Help talk:IPA/Portuguese and Galician/Archive 1

Affricates and R
Because so many articles specify them, I think we should have separate BP-specific lines for. It might take people a long time to find them in the footnotes. kwami (talk) 18:46, 13 June 2009 (UTC)
 * Shouldn't we be imposing a standard in regards to ? I'd suggest that we use  in both BP and EP and consider the other transcriptions deprecated. — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  19:54, 13 June 2009 (UTC)
 * Maybe, but for now lots of articles use [x], and IMO the point of this key is to explain the IPA symbols used in the articles. kwami (talk) 20:59, 13 June 2009 (UTC)
 * All right. Since you've been going through and converting IPA2 and IPA-all to language-specific templates (wonderful work, by the way), it shouldn't be too hard for us to convert the x's and h's to  afterwards.  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  21:18, 13 June 2009 (UTC)
 * I agree it would be better to standardize to . At the moment articles are using a variety of transcriptions, not just with regard to the trilled r. +Angr 13:11, 14 June 2009 (UTC)

Does Brazilian really differ from European in pronouncing a syllable-final rhotic as a trill? If so, we should find a source for it and have our Portuguese phonology article reflect that. — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi] 17:05, 14 June 2009 (UTC)
 * According to Stephen Parkinson, "Portuguese", in Bernard Comrie (ed.), The World's Major Languages (OUP, 1987), p. 267: "In many Brazilian Portuguese accents also fills syllable-final positions, invariably filled by /r/ [by which he means ] in European Portuguese". However, you have to keep in mind that it isn't much of a trill in BP; it's usually more of a [h] or weakly articulated [χ]. I can't vouch for the first "r" in "amor eterno" (maybe it gets resyllabified across a word boundary, I don't know), but the second word is definitely  for many Brazilians. That's certainly how I was taught to pronounce Brazilian Portuguese at university. If you go here and listen to her say roncar and tratar you'll hear the [χ] really clearly. For word-internal examples, listen to particularidade and partícula on this page and to reverberar on this page. The same speaker has many word-final examples too. +Angr 19:53, 14 June 2009 (UTC)

I don't quite see what you're talking about, but as a portuguese used to hearing brazilians, _amor eterno_ has first a simple flap and then a fricative. If _amor_ stood by itself it would have the fricative in most dialects (complete loss of -r is only usual in -ar, -er). _Tratar_ has first a flap and then a fricative or nothing. 85.240.106.120 (talk) 23:53, 22 July 2009 (UTC)

I've gone through and standardized the articles to, though I more-or-less remained neutral as to whether it appeared in the coda or not. — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi] 22:47, 6 July 2009 (UTC)

It varies too much among regional variants to be easily summarised.

Final /r/ is a flap when following word starts by a vowel, in all cases. When the next word starts by a consonant, or when there is no next word, practically anything can be expected: /h/, /χ/, /x/, /ʁ/, /r/, /ɹ/, no sound, no sound but affects the previous vowel (/amo:/ for "amor"). The boundaries between these are imprecise, I don't think there are reliable sources for them (there is no Brazilian phonological atlas), and, evidently, Brazilians of one region tend to represent utterances of other regions in a caricatural way). Impressionistically, I would say /r/ is typical of Rio Grande do Sul (but it could be of all Southern and Center-Western Brazil as well), /ɹ/ of São Paulo hinterland (perhaps adjacent regions of Minas Gerais). To me, people from Rio de Janeiro pronounce it, impredictably, as /h/, /χ/, or /x/; they also use it as in /amo:/ for "amor", but this tends, I believe, to be typical of colloquial or popular registers.

Post-vocallic /r/ is consistent with final /r/. Where final /r/ is pronounced as /χ/, post-vocalic /r/ will be pronounced as /χ/. This doesn't hold in the variants in which final /r/ is not pronounced, in which case it tends to be one of /h/, /χ/, /x/, or /ʁ/; never /r/ or /ɹ/.

Initial /r/ is the same as medial /rr/; they tend to be consistent with final /r/ except for the variants in which the latter is not pronounced, in which case they will, in any case, never sound as /r/ or /ɹ/, and the variants where the latter are pronounced /r/ or /ɹ/, in which they are, I think, usually /ʁ/.

Medial /r/ is always /r/, except for regions in which post-vocalic /r/ is /ɹ/, in which case it will also be /ɹ/.

So, for instance, "amor eterno, amor errado, amor perdido, querido amor", could sound (with IPA only for the rhotics, other phonemes as written) like:


 * 1) /amor etehno, amor ehado, amor pehdido, querido amoh/
 * 2) /amor eteχno, amor eχado, amor peχdido, querido amoχ/
 * 3) /amor etexno, amor exado, amor pexdido, querido amox/
 * 4) /amor eteʁno, amor eʁado, amor peʁdido, querido amoʁ/
 * 5) /amor eterno, amor eʁado, amor perdido, querido amor/
 * 6) /amoɹ eteɹno, amoɹ eʁado, amoɹ peɹdido, queɹido amoɹ/
 * 7) /amor etehno, amor ehado, amo pehdido, querido amo/
 * 8) /amor etehno, amor ehado, amo: pehdido, querido amo:/

(in the two final ones, /h/ could be replaced by /χ/, /x/, and perhaps /ʁ/, but certainly not by /r/ or /ɹ/.)

My own native variant is number 5. above; I find variants 1-4, 7 and 8 difficult to distinguish from each others. 200.198.196.129 (talk) 16:04, 14 August 2009 (UTC)
 * Wow, that is way more complicated than we need to involve ourselves with for a pronunciation key. Thanks for the info.  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  19:23, 14 August 2009 (UTC)

You can simplify that by saying that there are two main variants: one in which post-vocallic "r" matches medial "r" (5. and 6. above), and another in which post vocallic "r" matches medial "rr" (1.-4., 7., and 8. above). 201.86.167.173 (talk) 14:09, 15 August 2009 (UTC)

Isn't it better to include a note for the alveolar flap in words at the syllable-final as it has been done with the uvular fricative? Because many places of Brazil the syllable-final is not guttural, in Rio Grande do Sul state and São Paulo city for example it is a alveolar flap, and many cities in the interior of São Paulo state, south of Minas Gerais state, and someother places it generally is a alveolar/retroflex approximant, which is far more similar to alveolar flap than voiced uvular fricative, and also there are some non-rhotic accent.Luizdl (talk) 03:48, 5 October 2009 (UTC)

Diphthongs vs. hiatus

 * Diphthongs in Portuguese are constructed by vowel+semivowel or semivowel+vowel. So, the best way to transcrypt them to IPA is using the symbols "w" and "j" for semivowes. So, "mau" (and "mal" in BP) is better phoneticly symbolized by "maw".


 * Note that two vowels together are NOT always a diphthong. "Viúva", for example, has three syllables: vi-ú-va, which is a hiatus.


 * Can we stick to the IPA stress mark please, esp. on an IPA key? (The apostrophe looks like a glottal stop, or, if someone puts it before the vowel rather than the syllable, as sometimes happens, like an ejective consonant.) kwami (talk) 14:34, 24 June 2009 (UTC)


 * The fact is that in some Portuguese words the vowel is itself a single syllable AND the stressed syllable. Other examples: Jaú, Itajaí, heroína, Cabreúva sylx100


 * If we're transcribing these with /w/ and /j/ I see no point in listing them as diphthongs/triphthongs at all. They're just sequences of semivowels and vowels in various orders. +Angr 10:21, 25 June 2009 (UTC)


 * Agreed. Listing /w/ and /j/ under 'consonants' is more useful to English speakers than a page of diphthongs. kwami (talk) 22:00, 25 June 2009 (UTC)

Pan-dialectal representation
I propose that we have three standards for Portuguese.
 * 1) A pan-dialectal transcription that, like our English and Spanish keys, transcribes in such a way that readers can figure out the dialectal pronunciations from the key by knowing the differences. I see this as being an imperfect transcription that won't be able to account for some differences (see below).
 * 2) a European Portuguese transcription for Portugal-related articles that exhibit features that the pan-dialectal representation can't account for
 * 3) a Brazilian Portuguese transcription for Brazil-related articles that exhibit features that the pan-dialectal representation can't account for.

Some things that I see us having trouble with the pan-dialectal transcription include:
 * The affrication of /d t/ before front vowels in BP


 * I wouldn't worry about that one. It isn't universal, and in fact is quasi allophonic. Speakers easily switch according to register. —Preceding unsigned comment added by 85.240.106.120 (talk • contribs)


 * Coda-r. As Angr has helped us understand, In EP it is a flap and in BP it is a trill.  It's fairly minor, though, and even Mateus & d'Andrade (2000) gloss over it in transcriptions soon after they describe it.


 * As I mention above, it's a flap in BP if a vowel follows - though usually that will only happen with -or, as the others will either drop the -r or use a fricative plus a pause. —Preceding unsigned comment added by 85.240.106.120 (talk • contribs)


 * Vowels
 * According to the comparisons in Mateus & d'Andrade on pp 17-18, there is a vowel that, before "non-nasal" palatals, is in EP but  in BP (e.g. telha, lei).


 * Pronunciation of e before palatals varies a lot in Portugal. The central dialects have enh with [ɐ], elh with [e] and ej with [ɐj]. The standard used to follow that but has of late gone with [ɐ] for elh (but not for ej). —Preceding unsigned comment added by 85.240.106.120 (talk • contribs)


 * In BP, are allowed in pre-stressed position while in EP they become *


 * Rather, some of them become. EP keeps many pre-stress open vowels (padaria, corar, pregador, etc). It's in Brazil that some of those coelesce with their mid-closed conterparts. —Preceding unsigned comment added by 85.240.106.120 (talk • contribs)


 * Similarly, are allowed in post-stressed position in BP, while they become * in EP.
 * Final unstressed  is in BP but * in EP.
 * * EP unstressed can be deleted in "colloquial speech", but apparently not in BP.


 * Because in BP it is a 'full' i. In EP, even barred-u (a common allophone of u, as opposed to barred-i which is a phoneme in its own right) can be deleted. —Preceding unsigned comment added by 85.240.106.120 (talk • contribs)


 * is velarized in EP but not in BP. Coda-/l/ becomes [w] in BP.  We could use [l] normally and  in the coda with a note in the key that it's [w] in BP.
 * In much of EP, voiced plosives are lenited much like in Spanish. This doesn't occur in BP.

When we do use the Brazilian and European transcriptions, they should differ only in the areas that can't be accounted for in the pan-dialectal transcription. For example, Rio de Janeiro would be ) even if might be more phonetically accurate. — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  07:36, 2 July 2009 (UTC)


 * I wonder, if we're going to have specific BP and EP transcriptions, why wouldn't we be consistent and use [x] in BP? The reason we can gloss over so much in English is that our readers are familiar with it. That's not the case with Portuguese. Anyone familiar enough with Portuguese to know that /r/ is [x] in BP won't need the IPA to begin with, so our audience are those who know little or nothing of the language. For them, I think transcriptions like [x] would be helpful. kwami (talk) 07:59, 2 July 2009 (UTC)
 * I see two reasons why that might not work. The first is that using a separate character for what is, essentially, the same phoneme will give readers the false impression that  and  are two different phonemes.  The second problem is that  is one of several realizations of the "trill" rhotic depending on dialect.
 * I do get your point about speakers being familiar with Portuguese not needing to check out the IPA key, but it seems to me that a reader unfamiliar with the IPA and curious about the pronunciation should be able to look at the key here and understand that in Portuguese IPA transcriptions at Wikipedia, represents the sound that is a uvular fricative in some dialects and a velar one in others.  It's especially smooth if the article itself says "Brazilian Portuguese pronunciation..." and the key specifies which pronunciations are Brazilian.
 * That would make it similar to our Spanish key, where is used to represent the sound that is a velar fricative in some dialects and a glottal one in others,  represents a sound that is like English th in some dialects and s in others, etc. — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  09:40, 2 July 2009 (UTC)


 * "Similarly, [a e] are allowed in post-stressed position in BP, while they become [ɐ ɨ]* in EP." I'm pretty sure they aren't allowed in posttonic position in BP; rather, they become there. The advantage to using /w/ for coda-L in BP is that it shows it's a real merger with the /w/ element of diphthongs: pau and mal rhyme in BP, but transcribing them  and  wouldn't show that. Incidentally, Mateus and d'Andrade (2000) are not terribly reliable when it comes to BP: over at Portuguese phonology, a user recently corrected the BP pronunciation of abdómen from  to  (more accurate still would be  or possibly ); I checked the source and M&d'A really do use . Basically, they transcribe everything as in EP except for BP differences that are the immediate topic of discussion. And as far as I can tell they don't even mention Brazilian diphthongization before coda /s/ (see Portuguese phonology). +Angr 09:31, 2 July 2009 (UTC)


 * M & d'A may have some transcription flubs but they spend a few pages (pp 17-21) talking about stressed and unstressed vowels and highlighting differences between BP and EP. If that information is incorrect, then they've made a pretty elaborate error.
 * It seems to me that losing a clear indication of a merger is a small price to pay for a pan-dialectal transcription. We can clarify the merger in the key.  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  09:51, 2 July 2009 (UTC)
 * I'll double check what they say about posttonic a and e again when I get home later today. Maybe what I thought was true of all posttonic a and e in BP is true only of word-final a and e; at any rate I'm quite certain taça and gente are and  in BP, not  and . +Angr 10:06, 2 July 2009 (UTC)
 * Hmm, Angr seems to live on one of Jupiter's moons, where the days are longer. For anyone curious, here's a table showing what M & d'A claim to be the differences in unstressed vowels:

There's a mistake here - jure ends in barred-i in EP. And more to the point, in final position, -a can be either [ɐ] or [a] in BP (but not so with -e, as was being discussed).85.240.106.120 (talk) 00:16, 23 July 2009 (UTC)


 * It looks like, in this regard, BP is more conservative. — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  20:50, 4 July 2009 (UTC)

It may seem so, but it isn't. EP keeps 9 vowel phonemes where BP has only 7, though in terms of phones the count is 10 (EP has a barred-u) to 8 (a and inverted-a are the same phoneme in BP). What happens in unstressed syllables is that EP has reduced the mid-close vowels to closed (e/o > ɨ/u, except before nasals), losing a handful of contrasts, but it has kept the open vowels separate, and distinguishes i from ɨ, resulting in 7 distinct vowels (the 9 except e/o). Whereas in BP, the mid-closed and open vowels tend to merge, the distinction between the two a was lost, and there is no ɨ, resulting in a 5 vowel system.85.240.106.120 (talk) 00:30, 23 July 2009 (UTC)


 * No longer days here, I just forgot. But your table confirms my second suspicion, namely that what I knew to be true of word-final posttonic vowels in BP, I incorrectly believed to be true of all posttonic vowels. Mea . +Angr 06:09, 7 July 2009 (UTC)
 * P.S. M & d'A are doing their usual thing of transcribing BP differently from EP only with respect to the topic under discussion, rather than transcribing it holistically: mirar, murar, morar, pagar and pegar do not end in in BP, they end in . +Angr 06:14, 7 July 2009 (UTC)

Actually, they usually end in -á. (NB that this also happens in EP, only much less noticed.)85.240.106.120 (talk) 00:30, 23 July 2009 (UTC)

Errors and imprecisions
Someone please correct the following because the page's markup looks like salad to me:

- cabide only has a in BP, it has @ in EP

- likewise semáforo has barred-i in EP (it's before a nasal, but the nasal isn't in the same syllable)

- EP has plenty of unstressed E and O and a, it's BP that merges them with e/o. E.g., velhice, pregar (preach), corar, doninha, padeiro, activo.

- boneco has u in EP

The fact is that EP and BP have different phonemical systems (like AE and BE), and presenting their sounds in terms of realisations rather than correpondences may be misleading.

The mapping between phones is more or less the following (V is the 'archetypal' vowel, etymological phone if you like):

Stressed:

V     EP     BP a      a      a ɐ      ɐ      a ɛ      ɛ      ɛ e     e      e ɔ      ɔ      ɔ o     o      o u      u      u i      i      i ɨ      ɨ      i

Pre-stress: a     a      a ɐ      ɐ      a ɛ      ɛ      e e      ɨ      e (i when initial) ɔ     ɔ      o o      ʉ      o u      ʉ      u i      i      i ɨ      ɨ      i

Post-stress: a     ɐ      a (or ɐ when final) ɐ     ɐ      a (or ɐ when final) ɛ     ɛ      e e      ɨ      e (i when final) ɔ     ɔ      o o      ʉ      o u      ʉ      u i      i      i ɨ      ɨ      i

In addition, when a nasal follows in the same syllable, unstressed e/o are not reduced in EP. Post-stress ɛ/ɔ are rare, but they do exist in EP. And ʉ is not a phoneme, but a very common allophone of u. 85.240.106.120 (talk) 01:13, 23 July 2009 (UTC)


 * We're still figuring out the best way to represent Portuguese and the best way to present this in the key, so your effort and input is appreciated. However, the table you've created above implies a number of things that contradict sources.  For instance, it implies that EP contrasts  and,  and ,  and , in unstressed syllables.  What are the minimal pairs?  Is there really contrast in BP between  and  in final syllables? Your table also implies an additional phoneme  that contrasts with.
 * Or am I misunderstanding your charts? — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  08:03, 23 July 2009 (UTC)

No, you're not misunderstanding my charts:

- Yes, EP has all those contrasts. _Pregar_ 'preach' has, _pregar_ 'nail' has. _Corar_ 'blush' has, _decorar_ 'decorate' has. _Pazinha_ 'little shovel' has, _casinha_ 'little house' has. Those are simple popular words; learned words and sentences offer many more examples, e.g. _amora_ [ɐ'mɔrɐ], _a amora_ [a'mɔrɐ].

- I can't be as trustworthy about BP as I'd like to, but the instances of medial unstressed  are few for me to decide. I'm almost sure a word such as _celêntero_ would have 3. NB 'my' chart says BP final -e is [i] - though you can hear Cláudia Ohana sing 'do que viver assim' with que = [ke] (http://www.youtube.com/watch?v=lD74pEgS-vc, 1:45), it happens sometimes. BTW, this very passage offers a clear example of -r before a vowel (viver assim): it's a flap.

- The one thing I wrote misleadingly (and now corrected) was. NB I did say it wasn't a phoneme, the vowels under 'V' are rather the etymological vowel that would be present if there was no vowel reduction at play. You may observe that in stressed position all are kept in EP, while BP merges and  (in inherited words, the distribution of those two is etymological in EP, it's not just a case of EP having recently developed an  phoneme). In unstressed position, some of the 9 vowels are reduced in different ways in EP (to 7/6) and BP (to 5). The bit I corrected was that I was presenting in its own right, but I hope it is now clear that it is only the usual EP version of o/u in unstressed position.

- And I'm sorry that I put comments all over the page. I don't get along very well with this way of editing - and, in a non-technical view, I don't get along well with the editing process in Wikipedia, though I have no alternative to offer and respect the people who do work here - and worst of all I'm too used to the simplicity of ASCII IPA with a@eEi+oOuU or somesuch to be comfortable discussing phonology here. But since EP is my language, and I've been listenning to BP for many years, I couldn't help giving my input. It could be worse, I could have offered a comparison with galician as well! (Galician sticks to its 7 vowels in pre-stressed and stressed position and merges E/e and O/o in post-stress.)

85.244.235.39 (talk) 00:08, 24 July 2009 (UTC)
 * Feel free to continue using ASCII IPA on talk pages; it's only in actual articles where Unicode IPA characters need to be used. +Angr 05:29, 24 July 2009 (UTC)

Brazilian Portuguese is not uniform
The article uses "Brazilian Portuguese" (BP) as if that were a single entity. But the pronunciation varies considerably by region; for example, the pronunciation "djia" for "dia" is a distinctive feature of the accents of Rio de Janeiro (and perhaps Rio Grande do Sul), but not so much of São Paulo and other regions. Also the Northeast pronunciation has distinctive ways of realizing stress. --Jorge Stolfi (talk) 16:18, 8 August 2009 (UTC)
 * Yes, that's true. We should make it clear what we mean by "Brazilian Portuguese."  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  18:08, 8 August 2009 (UTC)

I am not aware of any region in Brazil in which "dia" isn't pronounced as "djia", except Pernambuco (Recife indeed; I am not sure if this is valid for the countryside). In São Paulo it is most definitely pronounced "djia".

I think trying to represent each and all of the different Brazilian pronunciations would make the article extremely overburdened. It would also increase dramatically the possibilities of misrepresentation. We have few Portuguese speakers here; I am not sure that we have speakers of all the most important variants of Portuguese Brazilian. Given the common phenomenon of exagerating local particularities (or, on the other hand, downplaying them), this would probably cause some edit warring. 200.198.196.129 (talk) 15:12, 14 August 2009 (UTC)

there's no such word as "ramsack", and why do all the examples have to be word-internal anyway?
Tehy should be word-internal because they resemble word-internal English consonants, not initial English consonants (which tend to be aspirated). 200.198.196.129 (talk) 15:18, 14 August 2009 (UTC)

Diphthongized vowel
It perhaps should be more formal if removing the "diphthongization" of some vowels, except, of course, when trying to represent a formal conversation, although it really is very common, it is considered a mispronunciation among the Brazilians.Luizdl (talk) 02:14, 12 September 2009 (UTC)

I mean, in some representations, for example, the IPA for "portugues", in Brazilian portuguese it ending with a "geis", it is very common in Brazil, but "ges" is also common and is more formal, and sometimes it occurs the opposite, words as "brasileiro" sometimes is pronounced "brasilero", but I never saw this representation on Wikipedia, both are informal.Luizdl (talk)
 * Oh, I see what you mean. We should go with the more common formal pronunciation.  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  02:28, 13 September 2009 (UTC)

Paíz
What's the other way on speaking "paíz"? —Preceding unsigned comment added by 187.2.99.64 (talk) 03:55, 6 December 2009 (UTC)
 * My understanding is that is more common, though  occurs in parts of Brazil.   or  are only if the next word begins with a voiced consonant.  Is there something wrong with meto? — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  05:57, 6 December 2009 (UTC)
 * Would be good to use the word paíz (country) because it contrast with pais (parents, fathers), perhaps we should set it in the following way:

The word "meto" is the first person of the verb "meter", which means "to introduce", generally its usage is not vulgar, but it may also be used as vulgarly as the English word f*ck (I think you got the word).187.2.102.164 (talk) 22:00, 6 December 2009 (UTC)


 * Can you think of another pair of words that don't end in a sibilant? — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  00:48, 7 December 2009 (UTC)


 * I can suggest these three words:


 * sábia (wise woman),
 * sabia (I knew) and
 * sabiá (Thrush) --Luizdl (talk) 02:04, 7 December 2009 (UTC)


 * The point of not having an example witha coda sibilant is to have an easily dialect-neutral example. Why exactly do we need to find examples that demonstrate the importance of marking stress?  Why can't we just have an example that shows how to use the stress marker in IPA?  — Æµ§œš¹ <span title="Representation in the International Phonetic Alphabet (IPA)" class="IPA"> [aɪm ˈfɻɛ̃ⁿdˡi]  02:42, 7 December 2009 (UTC)

Crera [ˈkɾeɾɐ] (he had believed (pluperfect)) and crerá [kɾeˈɾa] (he will believe (Future tense)) --Luizdl (talk) 03:27, 7 December 2009 (UTC)


 * Good find. My question still stands, though.  — Æµ§œš¹ <span title="Representation in the International Phonetic Alphabet (IPA)" class="IPA"> [aɪm ˈfɻɛ̃ⁿdˡi]  03:40, 7 December 2009 (UTC)


 * Was good to find examples for show the phonemic distinction of the stress, as it was done with some others phonemes like in "aranha" and "arranha".Luizdl (talk) 04:10, 7 December 2009 (UTC)

THE WORD PAÍS - SOME COMMENTS

1. Spelling. The correct spelling is país (pl.: países).

2. Pronunciation (in Portugal and some regions of Brazil). - when the next word begins with a voiceless consonant. - when the next word begins with a vowel. - when the next word begins with a voiced consonant.

Actually this rule applies to all instances of s at the end of a word (or syllable) boundary:

– isolated word / following word begins with voiceless consonant

- following word begins with voiced consonant

- following word begins with vowel

In some areas of Brazil s is always pronounced, regardless of its position in the word.

Pais ~ País. There is some inaccuracy in the transcription in the table above. In EP the pronunciation (pais) does not exist; it is  (monosyllabic). As for país, it is pronounced (disyllabic). Note the (isolated word).

In EP the diphthong is an allomorph of /ej/ in the pronunciation of Lisbon (by the way, that’s where I was born and am presently living) and some other southern regions (it tends to spread due to the prestige of the variety). /ej/ has two other allomorphs: [ej] (North) and [e] (South; note the monothongisation)

Analogica (talk) 03:52, 15 February 2010 (UTC)