Help talk:IPA/Italian/Archive 1

Archive 1

Archive 2

gg?

I'm not a linguist, so I hope I won't write anything [too] stupid. I have noticed that most consonants are listed in their "single" and "double" form [/l/ for "pala", and /ll/ for "palla"]. That's not the case for /g/. Is the lack of /gg/ [as in "agguato"] intentional, or was it overlooked?

Cheers, Rgiuntoli 08:44, 28 May 2007 (UTC)

Overlooked no doubt: it’s in the list at it:Fonologia dell'italiano, as is /dʤ/ (maggio, oggi, peggio). —Ian Spackman (talk) 05:27, 11 June 2008 (UTC)

IPA-ing apostrophed words

Hello,
silly question for which I have found no answer on wiki (yet): how does one render in IPA proclitic apostrophed words so typical of the Italian language?
E per chi mi leggesse dalla patria, come si trascrivono foneticamente le proclitiche apostrofate, come nell'esempio seguente?
Ex: pietra dell'acculata = [ˈpjɛ.tra dell ˌak.kuˈla.ta], [ˈpjɛ.tra dell’ˌak.kuˈla.ta], or [ˈpjɛ.tra del.lˌak.kuˈla.ta]?

Thanks! -- Francesco Campelli (talk) 19:48, 3 February 2009 (UTC)

My understanding is that the apostrophe is an orthographic convention in Italian that gives non-phonetic information. So, in short, there really isn't a way of indicating that. But if, in your transcribing, you also have the original orthography, it should be okay. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 21:41, 3 February 2009 (UTC)

By calling it a proclitic, you imply that it forms one whole with the following word. So the proper division would be [ˈpjɛ.tra delˌlak.kuˈla.ta]. −Woodstone (talk) 22:57, 3 February 2009 (UTC)

I just noticed the secondary stress marker. Italian has secondary stress? — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 18:14, 4 February 2009 (UTC)

Thank you both. Yes, we do have secondary stresses. And thanks for the indication of how to express a proclitic (which l' and dell' would be) and, by extension, enclitics too. I guess the same would apply to articles in French…l'amande or l'exploit… Best regards to both of you! Campelli (talk) 21:48, 4 February 2009 (UTC)

ɟ

In Italian there's the foneme ɟ as in ghianda and ghetto. —Preceding unsigned comment added by 87.14.197.41 (talk) 12:25, 5 January 2009 (UTC)

That's not a phoneme, that's an allophone of /g/ before /i/. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 06:33, 6 January 2009 (UTC)

Not even that; at least in my dialect, it's just a [ɡʲ], no more palatalized that anybody would expect before /j/; it is definitely not the non-nasal counterpart of the ɲ in spugna. If you pronounce gara, ghianda, spugna you see that the point at which the tongue blocks air is more-or-less the same in the first two (altough the front part of the tongue is more forward in the second), and about 1 cm more forward in the last one. --A. di M. (talk) 00:04, 23 February 2009 (UTC)

Request for clarification

If I understand correctly one should use phonemic transcription except that one should mark long vowels, velar n, and secondary stress, right? (And does anyone know of any Italian dictionary where to look up secondary stress? Or do we put it "by hear") --A. di M. (talk) 00:08, 23 February 2009 (UTC)

Since it seems like geminated /ʎ ɲ ʃ/ are also allophonic, that's another nonphonemic thing we're indicating. I'm not sure about the distribution of secondary stress. From the examples, it looks like secondary stress may be every other syllable preceding the primary stressed one but maybe native speakers or experts can clarify.

Also, we shouldn't be marking long vowels unless we explain here where they're found. I'd say we needn't do so since it's sort of redundant but if other's disagree I won't be violently opposed. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 05:35, 23 February 2009 (UTC)

Yes, I forgot /ʎ ɲ ʃ/. As for secondary stress, I don't think the rule is that simple (for example in televisione /televiˈzjone/, to my hear it sound like that there's a very slight stress on the first syllable, if anywhere). But I think it is ever more useless than the vowel length (about which I have no opinion), considering that even in Italy most dictionaries just show the primary one. --A. di M. (talk) 09:24, 23 February 2009 (UTC)

Ahh, well then I've removed the secondary stress from the guide. In addition to secondary stress having a dubious importance here, we'd need clarification on how to determine where to put it anyway. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 17:31, 23 February 2009 (UTC)

I added vowels length, since that's another phonetic detail that's in many of the templates, and I added it to the few that were missing it. (I think such allophony is worth indicating, since except for the mid vowels and /dz/, the only people who will need this won't know Italian.) Also an example of stress on a geminate consonant: I assume that's why we're writing [kk] rather than [kː], but I've been finding [aˈkka] and [akkˈa] rather than [akˈka] in the articles.

Oh, I removed the geminate z in Uffizi. I assume that was an error, but my apologies if that word is irregular. Also, at Autobianchi, do we really want a long [ŋː]? kwami (talk) 05:12, 2 May 2009 (UTC)

"ia" not a diphthong

Sorry for reverting without a comment: "ia" as in "via" is not a diphthong, as "via" is a two-syllable word. Goochelaar (talk) 13:32, 1 August 2009 (UTC)

Should we make use of the syllable boundary so that via would be transcribed [vi.a] or are the differences between hiatus and diphthong sufficiently clear with the use of j? — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 20:18, 1 August 2009 (UTC)

Correct IPA for "baccano"

I'm not really sure where to ask this...

Is ba.kka.no the correct IPA for baccano (the word for noise or din)? ~Itzjustdrama ^{? C} 00:21, 5 October 2009 (UTC)

This is a good place to ask. It's missing stress, and the syllable breaks are wrong. I can only assume that the [o] is correct. kwami (talk) 00:31, 5 October 2009 (UTC)

I think I found it: bakˈkano. ~Itzjustdrama ^{? C} 18:08, 5 October 2009 (UTC)

Yes, that's right. I sub'd the correct stress mark. kwami (talk) 21:23, 5 October 2009 (UTC)

Thanks! ~Itzjustdrama ^{? C} 21:49, 5 October 2009 (UTC)

Levi

Is Levi (as in Primo Levi) [lɛvi] or [levi]? I want to put this in so people don't call him Pry-moe Lee-vye as I did when I was little... Lfh (talk) 14:53, 7 January 2010 (UTC)

English approximations

I'm being bold and adding English approximations for most of the vowels, and some of the consonants, while keeping all the Italian examples. If the formatting or word choices are deemed inadequate, I'll understand if it's reverted - but let's see. Lfh (talk) 15:06, 21 January 2010 (UTC)

I'm gonna go out on a limb here and say we don't need to relist every consonant in its geminate form. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 17:45, 21 January 2010 (UTC)

I'm not a good enough linguist to check all of your choices; but I'm pretty sure the "o" in English "no" is the diphthong [əʊ̯]. It's a common mistake for English speakers learning Italian to insert unwanted diphthongs. I don't think English has a pure vowel that's exactly the same as the Italian "o"; the closest would be the vowel in "not". See Wikipedia:IPA for English. Jowa fan (talk) 04:41, 27 January 2010 (UTC)

Good point. Some of my choices are very approximate. I picked "no" because we use the same vowel in Wikipedia:IPA for Spanish; while at Wikipedia:IPA for German, we also use that vowel but specify Scottish English. "Not" probably isn't any better because most North Americans would have something like [ɑ] for that. But "no" may have to be tweaked. Lfh (talk) 08:23, 27 January 2010 (UTC)

I replaced "cat", "pan" and "task" with "scar", "spin", and "star" since the Italian /k/, /p/, and /t/ are not aspirated, unlike initial /k/, /p/, and /t/ in English. --Andrew C ^talk (afc0703) 01:33, 3 March 2010 (UTC)

IPA links

Recently, this page was edited to wikify the IPA characters to their corresponding articles. One grave error is the classification of the Italian /t/ and /d/ as alveolar, when they are in fact dental. I'll be changing the links shortly, but my question is: should the IPA symbols reflect the dental quality of these consonants? The diacritics are commonly left off of [t̪] and [d̪], especially in broad transcriptions. --Andrew C ^talk (afc0703) 19:42, 2 June 2010 (UTC)

IMO there's no need for diacritics. I also doubt they're dental: more likely alveolar but laminal as in French and Spanish. (All three are commonly called "dental", but in the case of F & S that's an error for "laminal".) — kwami (talk) 19:50, 2 June 2010 (UTC)

French and Spanish (and I believe Italian as well) are laminal denti-alveolar. But the diacritic is unnecessary as the distinction isn't contrastive in Italian. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 20:16, 2 June 2010 (UTC)

transclusions checked

I've gone through all transclusions of this template, and all the following should be correct: only the symbols used in this key are found; medial ʎ,ɲ,ʃ are always geminate; long vowels are marked long; all entries have stress; geminate C's are written double (Please verify Venice doesn't have a genitive [tts]); della, delli is not stressed; /n/ is velar before k,g; no j after tʃ,dʒ,ʎ,ɲ,ʃ unless redundant i in the orthography (gliV, gniV, as in pagliacci). Exceptions: 2ary stress; 'Ndrangheta (syll. n), Prisencolinensinainciusol (is that supposed to be English?) — kwami (talk) 01:48, 27 October 2010 (UTC)

Isn't it [paʎˈʎatʃi], and the i is just there to "soften" the preceding consonant(s)? - AlexanderKaras (talk) 11:09, 19 December 2010 (UTC)

I don't know. Why would it need to be "softened"? So, you're saying it's pronounced as if it were paglaci, with no i and one c? — kwami (talk) 13:04, 19 December 2010 (UTC)

mid vowels

The phonology article says that the mid vowels are in complementary distribution when unstressed. However, our examples in this key appear to violate the stated distribution. Can someone check this? — kwami (talk) 08:53, 26 January 2011 (UTC)

You are right, perhaps we could only indicate stressed vowels and add an explanatory note: close-mid and open-mid vowels are in complementary distribution when unstressed.

What do you think we call [j] and [w] semivowels, despite they are defined as approximants. We could list them on the consonant side of the guide, same as WP:IPA for French. If you don't think this is right, just revert my last edition. :) Jɑυмe (xarrades) 20:09, 26 January 2011 (UTC)

I think we should have unstressed mid vowels too, but add a note explaining their distribution. It just seems that our transcriptions violate that distribution.

Yes, IMO /j/ and /w/ are best put in the C section. There's no need to separate them out: the key is for people who do not know the IPA, and for such people, j and w are consonant letters. — kwami (talk) 06:16, 27 January 2011 (UTC)

Aspirated /k/?

Is /k/ in Italian aspirated or unaspirated? In Italian phonology, I can see that <c, k> is phonetically transcribed as [k], not [kʰ], but often I think I hear people pronounce it, in this language, like the word-initial /k/ in English and this feature is said nowhere, so I am confused. --125.24.12.231 (talk) 07:32, 29 July 2010 (UTC)

Not aspirated AFAIK. — kwami (talk) 13:05, 19 December 2010 (UTC)

Aspirated only in Tuscany, not in the rest of Italy. --87.3.90.252 (talk) 18:10, 15 June 2012 (UTC)

Issues

According to my dictionary (Oxford-Paravia; English-Italian, Italian-English with IPA transcription of words) there are not long vowels, so "primo" is simply [ˈprimo] and not [ˈpriːmo]. Also /ŋ/ does not exist, so it is [fango] and not [faŋgo]. Is it possible to cite a source which states otherwise? --87.3.90.252 (talk) 18:19, 15 June 2012 (UTC)

I believe the source(s) that state otherwise are cited in Italian phonology. It sounds like that dictionary is covering phonemic contrasts and avoiding more minor ones like lengthening and the velar nasal. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 03:11, 16 June 2012 (UTC)

I do not think so. As for English, the dictionary showes lengthening and the velar nasal. I wonder if those aspects are commonly accepted or not (for Italian language). --87.1.25.242 (talk) 08:44, 16 June 2012 (UTC)

The velar nasal is commonly accepted. As for the vowel lengthening, see this source. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 13:19, 16 June 2012 (UTC)

And Venetian?

Do we want to add Venetian to this guide? Venetian language seems to be saying that the two are nearly identical in their phonology but for a few things. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 18:42, 19 October 2012 (UTC)

Yes, we would surely need to add Venetian! But, is there anyone able to work on it? Maybe someone from vec.Wiki? --Checco (talk) 08:38, 22 October 2012 (UTC)

Single z: "z" or "ts"?

It was my impression that a double z as in "pizza" is pronounced "ts" to approximate sounding each letter separately, but a single z i just a z. Is that correct? My reason for asking is the article on Fabio Lanzoni, which indicates a pronunciation that I would expect for "Lanzzone." Monado (talk) 00:06, 24 March 2013 (UTC)

No. Zz is just a long z, like tt in spaghetti is a long t. Z may be either [ts] or [dz]. There's no way to predict which. — kwami (talk) 05:35, 24 March 2013 (UTC)

r and rr

There is a difference between r and rr. Shouldn't it be like the Spanish IPA page with ɾ and r?--77.0.253.160 (talk) 22:07, 28 August 2013 (UTC)

No, because in Italian it's analogous to p, t, k and pp, tt, kk. — kwami (talk) 00:42, 29 August 2013 (UTC)

Isn't "caro" with only a single tap and "carro" with a trill?--77.1.173.130 (talk) 22:50, 15 September 2013 (UTC)

No. They're both trills. — kwami (talk) 20:49, 29 September 2013 (UTC)

If we assume that a single-cycle trill makes a tap, then Italian /r/ can be a tap, but it can also be pronounced with 2 closures. — Lfdder (talk) 21:02, 29 September 2013 (UTC)

It doesn't. A single-contact trill is quite different from a tap in its articulation. In one the tongue is passive, in the other active. — kwami (talk) 21:06, 29 September 2013 (UTC)

The gesture for an equal-duration on-glide, closure and off-glide tap and single-cycle trill is (can be) exactly the same. — Lfdder (talk) 21:49, 29 September 2013 (UTC)

Ref? My understanding from Ladefoged was different. — kwami (talk) 21:50, 29 September 2013 (UTC)

See doi:10.1017/S0025100300005405, p. 40–41 and doi:10.1017/S0025100300005296, p. 2–3. The distinction seems to lie in that the tap involves a single "ballistic" movement that basically means the tongue bounces back, whereas the trill involves two distinct movements -- which is more or less what you said. However, it seems that some reserve this interpretation for the flap, and not the tap. — Lfdder (talk) 22:43, 29 September 2013 (UTC)

Request for comment: "moai" as a dubious example of an English suprasegmental

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.

---

Is there a better example of a suprasegmental in English than the [non-English] word "moai"? It seems than even the Polynesian etymology of this word is unclear, and using it as an English phonetic example (as is currently the case on this page) does not seem at all like a good idea— I am not comfortable enough with the concept of a suprasegmental to provide a suitable replacement on my own, but am hoping other editors will be able to assist with this (?). KDS4444^Talk 12:33, 19 April 2015 (UTC)

• The period (full stop) is used in the IPA to indicate a syllable break, as described in International Phonetic Alphabet#Suprasegmentals. I assume that the editor who added "moai" wanted an example of a syllable break between two vowels. A word like "rower" /ˈɹoʊ.ə(ɹ)/, "skiing" /ˈski.ɪŋ/, or "co-op" /ˈkoʊ.ɒp/ would serve that purpose without being so conspicuously foreign. —Granger (talk · contribs) 00:54, 20 April 2015 (UTC)

• Definitely best to replace moai with a more common word. I'd never heard of it before seeing it on the page, and that's probably true of many other readers. As Mr. Granger says, many more common English words have hiatus. — Eru·tuon 01:12, 20 April 2015 (UTC)

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.

Please help

What IPA correct for Italian name "Daniela"? Could you also help me with IPA for Italian surname "Galli"? I need this info for Dhany article. — Preceding unsigned comment added by 213.80.232.10 (talk) 07:24, 7 September 2014 (UTC)

 Done --95.251.19.191 (talk) 22:25, 24 August 2015 (UTC)

I assume Dhany is pronounced [ˈdaːni]. I am right? Peter238 (talk) 13:25, 16 September 2015 (UTC)

Southern American 'time' as an approximation of Italian /a/

I find it a bad choice. How many people from outside the North America are familiar with that pronunciation? Probably not that many. 'Art', on the other hand, has much more commonly a back-to-central vowel, and is therefore a better choice. Peter238 (talk) 12:03, 16 September 2015 (UTC)

Agreed. Most readers aren't amateur dialectologists. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 13:03, 16 September 2015 (UTC)

@IvanScrooge98: What's your argument for Southern American 'time'? Peter238 (talk) 15:17, 17 September 2015 (UTC)

I just found it was more similar to the Italian sound than it was fan in Modern RP (which I had previously added): I agree with restoring that, but I don't think art is really similar to [ä] (at least, not in most accents). イヴァンスクルージ九十八 _{（トーク）} 17:18, 17 September 2015 (UTC)

This is one of those sounds that there isn't a perfect equivalent for. Let's just use father. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 03:25, 18 September 2015 (UTC)

Anon

There's an anon (currently under the IP @151.20.99.127:, but before under many others) that's stalking @IvanScrooge98: (I know that's allowed... theoretically) and removing the syntactic gemination symbol from the articles. I've just warned him. Before, I reverted about 60 of his edits, but there are these transcriptions that he then changed:

• [1]: the edit summary tells me nothing about the [n]. Should it be short [ˈrɛntso] or long [ˈrɛnntso]?
• [2]: geminated or not geminated?
• [3]: geminated or not geminated?
• [4]: geminated or not geminated?

You can check the rest (mostly not reverted... as of now) here, if you want. I might've screwed up some more transcriptions, I'm not sure. Sorry for that if I actually did so. Peter238 (talk) 14:18, 8 October 2015 (UTC)

I checked all of these in DiPI Online and Dizionario d'ortografia e di pronunzia. All of the aforementioned edits are correct. Also, the anon chilled out and explained his behaviour, so I think the case is pretty much closed. Peter238 (talk) 15:24, 8 October 2015 (UTC)

English equivalents

I have made some bold changes to the English equivalents. Some are unclear or incorrect. Martin Hogbin (talk) 09:26, 21 October 2015 (UTC)

The sound in the English (AmE or BrE) word 'million' is not the same as the sound in the Italian 'gli'. It is not even a helpful approximation.Martin Hogbin (talk) 15:27, 21 October 2015 (UTC)

Have a listen to [[5]] for example. There is no sound like this in any English word, in any dialect of English, that I know of. Martin Hogbin (talk) 17:04, 21 October 2015 (UTC)

I have removed the English example 'million' as per our discussion at Talk:Palatal_lateral_approximant#That_was_quick.21 Martin Hogbin (talk) 17:53, 22 October 2015 (UTC)

Wiktionary: [6] (/ɲ/~/nj/) [7] (/ʎ/~/lj/) i.e. "canyon : ɲ = million : ʎ" :-| — Preceding unsigned comment added by 5.101.99.101 (talk) 08:33, 24 October 2015 (UTC)

I do not follow you. I see /ˈkænjən/ Martin Hogbin (talk) 12:15, 24 October 2015 (UTC)

/ˈkænjən/~/ˈmɪljən/

It's not /ɲ/~/nj/ and /ʎ/~/lj/, but [ɲ]~[nj] and [ʎ]~[lj]. You're mistaking phonemes for allophones. Actually, canyon on Wiktionary contains only the phonemic transcription /ˈkænjən/, no mention of [ɲ]. That said, you should read the whole discussion on talk:palatal lateral approximant#That was quick!. Peter238 (talk) 13:24, 24 October 2015 (UTC)

If [nj] is similar enough to [ɲ] to be used as an example of how "gn" sounds like in Italian, then why isn't [lj] similar enough to [ʎ] to be used as an example of how "gl" sounds like in Italian?

"[nj] : [ɲ] = [lj] : [ʎ]"

I said what I had to, I don't want to be insistent so you can choose to do whatever you like, I'm leaving. 5.101.99.101 (talk) 14:14, 24 October 2015 (UTC)

I don't have a problem with using million as an example word for [ʎ]. I already said that on talk:palatal lateral approximant#That was quick!, which I advised you to read. Peter238 (talk) 14:26, 24 October 2015 (UTC)

Following from our discussion on Talk:Palatal_lateral_approximant#That_was_quick.21 page, I suppose we need to ask ourselves what the purpose of this page is. Is it to give people a rough idea of what Italian sounds like or is it to help people to know what sounds Italians really say. I would agree that the sound in 'million' is probably the closest we have in English to the sound in the Italian 'gli' but, in my opinion, it is sufficiently different to be positively misleading to people who want to know what real Italian sounds like. If we want to help those people, we are better to leave it blank, and link to a sound file somewhere.

The problem is compounded by the, to my mind ridiculous, decision to use in Wikipedia the // form of the IPA, where a symbol can make a different sound in each language or even dialect, so ʎ can sound different in English and Italian. Crazy, utterly unhelpful and confusing. This, to my mind defeats, the original purpose of the IPA which was to have each symbol making a fixed sound, and, yes I have read the original book on the IPA and that was its original purpose. Martin Hogbin (talk) 13:25, 24 October 2015 (UTC)

Italian transcriptions on Wikipedia use brackets ⟨[]⟩, not slashes ⟨//⟩. English transcriptions, on the other hand, do use the slashes, because English has three main standard varieties (British (RP), American-Canadian (GA), Australian (GenAus)) that are equally valid, yet sound noticeably different from each other.

No, it's not "crazy, utterly unhelpful and confusing". You just don't understand what a phonemic transcription is. Strictly speaking, a phoneme is not even a sound, but more like a 'container' for sounds (sometimes just one sound). Besides, no IPA transcription (save for amateurs not knowing what they're doing) of English would write /ʎ/, because English does not have the /ʎ/ phoneme.

The original purpose of the IPA was not to have each symbol making a fixed sound, but to provide symbols for all of the sounds that are phonemically distinct in world's languages. For instance, IPA ⟨n, t, d, ɹ, l, r, ɾ⟩ can be dental, alveolar or postalveolar, depending on the language, whereas the symbols ⟨e, ø, ɘ, ɵ, ɤ, o⟩ can be used to transcribe the close-mid vowels [e, ø, ɘ, ɵ, ɤ, o] or true-mid vowels [e̞, ø̞, ɘ̞, ɵ̞, ɤ̞, o̞], depending on the language. In some cases, the symbols ⟨e, ø, ɘ, ɵ, ɤ, o⟩ can even be used to transcribe phonetically open-mid vowels [ɛ, œ, ɜ, ɞ, ʌ, ɔ] (that's rarely, if ever, the case in case of [ɜ] and [ʌ]). Peter238 (talk) 14:20, 24 October 2015 (UTC)

I think our purpose, at least as far as the "English approximation" column is concerned, is closer to giving people a rough idea of what Italian sounds like. This doesn't mean we can't guide people towards more nuanced, technically accurate information on the sounds of Italian. I believe Italian phonology accomplishes this, as do the individual phone pages (both of which we link to). If we feel like that would be insufficient, we could also provide links to sound files for individual Italian examples. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 15:19, 24 October 2015 (UTC)

Martin Hogbin: And where's the "consensus" for your version that you claim? [8]. I see people arguing above that it's intended to give a "rough idea" for laypeople, that million is about the best approximation one can get in English, which is also indicated by the column title "English approximation". If find the outright removal counter-productive.
Help:IPA for Serbo-Croatian (and many other IPA help pages) contain an even more explicit disclaimer that English approximations are in some cases very loose, and only intended to give a general idea of the pronunciation. See XXX phonology for a more thorough look at the sounds. Help:IPA for Polish provides elaborate notes on distinction between retroflex and alveolo-palatal consonants. In my opinion any solution is better than leaving an unhelpful empty space. No such user (talk) 15:27, 12 November 2015 (UTC)

Assimilated [m]

Context, the following note:

The nasals always assimilate their place of articulation to that of the following consonant. Thus, the n in /nɡ/~/nk/ is a velar [ŋ], and the one in /nf/~/nv/ is a labiodental [ɱ] (though for simplicity ⟨m⟩ takes its place in this list). A nasal before /p/ and /b/ is always the labial [m].

This may be correct phonetically; I am not a phonetician. However, in my estimation, it's not serving our readers well to transcribe Inferno (for example) as [imˈfɛrno]. An Anglophone who takes this advice literally will sound "more wrong" (to my non-native-speaker ear) than one who tries to render [inˈfɛrno].

I also kind of suspect that these are fast-speech forms, and that the "reference form" would indeed be [inˈfɛrno].

I would like to hear from native speakers about this. There may be some here; I haven't really spent much time hanging around this page. I'll also notify WP Italy and the language refdesk. --Trovatore (talk) 21:34, 11 November 2015 (UTC)

Hello, I'm a native speaker and I'll tell you my opinion about that. Both /n/ and /m/ are half-wrong and half-right, the only fully correct form is /ɱ/, if we choose to avoid using it and replace it with /n/ or /m/ there's no preference, maybe /m/ is graphically more similar to /ɱ/ than /n/ so we should prefer it. Yet I can't see the reason to choose a different IPA symbol if the sound is /ɱ/, specially if for Italian it's already used another nasal symbol which occours only before 2 consonants: /ŋ/. It's an allophone of /n/ before /k/ and /ɡ/ because they can never be preceded by /n/ or /m/, exactly as for /ɱ/ which is an allophone of /n/ before /f/ and /v/ because they can never be preceded by /n/ or /m/. This inconsistency looks strange, I wonder why if we use /m/ instead of /ɱ/ for "simplicity" we can't do the same thing for /ŋ/ replacing it with /n/. That's my opinions, I hope it was helpful. — Preceding unsigned comment added by 5.101.99.101 (talk) 11:33, 12 November 2015 (UTC)

So my concern is that the choice of [m] is actively misleading to English speakers, presumably the main users of English Wikipedia. Do you agree that an English speaker who renders inverno as [im'vɛrno] will sound more wrong than one who says [in'vɛrno]? --Trovatore (talk) 17:24, 12 November 2015 (UTC)

It's fine discussing with a kind user, they're not so common among Wikipedians :) Well, that's just my opinion, but if I'm forced to choose the most correct sound between 2 wrong sounds I'd choose the /m/. It's a labiodental, more similar to labials than to dentals. Think about the related frivcatives: /v/ is more similar to /β/ (Spanish, for example) than to /ð/ (or /z/, or /ʒ/). On my opinion, the fact is that both me and you are accostumed to Latin alphabet orthography, that is the letter sequence "nv" and "nf" instead of "mv" and "mf", where the nasal is anyhow /ɱ/, that's why maybe instinctively we tend to prefer N to M. Ah, and I assure you that not only English speakers read en.wikipedia, particularly when one's native language Wiki is not as exhaustive as this one (Italian case).

May I also ask, where are you from? Is it possible this is a regional issue? I lived in Padova. Last night I called a friend who's a native of Milan and who has taught Italian diction to opera singers, and asked her pronuncia, lentamente e con cura, la stagione dopo autunno, and to my ear she quite clearly said [in'vɛrno]. I did not hear even any [ɱ].

Although I'm a little confused about what [ɱ] is, exactly. In one place I read that supposedly [ɱ] is the first nasal in the English word "symphony"; if that's so, she most definitely did not say that. Somewhere else, I see that it's like pronouncing an [n] but with your upper teeth touching your lower lip, and that I suppose is possible; I'm not sure I would hear the difference. --Trovatore (talk) 17:30, 12 November 2015 (UTC)

Even if I live in Northern Italy my parents are from Tuscany, so... :) I know that a regional issue about nasals concerns not /ɱ/ but /ŋ/: Northern Italians tend to pronounce /ŋ/ instead of /n/ between vowels or, rarely, at the end of the words, because local dialects phonetics is like that... Or they even nasalise the following vowel! Instead, pronouncing words exactly as they're written is not correctly in diction: "un po'" (a bit) must be pronounced /umpò/ not /un.pò/ (it's an error I hear sometimes in TV dubbing), I think it's the same for "in-verno", pronouncing /n/ is wrong.

Yes, your example is correct: English "symphony" is exactly the same as Italian "sinfonia" regarding the nasal, it's /ɱ/ in both cases. Maybe you made such an example because graphically in English it's used M instead of N? It's what I've said above, since the phonetic is identic it must be an instinctive issue related to word spelling. At least, that's what I think :)

I add a note: I've read what you wrote about the "Sacra conversazione", and the IPA /ˌkonvɛrsaˈtsjoːne/ is wrong, except for the /ɱ/ issue, because there's no secondary stress (it's not a compound word), the E is not open since it's not stressed, and the Z is geminated: it should be /koɱversatˈtsjoːne/ (or /komversatˈtsjoːne/ with the current conventions). You can change it, if the information I gave you was useful!

Ah, let's drill down on the sinfonia case. I do not think that is true. But I don't know whether the disconnect is the way you pronounce "symphony" or the way I pronounce sinfonia. Can you say something about the position of the tongue when the nasal is created? --Trovatore (talk) 18:55, 12 November 2015 (UTC)

I hope this time the position of my reply is correct. Well, talking about "sinfonia": when I pronounce it, while the second N is definitely dental, the first one has the tongue almost in the same position as if I'm pronouncing an M. I've clearly perceived my tongue touching my palate just in the syllabe "ni". Watch the examples in Labiodental Nasal page: the English example is exactly "symphony" and it contains /ɱ/, while the Italian example is "invece" and even there you can find /ɱ/.

It's also worth nothing that the sound you're discussing is [ɱ], not /ɱ/ (which does not exist in Italian - see phoneme and allophone). Peter238 (talk) 18:39, 12 November 2015 (UTC)

I haven't understood what you mean. Is it wrong to write / / instead of [ ]?

Please sign your messages. Yes, in this case it is wrong because there is no phonemic /ɱ/ in Italian, only the [ɱ] allophone of the phonemes /m, n/ which occurs only before /f, v/. Peter238 (talk) 16:23, 13 November 2015 (UTC)

Ok5.101.99.101 (talk) 16:49, 13 November 2015 (UTC)

Maybe it will help if we compare some actual sound files to see if we know what each other is talking about.
Here is my attempt to follow the current guides, saying [im'vɛrno] and [mamˈfreːdi]: File:Inverno and Manfredi rendered with m.wav
Here is my attempt using the [n] instead: File:Inverno manfredi rendered with n.wav
I am aware that my accent is pretty heavy, but I think it's clear that the second effort sounds more correct than the first. --Trovatore (talk) 19:35, 12 November 2015 (UTC)

Oh, I just realized that I stressed the first syllables trying to make the point. Rats. See what you think anyway; it's a pain in the butt to make these and upload them, but I can try again later. --Trovatore (talk) 19:37, 12 November 2015 (UTC)

OK, fixed them now. If you hear the stress on the first syllable, try clearing your browser cache. --Trovatore (talk) 19:41, 12 November 2015 (UTC)

Well, you pronounce Italian words quite well! But I must be honest about your point: the first one sounds more correct to my ears, while the second one... It's like if it was forcedly pronounced in that way, I don't know how to explain. However, my ears are Italian, so this is just an Italian's sensation, probbly different from a native English speaker.

Oh, User:5.101.99.101, one more point I thought of: You make a convincing point about un po'. I was sort of nodding along to myself, thinking, yeah, that sounds right, hmm, how does that affect the discussion.
But that's before a p. As I mentioned in a related thread at the language refdesk, I think the current guidelines do make sense, before b and p. (I'd have to think about un boia. Maybe there's a more common word we can try?)
Where I tend to dispute the guidelines is before f and v. --Trovatore (talk) 21:10, 12 November 2015 (UTC)

My example was about diction errors due to hypercorrectivism, that is pronouncing words exacly as they're written, unnaturally. Because the natural sound of a nasal before a /p/ or /b/ would be /m/. Likewise, the natural nasal sound preceding /f/ or /v/ is /ɱ/. If I say "non va" (doesn't go) or "non fa" (doesn't do), I don't pronounce /non va/ (nor /nom va/) but /noɱ va/. On my opinion, it's wrong to spell Italian words both with /m/ and with /n/, instead of /ɱ/, though.

In response to the suggestion that we parallel the simplification of [ɱ] to [m] with the simplification of [ŋ] to [n], I think the main difference between the two is that the former pair never makes a phonemic contrast whereas the latter does for many languages, including English. In addition, the allophonic assimilation of nasals to a velar point of articulation is not universal (languages like Russian and Polish do not do this), so failing to mark this assimilation would be IMHO more marked than putting a labial pronunciation of a labiodental nasal. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 17:29, 13 November 2015 (UTC)

Polish does have the velar assimilation, it's just that its distribution is somewhat more restricted in certain dialects. Anyway, if we don't want to use the [ɱ] symbol, [m] seems to be a good substitute. Peter238 (talk) 18:00, 13 November 2015 (UTC)

I'm adding just a note about usage differences between [ɱ] and [ŋ]. There're some languages which don't have labiodental assimilations too, such as Chinese ("Lanfang"), the most spoken language on the planet. And there're also some African languages which have [ɱ] as an indipendent sound not only confined before a labiodental consonant.

5.101.99.101 (talk) 09:25, 14 November 2015 (UTC)

Citation forms versus fast speech

I wonder if it's possible we're talking past each other here because we don't agree on what the guides are supposed to describe (or prescribe).

We don't seem to have a good article on citation forms; it's just a redirect to an article that's not very specific about it, but my outsider's understanding is that a citation form is the way someone will tell you something is pronounced, when speaking slowly and carefully to give you all the details. It's what people think they say, perhaps rather than what they actually say.

Now, the question is, what are we trying to get at? To phonologists, no doubt, what people actually say when speaking naturally is the most interesting thing. But traditionally, pronunciation guides are not for phonologists; they're for people who want to know how the word is "supposed" to be pronounced, the way a native speaker would tell you it's pronounced when speaking slowly and carefully.

My strong preference is that our pronunciation guides should be citation forms. This is the usual convention, and it's what our readers are going to expect. Do we agree on that principle? If we don't, maybe we should talk about it. Whether we do or not, 5.101.99.101, will you comment on whether it would change your answer if we were trying to give citation forms? --Trovatore (talk) 18:01, 13 November 2015 (UTC)

Trovatore, I think I didn't understand what you mean in this last point. I don't know much about Wikipedian conventions, I've written my opinion about the [ɱ] issue because you asked for a native Italian speaker's point of view. Probably not 100% of Italians would agree with my opinion, but I believe that using [ɱ] would be the most proper solution and won't confuse readers. If we have to choose between 2 wrong spellings, [m] and [n] are about 55% and 45% (example). Current rules establish, since a lot of time, that [m] is correct, if we change them I'd rather choose [ɱ] than [n], or we can just keep [m]. About "citations forms", which I've first heard today, as I said I don't know (nor understand) enough about to have an opinion. — Preceding unsigned comment added by 5.101.99.101 (talk) 09:15, 14 November 2015 (UTC)

Well, for example, it's the pronunciation that you would typically find in a dictionary (if the dictionary uses IPA). Wiktionary, for example, tends to use the /n/ in these words, in the ones I checked at least, which was not a huge sample. Of course these are phonemic rather than phonetic, but I still think it's a good indication of the sound people "think they are trying" to make. --Trovatore (talk) 18:12, 14 November 2015 (UTC)

Ok, [ɱ] maybe is too complex for the average reader and since it's just an allophone occurring before [f] and [v] it's useless to use it, I've understood you point and I may agree. I just think that [m] is (fairly) more proper than [n] if that sound must be represented with a more simple nasal symbol, in fact it's the one which has been used so far.

5.101.99.101 (talk) 17:05, 15 November 2015 (UTC)

If we make a sort of algebraic sum of our valutations, it would be:

[m] 0 -1

[ɱ] +1 0

[n] -1 +1

In this way, the preference order would be:

1st [ɱ] 2nd [n] 3rd [m]

As you said, perhaps we might go with [ɱ] (if the others agree).

5.101.99.101 (talk) 10:20, 16 November 2015 (UTC)

I frankly think [m] is unacceptable. It is severely misleading to English-speaking readers; it would induce them to produce something like my "with m" sound files, which I think are just wrong.

I would strongly prefer [n], but if that is not acceptable, then we should probably go with [ɱ̩], which a reader may not understand, but at least gives a warning that there is some subtlety involved. --Trovatore (talk) 19:28, 15 November 2015 (UTC)

I think it's pretty clear at this point that [m] is closer to [ɱ] than [n] is. The articulators are much more similar and they are also acoustically similar, both to my own ears and to the native speaker who has offered their ear. I find it quite surprising that you would think [n] is more similar to [ɱ], which is sort of like arguing that [ð] is closer to [v] than [β] is.

The only reason I can think that you might think this is that the nasal assimilation is imperfect in careful speech. In such a case, /nv/ would be realized as [n͡ɱv] with either coarticulation or a rapid shift from dental to labiodental articulation. This sort of thing can happen in English words like symphony. It also happens with /nb/, which can be realized as [n͡mb]. But this is speculation. We have sourcing that characterizes the process in Italian as obligatory.If additional sourcing can be found that gets into the phonetic particularities of this phenomenon, it might be helpful. Until then, I don't see changing [m] to [n] before labiodentals as justified. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 06:04, 16 November 2015 (UTC)

I'm with Trovatore on this one. If we display purely phonemic representation, then the phoneme here is obviously /n/, using ⟨n⟩ is a clear winner – [ɱ] is an allophone of /n/ (and of /m/, which is irrelevant), and an English speaker will likely pronounce e.g. inverno with a [ɱ] anyway (and even a [n] would not be wrong). If we opt for more phonetic, narrow transcription, then it ought to be an ⟨ɱ⟩, as obviously the most correct. But /m/, being neither here nor there, is simply a rotten solution. We should agree whether we present broader (my preference) or narrower (fine with me as well) transcriptions, but IMO here /m/ is just invented on the fly. No such user (talk) 09:04, 16 November 2015 (UTC)

There is a spectrum between broad and narrow, though. English speakers are not necessarily going to pronounce inverno with [ɱ] if we tell them it's [n], since [ɱ] is only an allophone in more casual speech.

The reasons offered thus far for including ⟨ɱ⟩ here would also apply to transcriptions of Spanish and any other language that does this.

The main reason for not adopting ⟨ɱ⟩ is that it is unnecessarily confusing to readers. It is not baseless speculation, either. Look at how much confused user Trovatore was about what [ɱ] even is. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 15:50, 16 November 2015 (UTC)

@Aeusoes1: FWIW, my untrained ear hears [ɱ] in all five Italian examples at http://forvo.com/search/inverno/, and in all five English examples at http://forvo.com/word/invert/#en and http://forvo.com/search/inferno/en/, and they're all enunciating. I agree there is a spectrum, but I don't think there's a risk of confusion with either [n] or [ɱ]. Our transcriptions are not intended to transmit all really fine points, so with an /n/ an English speaker will likely produce [ɱ] anyway, and with a /ɱ/ they might wonder and investigate, learning something new. In my opinion using a /m/ has the highest risk to produce a "WTF moment" with both a layman ("I was taught Italians pronounce as they write, is this an error?") and an IPA-literate. No such user (talk) 16:39, 16 November 2015 (UTC)

Again, English does not have the obligatory assimilation of place of articulation with its nasals like Italian or Spanish does. English speakers can and do produce [n] before labiodental sounds. According to our sources, ⟨n⟩ would be the least correct phonetically. If it isn't, then the sound being produced in this context is not [ɱ]. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 17:34, 16 November 2015 (UTC)

I suspect the sources are talking about fast speech. The way I pronounce "nv" in inverno is the same as the way I say it in "invent", and definitely different from symphony. I suppose I can't prove to you that I'm not just pronouncing inverno wrong, but I really don't think so. --Trovatore (talk) 17:45, 16 November 2015 (UTC)

We can't base our changes on this sort of speculation. Putting [n] in front of [v] would be like putting [n] in front of [k]. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 17:54, 18 November 2015 (UTC)

I don't think you've come to terms with how utterly weird this looks. As No such user says, it's a serious WTF moment to have an m there. I have been talking to Italians, and they all reject the m utterly, when I describe the controversy. The one who has contributed here has not, but I suspect he/she is atypical.

I think I'll try soliciting input from Italian Wikipedia, if I can figure out where to post it. --Trovatore (talk) 19:37, 18 November 2015 (UTC)

If you talk to native speakers with little or no training in phonology or phonetics, you're going to get a more phonemic interpretation and careful speech. I'd recommend finding sourcing that talks about the issue rather than stumbling in the dark with ad hoc original research.

That's only between [ɱ] and [n]. Between [ɱ] and [m], it's more about typographic conventions than phonetic claims. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 22:27, 18 November 2015 (UTC)

My point is that we should be using "careful speech". --Trovatore (talk) 23:00, 18 November 2015 (UTC)

Then we would eliminate [ŋ]. I don't think we should do that. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 23:19, 18 November 2015 (UTC)

I don't think careful speech would eliminate ŋ, no. If you're a foreigner and ask a native how to pronounce ancora, you may very well get [aŋ'kɔra]. In fact I think [an'kɔra] is a bit unlikely. But you will absolutely not get [im'vɛrno] for inverno, and I think even [iɱ'vɛrno] is a bit unlikely, even from speakers who would habitually pronounce it that way when not thinking about it.

This goes to the whole point I raised at the start of this subsection, really. What is the purpose of our guides? Are they really for phonologists? I don't think so. Phonology is a fine and worthwhile discipline, but it is not the subject in primo piano in any but a tiny fraction of the articles in which pronunciation guides appear.

I think they're for readers who want to know "the right way" to pronounce the words. And giving the advice to pronounce inverno with an [m] is very much "original research". The sources specific to that point are all going to advise (admittedly phonemically rather than phonetically) that you use an /n/. --Trovatore (talk) 23:31, 18 November 2015 (UTC)

Put another way — are the guides descriptive or prescriptive? Descriptively, you can make a sort of a case for [m], though I think it's a bit forced. But prescriptively, it's utterly freaking bizarre, and certainly not supported by the sources that mainly provide prescriptive pronunciation guides, such as dictionaries and teaching materials for language learners.

I prefer that the guides be prescriptive; I think that's the most useful thing for our readers. But even if we take them as descriptive, I think the extreme untowardness of [m] interpreted prescriptively should weigh against it. That's why I would see [ɱ] or [ɱ̩] what's the difference between those two, by the way? as somewhat of an improvement, not because I make a huge distinction in the actual sound myself, but because it tends to defuse that this-is-just-plain-wrong reaction you're going to get with [m]. --Trovatore (talk) 01:01, 19 November 2015 (UTC)

The difference is that [ɱ̩] is syllabic (as in one pronunciation of German offen), whereas [ɱ] is not. Only the latter sound exists in Italian. Peter238 (talk) 01:28, 19 November 2015 (UTC)

Ah, thanks. --Trovatore (talk) 01:36, 19 November 2015 (UTC)

These guides should be more descriptive than prescriptive. Though if you mean "prescriptive" in the sense that we give deference to dictionaries, then I suppose that is also true.

As I see it, the difference between [ɱ] and [m] is level of detail. Using the former is kind of like using dental diacritics on [t] and [d]. It's more accurate, but for a pronunciation guide (particularly to a lay readership) it is likely to be more confusing than helpful. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 15:31, 21 November 2015 (UTC)

I mean "prescriptive" in the sense that readers are (I propose) more interested in what is the right way to pronounce this? than they are in if you record people on the street and break down the sounds on an oscilloscope, what IPA symbols does that correspond to?

So if I'm correct about that, which I think I am, then we absolutely should be reporting "careful speech", the way a native speaker will tell you something is pronounced when he/she is thinking about it and enunciating.

When you're trying to learn a language, you need to learn what people think they are trying to say, not what they actually say in relaxed speech. The latter will come naturally, through conversation and practice, but if you reproduce it when enunciating, you sound really bad. --Trovatore (talk) 20:12, 21 November 2015 (UTC)

We don't want to put pronunciations that people think they are trying to say. That's how you get weird or unnatural speech. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 22:44, 21 November 2015 (UTC)

Let's take an example.

I have lived much of my life near Santa Monica, California. I think I have a fairly "native" pronunciation. In relaxed speech, I pronounce the first word, roughly, [sæə], with two syllables, and the [æ] nasalized maybe you can help me find the symbol for the nasalized [æ]; I couldn't find it quickly. There is no trace of a [t], and even the [n] turns into a nasalization.

Now, do you think it would be responsible, in a pronunciation guide for foreigners, to transcribe the first word of "Santa Monica" as [sæə] with whatever you do to indicate the nasalization?

In my view, that's roughly analogous to the current use of [m] in these words. --Trovatore (talk) 01:32, 22 November 2015 (UTC)

Yes, I get your view. But you have so far declined the opportunity to provide sourcing to give your view any credibility. Without additional sources, we can't really go any further. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 09:21, 22 November 2015 (UTC)

There's plenty of sources for the /n/, if that's what you think is missing. Dictionaries mostly, which are not ideal sources for most things, but they are actually most likely the only available sources for IPA versions of specific words.

There is on the other hand no source at all, that I've seen at least, for the [m]. --Trovatore (talk) 18:39, 22 November 2015 (UTC)

I'm talking about your claims regarding the distinction between citation forms and normal speech. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 18:45, 22 November 2015 (UTC)

That's not something I propose to put in Wikipedia, per se. It's an explanation as to why the matter is important, not something to get formal about. If you want to get onto the formal, sourcing level, then I propose to use the pronunciations that can be found in monolingual Italian dictionaries from Italy. --Trovatore (talk) 18:51, 22 November 2015 (UTC)

But you are making claims of fact. Since your stance seems to depend on these facts being true, I'd like to see some substantiation for these fact claims. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 18:57, 22 November 2015 (UTC)

Well, here is where you need to talk to Italians. In real life, not online. Describe the question neutrally and see what they say. I challenge you.

However, it's not really true that there's nothing to discuss without establishing my factual claims. There is still the point that the [m] (albeit not the [ɱ]) seems to be a pure invention of this page. I would be a little happier with the [ɱ], because it makes it clear that this is some technical phonological thing, and hopefully would be less likely to make readers start thinking they should really try to use an [m] themselves when pronouncing these words.

Also, there's the general question of whether the guides should report careful or relaxed speech, on which you have taken a position, but not substantiated it. --Trovatore (talk) 19:05, 22 November 2015 (UTC)

Discussions with Italians would only provide anecdotal information.

I don't know what you mean by "substantiate" my position. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 19:52, 22 November 2015 (UTC)

It would provide anecdotal information that would help you understand why this is important and why the current situation is bad. It would not, of course, provide any information that could be asserted in Wikipedia per se, but that isn't the point here.

I'm talking about substantiating claims like: We don't want to put pronunciations that people think they are trying to say. That's how you get weird or unnatural speech. and These guides should be more descriptive than prescriptive. Why should they be more descriptive than prescriptive? Is that just your preference? My preference is very strongly the other way, and I've explained why. --Trovatore (talk) 20:32, 22 November 2015 (UTC)

Again, I understand your perspective. I just don't buy it without actual sources corroborating the factual claims you make. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 02:15, 23 November 2015 (UTC)

Please address the points I make in my most recent comments. --Trovatore (talk) 02:34, 23 November 2015 (UTC)

This is pointless without sources. I'm not going to repeat myself anymore. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 03:03, 23 November 2015 (UTC)

How about you give some sources for the things I challenged? --Trovatore (talk) 04:10, 23 November 2015 (UTC)

And by the way, right, I hope you won't repeat yourself anymore. That's what you did in your contribution of 02:15. What you did not do was respond to my points. I am waiting for that. --Trovatore (talk) 04:11, 23 November 2015 (UTC)

We might be talking over each other a bit. It seems like when you say "prescriptive" and "reference form" you are actually referring to phonemic representation. Your Santa Monica example thus doesn't reflect a different underlying phonemic representation as much as a different set of generative rules going from the underlying form to the surface form compared to more careful or standard pronunciations. [ˈsæ̃ə] (or what I suspect is [ˈsæ̃ɾ̃ə]) is still phonemically /ˈsæntə/. Extrapolating on this example, asking someone to pronounce Santa Monica carefully would produce something like [ˈsæ̃nˌtʰɐ], a pronunciation that is noticeably inaccurate. That people speak differently when they are paying attention to their pronunciation is such a commonplace in linguistics that overt statements to the matter appear only in introductory textbooks. Do you really need me to find such a statement or do you now recall this commonplace from your own linguistics instruction?

What makes this more complicated is that broad/phonemic and narrow/phonetic do not form a binary. There is a spectrum between broad and narrow. In general, we try to avoid transcriptions that are so narrow that diacritics are required. We also avoid transcriptions that are so broad that they require knowledge of the language's phonology to get right (see the WP:MOS/P for the general prohibition against broad phonemic transcriptions of foreign words).

Based on the sourced description of nasal assimilation in Italian, [nf] is just plain wrong for a narrow transcription, which is what we are striving for. Transcribing [n] before labiodental sounds would be an example of overly broad transcriptions. As a matter of consistency, it would also not make sense to use a broad representation of [n] before labiodentals but not before labials and velars, which is what you have been advocating.

The matter of citation form versus fast speech is a red herring, I think. Everyone here agrees that [ɱf] is correct phonetically. The disagreement is whether ⟨mf⟩ is a palatable substitute. I think it is, and believe [ɱ] to reflect a narrowness akin to the aforementioned diacritics. I don't dispute that ⟨mf⟩ could lead to a "literal interpretation" and make way for pronunciations with [mf]. However your assessment of [mf] as "more wrong" than [nf] goes against my own understanding of the acoustic properties of all three nasals, phonology and phonetics in general, and the phonology and phonetics of Italian in particular. This is where sources could help you convince me. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 06:59, 23 November 2015 (UTC)

I agree with Aeusoes that [m] sounds closer to [ɱ] than [n] does. If we want not to use [ɱ], [m] is a more reasonable substitute than [n]. Peter238 (talk) 05:38, 24 November 2015 (UTC)

There are a lot of issues here, and I will need to take a bit of time to sort them into a careful response. --Trovatore (talk) 06:24, 24 November 2015 (UTC)

Alright, let's make a start at least.

It seems to me the issues break down into the following (somewhat overlapping and non-exclusive) categories:

1. Is it factually correct that Italians use [ɱ] in careful speech?
2. Should we use phonemic or phonetic transcriptions, or something along the continuum?
3. Should we use careful speech or relaxed speech?
4. Should we prefer "descriptive" sources such as phonetics journal articles, or "prescriptive" ones such as pronouncing dictionaries and language-teaching materials?
5. And finally, assuming arguendo that [ɱ] is the desired recommendation, is it justified to substitute [m]?

As to point 1, I believe the answer is "no"; in careful speech, when enunciating and trying to explain to you how to pronounce it, they usually do not use [ɱ] in these words. I wasn't sure about that point for a while because I wasn't sure what [ɱ] was, exactly, but now that I think I'm reasonable clear on that point, I believe the answer is "no".

Now, if we were to assert that in Wikipedia, of course it would need a source. However it does not need a source just to bring it up in discussion as a motivation for other things. For anything that ultimately depends on its truth, it is reasonable to ask how to verify or falsify it (and I think I've given a clear suggestion on that point), but that is different from demanding the same sort of sourcing that would appear in article space.

In any case, even if I happened to be wrong about point 1, there would still be plenty to discuss, so trying to shut down discussion until I can find specific sources for point 1 is wrong in at least two ways.

Point 2 overlaps quite a bit with point 4, so I defer it.

Point 3 may be the real bone of contention. Aeusoes made a bit of a point, in the last intervention, about whether I had to be convinced that people speak differently when they are speaking carefully. Of course I do not have to be convinced of that. That's part of my point. The question here is, which is better for us to use?

In my opinion, careful speech is to be preferred. Among my reasons are (a) it is traditional, and (b) it is more useful for learners. I can go into more detail about point (b) at some other time; there's a fair bit I want to say about it and this is already long.

On point 4, I clearly prefer the dictionaries and teaching materials. Again, I believe these are more useful for people who want to know how to pronounce the word, themselves, without sounding "bad". They are less useful for phonologists, I suppose, and I am not at all anti-expert — but I would argue that, when one must choose, things that cater specifically to expert phonologists should be in phonology articles, not articles on a style of painting.

Point 5 I listed for completeness; I don't want to discuss it yet, as this is already way too long. --Trovatore (talk) 07:29, 24 November 2015 (UTC)

Perhaps one way to think about where we should fall along in the careful/relaxed continuum is appearances in the media. In a news broadcast or government speech, the speech isn't careful, but it isn't fully relaxed either. Speakers are devoting enough attention to pronunciation that they are representing what might be considered a "proper" or "standard" accent. As I said above, asking people to produce words in isolation can give forth incorrect pronunciations.

In other words, your methodology for getting careful speech will likely produce something different than prestige pronunciations found in media broadcasts. I gave an example with Santa Monica but I believe the same thing happens with nasal assimilation in Italian and Spanish.

We have every indication that this nasal assimilation is a feature of even prestige varieties of Italian. We already have sources that clearly identify the assimilation as a feature of Italian (and I can provide even more that say the same) and speculating that it is not a part of standard pronunciations (or that labiodental assimilation is not while labial and velar is) seems made up whole cloth. You keep acting like I'm appealing to our NOR policy when I ask for a citation, but I'm not. I am not motivated to accept your citation vs. relaxed argument without evidence. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 16:32, 24 November 2015 (UTC)

It's not quite clear what you're asking for a citation for in the last sentence. If you mean for how (say) inverno is pronounced in careful speech, then there are other sorts of evidence than citations. Why not talk to Italians in real life, as I suggested? I assume you want to know the truth, and not just "win", right?

But that is only point 1. We still have lots to talk about, whether you buy my claims on point 1 or not. For example, why should "media broadcasts" be the standard? That's still descriptive. In my view, descriptive phonological claims, even of prestige varieties, are not as useful to our readers as prescriptive pronunciations from pronouncing dictionaries. They are also less traditional as "pronunciation guides", and less sourceable (you can source these features of the language, but in general you can't source the narrow transcription of an individual word). --Trovatore (talk) 18:39, 24 November 2015 (UTC)

I still think this whole careful/relaxed avenue is DOA. There's no way that we wouldn't represent nasal assimilation in Italian (or Spanish, for that matter). Why don't you expand on your point 5 on why [m] is an unpalatable alternative to [ɱ]? — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 21:37, 29 November 2015 (UTC)

Why wouldn't we? How about you justify your claims that we "don't want" careful speech, or that we "want" a narrow transcription? --Trovatore (talk) 00:54, 30 November 2015 (UTC)

I already have. Pardon me for being curt, but I'm not interested in repeating my arguments because you chose to ignore them the first time.

You have referred to rationales for your point 5. Let's hear them. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 01:17, 30 November 2015 (UTC)

Not yet. I still want to know why we shouldn't use a broad transcription and/or careful speech. All that you've said is that careful speech is "unnatural". I don't agree that that means we shouldn't use it. Why shouldn't we use it? --Trovatore (talk) 03:29, 30 November 2015 (UTC)

You can reread what I've written to find what you're looking for. It looks like going on any further will be a waste of everyone's time. Unless someone else has something to contribute, this conversation is at an end. Regards. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 05:50, 30 November 2015 (UTC)

Syntactic gemination

We should not show the syntactic gemination (SG) symbol (*) in our transcriptions of Italian, except in articles that are specifically about Italian phonology, for three reasons: 1) cases are predictable except for a small closed set of function words; 2) the actual set of words varies by variety; 3) it is not a universal feature of standard spoken Italian. In detail (from Absalom-Hajek 2006):

1. Syntactic gemination is fully predictable — they occur after stressed final syllables, including stressed monosyllables (phonological SG) and after a small, closed set of unstressed monosyllables and some penultimate-stressed polysyllables. This closed set includes only function words and not nouns, which are the typical words for which we give pronunciations.
2. The actual set of exceptional words varies by variety of Italian, even among those varieties which show SG. But in any case, it does not include nouns, if I'm not mistaken.
3. Many varieties of standard Italian spoken outside central Italy do not show SG at all.

It is thus pedantic and unhelpful to show the pronunciation of the river Po as [pɔ*]; for those varieties which show SG, it will automatically trigger SG since it has final stress and is not a function word. --Macrakis (talk) 22:55, 8 December 2015 (UTC)

Thank you for showing common sense and starting this discussion. I agree that there's no need to transcribe SG when the last syllable is stressed, but I'm not sure about the rest. Peter238 (talk) 23:00, 8 December 2015 (UTC)

The closed set of words triggering SG that are not stressed on the final syllable are all function words, not nouns (fra 'among', ma 'but', come 'how', dove 'where', ...). We don't normally have Wikipedia articles about function words (English or Italian or anything else), so the situation is unlikely to come up. --Macrakis (talk) 23:47, 8 December 2015 (UTC)

4. Also, the only time the "*" annotation is useful is if the word is being composed with another word, e.g., "il Po superiore", which is presumably [il'possuperi'ore] rather than [il'po.superi'ore]. But if you know enough Italian to compose words like this, you don't need the "*". --Macrakis (talk) 02:53, 9 December 2015 (UTC)

Being predictable doesn't necessarily make something not worthy of our transcriptions. However, if there is dialectal variation in both the distribution and occurrence of this phenomenon, then that is a bit problematic. If we can't pick a particular prestige variety that does feature SG, then it seems like including it is not justified.

However, if we can identify one then incorporating SG in our transcriptions is motivated. I don't particularly like the asterisk to mark it, though. Why couldn't we just show the actual gemination if it occurs in our transcription? It would be like somehow marking French transcriptions with the final consonant that is normally elided except in cases of liaison. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 05:25, 9 December 2015 (UTC)

I do not like the asterisk symbol either, and I am Italian. It is a conventional symbol used in a few Italian vocabularies, others use the plus symbol, most of them do not use any symbol at all, but more important the International Phonetic Alphabet has never been using a symbol for syntactic gemination. If we want to speak about a sequence of words where the syntactin gemination, in standard Italian, is present ("perché lui" [perkelˈlui]) no problem, but adding a symbol every time there is an accented word or a tonic monosyllable is not useful and may cause misunderstandings, like in this case. — Preceding unsigned comment added by 151.20.126.192 (talk) 10:00, 9 December 2015 (UTC)

Ƶ§œš¹, it seems to me that predictability is a pretty important criterion. Otherwise we should be transcribing Alaska as /əˈlæskə(r)/ since, after all, prestige varieties of English (including the prestige variety, Received Pronunciation) regularly show intrusive R /əˈlæskər iz big/. For that matter, should we also include a special symbol //əˈlæskə*/ to show that a following "is" can become "'s" ("Alaska's big.")? This is fully predictable from the fact that the word ends in a vowel, but you say that isn't enough. There is also the geographic question. Someone in Florence might say [oltreˈpɔ ppaˈveːze], but locals would say [oltreˈpɔ paˈveːze]. It doesn't make sense to track that sort of thing since it is a regular phonological phenomenon. --Macrakis (talk) 15:09, 9 December 2015 (UTC)

In a phonemic transcription like that, yes predictability would mean we shouldn't include something. But Italian transcriptions are not (and should not be) phonemic. In a phonetic transcription, predictability is a less relevant criterion for exclusion. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 15:28, 9 December 2015 (UTC)

In looking back over the discussion, I see that I wasn't specific enough about the case in question. I didn't intend to discuss the transcription of syntactic gemination in the interior of phrases. That is, I see no problem with [kafˈfɛ lˈlatte]. I was concerned with cases like the Po (river) article, where the name of the river is transcribed as [pɔ*]. This is parallel to transcribing Alaska (and every other noun ending in 'a') as [əˈlæskə(r)] because in sentences like "Alaska is big", RP speakers say [əˈlæskər ɪz 'bɪg], which I hope no one is proposing. --Macrakis (talk) 17:29, 9 December 2015 (UTC)

I can't say I disagree with you on any point. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 02:38, 10 December 2015 (UTC)

I see that after a week no one else has joined the discussion to pronounce against the removal of the asterisk symbol. Might we say that we concur about that, then? If it is so, should and could we remove the symbol from the article about Po and in other similar cases? Feel logically free to object and to say your opinion if you do not agree. — Preceding unsigned comment added by 151.20.92.74 (talk) 22:59, 15 December 2015 (UTC)

Since the only proponent (as far as I can tell), User:IvanScrooge98, seems to be unwilling to discuss here despite notification, yes, I think we can consider that there's a consensus to remove * from transcriptions of Italian. Gemination in the interior of phrases should be notated by doubling the consonant as in [kafˈfɛ lˈlatte]. --Macrakis (talk) 16:20, 16 December 2015 (UTC)

Ditto. The consensus is obvious. Peter238 (talk) 19:54, 16 December 2015 (UTC)

Diphthongs

Hi, is it possible to add a note with all or most vowel combinations ending in [i̯] and [u̯]? — Jɑuмe ^(dis-me) 03:41, 9 July 2016 (UTC)

Move discussion in progress

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