Help talk:IPA for Catalan

From Wikipedia, the free encyclopedia
Jump to: navigation, search


I'm not familiar with Occitan as much as I am Catalan. Is it important that Occitan orthgraphy seems to differ with Catalan's? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 02:08, 28 June 2009 (UTC)

I've put an "under construction" tag to prompt editors and readers to understand that this page is by no means complete. Once the "and Occitan" was added, it's gotten into territory that I'm not familiar with but, for example, [ɕ ʑ] are probably not accurate when it comes to Occitan, nor is the pattern of lenition. If this is the wrong template, I'm not sure what to put instead. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 02:45, 15 July 2009 (UTC)
All right, I've removed the Occitan information. I believe Catalan and Occitan to be too different phonetically and orthographically from each other to warrant a single key for both of them. If editors wish to have a guide for Occitan, the as-yet uncreated Wikipedia:IPA for Occitan is the place to do this. Right now, {{IPA-ca}} is still designed with Occitan in mind and there are probably a few dozen pages that say "occitan pronunciation." This will have to be fixed. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 10:06, 7 November 2009 (UTC)

This table lacks the /v/ sound which is present in the Balearic Island and in the Land of Valencia[edit]

The /v/ sound exists in the Catalan spoken in the Balearic Island and the Land of Valencia. Could you include it on the list as Balearic and Valencian pronunciation. I am from Denia in Alicante and the Catalan-speaking people from here use it, thanks. (talk) 21:29, 11 November 2009 (UTC)

The /a/ Phoneme[edit]

Hello! I have realised that you have used the same symbol (/a/) for representing the stressed a's of both Spanish and Catalan. That is not possible!

I am only an amateur in this field, but any native Catalan speaker can see the enormous difference between the Spanish stressed "a" and its Catalan counterpart. It is very hard for us to learn the pronunciation of Spanish a's and vice-versa.

Another thing that exemplifies what I am saying is that Catalan ortography marks some tonic vowels with accents. The possible combinations are as follows: à, è, é, í, ò, ó, and ú. The ones with acute accent (called "closed" vowels) are pronounced like Spanish (witch only has 5 vowels: á, é, í, ó and ú), but the rest (with grave accents) are called "open" and have no Spanish equivalents.

So you should change the symbol for this phoneme in one of those articles.

Thank you very much! —Preceding unsigned comment added by (talk) 13:48, 26 March 2010 (UTC)

Symbols used for Catalan palatal fricatives and affricates[edit]

Replaced the symbols /ʃ/, /ʒ/, /tʃ/, /dʒ/ for Catalan palatal fricatives and affricates, as those are used everywhere for transcription of these same sounds in the charts of other languages (English, Spanish, Italian, etc) here in Wikipedia and everywhere else in scholarship etc. So I cannot see the need to insist in another set of bizarre symbols /ɕ/, /ʑ/, /tɕ/ and /dʑ/.Perique des Palottes (talk) 16:11, 21 April 2010 (UTC)

There is a tendency to represent a variety of postalveolar fricatives with <ʃ> and <ʒ>. I've seen it done with the retroflex consonants of Russian and Polish and the alveolo-palatal consonants of Korean and Japanese. The literature I've seen on Catalan varies in the characters it uses and I suspect that this is because the distinction between alveolo-palatal and palato-alveolar is pretty subtle (I really can't hear the difference). Interestingly, Wheeler (2005) uses <ʃ> and <ʒ> while referring to them as alveolo-palatal, suggesting that there may be typographic constraints. Thus, rather than base the use of <ɕ> and <ʑ> on what the IEC, the AVL, or anybody else uses, I found an article by a phonetician (Daniel Recasens) who describes the coronal and palatal consonants in a manner that's more detailed than the symbols indicate, telling me that he's studied it himself using acoustic data. It's possible that there is geographic variation but we'd really want sourcing that describes such variation before we change anything. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 16:42, 21 April 2010 (UTC)
If you read carefully Recasens paper ("Coarticulation...") you will see these same authors use freely the term alveolopalatal for any of the sounds [ʃ,tʃ,ʒ,dʒ,ʎ,ɲ] (the sounds conventionally represented in Catalan orthography as x-,-tx-,j-,-tj-,ll,ny), as opposed to palatal propper sounds like [j,ç]. And the same in other papers by the same author, as:

"An Electropalatographic Study of Alveolar and Palatal Consonants in Catalan and Italian" by Daniel Recasens et alt.

So, other than papers "telling you whatever" (your wording), please be so kind to cite the exact section that unequivocally states the Catalan sounds [ʃ,tʃ,ʒ,dʒ] are different in articulation to the analogous sounds in neighbouring Romance languages (other than length or gemination).Perique des Palottes (talk) 12:40, 20 May 2010 (UTC)
Unfortunately, I can't get ahold of Recasens 1990, which is where the phonetic study occurs. The source you've provided shows pretty clearly that Catalan /ɲ/ and /ʎ/ are the same as Italian's, but is silent about the postalveolar fricatives. A number of other works that Recasens has done talk about these consonants but not about the fricatives. But Recasens (1990) is the source to get, it seems. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 16:45, 20 May 2010 (UTC)
I got and read Recasens 1990 article Articulatory characteristics of palatals in JIPA 18, 1990, pp 267-280. There is no reference to Catalan sounds being other than postalveolars (same as Occitan and Italian ones). Our postalveolars [ʃ tʃ] are described with references to examples from Catalan, among others like Italian and French (as could be expected). Your alveolopalatals [ɕ tɕ] are described with references to examples from Polish and Mandarin Chinese (as could be expected - no Catalan here). Now, I challenge you to prove your case.Perique des Palottes (talk) 13:28, 2 June 2010 (UTC)
Looking through the work I do have by Recasens, I found this from a more recent article in JIPA "An electropalatographic and acoustic study of affricates and fricatives in two Catalan dialects" (37.2, 2007, p. 145):

While language-dependent differences in constriction location for /s/ always occur within the alveolar zone, the place of articulation for /ʃ/ (and for its voiced cognate) is less clear. According to the International Phonetic Alphabet, the phonetic symbol [ʃ] corresponds to a postalveolar fricative, and the term 'palatoalveolar' which is also assigned to this fricative corresponds to a lamino-postalveolar articulation (see Ladefoged & Maddieson 1996: 14f.). Other close fricatives are the dorsopalatal [ç], which is an allophone of /x/ in German and Norwegian (Simonsen&Moen 2004), and the alveolopalatal [ɕ] in languages such as Chinese and Polish, where it is produced with a lowered tongue tip, a high tongue dorsum position and a long constriction extending from the alveolar zone to well inside the palatal zone (Ladefoged & Maddieson 1996: 150–164)."

Already, we see that Recasens clearly understands the distinction between the different kinds of fricatives and is does not use alveolopalatal when he doesn't mean alveolopalatal. There's more:

Electropalatographic data for five speakers of Eastern Catalan reported in Recasens & Pallarès (2001: 84f.) reveal that Catalan /ʃ/ is articulated essentially at the postalveoloprepalatal zone for all speakers and involves much dorsal contact at both sides of the palate behind the constriction. Moreover, speakers seem to be having the tongue tip down during the production of this laminal or lamino-predorsal consonant, as suggested by the fact that the two frontmost rows of electrodes of the artificial palate remain completely unactivated and that the frontmost lateral contact has a V-like shape. Based on these data, we believe that Catalan /ʃ/ ought to be labeled 'alveolopalatal' which is consistent with the presence of other alveolopalatal consonants such as /ɲ/ and /ʎ/ in the language. Accordingly, we will use this term to refer to /ʃ, ʒ/ and to /tʃ, dʒ/ in this paper.

"It appears then that alveolopalatal articulations such as /ɲ/ in most Romance languages and /ʃ/ in Catalan cannot be possibly assigned a single articulatory zone: the IPA term 'postalveolar' would not be appropriate for Catalan /ʃ/ since the constriction for this consonant also occurs at the palatal zone, and the IPA term 'palatal' cannot be applied to /ɲ/ in languages or dialects where those consonants are alveolopalatal since closure location extends into the alveolar zone in this case. For all these articulations, closure or constriction location takes place not just at one articulatory zone but at two articulatory zones simultaneously. Based on these observations, it seems that the term 'alveolopalatal' would need to be included in the IPA chart, and that the phonetic symbols [ʃ] and [ɲ] could be assigned two possible places of articulation rather than just one, i.e. 'postalveolar' and 'alveolopalatal' for the former, and 'alveolopalatal' and 'palatal' for the latter."

It is very clear here that the authors mean what I have been saying they mean. This last paragraph very overtly pegs the fricative of Catalan as separate from the equivalent fricative of other Romance languages. BTW, could you email me the pdf of that 1990 file? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 18:06, 2 June 2010 (UTC)
Eh? This last paragraph very overtly pegs the fricative of Catalan as separate from the equivalent fricative of other Romance languages? All the whole explanation seems messy to me, just to be polite, as your interpretation does too.
Just to show you I am not getting personal, where could I leave that pdf image file of 1990 recasens' article for you? It is 213.464 bytes long.Perique des Palottes (talk) 13:47, 3 June 2010 (UTC)
Send me an email (there's an option on my userpage to do so). I'll respond to your email and then, in your response to my response, you can attach the pdf. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 17:32, 4 June 2010 (UTC
You will need to be more explicit.Perique des Palottes (talk) 08:54, 9 June 2010 (UTC)
Go to my userpage. Either at the bottom or the side, there should be an option to "email this user." Do it. 16:32, 9 June 2010 (UTC)
There is not such an option, or at least it is not shown anywhere. I am not going to pursue this much long. If you want to contact me directly, if you are not a "yahoo" you'll understand I've set up a temporary email address with the 2nd and 3rd words of my user nick here as d..p......s at Perique des Palottes (talk) 08:24, 10 June 2010 (UTC)
Keep in touch. Pdf should be going.Perique des Palottes (talk) 12:54, 23 June 2010 (UTC)
It seems as though you want the author to say something to the effect of "this sound in Catalan is different from this sound in Italian/Spanish/French." I think that they get closest to this when they say, "It appears then that alveolopalatal articulations such as /ɲ/ in most Romance languages and /ʃ/ in Catalan..." Remember, Recasens already determined in another article that /ɲ/ shares the same place of articulation in Catalan and other Romance languages. If he believed that "/ʃ/" of Catalan was the same as other Romance languages then he would have worded it like this "It appears then that alveolopalatal articulations such as /ɲ/ and /ʃ/ in most Romance languages..."
I've quotes all three paragraphs, though, because in them he's basically saying that the sound of Catalan is the same as the sound of Chinese or Polish: He says that [ʃ] is a (lamino-)postalveolar/palatoalveolar fricative, that Chinese and Polish have alveolopalatal fricatives, that the IPA term postalveolar isn't correct for the sound in Catalan, and that "Catalan /ʃ/ ought to be labeled 'alveolopalatal.'" Is that convincing to you? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 17:32, 4 June 2010 (UTC)
This is very confusing. I found the article by clicking on the IPA for "Xavi", and the symbol used for the initial sound, the one I want to pronounce, ʃ, has been replaced. This is not how Wikipedia should be. tharsaile (talk) 13:56, 15 July 2010 (UTC)
I've gone through articles that use {{IPA-ca}} and fixed the IPA. I don't know who changed the postalveolar fricatives. Was it you, Perique? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 19:36, 15 July 2010 (UTC)
It's so clear now! Many thanks to Aeusoes1 tharsaile (talk) 12:54, 30 July 2010 (UTC)

Recent changes[edit]

Perique, I undid some of your changes and I may have inadvertently undone things that I don't actually agree disagree with (though I hope not). I put two citation requests for information that I'd actually like to see included with proper citation at Catalan phonology. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 04:08, 22 April 2010 (UTC)

You undo things pretty liberally, I see.Perique des Palottes (talk) 07:27, 22 April 2010 (UTC)
Consensus is not what you think of what you have understood of Recasens. Consensus among all schoolars is those sounds are the same, where present, in Spanish or Italian or Catalan or English. So why do you insist in just patronizing about pages on Catalan language?Perique des Palottes (talk) 07:15, 23 April 2010 (UTC)
Moreover, you insist in undoing corrections to flagrant error in Catalan examples, blind is cec not *sek, sack is sac not sak, initial sound of Jordi is not affricate in Eastern Catalan, among others...Perique des Palottes (talk) 07:26, 23 April 2010 (UTC)
It wasn't my intention to undo corrections to flagrant errors, but when you mix uncontroversial edits with controversial ones, these things can happen.
When I speak of consensus, I'm talking about Wikipedia's policy. The statement that all scholars believe the sounds to be the same is clearly false; if you were simply being hyperbolic, you haven't even presented the view that these sounds are identical, much less one that contradicts or discounts Recasens's work with phonetic data. So far, you've simply speculated about typography. If you don't have sources, we can't even have a conversation. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 13:06, 23 April 2010 (UTC)
I see you were the one introducing the change for signs for postalveolar sounds in Catalan phonology article,Revision as of 03:59, 18 January 2008. I assume you took most of it from somewhere, maybe Wheeler? But then, where did you take the alveolopalatal thing from?Perique des Palottes (talk) 13:55, 2 June 2010 (UTC)
If I remember correctly, Wheeler cites Recasens in saying their alveolopalatal while using the characters ʃ ʒ. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 17:44, 2 June 2010 (UTC)
So you decided unilaterally to use those signs, which are used for the Chinese sounds, which by the way clearly are NOT the same as the Catalan ones, where all authors, including Recasens, Wheeler, and all others, keep using the same characters ʃ ʒ, as used for other Romance languages, as they are the more approximate ones.
So, why do you think Recasens himself carefully differentiates between examples with ʃ ʒ from Catalan and other languages, and examples with ɕ ʑ from Polish and Chinese? As in p.38-39 of Recasens, 1993, "Fonètica i Fonologia" published by Enciclopèdia Catalana (this one carries the same discussion on palatal flavours and coarticulations, analogous palatograms etc and Recasens labels these sounds with the nice name of "laminoalveolars palatalitzades" - p.36).
Same in Recasens, 1996, "Fonètica descriptiva del català" published by Institut d'Estudis Catalans.
Same in Recasens & Pallares, 2001, "De la fonetica a la fonologia" published by Ariel (here Recasens labels these sounds with the nice name of "laminal postalveolars" - p.27).Perique des Palottes (talk) 12:00, 4 June 2010 (UTC)
I think that there is a long enough tradition of transcribing these Catalan fricatives with the same IPA character as that used in Spanish and French, that even though phonetic data has come to show that they are (at least now, if not before) different enough that another IPA character would be more accurate, tradition has come in the way a bit. This is also the case with the "hard" postalveolars of Russian and Polish (which are retroflex), and the vowel of cup in English (which is closer to [ɐ] than [ʌ]).
This is probably why he so freely uses ɕ ʑ for Chinese and Polish but sticks to ʃ ʒ for Catalan even though he sees them as the same. I'm speculating though, and I suppose we could ask him ourselves. I chose to use the more accurate signs quite a while ago, just as I chose to use the more accurate signs for Russian. The latter decision had more involvement with other editors. So far, the only people who have opposed this for Catalan are the people who believe the signs to be inaccurate, which I disagree with. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 17:45, 4 June 2010 (UTC)
Catalan has too long tradition transcribing with /ʒ ʃ dʒ tʃ/. All universities across the Catalan Countries and elsewhere where there are Catalan Studies transcribe with these sounds/symbols. All dictionaries for Catalan/Valencian/Balearic use these IPA symbols.
English also sticks to some traditional sounds "symbols":
  • English /ʌ/ from cup is closer to [ɐ] in many dialects, as RP or Californian English, but /ʌ/ is still used for Standard English. Also some dialects pronounce /ɛ/ as [e] or [e̞] as you mention on the IPA for English dialects article.
  • /r/ is widely used instead of /ɹ/ in broad transcriptions of English. In most dialects of English and Standard English it is an approximant /ɹ/ (represented with an inverted ɹ).
So, would you mind changing /ʑ ɕ dʑ tɕ/ back to wider used and traditional /ʒ ʃ dʒ tʃ/?
May be as you see, it is more logical and realistic representing /ʑ ɕ/ (not /dʑ/ and tɕ/) than /ʒ ʃ/ but there is a larger usage of /ʒ ʃ/ in Catalan. So, is it not possible to add a note like Catalan [β ð ɣ] (fricatives or approximants), Spanish [β ð ɣ] (approximants, traditionally fricatives) and English /r/ (which is an approximant; /ɹ/), saying:
/ʒ ʃ/ are widely used instead of /ʑ ɕ/ in broad transcriptions of Catalan (I doubt about /dʑ/ and /tɕ/ existence in Catalan, which are /dʒ/ and /tʃ/).
I've been taught at university to use /ʒ ʃ dʒ tʃ/ in Catalan and I'll keep using them, however I accept [ʑ] or [ɕ] are more realistic than /ʒ ʃ/. But is the trilled r more realistic for English than the approximant /ɹ/?
I also reckon this article should be divided into several pronunciation patterns. Same as European Portuguese and Brazilian Portuguese, and Northern Dutch, Southern Dutch and Afrikaans.
Patterns could be:
  • Eastern Catalan (standard Eastern Catalan).
  • Western Catalan (standard Valencian).
  • Central
  • Western
  • Balearic
I'd be glad to help. Jaume87 (talk) 00:28, 21 August 2010 (UTC).

──────────────────────────────────────────────────────────────────────────────────────────────────── In regards to tradition, I should point out that we deviate against tradition for Polish and Russian by transcribing the hard postalveolars as retroflex, for Japanese with the transcription of the back high vowel, and even the system at WP:IPA for Irish looks unconventional (though I could be wrong about that one). I don't think we need to honor that tradition, and it seems so far that there isn't a strict respect for such tradition with IPA transcriptions . You can point out a few transcription choices that don't reflect phonetic accuracy, but you can hardly call the system at WP:IPA for English traditional. I'm warm to the idea of reflecting dialectal variation of Catalan, but I'm not sure how much. The key is, like Portuguese and Dutch, to represent only as much dialectal variation as you need. How many dialects do we need to represent? With Portuguese, it's two. Spanish is normally one, though sometimes two. Dutch is three. I'd assume two for Catalan, especially if we stick to the standard varieties. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 04:54, 21 August 2010 (UTC)

I know I can't call WP:IPA for English traditional. English phonetics are amongst the most innovatives, I think you misunderstood me.
However is it too difficult to allow Catalan to use /ʒ ʃ dʒ tʃ/ as English is allowed to use /r/ for its transcription of /ɹ/? Too long tradition in Catalan, too many manuals, books, sources, etc. Even Recasens himself utilizes them.
Concerning the division into several patterns. I think the best think should be 2 dialects, Eastern Cat. (Central, Balearic and Northern Cat.) and Eastern Cat. (Valencian and NW Cat.).
/v/ should be kept and should be well explained on notes where it occurs, and where it doesn't (betacism).
Jaume87 (talk) 00:33, 22 August 2010 (UTC)
The issue isn't one of difficulty but of phonetic accuracy. As I've said above to Perique, the Recasens sources clearly define the sounds as alveolopalatal even if the palatoalveolar characters are used. Recasens and others may be guided by tradition, but I don't see why we should do the same, especially when we don't do so for other languages. As far as the English rhotic, /r/ is used instead of anything else because it's typographically easier.
I'm worried that the way you've framed the dialectal division might be cumbersome. What would be wrong (or different) with picking Standard Valencian (spoken in the southern part of Valencia) and Standard Eastern Catalan (based on Barcelonan Catalan)? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 01:12, 22 August 2010 (UTC)
Honestly the difference between /ʒ ʃ dʒ tʃ/ and /ʑ ɕ dʑ dɕ/ is minimum. You could content more people by using the traditional symbols used in all Catalan universities, with a note saying "currently these sounds are /ʑ ɕ dʑ dɕ/". For Catalan is easier typographically as well. Also, /ʒ ʃ dʒ tʃ/ are easier to find than /ʑ ɕ dʑ dɕ/, plus has more analogy with Catalan neighbouring languages (Romance languages). Any Catalan speaker would't even notice the difference between /ʑ ɕ dʑ dɕ/ and /ʒ ʃ dʒ tʃ/.
Those are good, Standard Valencian (spoken in the southern part of Valencia) and Standard Eastern Catalan (based on Barcelonan Catalan). —Preceding unsigned comment added by Jaume87 (talkcontribs) 01:40, 22 August 2010 (UTC)
Those arguments are equally valid for Russian or Polish. Should we change those as well? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 05:15, 22 August 2010 (UTC)
I don't know very much in-depth about Slavic phonetics and languages, though I find them very interesting, especially Russian. But Catalan sibilants are much simplier. First, Catalan doesn't distinguish/contrast palatal and non-palatal consonants as Russian. And second, Catalan doesn't contrast retroflex vs. palatal (or more specific alveolo-palatal) as do Russian and Polish. There is a slight difference among these sounds, same in Basque with laminar /s̻/ and apical /s̺/.
I don't know much about Polish and Russian phonetics manuals, whether they use more /ʒ ʃ dʒ tʃ/ or /ʑ ɕ dʑ tɕ/. I certainly know about Catalan, and we use /ʒ ʃ dʒ tʃ/ all the time, Catalan speaker aren't familiarised with /ʑ ɕ dʑ dɕ/.
  • Polish set of sibilants; /s ɕ ʂ/, /z ʑ ʐ/, /ts tɕ tʂ/, /dz dʑ dʐ/.
  • Russian set; /s/-/sʲ/, /z/-/zʲ/, /ʂ/-/ɕɕ/, /ʐ/-/ʑʑ/, /ts/-/tɕ/.
Far more complex than Catalan. I can't oppose to Polish & Russian /ʑ ɕ dʑ tɕ/ can also be represented like /ʒ ʃ dʒ tʃ/ if these signs are more common in Poland and Russia. Also it'd be easier for many people to recognise /ʒ ʃ dʒ tʃ/ than /ʑ ɕ dʑ tɕ/.Jaume87 (talk) 16:56, 22 August 2010 (UTC)
Recasens himself sticks in his articles to differentiate clearly examples from Polish and Chinese (transcribed with alveolopalatal signs) as not the same sounds used in examples from Catalan (transcribed with postalveolar signs).
All Catalan manuals, dictionaries, webs, etc. stick to postalveolar transcription AND description for those sounds as well, e.g. Els sons del català The exact articulation of those sounds has quite dialectal variation (in palatalization, tenseness, lentgh, africation, and even, coarticulation). Also, depending on the dialect, there are more or less assimilation with neighbouring consonants (say, e.g. much more assimilations in Majorcan then in Barcelonan).
A similar case could be exposed for /s/, which has different articulation among different speakers, say, coronal vs. apical with more hissing, the so called 's pija', that is the standard French sound, which is different from Catalan or Spanish ones.
So, why not sticking to generally used signs?
You, aeusoes1, are the only introductor and defender of using those signs for Catalan.
Please concede at least that the rest of us, Catalan native speakers and phonetics graduates, might not be so wrong.
Perique des Palottes (talk) 12:18, 24 August 2010 (UTC)
I've explained above that I prefer the alveopalatal signs because of their phonetic accuracy for Catalan. While you and others have argued that the sounds aren't actually alveolopalatal (in defiance of sourcing) Jaume is the first, and only, person to acknowledge this phonetic accuracy and make the case that we shouldn't be phonetically accurate in this regard. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 12:51, 24 August 2010 (UTC)
You keep reading just what you please. Jaume above said "I doubt about /dʑ/ and /tɕ/ existence in Catalan, which are /dʒ/ and /tʃ/". He is not saying that using alveolopalatal signs are the good and only possible ones. He is conceding that there might be a lot of variation in articulations and in possible conventions to transcribe them. And so this might be why even Recasens, your champion, insists in using [ʒ,ʃ] for Catalan, as different as the Polish, Chinese, examples etc: because those same signs used for Occitan, Spanish, French, Italian, etc map sounds much more similar to Catalan ones than the others. Perique des Palottes (talk) 09:04, 26 August 2010 (UTC)
You could be accused of the same. Right after, he says "however I accept [ʑ] or [ɕ] are more realistic than /ʒ ʃ/." His argument is basically that we should use the palatoalveolar symbols, despite the phonetic accuracy of the alveolopalatal symbols, because that's how they're normally transcribed not because it represents a more accurate phonetic picture. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 12:06, 26 August 2010 (UTC)
I doubted about /dʑ/ and /tɕ/ (not any more), but I've never doubted Catalan palatal sibilants are closer to [ʑ]/[ɕ] ~ [ʒʲ]/[ʃʲ] (which are more "realistic"), however this doesn't mean we have to use those signs in our transcriptions since no one uses them, and that's the reason I defended the traditional symbols [ʒ ʃ] used by all sources, and I suggested to changed it. And although we use palato-alveolar signs now, we label them as alveolo-palatal. Jɑυмe (xarrades) 20:10, 25 April 2011 (UTC)

Correcting some recent changes[edit]

  • salv for final [f] writen 'v', Balearic 'I save' (present simple 1st person), seems a quite convoluted example
  • càstig has accent
  • tsar is a common name
  • xoc; clenxa are voiceless as 'cheek'
  • xilofon initial sound is as 'shine', not as 'zoo'
  • dolçment should be standard dolçament, for a sequence ç+voiced consnant better feliçment — Preceding unsigned comment added by Perique des Palottes (talkcontribs)
Sorry I missed an accent, I don't have a keyboard with accents so it is easier to miss something. Yes i know tsar is a common name i misused capital letters, however it is a loanword.
In Valencian and Balearic salv can be used as a synomym of excepte ("except", "apart from"). How about in Central Catalan?
We could use another word with v that sounds [f] in Standard Valencian as Balearic. However it has to be a loanword; leitmotiv, molotov.
I know xoc and clenxa are voiceless; and caixmir, peix blanc/blau are voiced. Jaume87 (talk) 21:29, 24 August 2010 (UTC)
The representation of the two standards is coming along nicely, though I don't understand what the slashes mean. Valencian, for example has "l / ɫ." Which is it that we're going to be translating it as. Even if there's free variation, we should pick one and go with it.
It gets even more confusing with the vowels. I suspect it has something to do with stress, but it's not clear. Because there's dialectal differences in what vowels appear under stress, we should probably have a stressed/unstressed distinction like that found at WP:IPA for Russian. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 23:39, 24 August 2010 (UTC)
Then we could just say /l/ (in intervocalic position) and [ɫ] in the coda. The reason i used l / ɫ was because of the 'geminate l' [ɫɫ] (which is dark in both WC and EC).
I don't think it is necessary to do a stress/unstress table for vowels as in Russian or English. English & Russian vowels are more complex and go further reductions. European Portuguese features further reductions as well, and they haven't got no such table there.
I'll delete the vowel harmony and only mention it as a note. But it is very common, mainly /ɛ/ which also appears in north-western Catalan and Andorran. Jaume87 (talk) 01:15, 25 August 2010 (UTC)

Further changes[edit]

Now it is divided into the two major standards of Catalan; are we allowed to use /ʃ ʒ tʃ dʒ/ instead of /ɕ ʑ tɕ dʑ/? We could solve this matter with a note. Jaume87 (talk) 19:25, 26 August 2010 (UTC)

Fair enough. I've tweaked the note to reflect information as I understand it. Now someone's got to go through and fix all the transcriptions, including at Catalan phonology. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 03:27, 27 August 2010 (UTC)

Too many notes[edit]

These pronunciation keys shouldn't swamp the reader with what are for them irrelevant details. Right now, we have way too many footnotes, almost as if it's a content fork for Catalan phonology (where that kind of detailed information should go, if it's properly sourced). — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 14:03, 5 November 2010 (UTC)

I know, these are too heavy notes. However most of what it is said here has already been proven with sources.
Most notes should be simplified, but I think they could be useful to improve Catalan orthography and phonology articles.
I was thinking about adding some other features like;
  • Majorcan ca:iodització in regular speech (akin to yeísmo, but with many exceptions).
  • [ɫ] dropping and l-vocalisation in regular speech in some words.
However, as you say there are too many notes, with irrelevant details to most readers. Jaume87 (talk) 16:41, 10 November 2010 (UTC)
Simplified. I won't add here Majorcan iodització in regular speech, but will add it to the Catalan phonology.
Added (for local transcriptions of La Franja) /l/ is palatalized to [ʎ] in consonant clusters; e.g. plou [pʎɔw]. Notes left are essential for local transcriptions. Jɑυмe (xarrades) 23:15, 24 February 2011 (UTC)

Linking (no brake) [‿][edit]

Why did you remove [‿]? Though this IPA symbol is not described in the Catalan phonology we can certainly use it. [.] (syllable break) is not very well explained in many languages' phonology studies; but we use it for hiatus in most Romance languages [ˈmi.o] (sp/it), [ˈmi.e] (ro/it), [ˈʁu.ɐ] (pt), [ˈti.ə] (ca).

Linking (absence of a brake) clearly occurs in Catalan/Valencian, both in formal and common regular speech. Enllaços fònics/vocàlics (also contactes vocàlics) and sinalefes:

Valencian (page 39, enllaços fònics)
  • mitja hora [midʒˈɔɾa], una hora [unˈɔɾa], quina hora [kinˈɔɾa], quinze anys [kinzˈaɲs], onze anys [onzˈaɲs], una escala [unasˈkala] (mandatory)
  • no el saludes [noɫ saˈluðes], diu que ara ve [diw ˈkaɾa ve] (facultative)
In Central Catalan this is even more prominent due to [ə] existence, which is dropped in many cases: (from page 24 on, combinaisons de voyelles dans le Catalan Central)
  • ho ha estat [wasˈtat], hi he estat [jesˈtat].
  • torna a entrar [ˈtornənˌtɾa], a casa ho faré [ə ˈkazu ˈfəɾe], etc.

These two sources prove the occurrence of synaloepha and linking (absence of a brake) in Catalan/Valencian. As documents say, there is not a rule to predict such occurrences, being facultive in most cases, however in other cases such contractions are obligated by IEC and AVL. I've never heard a native Catalan/Valencian speaker who pronounces una hora or mitja hora as it is spelled, but un'hora* and mitj'hora* as Italian elision ('), un'ora and mezz'ora. Catalan orthography is also an article which needs upgrading, currently it doesn't correspond with official Catalan orthography. Jaume87 (talk) 21:43, 14 November 2010 (UTC)

I don't doubt that the phenomenon occurs, but I don't think it's important for our readership. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 21:47, 14 November 2010 (UTC)
Then i'll just add [‿] to the chart, with no explanation. Jaume87 (talk) 22:06, 14 November 2010 (UTC) :)
I don't think we should utilize it at all. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 22:15, 14 November 2010 (UTC)
IMO it could be used in these particular cases where a vowel becomes silent, I don't see any reason for not using it. —Preceding unsigned comment added by Jaume87 (talkcontribs) 22:48, 14 November 2010 (UTC)
It's not whether it can be used but whether it should be used. I don't think it should. It has no phonetic meaning and so, phonetically, is superfluous and potentially confusing. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 23:14, 14 November 2010 (UTC)
I understand, then you can take it off, but this could've been useful to indicate linking and synaloephas. Jaume87 (talk) 23:28, 25 November 2010 (UTC)

Marginal vowels: [y] and [ø][edit]

We could add these two marginal vowels, what do you think? Jɑυмe (xarrades) 15:35, 23 April 2011 (UTC)

No problem from me. They're only in loanwords, right? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 17:18, 23 April 2011 (UTC)
Yes, in loanwords and linguistic interferences from Occitan (Aranese/Gascon, Languedocien) and French. Jɑυмe (xarrades) 18:59, 23 April 2011 (UTC)

Unnecessary notes[edit]

I am taking attention to reducing the number of footnotes in some of these IPA for X pages. These pronunciation keys are designed primarily for readers wanting to understand the language-specific IPA transcriptions they encounter in Wikipedia articles. We shouldn't swamp them with irrelevant details. Because this information may still be pertinent to the project, I have duplicated the notes below rather than try to find a place for them. This is irrespective of whether I think these claims are true or whether they are sourced. I will leave it to other editors to move the information to the appropriate article space or check that it already is. — Ƶ§œš¹ [ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ] 23:07, 28 December 2011 (UTC)

  • In some dialects (particularly North-Western Catalan and Central Catalan), /p t k/ and /b/ may be elided when occurring utterance-finally after a nasal consonant (e.g. amb, seient, sang), in the case of /t/, in some of these dialects, this has extended also after liquids (e.g. alt ,hort).
  • In regular speech in Valencian, intervocalic [ð] may be elided (e.g. fideuada/fideuà).
  • In most of Majorcan the velar plosives, /k/ and /g/, become palatal, [c] and [ɟ], before front vowels (/i e ɛ a/) and word-finally (e.g. guix [ˈɟiʃ], sac [ˈsac]) in some of these dialects, this has extended to all environments except before liquids and back vowels; e.g. sang [ˈsaɲс].
  • ‹l› is always dark [ɫ] in Eastern Catalan. In other dialects, like Valencian, it may vary allophonically with the alveolar lateral approximant, [l]~[ɫ], as it does in English.
  • In La Franja Catalan (Aragon) /l/ is palatalized to [ʎ] in consonant clusters, such as /bl pl gl kl fl/; e.g. plou [ˈpʎɔw].
  • In Catalan and Valencian an assimilation process occurs wherein two identical sibilants appearing in sequence within a word are reduced to a single consonant. For details, see Catalan orthography.
  • In some dialects, /dz/ is deaffricated to [z] in verbal forms ending in -itzar; e.g. analitzar [ənəɫiˈza]/[analiˈza(ɾ)]. Similarly, /ts/, which only occurs word-initially in loanwords (e.g. tsar, tsuga) is deaffricated in these dialects.
  • The pronunciation of words with the digraph ‹ix› varies; an absent [j] is generally more common in Eastern Catalan dialects (e.g. caixa [ˈkaʃə]) and [j]-retention is more common in Valencian and North-Western Catalan (e.g. caixa [ˈkajʃa]), though there are exceptions.
  • In Western Ribagorçan /θ/ may replace standard Catalan and Valencian /s/ (from medieval /ts/): cinc [ˈθiŋk].
  • In Catalan and Valencian occurs synalepha (that is the merging of two syllables into one). A synalepha may be produced either by elision (when combining two non-high vowels): una hora [ˈunˈɔɾə]/[ˈunˈɔɾa]; or by diphthongization (when combining any vowel with high vowels); e.g. posa-hi [ˈpɔzəj]/[ˈpɔzaj].
  • While in most dialects /a/ is central [ä], it is front [a] in many Balearic dialects.
  • The mid-open vowels /ɛ/ and /ɔ/ are lower in Valencian and most Balearic dialects, that is, in these dialects the phonetic realization of /ɛ/ approaches [æ], while /ɔ/ is as low as [ɒ].
  • In most of Balearic, especially Majorcan, /ə/ can appear in stressed position; e.g. sec [ˈsəc]. This corresponds to stressed /ɛ/ or /e/ in other dialects.
  • Many Valencian dialects feature a sort of vowel harmony when a syllable with stressed /ɛ/ or /ɔ/ precedes another syllable with unstressed [a] or [e]; e.g. dona/es [ˈdɔnɔ(s)]~[ˈdɔnɛ(s)] and terra/es [ˈtɛrɛ(s)]~[ˈtɛrɔ(s)].
  • In some Valencian and North-Western Catalan dialects there are some instances where unstressed ‹e› and ‹o› may be reduced:
  • Unstressed ‹e› merges with [a] before a nasal or sibilant consonant (e.g. enveja, espill, eixugar), in monosyllabic clitics, and in some environments before any consonant (e.g. terròs, clevill, trepitjar). Likewise, unstressed ‹e› merges into [i], in lexical derivation with -eixement/-aixement (e.g. coneixement). In some subvarieties /e/ is raised to [i] in all instances when in contact with palatal consonants; e.g. senyor [siˈɲo(ɾ)].
  • Unstressed ‹o› merges with [u] before a bilabial consonant (e.g. cobert), when it precedes a stressed syllable with a high vowel (e.g. conill), in contact with palatal consonants (e.g. Josep), and in monosyllabic clitics. Lleida
  • In North-Western Catalan word final unstress ‹a› and ‹e› may reduce to [ɛ]; e.g. Lleida [ˈʎejðɛ], dona/es [ˈdɔnɛ(s)]; terra/es [ˈtɛrɛ(s)].
  • In most of Valencian, preposition amb merges with en. Also, some verbal forms ending in unstressed ‹a› are pronounced as [e] (e.g. verbs in third person singular: cantava, cantaria, canta, thus in Valencian first person singular (jo) cantava [kanˈtava] contrasts with third person singular (ell) cantava [kanˈtave], a similar process also occurs with gender neutral words; e.g. artista (m.) [aɾˈtiste] and artista (f.) [aɾˈtista]).
  • Marginal vowels are found in Northern Catalan in loanwords and interferences from Occitan and French.
  • Sonorants /m n l ʎ/ (except /ɲ/) are the most commonly geminated consonants (e.g. setmana, cotna).
  • /ʎː/ does not occur in Valencian and Balearic Catalan (e.g. motlle/motle). In north-western and eastern varieties plosives /b/ and /g/ may also be geminated in certain environments, instead of usual lenition (e.g. poble [ˈpɔbːɫə], regla [ˈregːɫə]). Moreover affricates are particularly long in intervocalic position in many dialects; e.g. metge [ˈmeddʒə].
  • Marginal consonants are found in loanwords, largely from Spanish but also from Caló, Arabic, Hebrew, Greek, English, German, etc.
I agree that notes are still excessive and redundant. But, I do not agree all this info should be (re)moved from here, like [ʃ] (esp. eastern dial.) vs. [jʃ] (esp. western dial.), silent plosives (Catalonia), palatal consonants (Balearic), and geminate consonants.
As you say, this information guide is designed for readers wanting to understand the IPA transcriptions of Catalan, i.e it is intended to help users to transcribe Catalan. How could (some of) these movements/removals help users to transcribe Balearic? Remember, we don't transcribe in just one variant. If someone wants to transcribe the name of a Majorcan place, person, etc., like Capdepera [ˌcapdəˈpeɾə ~ ˌcaddəˈpeɾə] or Berenguer d'Anoia [bəɾəɲˈɟe ðəˈnɔjə], won't know c and g are palatal in those words without going through other articles. — Jɑυмe (xarrades) 15:01, 29 December 2011 (UTC)
Categorizing footnotes
  • Balearic palatal consonants
  • Balearic stressed /ə/
  • [ʃ] vs. [jʃ]
  • Geminate consonants
  • Dark l
  • Vowel shifts
  • Valencian vowel harmony
  • Rare vowel reductions in Western Catalan
  • Synaloephas, etc.
We're transcribing for Balearic, too? That's not apparent from the guide. It looks like you've re-added notes about:
  • Elision of plosives
    • This is more relevant to editors, not readers. The addendum that some dialects extend this environment is irrelevant if we aren't transcribing it that way.
  • Balearic dorsals
    • This is only relevant if we are indeed encoding for Balearic varieties, which I'm not sure is a good idea. Can you point to a place where you think this is necessary?
  • ‹ix›
    • This is only important to editors and only if the variation is reflected in our transcription practices.
  • dark l
    • The chart itself expresses with the examples the way we transcribe l. That should be good enough.
  • the source of marginal sounds
    • In addition to being kind of obvious, it's not that important to readers.
  • gemination
    • This is problematic in two ways, one is that labeling something "gemination" clarifies most of what this note explains. The other is that it validates two different methods of transcribing gemination. We should pick one method and go with it. Are we doubling the letter or using the length mark?
I do want to admit that I'm being more strict on the notes than I otherwise would be because there are already so many of them.
The difference between things relevant to editors and readers may prompt us to consider either putting editor-centric information to the talk page (as has been done at WP:IPA for Russian) or putting it in a collapsible table. — Ƶ§œš¹ [ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ] 00:12, 30 December 2011 (UTC)


How do you pronounce David Pujadas in Catalan? — Preceding unsigned comment added by (talk) 08:50, 5 May 2012 (UTC)

[dəˈβit puˈʒaðəs] (in standard Central Catalan). Jɑυмe (xarrades) 21:13, 28 September 2012 (UTC)

[ɲ] in <<punxa>> and <<menjar>>?[edit]

Is this true? In my university class in Barcelona we were told it's ['puṉʃə] and [məṉʒ'a], and one student was chastised for constantly asking it could be [ɲ]. We were also told that /l/ is [ḻ] and not [ʎ] before the palatal fricatves. Is this incorrect? This page from the UB claims the same: (; look under Taula de sons (Sound table) --> Central --> Endarrerit (Moved back). saɪm duʃan Talk|Contribs 18:32, 17 April 2013 (UTC)

Yes it is. It's a very similar sound (if not the same) to the n and l in the Spanish words concha and colcha. These types of consonants don't have specific symbols, and the closest ones are [ɲ] and [ʎ], that's why they're used here. Jaume IPA-va:/ˈame/ (xarrem?) 23:24, 4 September 2014 (UTC)