Readers who don't understand the sentence you quoted have little business reading this page. To me, your complaint sounds equal to "why do we use IPA on this page".
 * Also, it's obvious that ng has two different pronunciations in Italian - one is (as in Inzaghi), the other is  (as in mangiare). The difference is in the pronunciation of what follows the nasal (which assimilates to the place of articulation of the following consonants in all cases, save for the approximants  and ). In Spanish $⟨ng⟩$ is also pronounced in two different ways:  and  (or just  and, with the preceding vowel being nasalized - I think this is very usual in the case of ). Kbb2 (ex. Mr KEBAB) (talk) 09:02, 19 May 2020 (UTC)

Isn't "palatalized postalveolar" alveolo-palatal?
The article states are "palatalized laminal postalveolar  before " and  are "alveolo-palatal". But aren't "palatalized laminal postalveolar" and "alveolo-palatal" the same thing? In fact our articles about alveolo-palatal nasal and lateral give ⟨n̠ʲ, l̠ʲ⟩ as impressionistic transcriptions in addition to ⟨ɲ̟, ʎ̟⟩.

And since and  do not contrast in codas (save for geminates), transcribing the former preceding  as ⟨ɲ, ʎ⟩ would not only pose no problem but achieve the same phones being transcribed with the same symbols. So wouldn't incinta, mangiatoia be better transcribed as than as ? We already use not only ⟨ŋ⟩, which DOP makes note of and which may be phonemic marginally in loanwords, but also ⟨ɱ⟩, which AFAIK no dictionary makes note of and which is incredibly rare to be found phonemic cross-linguistically. (I'm not inclined to using ⟨ʎ⟩ for, as in for calcio, Golgi, however, as far as Help:IPA/Italian transcriptions are concerned. That would be too confusing.) Nardog (talk) 20:31, 24 June 2020 (UTC)
 * Makes sense to me. — Æµ§œš¹  [lɛts b̥iː pʰəˈlaɪˀt] 03:56, 25 June 2020 (UTC)
 * I don't seem to have easy access to Canepari's Manuale. Can anyone supply a good source for the claim /n/ → [ɲ]/_[tʃ]? Barefoot through the chollas (talk) 01:54, 2 July 2020 (UTC)
 * Hmm, I seem to have jumped the gun here. Rogers & d'Arcangeli (p. 118) explicitly say it's [n] (whatever its precise value is) before /ʃ, tʃ, dʒ/. I'll revert the change. This means we indeed need better sources about the realization(s) of /ɲ, ʎ/ though.
 * As for Canepari, I don't have access to the Manuale either, but here's an excerpt from his Italian Pronunciation & Accents (2018). He says /ɲ, ʎ/ before /ʃ, tʃ, dʒ/ are "postalveo-palatal", which seems to correspond to palato-alveolar in ordinary terminology. Nardog (talk) 03:24, 2 July 2020 (UTC)
 * Thanks for cleaning up Francesco. Thanks also for access to Italian Pronunciation... I don't seem to have located the text in question yet, but I'll keep looking. I also haven't come up with a context in Italian in which /ɲ, ʎ/ can come into contact with any of /ʃ, tʃ, dʒ/. Barefoot through the chollas (talk) 12:37, 2 July 2020 (UTC)
 * There are a few French borrowings in Canepari's Dizionario di pronuncia italiana that end in : champagne/Champagne, guigne, hors ligne, Issogne, Montaigne. According to him, only Issogne has a variant pronunciation in . Love —LiliCharlie (talk) 13:38, 2 July 2020 (UTC)
 * Good finds. Makes sense I suppose for Issogne, as valdostano in origin, Italianized. The question for the French borrowings, then, is what happens in champagne gelido and the like. My guess, not totally WA but not trustworthy as fact, is that, especially in light of the traditional Italian form sciampagna, monolingual Italians come up with lengthened or semi-lengthened [ɲː] or [ɲˑ], and a bit of a vocale neutra or even [a]. If accurate, that and the need to illustrate with non-nativized gallicisms would mostly serve to reinforce the observation that the contact is at very best an unhappy one. Barefoot through the chollas (talk) 14:47, 2 July 2020 (UTC)
 * Sorry, I'm not a native English speaker, so my English is too poor to understand you. What on earth does "not totally WA" mean? I searched here and there and here and there and here and found over a hundred definitions of "WA" but none seemed to fit. And to my surprise none of them contained the word "without." Love —LiliCharlie (talk) 16:07, 2 July 2020 (UTC)
 * Charlie, sorry -- I didn't intend to be cryptic. WAG 'wild-ass guess'. Ergo, my guess, not totally WA = my informed/educated (but non-native) guess. Barefoot through the chollas (talk) 16:59, 2 July 2020 (UTC)
 * I also haven't come up with a context in Italian in which /ɲ, ʎ/ can come into contact with any of /ʃ, tʃ, dʒ/. The question isn't whether /ɲ, ʎ/ occur before /ʃ, tʃ, dʒ/ (I'm sure they don't) but whether /n, l/ before /ʃ, tʃ, dʒ/ share the same articulatory configurations as /ɲ, ʎ/ in other environments. Nardog (talk) 21:01, 2 July 2020 (UTC)
 * Okay. Back to what triggered my presence here, it's difficult at best, even making the effort, to articulate the /n/ of Francesco as [ɲ]. Barefoot through the chollas (talk) 00:35, 3 July 2020 (UTC)

I've managed to get a copy of the 1992 edition of the Manuale. Canepari describes as palatale/i by default and   as postalveopalatale/i before  which are characterized as postalveopalatolabiale/i, i.e., they have the same place of articulation as a nasal or lateral that precedes them but are labialized. The midsagittal sections look the same as the following in his Natural Phonetics and Tonetics: #42*/48 on p. 172, #48 on p. 181, #80 on p. 189, and #20*/23 on p. 202, except that a lateral airflow is indicated by three horizontal lines that are not vertically aligned rather than by means of an arrow, and the consonants here marked with an asterisk are represented by the letters used for #44 on p. 172 and #21 on p. 202. I hope this isn't too confusing. Love —LiliCharlie (talk) 10:42, 3 July 2020 (UTC)

'Ndrangheta
According to article 'Ndrangheta the Italian pronunciation of the organization is which violates the rules we give in the Phonotactics section of this article. Three questions: 1. Is it correct in Standard Italian to pronounce the $⟨N⟩$ as ? 2. If so: Does the constitute a separate syllable (i.e., the nucleus) or is it part of the consonant cluster of the onset? 3. How can we handle this and other (more or less) isolated cases that occur in foreign loans here? Love —LiliCharlie (talk) 22:18, 16 August 2020 (UTC)
 * Charlie, 1) yes, it's /n/ → [n]; 2) yes, it's a syllabic consonant in a standard citation version, thus more accurately (although some people will automatically start out with a bit of schwa); 3) the only way to handle these things that I know of is report the pronunciation as it is in Italian; if there's space, or reason to, perhaps point out that they're not Italian in origin (this one's most immediately Calabrese, supposedly ultimately Greek). Barefoot through the chollas (talk) 01:41, 17 August 2020 (UTC)
 * Interesting. So there is no variation for the citation form in Standard Italian, except for an optional prothetic schwa, and pronouncing the word with a non-syllabic sounds wrong? (This is what happens in German, the language I know best. My copy of Duden. Das Aussprachewörterbuch  (2005), which marks syllabic sonorants [that frequently occur] throughout, gives pronunciations such as  for Mbovamb where it is clear that the stress falls on the syllable with the long vowel  which would be shortened if unstressed.) Love —LiliCharlie (talk) 03:21, 17 August 2020 (UTC)

Just a pair of considerations by the usual native speaker. Actually "'ndrangheta" isn't an Italian word, is a Calabrian word. It's also used among Italian speakers and included in Italian dictionaries, but it remains a "foreign" word, like "birthday". The average Italian will always say "andrangheta", adding a syllable, because words starting with /nC/ don't exist in Italian, indeed in some dictionaries you can find "an-dràn-ghe-ta" too. I'll tell you my opinion, which you're free to ignore if you don't agree: it should be used the template "IPA-itdia" for the Calabrian version, then the template "IPA-it" for the Italian version but with a sentence such as "partially adapted in [nˈdraŋɡeta] and fully adapted in [anˈdraŋɡeta] spelled andrangheta", possibly linking a source.--151.64.155.9 (talk) 07:49, 17 August 2020 (UTC)
 * It's a judgment call on register, but I wouldn't go so far as to recommend initial /a/ or even just [a] for Standard Italian approximation of the Calabrese (which may be a re-parsing of the frequent [lanˈdraŋɡeta] with definite article as l'andrangheta rather than la ndrangheta), but otherwise pretty much agreed. Reporting both [nˈdraŋɡeta] (with or without syllabic mark under [n]) and [ənˈdraŋɡeta] seems reasonable at first glance, but the second could lead readers to believe that the schwa is or should be always present, i.e. even [la ənˈdraŋɡeta]. (Charlie, I don't know what you mean by pronouncing the word with a non-syllabic [n]. It seems I can't do it.) Barefoot through the chollas (talk) 13:53, 17 August 2020 (UTC)
 * @Barefoot through the chollas: I meant (articulatorily) prenasalized stops. To get an impression download Cleghorn & Rugg (²2011): Comprehensive Articulatory Phonetics (audio and e-book uploaded by one of the authors), read p. 292 and listen to exercises 25.9–25.10. (The print version of the book is extremely inexpensive. Consider buying a copy or two to support the authors.) Love —LiliCharlie (talk) 17:49, 17 August 2020 (UTC)
 * I can do, hear, etc. /nd/ realized with pre-nasalization of /d/ (infelicitous attempt at IPA: [ⁿd]; better, but still not good, with ligature). I was referring to non-syllabic [n] as presumably would be represented [ˈndraŋɡeta]. Barefoot through the chollas (talk) 19:10, 17 August 2020 (UTC)
 * Whether segments are perceived as syllabic or not is determined by phonotactic expectations more than by peaks and troughs in prominence. I once had a student from Spain who insisted that German stets  is a trisyllable, whereas German native speakers agree the word is monosyllabic. Similarly,  and  do not necessarily sound different, and much depends on the tradition the transcriber was trained in. (If my memory serves me right,  hinted at problems to distinguish between  and  on a purely etic level in our recent discussion on Help_talk:IPA/Italian). Love —LiliCharlie (talk) 19:58, 17 August 2020 (UTC)
 * I'm going more by articulation than acoustics (although while the possibilities in the space obviously form a continuum, the poles of a syllabic [n] and a prenasalized /d/ do sound different to me). In any case, all of this is pretty much moot for the purpose at hand: a single Standard Italian rendition of 'ndrangheta in a broad phonetic transcription. Barefoot through the chollas (talk) 21:15, 17 August 2020 (UTC)

I concur with LiliCharlie. When an onset illegal in the native vocabulary is found in English, as in vlog, vroom, tsunami, Dvorak, etc., we regard it as having a marginal onset, not as beginning with a syllabic consonant. I've entertained the thought that, had it not been for the knowledge of the historical relationship and free/dialectal variation between [ən] and [n̩], what we're used to thinking of as [n̩] in English might have been regarded as simply a coda /n/ (especially following fricatives as in dozen) and English phonotatics might have had a much longer list of legal codas—as much as it would run counter to sonority hierarchy and as much as there is often a psychological reality of syllabicity in speakers' minds, e.g. when they're asked to clap for each syllable. In other words, I suspect the concept of a syllabic consonant is sometimes employed to satisfy (or simplify) phonotactics rather than to account for an actual, demonstrable difference in duration etc. This of course doesn't hold much water in more clearly syllable- or mora-timed languages. Nardog (talk) 17:48, 18 August 2020 (UTC)
 * Fully agree that reports of English word-final syllabic /n/ are often doubtful. That aside... Dvorak is genuinely problematic in English, thus common schwa-insertion and descriptions such as "Dvořák's name is pronounced 'Devor-jacques'". The ease with which children produce vroom with onset [vr] illustrates what the plenitude of /fr-/ forms in English suggests: the switch to voicing for /vr-/ is not phonologically troublesome, just highly infrequent, mostly by accident of historical phonology. Tsunami can invite some bickering, with normal [su-] possibly fueling a claim that /ts/ onset is banned, a claim at least wounded in flight, if not definitively shot down, by native speakers of English being able to produce onset [ts] easily to deliver a trouble-free [ts]unami. 'ndrangheta bears some resemblance to Dvorak in that the onset /nd/ (or /ndr/) is "genuinely problematic" in the terms above, thus /n/ is treated as syllabic as in [n̩ˈdraŋɡeta] or, arguably slightly down-register, liberated from its syllabic solitary confinement with a bit of vocalic company: [ənˈdraŋɡeta]. Which winds back to Charlie's original questions: Standard Italian approximation of the Calabrese is [n̩ˈdraŋɡeta]. (BTW, It's just occurred to me that it may not be common knowledge that Italians are not totally unaccustomed to hearing, and in the appropriate contexts producing, initial syllabic /n/; Romanesco ndo (or 'ndo) 'where' as in ndoè 'where is' is pretty much known to all.) Barefoot through the chollas (talk) 19:49, 18 August 2020 (UTC)
 * Whether the words I mentioned are actually pronounced with such complex onsets by native speakers of English is immaterial here. What's your source or evidence that the citation form of 'Ndrangheta in Italian (or in Calabrian for that matter) begins with a syllabic consonant as opposed to a marginal onset (or to an "extrasyllabic" consonant like the "impure /s/")? Nardog (talk) 20:23, 18 August 2020 (UTC)
 * Source is knowledge of Italian and Italian phonology. A study by a good phonologist of the phonetics/phonology of 'ndrangheta would be welcome. I'm not aware of anyone having bothered to publish on the topic, but the info may be buried in broader text. In principle, yes, English (German,. etc.) are immaterial here. But your examples vl-, vr-, ts- in English are helpful, precisely because of the way native speakers pronounce them. That [vr], [vl] and, to only a slightly lesser extent [ts] are not problematic as English onsets illustrates that the phonologically non-Italian onset */nd(r)/ and marginal onsets such as /vl-/, /vr-/, /ts-/ in English are quite different in kind. You raise a point of interest, though, for the result: nd(r)- is not an Italian onset, so should the /n/ be treated structurally as extrasyllabic in Standard Italian? Good phonology question (although either way the Standard Italian phonetic form is pretty clearly [n̩ˈdraŋɡeta], which is the point for the transcription in question). Barefoot through the chollas (talk) 21:13, 18 August 2020 (UTC)
 * I'm less sure than you that "the Standard Italian phonetic form is pretty clearly [n̩ˈdraŋɡeta]". The English, French, and Madagascan wiktionaries all have transcriptions with ⟨ˈndra⟩, so the only sources with IPA transcriptions I know don't support your claim, even if they are not reliable encyclopaedic sources. — Is the English "phonetic form" of Dvorak "pretty clearly" with an optional epenthetic schwa? Love —LiliCharlie (talk) 04:18, 19 August 2020 (UTC)
 * I said Whether. What I mean is that while most English speakers indeed pronounce tsunami with /ˈs-/, I only mentioned the word to illustrate the fact the pronunciation with /t/ in English—regardless of whether it's actually used in spontaneous speech or only theoretical/prescribed—is analyzed/represented as /ˈts-/, not /tˈs-/ or /t̩ˈs-/. Nardog (talk) 09:18, 19 August 2020 (UTC)
 * Yes, that's the point: extremely low frequency per se does not imply lack of phonological fit. Invent a name for the new tech company that you're about to launch: tsuba. Any problem in the anglosphere pronouncing it ['ts]uba? Nope. What if you want to insist on ntuba instead? Oh my.... Barefoot through the chollas (talk) 14:19, 19 August 2020 (UTC)
 * Charlie, I really don't have anything more to add. You asked questions, I answered, and if you read between the lines, unnamed IP 151 (etc.) gave a native speaker's impressions corroborating that an initial cluster ['ndr] is at the very best big trouble. The German Wikipedia page for 'ndrangheta offers the compromise [] (with a footnote not very helpful for Italian), but I don't know of any reason to believe that [] is a variant in the realm of Standard. Given all the evidence and the fact that nasals are universally the most likely to manifest syllabicity when stranded, I don't see any reason not to report [n̩ˈdraŋɡeta] (or [nˈdraŋɡeta]) in the Wikipedia article (and I don't see any justification for reporting another form instead). -- The onset problem seems to interest you. It might be even more fun to do battle with dictionaries that report Italian tmesi as /ˈtmɛzi/! Barefoot through the chollas (talk) 17:07, 19 August 2020 (UTC)
 * Alright. I don't see how our sourceless discussion can lead to a modification of this article, so let me return to the starting point. Should we avoid the problem and alter the first sentence of 'Ndrangheta to something like "The 'Ndrangheta (... Italian: la 'Ndrangheta ...)..."? Love —LiliCharlie (talk) 02:36, 21 August 2020 (UTC)
 * Charlie, I have to be honest and say that I don't see a problem that needs avoiding. The mark for syllabicity might be added under the [n], but it doesn't seem necessary in a broad transcription, and for some readers it might just add confusion. The transcription as is matches the IPA on the Italian page, and no one seems to have questioned it (years ago one person wrote about pronuncia, but seemed more worried about the orthographic apostrophe). Bref, this strikes me as a pseudo-problem. Barefoot through the chollas (talk) 07:44, 21 August 2020 (UTC)

LiliCharlie's proposal "( Italian: la 'Ndrangheta )" is a solution which cuts the head off the snake.--151.64.158.77 (talk) 12:40, 21 August 2020 (UTC)
 * It's a helpful guide to pronunciation with the article, and some readers will perhaps realize that the syllabification can be extended to preceding vowels in general. Unfortunately, it doesn't provide what's primarily called for, i.e. pronunciation of 'ndrangheta in isolation. -- While the 'ndrangheta article really needs only the citation form, some more elaboration in the Onset portion of the Italian phonology article re partial assimilations (incorporations) might be in order. Other than the bit on /sC/, troublesome onsets are given short shrift, so that tmesi, for example, is presented as though /tm-/ is uncontroversial as a genuine onset, no less problematic than /fr-/ (and /nd(r)/ isn't mentioned). Barefoot through the chollas (talk) 14:35, 21 August 2020 (UTC)

A strange error in the coda consonants’ list
I refer to this paragraph. Actually, /s/ occurs many times as a coda before the fricatives /f/ and /v/. Just a few examples: Kapuov (talk) 06:12, 12 March 2021 (UTC)
 * asfalto — /as.ˈfal.to/ — “asphalt”;
 * le sfuggí — /le‿s.fud̚.ˈdʒi/ — “he/she/it escaped her”;
 * trasvolare — /tras.vo.ˈla.re/ — “to fly across”;
 * disvalore — /dis.va.ˈlo.re/ — “lack of value”, “disvalue”.
 * Since none objected, for now I erased the wrong exception and removed the relative note. Kapuov (talk) 15:51, 27 March 2021 (UTC)
 * Was it really a mistake though? The last two examples are contain, not (AFAIK), and le sfuggí clearly doesn't have coda  in phonology, it has onset  since it belongs to one word (sfuggí). You'd have to provide convincing evidence for the resyllabified phonetic . And since the  onset is permitted, is asfalto really  or ? Per Krämer, it's the latter. I've reverted your changes. Sol505000 (talk) 09:21, 4 April 2021 (UTC)
 * A couple of issues here. One, whether the sibilant in trasvolare and disvalore is /s/ realized phonetically as [z], or /z/ realized [z]. The most widely accepted analysis seems to be /s/ → [z] as a simple assimilation of voicing, since the /z/ analysis would require positing two prefixes, /tras/ (for e.g. Tra[s]tevere) and /traz/, for which there is no non-circular evidence; similar for dis-, which is more productive, and assumed by its phonetically banal behavior to be /dis/. The other is the syllabification of e.g. /sf/. See the discussion under Onset in the article, beginning "As an onset, the cluster /s/ +...". Most evidence points to /as.fal.to/. Word-initially, lo studente rather than *il studente likewise argues for s.C rather than sC. Another way of saying this is that [sC] onset is permitted phonetically IFF there is no vowel preceding, sort of parallel with [ts] or [dz] permitted initially IFF a vowel does not precede, in which case the underlying geminate is liberated, no longer blocked (e.g. [dz]aino, but lo[ddz]aino). Barefoot through the chollas (talk) 16:22, 4 April 2021 (UTC)
 * Thank you, Barefoot through the chollas.
 * Sol505000, I advise you, before any inconsiderate act, to open an Italian dictionary (for the phonetic syllabifications) and to read the article whose we’re talking about, at least («The phonetic distinction between [s] and [z] is neutralized before consonants and at the beginning of words: the former is used before voiceless consonants and before vowels at the beginning of words; the latter is used before voiced consonants (meaning [z] is an allophone of /s/ before voiced consonants)» ).
 * Kapuov (talk) 18:57, 4 April 2021 (UTC)
 * Just a side note: I would say be very careful in consulting traditional dictionaries for syllabification. Their charge is a broad one with regard to transcription. In the specific case of syllabification of /sC/, the consideration of prescriptive hyphenization norms in script can override presentation of authentic phonetics or phonology. Introductory pages to the dictionary should explain the principles adopted, but absolute clarity even there seems not to be the norm. The most accessible source that's trustworthy is Canepari's dipionline, but his phonemic representations indicate syllabification only if stress is shown at the relevant syllable boundary: /fesˈtina ˈlɛnte/, /trasˈkorso/ but /ˈfɛsta/, /trasfiˈɡuro/. Barefoot through the chollas (talk) 16:28, 5 April 2021 (UTC)

About the gemination of /ɲ/, /ʃ/, /ʎ/, /dz/ and /ts/
From the first note of this paragraph: «/ɲɲ/, /ʃʃ/, /ʎʎ/, /ddz/, /tts/ are always geminated word-internally». This sentence is tautologic: /ɲɲ/, /ʃʃ/, /ʎʎ/, /ddz/, /tts/ are geminated by definition. Maybe the intended meaning was: «/ɲ/, /ʃ/, /ʎ/, /dz/, /ts/ are always geminated word-internally». This is a reasonable sentence, but also a wrong one. In fact, these phonemes are always long whenever preceded by a vowel and followed by either a vowel or an approximant, regardless of their position in the word (e.g.: oggi sciopero — [ˈɔ̟d̻̚.d̻ʒi‿ʃ.ˈʃɔ̟ː.pe.ɾo] — “today I strike”, azione — [ät̻̚.ˈt̻sjoː.n̺e] — “action”); and they are short in all the other cases, even word-internally (e.g.: minzione — [min̻ˑ.ˈt̻sjoː.n̺e] — “micturition”, una zanzara — [ˌuˑ.n̺ä‿d̻̚.d̻zän̻ˑ.ˈd̻zäː.ɾä] — “a mosquito”).

Kapuov (talk) 12:53, 8 March 2021 (UTC)
 * Good point, and well illustrated. Barefoot through the chollas (talk) 15:57, 8 March 2021 (UTC)
 * Though the cited source does say "in the interior of a word", "intervocalic" would clearly be more accurate here. Replaced with a more recent source. Nardog (talk) 10:16, 10 March 2021 (UTC)
 * Still a bit of a mess in the article: "/ɲ/, /ʃ/, /ʎ/, /dz/, /ts/ are always geminated intervocalically" reads like a claim that e.g. foglio contains /ʎ/, which is realized phonetically [ʎː] (or [ʎʎ]). As above, the standard view is very straightforward: /ʎʎ/ is realized [ʎː]. Nothing is (inexplicably, magically...) "geminated"; the phonological unit in question simply IS a geminate realized as such. Barefoot through the chollas (talk) 14:45, 28 April 2021 (UTC)
 * The wording is correct - are not phonemes of their own but geminated, i.e. not one phoneme but two (and I don't know whether the affricates are uniformly treated as one phonological unit in Italian. In German, the analyses vary. If they're treated as  +  etc., then  contain three phonemes). Sol505000 (talk) 22:47, 13 April 2022 (UTC)
 * Mountains of evidence to the contrary for Italian. But don't feel bad. Phonologists who specialize in Italian are known to drive themselves a bit batty with this, at least at the level of terminology -- and that has its effect on conceptualization. The topic merits a good monograph, but I'll keep this very brief with just one set of examples. Tedious fleshed-out version available if you need it.
 * There is no singleton phoneme /ʃ/ in Italian. There can be no structure */paʃe/, for example, because "/ʃ/ is always long" (obviously mis-stated ingenuously for the nonce). The phonetic form [paːʃe] pace does exist for millions of Italians (alongside non-weakened [paːtʃe]), but that phonetic [ʃ] results from application of an allophonic weakening rule /tʃ/ → [ʃ]/V_V, not as a realization of /ʃ/. The verb pasce exists as well, pronounced [paʃːe]. Thus pasce [paʃːe] and pace [paːʃe] can contrast on the basis of internal consonant length (vowel length is epiphenomenal, and not distinctive), but not as geminate/long vs. singleton/short articulation of the same phoneme. The form pasce co [lntains a phoneme intrinsically long -- geminate, if one insists -- not /ʃ/ that somehow lengthens/geminates. Perhaps this clarifies my objection above to the claim "/ɲ/, /ʃ/, /ʎ/, /dz/, /ts/ are always geminated intervocalically", i.e. in strict terms, there are no instances of /ʃ/ being geminated (undergoing a process from singleton/short to lengthened or geminate) because /ʃ/ doesn't exist. Barefoot through the chollas (talk) 20:20, 14 April 2022 (UTC)
 * So which phoneme(s) is/are at the beginning of the verb sciare? Sol505000 (talk) 21:33, 14 April 2022 (UTC)
 * Excellent question, to which I'll give a dodgy answer at first: the same structure as in pasce above. Let's use a masc. noun, since it selects articles. sciocco 'idiot': lo sciocco 'the idiot'. Why lo, not il? Because the bit represented by orthographic sci is inherently (phonemically) ambisyllabic. Article il /il/ would produce an inadmissible consonant cluster in syllabic terms [lʃ.ʃ...]. But give it a chance, i.e. a vowel in front of it, and voilà, it can spread out its length to close the syllable of the preceding vowel: [loʃːɔkːo]. On the other hand, if you say just the word sciocco, the length of the initial consonant long by nature is blocked, it degeminates phonetically down to simple [ʃ]: [ʃɔkːo]. It's basically the same principles as those at work with e.g. lo studente or lo psicologo, pretty obvious in the latter: [ilps] is pretty much outlawed in normal Italian phonotactics. If the [p] isn't just dropped altogether, the problem is solved with [lop.si...], an acceptable sequence. Barefoot through the chollas (talk) 22:18, 14 April 2022 (UTC).
 * The phonemic structure of words is analyzed according to how they are pronounced in isolation, rather than in running speech (and lo sciocco counts as running speech to me, a short phrase. The definite article isn't glued to the word no matter what, no?) Since sciocco alone is never pronounced with a geminate, why claim its existence on a phonemic level in the first place? I'd like to know what Krämer 2009 has to say about that.
 * Though I admit that, for example, the German word Bild 'picture' can be analyzed as  due to the fact that the voiced/lenis  resurfaces in the plural form Bilder  'pictures', something that does not happen to a final . So maybe sciocco does have a phonemic geminate after all, as it affects the preceding definite article in a way that non-geminates don't. Then again, there's simply no contrast in the word-initial position, as far as I can see anyway. The initial  (or, whatever) simply requires lo as the definite article. I've read about the impure s, and to claim that sciocco starts with a geminate because of that looks like circular reasoning to me. Sol505000 (talk) 21:22, 23 April 2022 (UTC)
 * The phonemic structure of words is analyzed according to how they are pronounced in isolation, rather than in running speech. True, unless -- as you're quite right to observe that German Auslautverhärtung suggests -- further observation reveals clues in running speech that the structure in question isn't as straightforward and simple as it might seem at first glance. "Linking R" in some non-rhotic (rhotic-suppressing, actually) varieties of English: tuner and tuna in isolation are homophones, also if a consonant follows them; but follow tuner with amp and voilà, it's [ˈtjuːnər æmp] (but no [r] in tuna and..., tuna in..., etc.). -- "Impure /s/" is historically a syllabification problem, with results in Italian similar to the geminate-by-nature phonemes. Here's one students enjoy as an entertaining exercise in the necessity of sometimes looking beyond the individual word for phonological understanding: using Italian lo studente (not *il studente) 'the student' as a clue, explain why native Spanish-speaking DJs say ¡Música non-estop! with the extra vowel. Extra points for anyone who explains in light of that how/why French acquired the initial /e/ of école 'school'. Barefoot through the chollas (talk) 00:34, 24 April 2022 (UTC)
 * Sorry, forgot about this: The initial (or, whatever) simply requires lo as the definite article. Why? What caused or causes that? Barefoot through the chollas (talk)
 * I think that enough research has been done on "linking r" ( just to name one source. See also ) to disprove the analysis . And etc. is out of question in non-rhotic speech that has intrusive r - or, conversely, no linking, whether etymological or not. There's nothing "suppressing" about full non-rhoticity, you can say that about those (semi)-rhotic speakers who feel pressured to conform to a non-rhotic model of speech in England, Wales or the southern part of New Zealand (or, as a (semi)-rhotic expat, anywhere in the world where the local standard is non-rhotic).
 * When did lo begin to be used before initial ? Was there at all, or was it some other consonant (or a sequence of consonants)? Sol505000 (talk) 01:03, 24 April 2022 (UTC)
 * Not easy to judge, but it looks as though there's a good bit of confusion or at least imprecision going on there between non-rhotic (rhotic-suppressive) types with linking R and types with what's called intrusive R. There are very good sources available, but it doesn't really matter here. The point is that speakers with tuner and tuna behaving as described have phonemic final /r/ in tuner, revealed phonetically IFF a vowel follows, i.e. in utterances beyond the word level -- classic linking R. Another example from English that the individual word may not be sufficient is tutor = Tudor in North America, frequently homophones. Presumably, tu[th]orial tells the tale.
 * Historical sources of long-by-nature /ɲɲ/, /ʃʃ/, /ʎʎ/, /ddz/, /tts/ vary, and settling into definite article selection (essentially, which syllable of e.g. illu) didn't happen overnight. I don't have the book with as me here, but I'm sure that Maiden's A Linguistic History of Italian offers a clear description. For modern Italian, it's clear: showgirl is presumably a fairly recent Anglicism in Italian. Short [ʃ] in isolation: [ʃoˈɡɛrl], but it has length(ening) if preceded by a vowel, most frequently /a/ of articles, thus una or la [ʃʃoˈɡɛrl], same as the medial consonant of fascia [ˈfaʃʃa], which, always between vowels, can't escape its long nature. Barefoot through the chollas (talk) 06:51, 24 April 2022 (UTC)

Open vs Closed E and O
Is there any rule for determining whether an E or O is open or closed? When I learned Italian in the early 90s, we were only taught the open versions. Is that common in the US? Since it was obvious to me that there were at least 2 different versions of E and O, I would always ask my teacher about it, and she would just yell at me, insisting that there is only one pronunciation for each letter. I posted this question a while ago on the Talk page for Italian Orthography, but didn't get much response. Mitsguy2001 (talk) 16:37, 12 April 2022 (UTC)
 * Is there any rule for determining whether an E or O is open or closed? In stressed syllables, no (Bertinetto & Loporcaro 2005:137). When I learned Italian in the early 90s, we were only taught the open versions. Is that common in the US? Not that I know of, and certainly not in my experience. Perhaps you could clarify what was not answered in the rather extensive response you got to the same questions posed under Italian Orthography, including the references that you were given, especially the easily accessible article by Bertinetto and Loporcaro, pp. 137-138. Barefoot through the chollas (talk) 14:23, 13 April 2022 (UTC)
 * Mostly that I strongly expect that there is some rule for determining open vs closed vowels, even if it has many exceptions. And since you don't agree that beginners in the US are taught only the open versions, how are beginners usually taught about E and O?  Mitsguy2001 (talk) 15:03, 13 April 2022 (UTC)
 * Natives don't have any problems understanding speakers who don't make the distinction. Millions of speakers of Italian don't. This is probably why your teachers didn't bother to teach you that.
 * Just use a pronunciation dictionary if you want to learn to distinguish between them. For verb conjugation, you'll probably have to look for other materials to learn the patterns (I imagine the same is true in the case of Portuguese). Sol505000 (talk) 15:57, 13 April 2022 (UTC)
 * Your expectation is met to some extent historically: in central Tuscany Latin stressed Ĕ in open syllables developed as /jɛ/ or occasionally /ɛ/ (Latin FĚRUS > fièro), as /ɛ/ in closed syllables (SĚPTEM > sètte); similar for o. Without knowing the Latin, however, even this simple bit is unknowable and unguessable coherently, i.e. not rule-governed. -- Off topic, but if you're interested in the historical approach, a very accessible source is Peter Boyd-Bowman's From Latin to Romance in sound charts, a nicely affordable little paperback. Some may claim that that sort of knowledge is esoteric, but it can be practical. I was once in the hinterlands of northern Mexico and needed to say 'scythe' or 'sickle', either would do. Because I knew falce in Italian, and the regular historical phonology of Spanish, I was able to guess hoz without much hesitation. My interlocutor had no idea that I had assembled the term on the fly. Barefoot through the chollas (talk) 17:21, 13 April 2022 (UTC)
 * Thank you. It sounds like open vs closed vowels is not something that needed to be taught in an introductory course.  And even though the closed version seems more common, I guess we were taught the open version, since they were afraid that we'd pronounce the closed E like an English long A, and the closed O like an English long O.
 * I guess my question was if there is any rule of thumb, such as open vowels in closed syllables, or closed vowels in open syllables. But I'm guessing that is not the case at all?  Or is that a rule that is true more often than not?  I have read that rules exist but are not consistently followed.  Whereas you suggest that no rules exist at all. Mitsguy2001 (talk) 18:49, 14 April 2022 (UTC)
 * All 7 vowels can be long or short. Quality has nothing to do with it, it's an allophonic effect (stressed vowels are long before single consonants followed by a vowel). Sol505000 (talk) 19:18, 14 April 2022 (UTC)
 * I recall having heard the open-syllable vs. closed-syllable vowel quality tale in my undergrad days, back in another millennium. Didn't take long to find that it didn't work. From Treccani vocabolario: lèttera (o léttera), pénna, légge s. f., lèggere v. tr, carézza, èssere, coténna, fésso, gèmma, marémma, méttere, millènne, principéssa, pètto, io pèttino, détta (atto del dire), dètta (debito), fétta, sètta, vétta... Barefoot through the chollas (talk) 21:36, 14 April 2022 (UTC)
 * In any case, it seems that, at least in theory, open vs closed E and O have their roots from the original Latin, but in modern Italian, the distinction seems to be somewhat lost. Is it the same in French and Portuguese, or are they more accurate and/or consistent when it comes to open vs closed E and O? Mitsguy2001 (talk) 16:22, 15 April 2022 (UTC)
 * If you're actually interested in this stuff, Boyd-Bowman's book is a good place to start your explorations. Barefoot through the chollas (talk) 16:41, 15 April 2022 (UTC)

I guess none of this discussion is going to lead to update to the article. Mitsguy2001 (talk) 16:55, 15 April 2022 (UTC)
 * If you think something needs updating, make specific suggestions. Better yet, consult the appropriate scholarly references and propose your own updating. In either case, keep in mind the purpose and scope of encyclopedia articles. Barefoot through the chollas (talk) 17:23, 15 April 2022 (UTC)
 * I definitely don't have the knowledge to update this article. Mitsguy2001 (talk) 17:27, 15 April 2022 (UTC)
 * The article should definitely tell the readers about the distribution of the close and open mid vowels like French phonology does, especially if there aren't that many rules. AFAICS only the open and  occur in the diphthongs, so that  and  are not possible in Italian (though Bertinetto & Loporcaro do list ). Sol505000 (talk) 18:13, 15 April 2022 (UTC)
 * Good point that the diphthongs need work in a sweeping cleanup of that section. But since there's nothing even faintly like the French loi de position for Italian, there can be no account resembling the French for prediction of open/close(d) in general (see the brief list above from Treccani; it could go on and on). FWIW, /ei/ is genuine Italian, /ou/ is not. Barefoot through the chollas (talk) 19:09, 15 April 2022 (UTC)
 * It seems that it's common, even in Italy, to not teach the difference between open vs closed E and O vowels. Should it maybe be mentioned that at least for some speakers and/or some regions, there is no real difference between the open and closed E and O vowels?  I'm still curious: are French and Portuguese the same way where open vs closed E an O is mostly ignored, or do people pay more attention to it in those languages? Mitsguy2001 (talk) 16:15, 21 April 2022 (UTC)
 * If you're still understandable when you fail to distinguish between the open and close mid vowels, I guess that the result is considered "good enough" by most teachers, who feel that maintaining some of your foreign accent is inevitable. Well, maybe... if you're learning Danish. Or a heavily tonal language. Italian is neither, and it's a 19th-century approach. I always try to nail the vowels of the language I'm learning. The result isn't perfect, but at least I have my own definition of "good enough" :P
 * $⟨è⟩$ and $⟨ò⟩$ seem to always stand for the open-mid vowels. I imagine that those letters are discussed at some point in a course - but there is a difference between the letters $⟨è⟩$ and $⟨ò⟩$ and the sounds and, which are more often written without the diacritic (so the same as the close-mid  and ). That's what they probably fail to mention. Then again, maybe not. In my experience, most Poles who speak German can't tell  and ,  and  as well as  and  (other vowels aside) apart. Sol505000 (talk) 21:40, 23 April 2022 (UTC)

There seem to be many sources, all written in Italian, that explain when an E or O is open vs closed. So it's not entirely true that there is no rule. But it's also not a simple rule of thumb that can be posted here. It depends on what endings and parts of speech, and there seem to be exceptions. Seems like it's something that an advanced native Italian speaker would be concerned about, not a foreigner learning Italian. Mitsguy2001 (talk) 16:40, 29 July 2022 (UTC)
 * I see no reason not to describe at least the main rules here. This is not a course on Italian pronunciation aimed at foreigners but an actual description of the phonology of Standard Italian. Sol505000 (talk) 14:14, 31 July 2022 (UTC)
 * Is this something we would thus need to rely on dictionaries to confirm? The alternative would be to delineate what these morphological those rules and exceptions are so that editors would be consistent, but that might not be feasible. — Æµ§œš¹  [lɛts b̥iː pʰəˈlaɪˀt] 18:18, 1 August 2022 (UTC)
 * Once the principles that Mitsguy reports having found are revealed, descriptive dictionaries -- or better, Canepari's Dipi -- would be a touchstone for testing the principles. Barefoot through the chollas (talk) 18:43, 1 August 2022 (UTC)
 * I found many sources by doing a google search on "e o aperta chiusa" (without the quotes), including an Italian Wikipedia page: Dizione della lingua italiana - Wikipedia
 * I won't post links to the other sources since I cannot vouch for the accuracy of any of them in particular. But, in an case, my point is that it's wrong to say that there are no rules, but there are far too many rules to post here. Mitsguy2001 (talk) 16:23, 2 August 2022 (UTC)
 * Mitsguy, it's helpful to post the sources you've found, as long as they're serious and informed, not obviously amateur hour stuff. Posting them doesn't mean you vouch for them. They can be tested against Canepari, who's pretty much guaranteed to present reality rather than wishful thinking (or worse). Try some yourself. Read the intro to be able to understand the coding (click on Guida al Dizionario on the main web page), then do some searches. Try bestia and liquore, for example. https://www.dipionline.it/. Barefoot through the chollas (talk)
 * Honestly, there are too many sources that come up to list. I did include a link to the Wikipedia page.  You can use that or whatever sources it references. Mitsguy2001 (talk) 20:53, 2 August 2022 (UTC)
 * This is your topic. You've been flirting with it for months on Wikipedia talk pages. You've been referred to the publications of expert phonologists who explain the situation clearly. You've been introduced to Italy's premier phonetician, you have the url to his online phonemic dictionary, and with that a link to his personal web page containing among other things a full bibliography of his book-length publications on Italian pronunciation. A mixed metaphor of leading a horse to water served on a silver platter comes to mind. Barefoot through the chollas (talk) 22:16, 2 August 2022 (UTC)
 * That was completely uncalled for and it added nothing to this page. Mitsguy2001 (talk) 01:36, 3 August 2022 (UTC)
 * Did you even check this? https://www.dipionline.it/ Barefoot through the chollas (talk) 01:50, 3 August 2022 (UTC)
 * Yes, and that is only to look up specific words. It doesn’t give rules, so it’s useless for this article. Mitsguy2001 (talk) 01:53, 3 August 2022 (UTC)
 * Actually, it's enormously informative for this article. Looking up just the two words I suggested, bestia and liquore, teaches valuable lessons about searching for "rules" governing the vowels in question. Barefoot through the chollas (talk) 02:00, 3 August 2022 (UTC)
 * It'd be WP:OR to come up with rules based only on dictionary entries. We need something explicit. Sol505000 (talk) 07:15, 3 August 2022 (UTC)
 * I agree, that would be original research. I think the best bet is to include a link to that Italian Wikipedia page. Mitsguy2001 (talk) 13:16, 3 August 2022 (UTC)
 * Might not do any harm. It's my understanding that Wikipedia articles are not deemed acceptable sources, but there shouldn't be anything wrong with a link to it. As a similar illustration of non-authoritative prescriptivism, perhaps classical diction such as this could be linked:
 * http://www.attori.com/dizione/Diz03.htm
 * Before going any further, though, there are two things to be considered.
 * The first is illustrated on p. 137 of Bertinetto & Loporcaro's article: F[lorentine]I and R[oman]I contrast in the stressed vowel of each of lettera, scendo, storpio and colonna (the list could continue at length), and, importantly, "S[tandard]I accepts each variant except for the last" (Roman open "o" in colonna). The clear lesson is that the crux is lexical, not systematically phonological, and that the phonological structure of lexical items can vary (no surprise; cf. the infamous Mary, merry, marry contrasts or not in AmE).
 * The second thing to consider is more crucial: in phonological terms there is nothing problematic about Italian mid vowels. Problematic aspects surface only in relation to orthographic conventions. The dictum "Italian is a phonetic language" (by which is meant that the orthographic system used to represent Italian is phonemic) fails for the mid vowels. That's a problem of orthography, not phonology, thus the obvious question: How much discussion of it is appropriate in an article entitled Italian phonology? Barefoot through the chollas (talk) 14:45, 3 August 2022 (UTC)
 * It seems that the 3 issues are:
 * 1. People in different regions have differences in whether they pronounce a certain vowel as open or closed, and nobody speaks a version that is completely standard.
 * 2. There is no orthographic distinction between an open vs closed E or O.
 * 3. There are very few words that are disambiguated solely be the difference between an open or closed E or O, and those that are disambiguated that way either can be determined based on context, or are sufficiently rare.
 * 4. The difference between the open vs closed vowel is small, so you're understood either way. Mitsguy2001 (talk) 16:16, 12 August 2022 (UTC)
 * 1. Probably true, although "nobody speaks a version that is completely standard" is difficult to test.
 * 2. True, thus this entire discussion turning more on orthography than phonology.
 * 3. and 4. Mostly true, assuming that the very frequent words [ɛ] and [e] are clear by context. Barefoot through the chollas (talk) 17:26, 12 August 2022 (UTC)
 * Sorry, that last is badly stated: assuming 'is' and 'and' are clear by context if not pronounced [ɛ] and [e] respectively. Barefoot through the chollas (talk) 17:51, 12 August 2022 (UTC)