Appendix talk:Portuguese pronunciation

Canonical transcription

@Ungoliant MMDCCLXIV olá! nice to see you editing again, estava com saudades. I'm just learning IPA basics and had a look how our entries are transcribed. Just wondering, what makes a transcription "canonical" (first column)? What are the standard works for Portuguese? – Jberkel 00:10, 15 February 2019 (UTC)

Hey @Jberkel. These transcription standards are canonical only from the perspective of Wiktionary. There are no established phonological transcription systems for Portuguese in the same way that something like Upton’s Scheme exists for English. In fact, not even the use of IPA is as well established, since the alphabet provides the tools to transcribe underlying segments with little ambiguity for a variety of dialects.

When I wrote this page, I wanted to codify existing practices so that transcriptions tagged with {{a|Brazil}} or {{a|Portugal}} would be consistent. The majority of correct European transcriptions at the time had been added by Liuscomaes (talk • contribs), and most Brazilian transcriptions had been added by me.

Both schemes strongly “hug the phonetic ground”. I remember some people complaining about Liuscomaes’ use of ɣ and ð for what are usually considered allophones of /ɡ/ and /d/. My own scheme is not without similar issues; for example, [d] and [d͡ʒ] are often considered allophones, as are [a] and [ɐ]. However, our “every word in every language” policy means that we must include special cases like DJ, azerbaidjano and caminha that can be ignored by people writing short articles or simplified abstract systems.

If you’re interested in reading more about the phonology of Brazilian Portuguese, the names that pop up most often in citations are Leda Bisol and Milton Azevedo. I do not know who the foremost researchers on European Portuguese phonology are.

— Ungoliant ^(falai) 03:02, 15 February 2019 (UTC)

Hi! I am a native speaker of Brazilian Portuguese and have been editing articles of words in the language, so I noticed quirks like this. [d] and [d͡ʒ] are considered allophones, yes, but of /d/ to my understanding. And /d/ is a phoneme separate to /d͡ʒ/, which came from those loanwords that have it. --Gmestanley (talk) 02:46, 1 April 2022 (UTC)

Accent in Portugal's Norte

Like in Galician, in the Portuguese region of Norte, it's customary, to pronounce /v/ as /b/. Should those case be dealt in any way? - Sarilho1 (talk) 22:24, 4 August 2019 (UTC)

The letter R

Should we really be citing every single way of pronouncing the R in every single word? I think it could be simply noted somewhere in this appendix page, and left out of the actual pages. 189.76.49.102 20:12, 9 July 2020 (UTC)

• I mean, we don't note every type of pronunciation in the English IPA transcriptions either, so why should we do that here? 2804:22C:F650:B900:39A3:8B1D:5229:4A98 06:34, 26 August 2020 (UTC)

Intrusive nasal labio-velar/palatal approximants' criteria

There are actually set situations were we add nasal /j/ and /w/ into words. I'll be explaining them informally. So, nasal /e/ and /o/ happen when they come before an N or an M. In that case, the nasals will be added. However, if AFTER the N or M, there is a vowel, then that doesn't happen. So on words like ônibus, Emily, or Jenifer, they don't get added, but they do get added on words like ontem, concorrência, and expoente. 2804:22C:F650:B900:FDEA:65B4:EA00:7609 18:55, 26 August 2020 (UTC)

• By the way, this happens every time said words appear, and some words become unrecognizable if the nasal approximants aren't pronounced.

Some sugestions/corrections

Let me start by saying that I'm not that much into programming, so the current pt-IPA module or template or whatever is called, that is currently being used in wiktionary, is geting quite a bit difficult for me to use and/or modify. I usually do copy-paste, and by tinkering here and there I can have some idea about it's functioning. My first suggestion regards the phononemic notation, which I think is currently being used in a more phonetic way instead of more phonemic; many allophonic pronunciations are indicated in the phonemic notation, when they should be used only in the phonemic one. For instance, the diphthongs /ou̯/ and /ei̯/ should be always written that way in the phonemic transcription, their eventual monophthongisation should be left for the phonetic transcription; one advantage of that is that we could do away with all the annoying parenteses. One example, the word janeiro is currently like this:

(Brazil) IPA(key): /ʒaˈne(j).ɾu/ [ʒaˈne(ɪ̯).ɾu]

(Southern Brazil) IPA(key): /ʒaˈne(j).ɾo/ [ʒaˈne(ɪ̯).ɾo]

(Portugal) IPA(key): /ʒɐˈnɐj.ɾu/

(Northern Portugal) IPA(key): /ʒɐˈnej.ɾu/

(Central Portugal) IPA(key): /ʒɐˈnej.ɾu/

(Southern Portugal) IPA(key): /ʒɐˈne.ɾu/

A much more elegant solution woul be something like this:

(Brazil) IPA(key): /ʒaˈnei̯.ɾu/ [ʒaˈne(ɪ̯).ɾu]

(Southern Brazil) IPA(key): [ʒaˈne(ɪ̯).ɾo]

(Portugal) IPA(key): /ʒɐˈnei̯.ɾu/

(Northern Portugal) IPA(key): [ʒɐˈnei̯.ɾu]

(Central Portugal) IPA(key): [ʒɐˈnɐi̯.ɾu]

(Southern Portugal) IPA(key): [ʒɐˈne.ɾu]

Or, if you wanna be more accurate with the diphthongs they could be rendered with [ɪ̯] and [ʊ̯] in the phonetic transcription. I only kept the parenthesis in the brazilian version because I don't have enough information on the matter; it is my general belief that most speakers monothongize them, but I also hear them occasionaly being pronounced as diphthongs; I don't know if it's dialectal or an artificial pronunciation.

You may also have noticed that I didn't use the symbols /w/ and /j/ for the diphthongs - that's a pet peeve of mine, I dont think those symbols should ever be used in the description of the Portuguese language because they represent consonants, not vowels, as I already explained here.

Other things that shouldn't be included in the phonemic transcription are the vocalization of the /l/ and the palatisation of /d/ before /i/ - these are all allophones and should be left for the phonetic trancription.

Now for the corrections. The southern pronunciation has the following errors:

• The diphthong [ei̯] is always turned into a monothong; that's not the case. It remains as [ei̯] in the following conditions:

Before other vowels: areia, meio

In the last syllable of a word: sei, seis, répteis, alcancei

Usually before t: leite, feito - this last one may vary, I should say that I'm from Alentejo Litoral and people from the interior migh pronounce these words with monophthongs; even I have an exception tho that rule, only with the word deitar, which I pronounce as [deˈtaɾ] (maybe because the "ei" is untressed in many conjugations and then it spread to the stressed forms as well?)

• /b/ /d/ /g/ are transcribed as [β] [ð] [ɣ] between vowels - those sounds are never pronounced as fricatives in the South, always [b] [d] [ɡ]

• This one I don't know if its widespread in the south, but in here we still pronounce the diphtong /ɛi̯/ in words like idéia, Andréia, diarréia, boléia (and all the words with the greek suffix -éia); in fact I used to say "verborrêia" instead of "verborréia", untill I realised it's the same suffix as in "diarréia". Thanks a lot, Portuguese spelling!

• The diphthong /ai/ is turned into [a] before "x"; caixa /ˈkai̯ʃɐ/ > [ˈkaʃɐ]. this also happens in Brasil and in many people in the Centre as well, including Lisbon. Note however, that it still behaves phonemically as a diphthong, as it does'nt suffer vowel reduction when unstressed: encaixar /ẽkai̯ˈʃaɾ/ > [ẽkaˈʃaɾ], and not *[ẽkɐˈʃaɾ]. The same happens with "ou" and '"ei" when monothphongized (is this a word?) - they are never reduced when unstressed.

• Of course, we still distinguish "ãe" from "em" /ẽĩ̯/

In the special case of the word têm, in careful or slow speach it would be /ˈtẽĩ̯ẽĩ̯/, wich in normal speach becomes [ˈtẽẽĩ̯] wich can further become [ˈtẽẽ] or [ˈtẽː] when unstressed.

• /e/ remains unchanged befor palatals - coêlho, sênha, bochêcha

Regarding other dialects: the northern dialect has "ch" as being still pronounced as an affricate, whis is very misleading; people from the northwest (Porto, for example) certainly don't pronounce it that way; that pronunciation - if it still even exists - must be limmited to some areas in the northeast; the module makes it seem that it's widespread in the whole north, which is not the case. Another feature of the northern pronunciation that's made to seem widespread is the merger of "v" with "b"; now, while certainly that's their natural pronunciation, it is a heavily stygmatized feature - it shouldn't be, everyone should speak the way they learned how to, but, the fact is that many northern speakers do end up making the distinction. I think the solution to this would be to indicate both variants, for instance Vila /ˈvilɐ/ > [ˈviɫɐ] ~ [ˈbiɫɐ].

Oh, one final note about southern pronunciation: /l/ is generally velarised in all positions, as is general the case throughout mainland Portugal, although, I can't say for certain that there aren't speakers who don't velarised it before vowels, Alentejo is a very big place!

Turns out that wasn't the final note; regarding the rolled r, it is still quite common, and it always makes my blood boil whenever I read anybody (it's everybody!) claim that it's a "rural feature"; tell me this, would you be able to tell if a speaker is rural or not by the way they pronounce the R's? Of course not! The correct analysis regarding the use of the guttural or the rolled r is not of rurality; is generational: younger speakers tend to use the guttural one, which hapens to be my case, and I'm from a rural village; you wouldn't be able to tell that i'm a labrego by the way i pronounce my Rs, you'd be able to tell that I am one by all the other features I described above!

Some few more things (will this ever end?!): one more error in the southern pronunciation is that it adds an [i] to the end of words that end with r or l. Again, this might be a feature of some southern speakers, but it's not widespread throughout the whole south. The final note on southern pronunciation is that many people pronounce the pronoun 'lhe' as 'le', as in "eu dei-le o livro"; but again I don't know how widespread is this.

That's all for now, I hope anyone will read all this! Sérgio R R Santos (talk) 16:37, 22 August 2024 (UTC)