Help talk:IPA/Spanish/Archive 1

Archive 1

Archive 2

Archive 3

→

Archive 5

Untitled

Orignal Source : es:Transcripción fonética del español con el IPA

Please tell me if this page is worth translating. I translated the first few lines quickly, and im not sure if this page is needed. I looked at IPA and this seems to be the spanish equivalent page.

It was requested that this page be translated at : Wikipedia:Translation_into_English/Spanish#International_Phonetic_Alphabet_for_Spanish

Please give me info on if this page is useful, if it is'nt I can stop translating and save myself the effort of doing the whole page.

Bearingbreaker92 03:16, 13 November 2006 (UTC)

This page is useful, but right now it favors one dialect (not sure which). I think that it can mimic IPA chart for English in its presentation of alternate dialects. I think some of the examples would be incorrect for Castilian Spanish (like the uvular fricative in hijo). Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 19:56, 3 January 2008 (UTC)

Sound files

This page originally had sound files of lists of words, which made it difficult to add or remove words from the list. I've removed the sound files and revamped the page. I think newer sound files would be very helpful though, and if a Standard Spanish speaker would like to contribute, I think it's best to make sound files of individual words rather than lists of words. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 22:16, 5 March 2008 (UTC)

New world Spanish pronunciation

This all started with the article Mexico, which has a sound file, and an IPA entry. Those two contradict each other in that the IPA entry suggests a spelling of ['mexiko], while the sound file makes it sound more like ['mehiko] or ['meçiko]. After some digging around, I found this page: Toponymy of Mexico, which states: "In present-day Spanish, México is pronounced ['mexiko] or ['mehiko], the latter pronunciation used mostly in dialects of the Caribbean and some places in South America where /x/ has become a voiceless glottal fricative (/h/)." Now, while the sound does not sound like a pure [h] to my ears, I'm by no means an expert. However, it seems that neither [h] nor [ç] are mentioned in the IPA table for Spanish. Is this an oversight, or a deliberate decision? —Preceding unsigned comment added by 134.130.57.186 (talk) 20:52, 30 June 2008 (UTC)

While we're trying to have a pan-dialectal representation of Spanish, depending on circumstances, a dialect-specific transcription (i.e. one that differs from this help page) is acceptable. In Mexican Spanish, /x/ is [h]. Before /i/ this is probably a voiceless [j] which sounds a lot like [ç]. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 21:05, 30 June 2008 (UTC)

Which leads me to the the question: should I then still use the IPAes template, or just a generic IPA template? I can't really decide. It's Spanish alright, but it contains a letter that is not listed on this page (that the IPAes template links to). So the generic IPA template might be more applicable. I just don't know whether there's any precedent that following would be wise. --134.130.57.186 (talk) 00:33, 1 July 2008 (UTC)

Probably the best thing to do is use the IPA-es template when this help page can address all sounds represented so that a reader who knows nothing of Spanish phonology or of the IPA can click on the link and, with some careful study, understand what is being represented on a particular page. Otherwise, I suppose using {{IPA-all}} might be a better idea. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 04:30, 1 July 2008 (UTC)

OK, I went ahead and edited the page accordingly. Thanks for your input! I think the procedure of when to use which template is a lot clearer to me now. --Flosch (talk) 11:58, 1 July 2008 (UTC)

[ia] or [ja]

This Spanish [i] is a weak vowel and [a] is a strong vowel, therefore 'ia' is pronounced [ia] not [ja] (I believe) Xorothal (talk) 13:02, 1 July 2008 (UTC)

While ia is phonetically [i̯a], we are using [ja] for a number of reasons, the most compelling of which is that it helps distinguish between [i̯a], and [ˈi.a] (as well as [ʝ̞a]}}. As far as I know, the i of ia is no different than that of io or ie. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 18:47, 1 July 2008 (UTC)

Reloj

Is reloj really pronounced /reˈlox/? I thought the j was silent in this word and it was pronounced as if spelled reló. —Angr 18:14, 7 April 2009 (UTC)

Perhaps in less formal varieties or in rapid speech. I've never heard of speakers deleting final /x/, though. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 18:46, 7 April 2009 (UTC)

Hmm. I thought it was a phonological rule of Spanish that only coronal consonants are allowed to end a word. If reloj is /reˈlox/, it would be the only Spanish word I know of (not counting proper names and recent loanwords like esnob) that ends in a noncoronal consonant. —Angr 20:37, 7 April 2009 (UTC)

In fact, Robert M. Hammond in The Sounds of Spanish: Analysis and Application (Cascadilla Press, ISBN 1-57473-018-5) writes (p. 226), "word-final /h/ is systematically deleted from this word [viz. reloj] in unaffected speech". (He uses /h/ as a cover symbol for all the /x/ - /χ/ - /h/ varieties this sound has across Spanish accents.) —Angr 20:45, 7 April 2009 (UTC)

J. I. Hualde says reloj is pronounced /re'lo/ "by many speakers" (thus implying that many others do say /re'lox/), and points also to a few rarer words like boj and carcaj. —Angr 20:51, 7 April 2009 (UTC)

Ahh, nice sources. I think for our purposes as a transcription guide that we should include final /x/ even if it's deleted by some/many speakers. If relojes is more consistently pronounced with an [x], then there's no doubt that it's present underlyingly. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 21:01, 7 April 2009 (UTC)

Yeah, although it might be worth adding a footnote pointing out that the /x/ is often not pronounced in this word. —Angr 21:20, 7 April 2009 (UTC)

I think it's worth mentioning that final 'j' is indeed pronounced in some Spanish vernaculars, but the number of speakers who delete the sound is probably quite larger. I've never found any source about which varieties delete the sound and which ones don't. --Jotamar (talk) 16:57, 14 April 2009 (UTC)

What we've got here is informative enough. Greater detail should go at Spanish phonology. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 20:57, 14 April 2009 (UTC)

Diphthongs

I propose that the pairings listed as diphthongs (ja, je, jo, ju, and wa, we, wi, wo) are not actually diphthongs and should not be listed here. Diphthongs are a pairing of two vowels acting as vowels. i and u are semi-vowels and, in these cases, are acting as consonants--not vowels--and therefore do not constitute a diphthong, but rather a syllable of a consonant and a vowel. —Preceding unsigned comment added by Antigravity711 (talk • contribs) 18:13, 17 June 2009 (UTC)

While that's true for French and Romanian, sources that I've found say otherwise for Spanish. If it were the case that the high vowels act completely outside of the syllable nucleus in these instances, then exceptions to their formation would include complex onsets. In From hiatus to diphthong by Chitoran and Hualde (2007) it is said of Spanish that exceptions to diphthongization (and therefore occurrences of hiatus) happen across morpheme boundaries (boquiancho) or if the word is "paradigmatically related to another word where the stress falls on the high vowel" (liamos). There's also a tendency for hiatus when it would otherwise create a diphthong in the first syllable of a word (biólogo) and a "preference for hiatus in stressed or pretonic iV sequences" (diálogo, dialogo). There is no mention of complex onsets as being a factor in barring diphthongization. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 19:40, 17 June 2009 (UTC)

arrangement

Per the recent reverts, which philosophy should we follow for the arrangement of the non-Latin symbols, such as ɣ? Anyone who knows that ɡ~ɣ are allophones probably won't be coming to this page anyway; in fact, I doubt that many people who know that ɣ is a velar fric would come here. Wouldn't it make sense therefore to order it with graphically similar letters like v and y? kwami (talk) 21:15, 23 June 2009 (UTC)

I don't think you have to know that ɣ is an allophone of g or that it's a velar fricative. You just have to know it's a gamma, and probably anyone who took advanced mathematics or physics, or was a member of a fraternity or sorority, knows that. And once you know it's a gamma, you're probably going to look for it near "g". (I'd also put θ after t rather than treat it as a kind of o.) +Angr 21:49, 23 June 2009 (UTC)

I'm not adverse to having it in both places. We might get some people removing the duplicates at first, but it would be to help both readers familiar and unfamiliar with the non-Latin characters, not avoid edit conflicts. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 22:34, 23 June 2009 (UTC)

inconsistancies

I've been redirecting 'IPA2' and 'pronunciation' templates to IPA-es, and there are several very common inconsistencies that AWB could replace automatically: downtacks under the frics and mid vowels, and ɰ for gamma. Also, a lot of articles on American places lack θ and ʎ, but that would require a more considered review, as well as agreement as to when we should have an American transcription. (In Argentinian articles some have ʃ for /x/, but maybe those should stay?) kwami (talk) 07:51, 24 June 2009 (UTC)

Removing undertacks and ɰ automatically sounds sensible, though we may need to take θ and ʎ on a case by case basis and, if need be, formulate a more abstract policy on when to have an American transcription. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 08:37, 24 June 2009 (UTC)

ñ versus "canyon"

English speakers tend to make the mistake of equating "ñ" with "ny". This article acts as if they are identical. The problem is that the sound simply doesn't exist in English. Any ideas here? —Preceding unsigned comment added by Mavasher (talk • contribs) 18:44, 29 June 2009 (UTC)

What if, instead of "English equivalent" the column is labeled "English approximation" — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 19:17, 29 June 2009 (UTC)

Yes, better. kwami (talk) 21:06, 13 July 2009 (UTC)

Vowel sounds

The English examples for the closest vowel sounds seem quite far off to me, particularly for "a". I would says that the "a" in "tanque" is much closer to the "ɑ" in "father" or the "ʌ" in "gun" than it is to the "æ" in "tank", except in the Texan dialect.Gregcaletta (talk) 08:34, 4 August 2009 (UTC)

Which dialect do you speak? Tank is very close in RP, and pretty close in GA. kwami (talk) 10:13, 4 August 2009 (UTC)

The vowel of tank is [æ], and the vowel of gun is [ʌ~ɐ] and the vowel of father is [ɑ]. [ɐ] is probably closest to Spanish [ä] as far as formant values go, though my ears tell me that [ɑ] is closest and [æ] comes a close second. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 20:20, 4 August 2009 (UTC)

My natural accent is Australian, but I know the English accent well. I picked these examples because I believe the "a" sound is pronounced almost the same in the UK and the USA, except in the mid-west and south where they tend to draw out the vowels in "gun" and "tank". I don't know what RP or GA means. Is there an IPA page which has recorded sound files for each of the sounds? That would be really useful.Gregcaletta (talk) 02:39, 14 August 2009 (UTC)

I'm also going to add some alternative examples. You can remove them if you like, but please give some kind of an argument. It's a tricky issue because it's difficult for us to "discuss" what our ears are hearing, though I think I have consensus with Aeusoes1, that "father" and "gun" are quite good examples.Gregcaletta (talk) 02:42, 14 August 2009 (UTC)

Let's discuss it here first. Spanish [a] may be close to *Australian* u, but not British u. ("RP" is "Received Pronunciation" -- BBC English, and "GA" is "General American" -- US news broadcast English). In Strine, gun is [gan], but in both RP and GA it is [gʌn], which is quite distinct, no closer than Strine hot. In RP, it is bran which has the [a] vowel. In GA, nothing is terribly close; [a] is about equidistant between bran and bra.

Strine hot is not a bad approximation of Spanish [o] -- though thought is better -- but again, hot is not at all a good example for RP. And again, there is no close equivalent in GA. kwami (talk) 04:59, 14 August 2009 (UTC)

I think the Spanish "a" is actually pronounced anywhere between [ɑ] and [a]. "Gun" in my opinion is pronounced [gɐn] in most dialects, although [ʌ] is often used as an approximation. I would say it is pronounced to rhyme with "nut", which is pronounced with [ɐ] in RP and Californian (http://en.wikipedia.org/wiki/Near-open_central_vowel), and I believe that is how I too pronounce it; I have lived in England (and Swine is a made-up language). Ƶ§œš¹ asserted that that [ɐ] is the closest approximation to the spanish "a" in "tanque", and I agree

It may perhaps be the closest in formant space, but the [a] of Mexican Spanish certainly doesn't sound at all close to the [ɐ] of California English. kwami (talk) 09:55, 24 August 2009 (UTC)

Well, that opens up a new debate on whether we are talking about Mexican Spanish or Continental Spanish, although I think you would agree that we should just use approximations which cover both dialects. We seem to have an a posteriori disagreement here, which can't really be resolved by discussion, so as a compromise, why not just include both your examples and mine?Gregcaletta (talk) 02:48, 1 September 2009 (UTC)

We could always refer to the works of other authors. I've seen father used as an example in instructional texts. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 03:00, 1 September 2009 (UTC)

I'd definitely vote for "father" over "tank". It's a pity we can use Australian English because father is actually pronounced with [a] in Aussie instead of [ɑ] and bed is pronounced with [e] instead of [ɛ]. I still thing "bed" is better than "pay" for "e" because it doesn't involve a diphthong.Gregcaletta (talk) 04:09, 1 September 2009 (UTC)

I'd say bed is better not just for Ozzie English, but British as well. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 06:09, 1 September 2009 (UTC)

Agreed. -- Gregcaletta (talk) 07:04, 1 September 2009 (UTC)

2ary stress?

Some articles mark secondary stress, us. in alternating syllables ahead of the primary, but s.t. in initial syllables. Is this s.t. we should formalize? kwami (talk) 08:52, 7 September 2009 (UTC)

My understanding is that secondary stress occurs with cliticization as in miralo /ˈmiɾaˌlo/ and guardándoselos /garˈdandoseˌlos/. Otherwise, AFAIK, secondary stress in the way you describe is incorrect. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 20:14, 7 September 2009 (UTC)

Okay, I'll start removing them then. (If I don't forget.) kwami (talk) 20:29, 7 September 2009 (UTC)

This article should add more sounds

This article is so wrong at the moment. http://es.wikipedia.org/wiki/Transcripci%C3%B3n_fon%C3%A9tica_del_espa%C3%B1ol_con_el_AFI Phonetic transcription for the Spanish language with IPA (article in Spanish)

Other symbols used in transcription of Spanish pronunciation

• 1st. Regarding stress, add a second stress for adverbs ending in -mente plus other long words.
• 2nd. Add also (~) for vowel nasalisation (PHONETIC vowel nasalisation).

IPA is used for phonetics, and this article doesn't show the Spanish phonetics, and what many people do study on the Spanish universities. Studies as Filología Hispánica, Hispanic studies goes really in-depth the Spanish pronunciation which is different from what you get on here.

Here you can only see the Spanish allophones ð, β, ɣ and z. Even, ð, β, and ɣ are fake, they are not represented with the undertack [β̞], [ð̞], [ɣ˕]. (I know you clarify it after). However there are many more allophones than just those, pronounced by the Spanish speakers.

Other allophones as [β̞], [ð̞], [ɣ˕] in Castilian Spanish.

• [χ]
• [ɰ] rather than [ɣ˕]
• [ʤ] emphatic allophone
• [ɟʝ]
• [ð] (different from [ð̞])
• [v]
• [z]
• [N] archiphoneme ([n], [m], [n̪], [n̟], [nʲ], [ɲ], [ŋ], [ɴ])
• [L] archiphoneme ([l], [l̪], [lʲ])

Special features in southern Spanish dialects as Andalusian and Canarian Spanish, Caribbean Spanish (Venezuela, Panama, Costa Rica, Puerto Rico, Cuba, Dominican Republic, coast of Colombia) and Southern Cone Spanish (Argentina, Uruguay, Chile)

• [h]
• [vowel opening]
• [ʒ] and [ʃ] in Southern Cone Spanish.

149.254.49.34 (talk) 23:39, 14 February 2010 (UTC)

This isn't an article, it's a guide to pronunciation for people who don't speak Spanish (and a guide to adding IPA transcriptions for people who do). The kind of detailed information that you want to add belongs at Spanish phonology. Lfh (talk) 16:49, 23 February 2010 (UTC)

However on the Wikipedia:IPA for English article you mention the R-coloured vowels which does not exist (only when following a vowel) in the South of England as in fewer other English-speaking areas. Even the English "approximant r" [ɹ] is turned into [r] to ease one a better visual comprehension ;)

In Spanish there are differences regarding other consonants; "s", "ll"/"y", "c"/"z". Many dialects drop "s" before a consonant or at the coda positon, as would do English on certain dialects with an "r". Also, diphtongs should be indicated with a diacritic in Spanish because the vowels are pronounced quicker and at one sound, [ai̯], [au̯], [ei̯], [eu̯], [oi̯], [ou̯] and [ui̯], which is only found in muy, but can be pronounced as well as [wi].

In Spanish is mentioned, "i" and "u" can be either semivowels, [i̯] and [u̯], or semiconsonants, [j] and [w].

And it is totally wrong to affirm on this article the last "r" of alrededor is pronounced as a "trilled r". Even, the "r" on eterno is pronounced as a trilled on syntactic phonology, as would be pronounced the "f" as [v] on the word Afganistán.

With the current transcription, one should find very hard to understand a person from Andalusia, Argentina, Venezuela, Cuba, Bolivia, Colombia, Ecuador, Puerto Rico, El Salvador, Honduras, Panamá, etc.

On the Spanish language article, there is a coloured map with the main dialects. This article should be divided as those colours (main dialects & features). Northern Spain (Northern and most of Central Spain, including the Catalan/Valencian/Balearic area, Galician area, and Basque/Navarre area - Andalusian (Murcian, Canarian, Extremaduran Ceuta & Melilla) - Mexican & Central American (Mexico, Guatemala...) - Caribbean (Puerto Rico, Cuba, Venezuela, Panama, Coast of Colombia, Coast of Mexico, Dominican Republic) - Andean (inland Colombia, Peru, Bolivia, Ecuador) - Río de la Plata (Argentina, Uruguay...) - Chilean.

149.254.57.166 (talk) 16:33, 17 March 2010 (UTC)

That information is already at Spanish phonology. This is a rougher guide than perhaps you were expecting, but it's still accurate. This includes the distribution of the trill). — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 17:24, 17 March 2010 (UTC)

Yes, but as this guide mentions [β], [ð], [ɣ] (which should be [β̞], [ð̞], and [ɣ̞] or better [ɰ]) and [z] you should get other possible ways of pronunciation in Standard Spanish [v], [ð] (different from [ð̞]), [χ]...

Vowel phonetic nasality with (~), as double stress for long words, etc.

And even a dialectal differentiation, as some phonemes are pronounced different in the Spanish-speaking countries of the Americas than in Spain, Northern Spain and even Andalusian-Canary Islands, which is closer to the Caribbean Spanish. And the sibilant dropping should be also mentioned as possible for so many countries and areas.

Alrededor is pronounced [a̠lre̞ð̞e̞ð̞o̞ɾ], the last "r" is not a trill. And even eterno, is very doubtful to be pronounced with a trill, it is more accurate to be transcribed as [e̞t̪e̞ɾno̞].I've never heard a standard realisation of an "r" at the coda position pronounced as a trill, It may be common in some Caribbean and Andalusian dialects, as they may pronounce the r at the coda position with a trill, [r]. While other Caribbean and Andalusian dialects may pronounce the "r" at the coda position as [l]. You need the diacritics for Spanish if you want an accurate transcription. Some Spanish vowels and consonants aren't the same as the standard IPA vowels & consonants. Therefore, [e̞], [a̠], [o̞], [t̪], [d̪], [ð̞], [β̞] and [ɰ].

It is not enough to be mentioned at the bottom of the page, some vowels and consonants needs to be represented with their respective diacritic on this chart so people can see it is not pronounced as would be in English or whichever language where they are alveolar plosives [t] and [d]. As it says, the Spanish [t̪] and [d̪] aren't alveolar plosives but dental plosives.

Jaume87 (talk) 16:32, 18 March 2010 (UTC)

It's best that we stick to Castilian Spanish since other dialects' pronunciations are predictable with a Castilian transcription (that is, it makes all phonemic contrasts other dialects make, as well as a few others).

We had a discussion earlier (see above) where we decided not to use the diacritics since it's unnecessary detail, especially to English-speaking readers. If we incorporate all those sounds, it will be too unwieldy. Like I said, it's a rough guide and so there's going to be some glossing over of phonetic particularities. Everything you said is mentioned at Spanish phonology, including variation on pre-consonantal r. To make a case that we should change coda-r to a tap, you should find sources for Spanish phonology that detail usage (for example, if this were much more common). — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 22:30, 18 March 2010 (UTC)

The Spanish /e/ sound

I noticed that someone had chosen the word bed to represent the /e/ sound in Spanish, but that is only useful if you speak the Australian, New Zealand or South African English dialect, where the /e/ in bed sounds like the /e/ in ballet, and therefore is a good match for the Spanish /e/. Skol fir (talk) 19:27, 12 April 2010 (UTC)

Or Received Pronunciation. The Spanish vowel is a mid vowel, not close-mid. Comparing it to a diphthong/long vowel isn't as helpful as comparing it to the vowel of bed. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 20:17, 12 April 2010 (UTC)

I am afraid it's more complicated than that. :) See this quote from Pronouncing the Spanish E.

The simple vowel e can be rendered anywhere across a range of tongue heights, from roughly mid-low (or mid-open), resembling what you hear as 'por-KEH,' to mid-high (or mid-closed), resembling what you hear as 'por-KAY.' The key feature of the simple vowel e is that it is pronounced somewhere within that range of tongue height and that the tongue does not change height or shape during the course of pronouncing the vowel. Standard Spanish does not distinguish between words based on how open or closed the vowel e happens to be pronounced. You may hear a more open pronunciation more often in closed syllables (syllables that end in a consonant), and you may hear a more closed pronunciation more often in open syllables (syllables that end in a vowel).

For example, if you listen to a Spanish speaker carefully, you will see that the two 'e's in nueve are pronounced slightly differently, the first being closer to "eh", the second to "ay."

Therefore, it would be more correct to say the Spanish /e/ lies between jester and paste, not to artificially fix it to the /e/ in bed. I am apparently not the only one who wanted to represent the sound closer to pay or taste, as you can see from the history of this project page.

Skol fir (talk) 03:50, 13 April 2010 (UTC)

Are we really going to take phonetic advice from Gerald Erichsen, someone with no training in phonetics? He cites "Mim100" for that particular claim who is even less of an expert in Spanish. I doubt the variation is that significant. It's actually unhelpful to tell people that a certain vowel is pronounced between the pronunciations of other vowels.

The reason people want to say Spanish e is the vowel of English pay is because they use the same character in the IPA ([eɪ] and [e]). If we're going to pick one, I'd go with the vowel of bed since it's short, monophthongal, and no further away from the vowel height that the first element of pay is. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 04:34, 13 April 2010 (UTC)

Gerald Erichsen is not giving phonetic advice, so he does not need to be an expert. He is a guide for about.com, which is completely different. From his background you can see that he has a keen interest in Spanish, as he has personally studied the language and obviously knows a thing or two about it. The person who actually wrote the quote above was not "Mim100", but "wrholt" in reply to "mim100," and Gerald Erichsen seems to have a problem distinguishing the "from" and "to" parts of his forum threads. If you check on "wrholt" he is no slouch, and has a college degree with a minor in Spanish. However, again no expertise in phonetics. :)

I was not trying to convince you that these guys have all the answers. I just wanted to make you aware that it is not as simple as you apparently want it to be.

Vive la différence!

Skol fir (talk) 05:10, 13 April 2010 (UTC)

Aeusoes1 stated, "The reason people want to say Spanish e is the vowel of English pay is because they use the same character in the IPA ([eɪ] and [e])." That sounds to me like a weak justification. People who speak Spanish would not care about the IPA characters. They are basing their opinion on experience of how the /e/ actually sounds, not what the IPA character looks like!

Look at another worthwhile opinion—in the world of democracy we all should have the ability to voice our opinions so no one can dictate what others should think!. This website is from Laura K. Lawless, who is a qualified Spanish Teacher. In the section Vocales - Spanish Vowels we find:

letter: E, sound: [ay], English examples: they, hey, Spanish examples: que, tres, ella.

IMHO phonetic terminology does not tell a person how to actually pronounce the letter. That comes from practice and living in a Spanish-speaking country. I always remain open to other's opinions, as I might learn something by keeping an open mind.

Skol fir (talk) 06:28, 13 April 2010 (UTC)

At a Wikimedia Project called WikiBooks, the following line appears in the book entitled Spanish/Pronunciation:

"E ... Like e in ten, and the ay in say." As I said, I am not the only one who thinks that the "ay" sound is also a valid approximation, particularly for Spanish words ending in an "e," such as leche or nueve.

Skol fir (talk) 17:10, 13 April 2010 (UTC)

BTW, Aeusoes1, you helped make my case at the top of this section, when you said "The Spanish vowel is a mid vowel, not close-mid." So then it is not open-mid either, is it? It is precisely somewhere in between and that must be the sound of the /e/, that is hard to duplicate in English, although we have many different sounds for the "e".

Yet, leaning toward my interpretation of the Spanish /e/, we have the Wikipedia article Close-mid front unrounded vowel which shows in the lower table an example of the Spanish /e/, namely bebe, explaining in that section that "Many languages, such as Spanish, Japanese, Korean, Greek and Turkish, have a mid front unrounded vowel that is clearly distinct to speakers from both the close-mid and open-mid vowels. A number of dialects of English also have such a mid front vowel. However, since no language is known to distinguish all three, there is no separate IPA symbol for the mid vowel, and [e] is generally used."

Curioso!

Skol fir (talk) 18:59, 13 April 2010 (UTC)

(edit conflict)At the risk of sounding repetetive, I'll break it down how I see it:

1. Spanish /e/ is a mid vowel, one that stands between cardinal [e] (2) and cardinal [ɛ] (3). If we assume the two English mid vowels to differ only by vowel height and to be at the cardinal points, the choice would be arbitrary. However, neither is true for English dialects.
2. Dialectal variation is such that one can't make an accurate generalization about whether such a vowel is present in English
   1. Some accents, such as Scottish English and Northern English, the face vowel is basically cardinal [e] (though longer); this same vowel is more like [e̞ɪ] in southern England (including RP) and in GA, which begins with a vowel between cardinal 2 and 3 position.
   2. Some accents, such as GA and RP, pronounce the vowel of dress as [e̞] or [ɛ̝]. It is closer to cardinal [ɛ] in Scottish English and cardinal [e] in Australian English.
3. There are several reasons that the face vowel is a bad choice.
   1. It is long, unlike the Spanish vowel.
   2. It is often diphthongal, making it more like the vowel of peine (which, by the way, forms a minimal pair with pene)
   3. It varies considerably between some varieties, and can be very different from either cardinal 2 or 3; being [ɐɪ] in Cockney and [æɪ] in Australian English.
4. There are several reasons that the face dress vowel is a good choice
   1. It is short, like the Spanish vowel.
   2. It is often monophthongal, also like the Spanish vowel
   3. It varies little between dialects and is itself often somewhere between the two cardinal points

That's how I see it. Maybe others can weigh in here. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 19:41, 13 April 2010 (UTC)

Agreed, except I think you mean the dress vowel in point 4. — ˈzɪzɨvə (talk) 20:00, 13 April 2010 (UTC)

Oops! — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 21:17, 13 April 2010 (UTC)

If we can't find an English equivalent, then what about the French "é" or /e/ as in the word blé? To get closer to the mid vowel there is also the "ê" or /ɛː/ as in fête. As far as I know, that's a monophthong, and very close to the Spanish /e/ in mesa. Skol fir (talk) 22:39, 13 April 2010 (UTC)

Using examples in other languages is fine when there's no English equivalent but we clearly have them for the Spanish mid front vowel. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 23:07, 13 April 2010 (UTC)

At the Wikipedia article Spanish phonology there is a table for the Vowels which I will recreate here:

	Front	Central	Back
Close	i		u
Mid	e̞		o̞
Open		ä

I see that for the vowel /e/ they chose the [e] sound with a downtack diacritic [ ̞]. That is exactly what I was trying to say in my discussion above, that we should be closer to the [e] than the [ɛ]. However, in the interest of closing this discussion, I will go along with the choice you have (bed), but it might warrant a footnote explaining what you said above, that the "Spanish /e/ is a mid vowel, one that stands between cardinal [e] (2) and cardinal [ɛ] (3)." That would make me less uneasy about directing people to say mesa like "mess-ah". You see my point? Skol fir (talk) 18:27, 14 April 2010 (UTC)

Ƶ§œš¹, in this discussion thread I quoted three Wikipedia articles that support using the [e] sound to represent the Spanish /e/, namely Spanish_phonology, Close-mid_front_unrounded_vowel and Spanish/Pronunciation. Why is it that you insist on being different from all of these? Is is just fear of using the "ay" sound because it may be diphthongal? I understand that when only the phonetic symbol is given without having to provide an English example, it is easier to present as closer to [e], in fact as [e̞]. That is why I said that there is no true equivalent in English for the actual sound in Spanish.

Skol fir (talk) 19:39, 14 April 2010 (UTC)

You have, just as I said of others, assumed that Spanish e is more close than open because of the character that is being used. I've given my reasons above. The three articles you cite simply reinforce my point #1. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 20:09, 14 April 2010 (UTC)

So, as I understand it, even though you use the [e] symbol in the table, you mean to use the [ɛ], because that is the correct symbol for the word bed. Maybe that is what confused me all the time. However, by keeping the [e] in the table you actually split the difference, and allow people the choice of going closer to that sound.

I still think a footnote is required here (after the word "bed"), as I mentioned above. Would you agree to that? Skol fir (talk) 20:30, 14 April 2010 (UTC)

Sorry, I should've responded to your suggestion of a footnote before. I agree that a footnote may be in order. That's an interesting interpretation of my stance. I didn't think of it that way, but I suppose that may be how some interpret it. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 23:03, 14 April 2010 (UTC)

That's fine, no worries. I distracted you with another message after that suggestion about the footnote. If you want to put the following information into a footnote, go ahead. My suggestion is:

Footnote

"The Spanish /e/ is a mid vowel, one that stands between cardinal [e] (2) and cardinal [ɛ] (3). Spanish has a mid front unrounded vowel that is clearly distinct to speakers from both the close-mid and open-mid vowels. A number of dialects of English also have such a mid front vowel. However, since no language is known to distinguish all three, there is no separate IPA symbol for the mid vowel, and [e] is generally used."

How about that? Skol fir (talk) 03:08, 15 April 2010 (UTC)

I think a footnote should be tailored more for English speakers. Something like:

"The Spanish e doesn't quite line up with any English vowel, though the nearest equivalents are the vowel of pay (for most English dialects) and the vowel of bed; the Spanish vowel is usually articulated at a point between the two.

We have to be careful, though, that this sort of footnote parsing not be too common. The difference is really one of minor phonetic particularities and we really ought to have reduced verbage on these guides so users don't get overloaded with information. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 04:36, 15 April 2010 (UTC)

You wrote, "I think a footnote should be tailored more for English speakers." Exactly!...if someone was planning to travel to Spain for their holidays, and wanted to check in Wikipedia for pronunciation, it would be beneficial for them to be aware of this subtlety for the /e/, since it is one of the most common letters in Spanish. I know from experience that the Spaniards are more friendly and helpful if they see that you are trying your best to speak their language "correctamente," not speaking with a heavy American—or Canadian for that matter—accent.

We can go with your version of the footnote. I like it. I will put it into place beside bed.

Skol fir (talk) 14:45, 15 April 2010 (UTC)

 Done

You might also be interested in the frequency of the letter E in the Spanish version of the game "Scrabble." As found in Wikipedia at Scrabble letter distributions:

Spanish-language sets sold within North America (known as Scrabble - Edición en Español) use these 103 tiles:

• 2 blank tiles (scoring 0 points)
• 1 point: A ×11, E ×11, O ×8, S ×7, I ×6, U ×6, N ×5, L ×4, R ×4, T ×4
• 2 points: C ×4, D ×4, G ×2
• 3 points: M ×3, B ×3, P ×2
• 4 points: F ×2, H ×2, V ×2, Y ×1
• 6 points: J ×2
• 8 points: K ×1, LL ×1, Ñ ×1, Q ×1, RR ×1, W ×1, X ×1
• 10 points: Z ×1

You can see the importance of /e/. Skol fir (talk) 15:05, 15 April 2010 (UTC)

Voiced s and th?

I'm not an expert on phonetics, but a couple of rows in the table struck me as odd.

I'm a native speaker of Spanish (from Madrid), and am surprised by the entry on /z/. May I ask the source for the information that words such as "rasgo" and "mismo" are pronounced with a voiced s? In what regions does that hold? I can confidently say that all Spanish people I know always pronounce a voiceless /s/. Actually, most Spanish learners of English have trouble with the /z/ phoneme since it doesn't exist in our native tongue.

Also, is it really true that the "d" in "arder" or "ciudad" is akin to the "th" sound in "they"? While I concede they sound somewhat differently from the usual "d", it seems to me it's just a relaxed attempt at pronouncing /d/, and the tongue is always on the palate or maybe the upper teeth; whereas, if I'm not mistaken, the English sound in "they" is uttered with the tongue between the teeth, isn't it? (By the way, most Spanish speakers would also pronounce the English "they" as "dey").

On an unrelated note, the entry for /x/ shows the non-existent word "suje", maybe it was meant to be e.g. "sujetar"? —Preceding unsigned comment added by 192.16.196.218 (talk) 14:50, 19 April 2010 (UTC)

Someone else will be able to answer you more fully, but in the meantime, you might look at Spanish phonology#Consonants. That page (as well as a footnote here) explains that Spanish [ð] is an approximant, so it is different from the English voiced fricative "th". It also provides a source for the [z] claim, in ref 20. Lfh (talk) 21:12, 19 April 2010 (UTC)

Spanish phonology cites Martínez-Celdrán et al (2003) for the claim of voicing Spanish /s/ to [z] in Standard Castilian. The Spanish d of nada and the English th of another have phonetic differences, though they're the closest you're going to find to equivalents.

It's interesting, both of your concerns have to do with a lack of phonemic contrasts in Spanish that allow for allophony between sounds that English speakers contrast. Take /s/, for example. Although [z] is an allophone, its lack of contrastiveness in Spanish means that when a Spanish speaker hears [z] in mismo, they perceive it as the same sound as the [s] in siempre. The specific context where it gains voicing is next to a voiced consonant, which is common crosslinguistically.

With the d of nada it is (in the minds of Spanish speakers) exactly as you say, a relaxed pronunciation. The process is called lenition and results (as this page describes) pronunciations of b d g where the articulators approach each other but don't touch. Because this happens mostly between vowels, Spanish learners of English may still have difficulty with the /ð/ of English they since that word begins with a sound they normally have between vowels. They may also have trouble pronouncing a non-relaxed [d] between vowels. Non-native pronunciations of English#Spanish describes this with some relevent citations. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 03:58, 20 April 2010 (UTC)

Interesting. But I still disagree this is the standard pronunciation or the most common one. I'd be surprised to learn Martínez-Celdrán et al.'s paper actually claims that.

Even though I'm Spanish I can tell a voiced /z/ apart from a voiceless /s/. Obviously I know what I may write here is not verifiable, but I'm saying "mismo" and "rasgar" right now and can hear a voiceless s distinctly, and am sure many others say them like that too. If I try to speak fast, a /z/ sound does seem to sleep in occasionally, but not on a regular basis. So while it seems correct that in fast conversation a /z/ sound is sometimes made, I think it's misleading to claim this is the standard pronunciation as if it were a general rule rather than an unintended side effect that does *not* always happen. In particular a foreign learner of the language shouldn't strive to make this distinction because it's not "real". It would be similar to teaching students that the correct pronunciation of "you" is "ya": it's good for them to know in order to understand colloquial speech better, but it's neither the pronunciation as shown in the dictionary nor what one would hear from a native trying to speak clearly.

(On the other hand I guess it might depend on the region too. All three authors of Martínez-Celdrán paper are from the University of Barcelona, and the phoneme /z/ does exist in Catalan. Maybe the phenomenon is more widespread there. I'm just guessing here, maybe it's nonsense since I suppose the authors studied the speech of people of different backgrounds anyway, but I don't have access to the text). —Preceding unsigned comment added by 87.210.249.148 (talk) 21:18, 20 April 2010 (UTC)

More importantly, why do we have so many allophones [β, ð, ɣ, ŋ, z] listed here? Isn't the general practice on these WP:IPA for XX pages to only list phonemes? Though I think [ŋ] is listed in many cases where it's a mere allophone. I guess [β, ð, ɣ] make some sense, since they may be more common than their phonemes' "plain" allophones, [b, d, g]. Then what's the reason again for excluding [ɟʝ]? But if we list [z], we should also list [v], since they exist in the same environment. None of these are listed separately in the consonant chart at Spanish phonology. — ˈzɪzɨvə (talk) 22:03, 20 April 2010 (UTC)

Martínez-Celdrán et al are talking specifically about Castilian Spanish and they say, "...[s], becomes voiced before a voiced consonant. For instance, rasgo 'feature' [ˈrazɣo]..." Our article on Spanish phonology says there's some free variation in this regard.

As for Xyzzyva's issue, these IPA for X pages don't typically list only phonemes (which the table at Spanish phonology features), though in some cases it's pretty close (such as with French). What we've done here at IPA for Spanish is detailed above. I suppose it's fair to treat [z] and [v] the same, though I'm somewhat skeptical of the saliency of the latter; the source that backs it up isn't exactly a peer-reviewed journal. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 00:28, 21 April 2010 (UTC)

No one is denying the existence of the phenomenon. My point is that however common it might be, it does not always happen and is not the "standard" pronunciation, so why include it in the table?

The following extract, by T. Navarro Tomás (Manual de pronunciación española, CSIC, Madrid: 1982) deals with the similar phenomenon of voicing the letter "z" in words such as "hallazgo". While it does state that in colloquial speech the letter usually becomes voiced, it also says that a clear pronunciation of the word prevents this from happening:

"Sonorización de la z: Interdental fricativa sonora; ort. z, fon. /z/ (con un puntito debajo; en caracteres IPA: /ð/). En la conversación ordinaria, la z débil final de sílaba o de palabra, en contacto con una consonante sonora siguiente, se hace también sonora, resultando un sonido muy semejante al de la th inglesa en their, this. La pronunciación lenta, fuerte o enfática impide parcial o totalmente esta sonorización. Ejemplos: juzgar /xuðˈɣar/, hallazgo /aˈʎaðɣo/, etc."

Which renders

"Voicing of /θ/: voiced dental fricative. In ordinary conversation, a weak /θ/ at the end of a syllable or a word, immediately followed by a voiced consonant, becomes voiced as well, resulting in a sound very similar to that of English "th" in "their, this". A slow, strong or emphatic pronunciation precludes this voicing partially or completely. Examples: juzgar, hallazgo, etc."

Also, the following is my translation of a passage of Máximo Torreblanca available at http://213.4.108.140/lengua/thesaurus/pdf/41/TH_41_123_102_0.pdf (page 5):

"In the notheast of the province of Toledo, almost all voiceless consonants of standard Spanish become voiced at the start of a syllable, a fact verified experimentally [...] Voicing depends on the way of pronunciation and the phonological environment. The faster or more relaxed the pronunciation, the higher the likelihood that /s/ and /θ/ have voiced realizations at the start of a syllable [...] Voicing occurs more frequently among young speakers than among older ones, which proves unequivocally that we are dealing with a modern phenomenon. [...]

[In Ávila and Cáceres], in colloquial (not slow or emphatic) speech, phonemes /s/ and /θ/ are voiced at the start of a syllable, in every word. "

(By the way, the fact that Martínez speaks explicitly about Castillian Spanish has nothing to do with whether the speakers they analyze are Catalan or not. I was just hypothesizing the phenomenon might be more widespread in Spanish speakers from Catalonia. But it's only a wild guess, we can just drop this particular issue). —Preceding unsigned comment added by 192.16.196.218 (talk) 15:48, 21 April 2010 (UTC)

I'm confused by your translation of Torreblanca. It looks like he's talking about a dialectal phenomenon of voicing voiceless consonants at the beginning of a syllable, which is not what we're talking about here. I'm not sure how to go about discovering how common this phenomenon is. This website with some isoglosses doesn't consider it important but that could just be because it's not a regional issue. I would think that, having done a bit of research to flesh out Spanish phonology that I would have found something that discusses this. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 01:48, 22 April 2010 (UTC)

It looks like you agree that the pairs /s, z/ and /d, ð/ do not form phonemic distinctions, but are situation dependent allophones. This is similar to pairs /t, tʰ/ and /p, pʰ/ in English. In the latter case no one oppposes using only one symbol to represent the whole phonemic range. So I propose to do the same in Spanish. Use only /s/ and /d/. −Woodstone (talk) 03:25, 22 April 2010 (UTC)

This isn't and IMHO shouldn't be a phonemic transcription. The lenition of voiced plosives is one of the more notable aspects of Spanish (and, more generally, Iberian) phonology. We would be remiss to omit them. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 04:11, 22 April 2010 (UTC)

The aspiration of leading unvoiced plosives is a chacteristic of English phonology. Yet we do not express it in our IPA renditions. Why treat Spansih differently? −Woodstone (talk) 04:33, 22 April 2010 (UTC)

English readers are familiar with English phonemes. They're not assumed to be familiar with Spanish phonemes. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 04:44, 22 April 2010 (UTC)

Treatment of [θ] in Southern Spain dialects

I think the current state of the article with regard to [θ] ("only in the Central and Northern Peninsular Spanish dialects; otherwise, like /s/ in other dialects") is an over-simplification of the reality, and this page should follow the description of the cases done in the Andalusian Spanish article, which is quite more accurate. Not mentioning "Distinción" —which happens on the most populated areas (cities and suburban areas) in Andalucía, Extremadura, Murcia, etc.— as well as "Ceceo" and pretending that only "Seseo" happens, is forgetting a representative number of speakers. 74.125.57.36 (talk)

What, specifically, is the wording you'd like to see? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 21:56, 28 April 2010 (UTC)

IPA conventions for [ɾ], [w]

What are our conventions for when to use [ɾ] and when [r] where they don't contrast? [ɾ] after plosives and [r] elsewhere? I'm reviewing all the transclusions of this template, but forgot to code for this. There are entries where [ɾ] is found finally, at Eduardo Duhalde for example, so this should probably be corrected.

Also, we don't give examples of [w] forming an onset of its own. How should names like Chihuahua be transcribed?

Also Néstor Kirchner. Are we okay with the [ʃ] there? If that is the local pronunciation, perhaps we could keep it but mark it 'local'.

And as for the voicing of s, f, θ, before voiced Cs, are [ɾ, r] exceptions? Or exceptions when initial, as in "frío"? — kwami (talk) 07:23, 3 May 2010 (UTC)

Spanish phonology#Alternations describes the distribution. The only lectal variation I know of is of a preconsonantal rhotic that is a trill in more formal Spanish and a tap in more casual speech; I've gone with the trill (and even gone through what uses this template to edit accordingly) but I'm sure this isn't consistently applied by other editors.

I would transcribe chihuahua as [tʃiˈwawa]. I'm pretty sure that [w] can be in the syllable onset, though this starts to get into a contrast between [w] and what our Spanish phonology article transcribes as [w̝].

AFAIK, [ʃ] for Castilian [tʃ] is not a dialectal feature of Argentina (that is mostly a feature of Western Mexico). Even if it were, since it's predictable, we should stick to Castilian. That said, if enough words can be shown to contain [ʃ] due to borrowing from Nahuatl or English, we ought to consider allowing for that in this guide. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 08:20, 3 May 2010 (UTC)

Okay. Still not clear on fricative voicing. I've never noticed voicing in say flor or isla. By "before voiced consonants", do they really mean "voiced obstruents"? (I'm assuming we should have a [z] in Aguas Buenas.)

Also, on s.t. like Ancient footprints of Acahualinca, since we don't list 2ary stress in the key, should we mark 1ary stress twice? — kwami (talk) 17:58, 3 May 2010 (UTC)

I don't know anything about secondary stress in Spanish other than when clitics are added (búscalo [ˈbuskaˌlo]). If I were going to transcribe Acahualinca, I'd have only one stressed syllable.

As far as voicing of fricatives, mismo is given as an example of a word with a fricative preceding a nasal, so it's not just obstruents. It could be that liquids are exceptions to this voicing phenomenon. Syllable boundaries may also be a factor. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 21:21, 3 May 2010 (UTC)

I'd consider nasals to be obstruents, but yeah, are approximants and rhotics exceptions? (Obviously [j] and [w] will be.) — kwami (talk) 23:04, 3 May 2010 (UTC)

According to the pdf doc that's used for one of the refs, ca. p 5, it says that the voicing only occurs in coda position. Also that /s/ is voiced before /r/, but also de-sibilized, so that it's [ɹ]. Doesn't mention /l/, so I think I'll code that as an exception for now, so that we don't have [izla] for isla. Also that /ʝ/ is [dʒ] after /n/ and /l/. Should we add that to our key? — kwami (talk) 23:19, 3 May 2010 (UTC)

Well, Spanish phonology says it's [ɟʝ], not [dʒ]. Ever since being convinced we need the voiced obstruents' approximant allophones, I have thought we should add the obstruent allophone for /ʝ/. So yes, I agree. But man, those things seem to almost always be approximants. Anyone aware of any experts treating the phonemes as /β, ð, ʝ, ɣ/, with the plosives as the allophones? — ˈzɪzɨvə (talk) 02:41, 4 May 2010 (UTC)

It all depends on dialect. [ɟʝ] is typical of many places, though [ʒ] and [ʃ] also occur in some places. The note already mentions this allophony but the guide doesn't encode it since Kwami and I figured it was too much detail.

I want to say Harris (1969) implies that the voiced obstruents are normally spirants that are fortified but I can't find it. He does say (p. 44) that "voiced continuants [β ð ɣ] become voiceless before voiceless obstruents and in final position" and that voiceless stops become voiced before voiced obstruents (including nasals), which I believe is something Kwami was wondering about. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 18:46, 4 May 2010 (UTC)

do we really not want local pronunciations?

I just changed the ll in Cumbia villera from local ʒ to Castillian ʎ. I'm wondering if that's really want we want to be doing. Spanish orthography is so phonemic that the IPA rarely adds anything for the reader who knows any Spanish (with the exception of Catalan, Galician etc. names), so it's primary value will be for those unacquainted with Spanish. But IMO such people will be better served by the local pronunciation; if they don't know which sound ll represents, they're not going to know when it's locally ʒ. They may also hear the local pronunciation, and be confused by a contradictory transcription. Any thoughts? — kwami (talk) 18:33, 3 May 2010 (UTC)

This is quite a can of worms. How phonetically accurate do we want to be? How local do we want to get? I'd hoped this Castillian transcription would effect a pan-dialectal transcription that could circumnavigate some difficult questions about when local is appropriate and when it's not. This also relates to our policy on English; if a pan-dialectal transcription isn't applied for Spanish, why do so with English? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 21:16, 3 May 2010 (UTC)

Because our readers know English, but not Spanish. That's why we have a phonetic rather than phonemic transcription. Perhaps we can have a 'locally X' in cases where the difference is particularly great. — kwami (talk) 23:16, 3 May 2010 (UTC)

I advise to explain better to avoid misunderstandings. Spanish tends to be a voiceless language, regarding consonants.

• There are speakers who pronounce /s/ as [z] (and /θ/ as [ð], only found in dialects from Northern Spain) before voiced consonants. Found in Castile and Leon, Spanish bilingual areas (excluding the Basque Country), dialects from Mexico, Colombia, Bolivia, Ecuador, Paraguay.
• There are also speakers who pronounce them voiceless (before voiced consonants), as could be in Castile-La Mancha, Extremadura, Madrid, plus in some dialects from Latin America.
• And there are many other speakers who drop /s/, /x/ (and /θ/ in some areas from Andalusia and Murcia) before all consonants, or speakers who pronounce them with an aspiration [h]; as in the Canary Islands, Panama, Venezuela, Puerto Rico, Cuba, coast of Colombia, some Argentinian dialects.

The voiced-voiceless-dropping/aspiration depends on the dialect; the area, the person, etc, and it should be mentioned.

Catalan bilingual people who speak Spanish get voiced all consonants before voiced consonants.

Myself, I am from the north of Alicante, and as the rest of the speakers from my area (Denia-Gandia) we pronounce [z] before voiced consonants and sometimes between consonants, we also distinguish /b/ and /v/, and this can be heard when speaking Spanish. 100 km south my area the major dialect is dropping /s/-/x/ before consonants with seseo, and further south, in Murcia and Almeria, the major dialect is dropping /s/-/x/-/θ/ before consonants, with distinction (/θ/ y /s/). Inland Valencia province mixes Valencian phonetics with Spanish phonetics, dialecto churro. Different phenomena happen in other dialects, etc. Such a shame the Royal Spanish Academy is so centralised, with no consideration about other pronunciations. However all of this have been proven in many Spanish universities, but one's got to do an in-depth study of the Spanish phonetics and phonology, as Spanish dialectology.Jaume87 (talk) 16:11, 4 May 2010 (UTC)

(Jaume, I've moved your comment down here since this seems to be where you wanted to put it)

I think including especially divergent local pronunciations is something we already allowed for, though not as they are linked to here and possibly in addition to the Castilian. Should we include sounds from such dialects in the guide? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 18:52, 4 May 2010 (UTC)

As a respect to all the Spanish speakers and dialects, the most widely used ways of pronunciation/major dialects should be mentioned. I don't think with the current IPA transcription for Spanish, anyone could get familiar with Andalusian, Extremadurian, Murcian, Canarian (southern dialects from Spain), native-Valencian/Catalan speakers whose pronunciation influences on their way of speaking Spanish, as Galician, or the many dialects spoken in Latin America. More than 400 million people speak Spanish, with some major dialects and plenty of subdialects, as some bilingual speakers. IPA chart for English dialects is so good, having all vowels alternation with the major English dialects. How would an English speaker who goes to Seville understand Spanish? Or whoever goes to Buenos Aires? In Buenos Aires there is zheísmo or sheísmo, /s/ aspiration + vowel lowering.

Spanish dialectal features:

• Where merges /ʝ/ and /ʎ/ into /ʝ/ (most of the Spanish dialects), yeísmo. Where it doesn't merge?, distinción . Mainly in rural areas of Spain from North to South, as Catalan and Galician bilingual speakers. In Andean Spanish, Bolivia, Paraguay, some dialects in Peru and Ecuador, as few in Colombia and rural areas in North-western Argentina. Where /ʝ/ and /ʎ/ merges into /ʎ/?, lleísmo. Some rural areas in Spain and Andean Spanish.
• Where merges /s/ and /θ/ into /s/ (most of the Spanish dialects), seseo. Where it doesn't merge?, distinción. Mainly in continental Spain, as very few speakers from Mexico, Central America and Peru. Where merges /s/ and /θ/ into /θ/?, ceceo. Few areas in Andalusia.
• Río de la Plata Spanish (Argentina and Uruguay) features /ʒ/ or /ʃ/ for the /ʝ/ and /ʎ/ merging. Zheísmo or Sheísmo. Yo [ʒo̞] or [ʃo̞].
• Mexican and Guatemalan Spanish includes /t͡ɬ/ and /t͡s/. Tlapalería [t͡ɬapale̞ˈɾia], náhuatl [ˈnawat͡ɬ], Coatzacoalcos [ko̯at͡saˈko̯alkos], xoloitzcuintle [ʃo̞lo̞i̯tsˈkwĩn̪t͡ɬe̞].
• /tʃ/ deaffrication occurs in some areas. In Seville, Cadiz, Chile, Northern Mexico. Different than /ʃ/ borrowing ; x for Catalan, Basque, Nahuatl, Mayan, Galician, Portuguese borrowings, as for Basque, Galician, and Catalan/Valencian places in Spain, Mayan, and Nahuatl in Mexico and Central America. And sh for English and Japanese.
• [x] and [χ] reduction to [h] in many dialects (mainly in Canarian and Caribbean Spanish and some Andalusian subdialects). Few dialects has kept the old aspiration of the Latin f-. Hartar [haɾˈt̪aɾ], heder [he̞ˈð̞e̞ɾ], halar [haˈlaɾ]. [h] from borrowings, hámster [ˈhãnst̪e̞ɾ]/[ˈxãnst̪e̞ɾ], holding [ˈho̞l̪d̪ĩn]/[ˈxo̞l̪d̪ĩn], hall [ho̞l]/[xo̞l], música house [ˈmusika hau̯s]/[ˈmusika xau̯s], Hiroshima [hiɾoˈʃima]/[χiɾo̞ˈʃima].
• In some areas of the land of Valencia and in the Balearic Islands occurs hypercorrection regarding /b/ and /v/, as such difference doesn't exist in standard Spanish. Also many bilingual people who speak Spanish, and individual speakers with the knowledge of this sound can be heard, from Spain to South America. Also some singers may use it when singing in Spanish.
• Voicing/Voiceless/Dropping-Aspiration-Velarisation-Gemination of final syllable /s/, (/θ/), (/x/).
  • Voicing, pronouncing /s/ and /θ/ as [z] and [ð] before voiced consonants. Where /s/ and /θ/ merges into /s/ would be only pronounced as [z] before voiced consonants. [ˈmizmo̞]. It always stays voiceless before voiceless consonants. Española [e̞spaˈɲo̞la].
  • Voiceless, pronouncing /s/ and /θ/ as /s/ and /θ/ before voiced consonants. Where /s/ and /θ/ merges into /s/ would be only pronounced as /s/ before voiced consonants, Mismo [ˈmismo̞]. And also stays voiceless before voiceless consonants. Española [e̞spaˈɲo̞la]
  • Aspiration, pronouncing /s/ as [ʰ] or [h] before any consonant. Mismo [ˈmi̞ʰmo̞], española [ɛʰpaˈɲo̞la].
  • Dropping, elision of /s/, (/θ/), and /x/ before any consonant. Mismo [ˈmĩ̞mo̞], española [ɛpaˈɲo̞la].
  • Gemination of the following consonant, /s/ (and /θ/) before any consonant. Mismo [ˈmĩ̞mmo̞], española [ɛppaˈɲo̞la].
  • Velarisation, /s/ before any consonant is turned into [x] or [χ]. Mismo [ˈmi̞xmo̞], española [ɛxpaˈɲo̞la] (Few dialects in Madrid and Castile-La Mancha).
• Andalusian, Murcian and Caribbean dialects may exchange /l/ <-> /ɾ/. Sartén [sal̪ˈt̪ẽ̞n]/[sal̪ˈt̪ẽ̞]. Él/El [e̞ɾ].
• Andalusian, Murcian and some Caribbean dialects may drop all consonants at the end. Comercial [ko̞me̞ɾˈθja]/[ko̞me̞ɾˈsja], comer [ko̞ˈmɛ], vez/ves [bɛ], ven [bẽ̞].
• Some speakers may pronounce [f] as [ɸ]. Fuera [ˈɸwe̞ɾa]. And in South America (from areas of Mexico and Central America, to Ecuador and Bolivia) many speakers pronounce [r] as [z], or [ʒ]. Perro [ˈpe̞zo̞].

Some features to add to Castilian Spanish:

• All allophones should be included if you include [ð̞], [β̞], [ɣ̞] (should be [ɰ]), [ŋ], [z].
  • [χ] in Spain. Gitano [χiˈt̪ano̞], jirafa [χiˈɾafa], junco [ˈχũŋko̞].
  • If you mention [ŋ], what about the other /N/ realisations as [n], [m], [ɱ], [n̪], [n̟], [nʲ], [ɲ], and [ɴ]. The Spanish /n/ and /m/ can be any of these previous sounds depending on the following consonant/vowel. Álbum [ˈalβ̞ũn], plural Álbumes [ˈalβ̞ume̞s]. Tándem [ˈt̪an̪d̪ẽ̞n], plural Tándems [ˈt̪ãn̪d̪ẽ̞ns]. Un loco [ũnˈlo̞ko̞], un poco [ũmˈpo̞ko̞], un coco [ũŋˈko̞ko̞], un foco [ũɱˈfo̞ko̞], un zoco [ũn̟ˈθo̞ko̞], un duque [ũn̪ˈd̪uke̞], un yunque [ũɲˈɟʝũŋke̞], un cheque [ũnʲˈt͡ʃe̞ke̞], un jeque [ũɴˈxe̞ke̞], un gnomo [ũnˈ(n)o̞mo̞]. You can find more examples as the previous ones in one single word.
  • [ɰ], neither [ɣ] nor [ɣ˕].
  • [ð] (only in Castilian) and [v]. Few cases. Same voicing as [z].
  • Emphatic [ɟʝ] and [d͡ʒ].
  • Vowel nasalisation [~] should be added, being more prominent in some dialects. Antes [ˈãn̪t̪e̞s], [ˈãŋt̪ɛʰ], [ˈãt̪ɛ].
  • Vowel lowering when aspirating, velarising, dropping /s/ before consonant, and /s/ elision with the gemination of the following consonant. (Dialectal)
  • Not only [e̞] is a mid vowel, but also [o̞]. Depending on the position they may be raised or lowered, some dictionaries mention this on their guide for Spanish pronunciation. The pronunciation guide for Arpitan and French on Wikipedia includes /o̞/. Spanish should have [e̞] and [o̞] instead of /e/ and /o/ on this guide.
  • Double stress on some long words, words end in -mente. Fácilmente [ˈfaθilˌmẽ̞n̪t̪e̞]/[ˈfasilˌmẽ̞n̪t̪e̞], sobresdrújula [ˌso̞β̞ɾe̞zˈð̞ɾuχula]/[ˌso̞β̞ɾe̞zˈð̞ɾuxula]/[ˌso̞β̞ɾe̞sˈð̞ɾuχula]/[ˌso̞β̞ɾe̞sˈð̞ɾuxula]/[ˌso̞β̞ɾɛʰˈð̞ɾuhula]/[ˌso̞β̞ɾɛˈð̞ɾuxula]/[ˌso̞β̞ɾɛð̞ˈð̞ɾuχula]/[ˌso̞β̞ɾɛxˈð̞ɾuχula].
  • On the French and Italian guide of pronunciation appear [w] and [j] on the consonant side, as for diphthongs. Spanish phonology mentions [w] and [j], and [u̯] and [i̯]. Saharaui [saxaˈɾawi]/[sahaˈɾawi], hueso [ˈwe̞so̞], deshuesar [de̞zw̝e̞ˈsar] and paranoia [paɾaˈno̞ja], emparanoiar [ẽ̞mpaɾano̞ˈjaɾ], meiosis [me̞ˈjo̞sis]. Note some dialects may add an epenthetic g, [(g)w]/[(ɰ)w], for words with hu.
  • Hiatus diphthongation, [o̞], [e̞], [a] -> [o̯], [e̯], [a̯] -> [j], [w] and elison.

Jaume87 (talk) 22:29, 10 May 2010 (UTC)

That's quite a bit of detail. More detail than I think most readers would care for if they're just trying to find the pronunciation of Spanish words for articles on Spanish topics. What do others think? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 06:02, 11 May 2010 (UTC)

In my opinion we should take the same approach as in English and try to use a pandialectic representation. Distinctions made in any of the major dialects should be kept. A set of rules should be defined how to derive a specific dialect from the general form. −Woodstone (talk) 06:14, 11 May 2010 (UTC)

Outside of NPOV concerns, that's pretty much what we've got right now by using Castilian. We've added [ʃ] already, which isn't found in Castilian AFAIK but is present in other dialects especially in loanwords; we could add [tɬ] and [ts] as sounds that may also appear in loanwords. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 19:11, 12 May 2010 (UTC)

Varieties of spoken Spanish should be recognised in IPA renditions. Although -in most cases- this can be done through the employ of a standard rendition (whatever the chosen standard may be) and some explicit set of rules to derive regional pronounciations, as Woodstone suggests, there are controversial exceptions: names of persons and/or locations. These should be rendered in the local variety by default. As seen in recent editions, it would be absurd to call Argentine businessman Mauricio Macri "Maurizio", or the Cataratas del Iguazú, "Iguazú" instead of local pronounciation "Iguasú". Although one and the same language, standard spoken Spanish do vary between regions: it is not a matter of "standard vs. dialect", but of diverse national/regional standards, all equally "correct", as even RAE aknowledges. Salut, --IANVS (talk | cont) 19:56, 12 May 2010 (UTC)

IMO local pronunciations are fine, as long as we mark them as local. Otherwise we imply that Spaniards pronounce z as s in some words. Which for all I know the might, but we should be explicit about it. And I can add an exception to AWB if the template has a 'local' tag; otherwise your changes will just get reverted the next time I make a pass over them. — kwami (talk) 20:10, 12 May 2010 (UTC)

Okay, but how local are we getting? Is it just variation in regards to mergers? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 21:03, 12 May 2010 (UTC)

AFAICT, Ianus has only changed theta to es. — kwami (talk) 21:50, 12 May 2010 (UTC)

Yes, but Jaume has offered that we do more than that. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 23:44, 12 May 2010 (UTC)

All transclusions checked

All transclusions will in a few minutes be in line AFAIK with this key at a low level of analysis. Not covered are substitutions of one phone for another that's listed here, such as transcribing z as [s] or ll as [ʝ]. I've caught a few dozen of those, but there must be many more.

Remaining issues: voicing of frics before l, ll, r, rr (isla, Afrancesado); /ʝ/ as [dʒ] after /n, l/.

Should /k/ be [g]/_C ? That would seem to be justified if we're voicing the fricatives.

I also separated vowels with a period when there's no stress mark between them, as in Maria.

Should I post the AWB code for others to use to keep this project in shape? I'm planning on moving on to other languages, and probably won't be back here. — kwami (talk) 23:41, 3 May 2010 (UTC)

Now checked for Latin American conflations for all articles with z, ce, ci, ll, y in their title, which should be the majority. — kwami (talk) 02:26, 12 May 2010 (UTC)

Questions

Should the i in Eliézer Alfonzo be transcribed [j] or [ʝ]? Should the b in subterraneo be [β] (in which case, what should the t be?), [b], or [p]? — kwami (talk) 01:30, 12 May 2010 (UTC)

Eliézer: [eˈljeθer]. Subterráneo: [suβteˈraneo]. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 02:24, 12 May 2010 (UTC)

Thanks. No devoicing then. — kwami (talk) 22:01, 12 May 2010 (UTC)

Technically, there is, but we aren't transcribing that right now, which further puts to question our practice of voicing fricatives before voiced obstruents. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 23:43, 12 May 2010 (UTC)

We and wi sounds

I'm not sure if it's valid, but the article is missing the we sound like in weather and wi like in wisdom in the diphthongs column which translate as güe as in nicaragüense or ambigüedad and güi as in güicoy or pingüino. This is usually forgotten but is correct in use as it is one of the ways that the letter "w" is replaced in the language.Frapoz (talk) 16:04, 12 May 2010 (UTC)

I think you just missed them when you were looking. They're there. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 17:09, 12 May 2010 (UTC)

Monophthong for the "a" sound in Spanish

This was also cross-posted on the following talk page:User talk:Kwamikagami... awaiting a reply for further action. Skol fir (talk) 21:02, 23 March 2010 (UTC)

Hello Kwami:

I have been researching the "a" sound in Spanish for about a week, and I thought I would improve your example of "father" with another example, namely "tar". The reason for this was simple. "Father" is a rather ambiguous example because in some regions of North America, it is pronounced more like "Fo-ther", where the "a" and "o" sound have merged. This is called the Father-bother merger. If you are a linguist, you may be aware of this. I just wanted to make sure that everyone who comes to the Wikipedia:IPA for Spanish has another example to follow that is not ambiguous and is often used as the only example for the "ah" sound in Spanish. Here is a case in point: [1], where you can see that the only example given is the word "car."

I think that the word "tar" is precisely the example you need to avoid ambiguity. I have no idea why you would so quickly dismiss it, unless you felt offended that anyone would question the word "father" as the sole example for this sound. I thought Wikipedia was supposed to be unbiased. It looks like some editors like to wield their "power" to strike out sensible additions without any good reason.

I did not see any mention of Spanish on your user page. Since when have you become an expert on this language? :) Just curious. Skol fir (talk) 19:04, 23 March 2010 (UTC)

If you read Father-bother merger more closely, you'll see that North American accents merge /ɑː/ with /ɒ/ into [ɑ(ː)], the father vowel. No one pronounces father with the /ɒ/ vowel, so inasmuch as /ɑː/ is a suitable approximation for Spanish /a/, father will work. That said, I think English speakers perceive /a/ as somewhere between /ɑː/ and /æ/, so maybe it wouldn't hurt to triangulate between them (such as "between father and fatter")? On a side note, I continually think it would be best to work out these English approximations at a central location, and not redebate them at each WP:IPA for XX page. WP:IPA would probably be the place. — ˈzɪzɨvə (talk) 21:29, 23 March 2010 (UTC)

I did read further in the internal link, and found the reference mentioning that in the Merriam-Webster pronunciation guide you will see the [ä],

"as in bother, cot, and, with most American speakers, father, cart. The symbol \ä\ represents the vowel of cot, cod, and the stressed vowel of collar in the speech of those who pronounce this vowel differently from the vowel in caught, cawed, and caller, represented by \o\. In U.S. speech \ä\ is pronounced with little or no rounding of the lips, and it is fairly long in duration, especially before voiced consonants. In southern England \ä\ is usually accompanied by some lip rounding and is relatively short in duration. The vowel \o\ generally has appreciable lip rounding. Some U.S. speakers (a perhaps growing minority) do not distinguish between cot--caught, cod--cawed, and collar--caller, usually because they lack or have less lip rounding in the words transcribed with \o\. Though the symbols \ä\ and \o\ are used throughout this dictionary to distinguish the members of the above pairs and similar words, the speakers who rhyme these pairs will automatically reproduce a sound that is consistent with their own speech. In words such as card and cart most U.S. speakers have a sequence of sounds that we transcribe as \är\. Most speakers who do not pronounce \r\ before another consonant or a pause, however, do not rhyme card with either cod or cawed and do not rhyme cart with either cot or caught. The pronunciation of card and cart by such speakers, although not shown in this dictionary, would be transcribed as \'k[a']d\ and \'k[a']t\. Speakers of r-dropping dialects will automatically substitute \[a']\ for the transcribed \är\. (See the sections on \[a']\ and \r\.)"

Interesting, isn't it? BTW I personally pronounce "godfather" as "gawd-fawther", and that means that for me, "father" is not the right example for the Spanish "a." This is my first foray into this discussion, so pardon me for putting it in the wrong talk page.

Skol fir (talk) 22:12, 23 March 2010 (UTC)

But presumably you have the same vowel in tar as you do in father, so the only thing that adds is the confusion caused by /r/. Meanwhile, other speakers have the Spanish vowel in father but AFAIK not in tar. So father is the better example for some speakers, equal to tar for others, but AFAIK tar is not better for any, only worse for some. I may be wrong though. kwami (talk) 22:56, 23 March 2010 (UTC)

Kwami, tar and father are not the same for me, because I always pronounce father as "faw-ther", while tar is more like your interpretation of father or papa. See the comment below about the "Canadian Shift."

Skol fir (talk) 04:47, 24 March 2010 (UTC)

Skol fir, I saw from your userpage that you are Canadian, and from further research, have discovered that [Western?] Canadians do in fact merge the vowels in question to [ɒ], as part of a process called the Canadian Shift. I was totally unaware of that. Additionally, such speaker have in the same process backed /æ/ to [a], so that for you the Spanish vowel in question would actually be pretty well approximated by fatter. Again, this highlights the fact that it can be difficult to get dialect-neutral English approximations for many foreign sounds. — ˈzɪzɨvə (talk) 23:02, 23 March 2010 (UTC)

It's interesting that you should mention that, Xyzzyva, since I just read about this "Canadian Shift" yesterday, before you pointed it out. I haven't had the time to really study that particular vowel change, but most likely it explains why I (as a Canadian) have adopted an "a" in father that resembles the [ɒ] sound. Thanks for drawing my attention to this fact, because now I am really motivated to look into it further. Skol fir (talk) 05:09, 24 March 2010 (UTC)

I found the following chart most useful for this discussion: IPA chart for English dialects. The diaphoneme [ɑ/ɒ] for Canadian English covers all of the following examples: father, wasp, law, while American English has the [ɑ] for father and wasp, and [ɔ] for law. I gather [spoken as gæ-ðer] that the [ɒ] sound as in law is the final destination of the "Canadian Shift." The Irish further distinguish the [ɑː] for father and the [ɑ] for wasp, with [ɔː] representing law. Who would have thought? :)

Skol fir (talk) 06:18, 24 March 2010 (UTC)

It's a miracle. We've helped someone! — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 06:57, 24 March 2010 (UTC)

"On a side note, I continually think it would be best to work out these English approximations at a central location, and not redebate them at each WP:IPA for XX page." - But having a central discussion wouldn't stop people redebating them locally as well. And is it a good idea to stop them? Sometimes, people with a good knowledge of the language in question challenge our English examples and replace them with something that looks less intuitive, but sounds better to their ears. See e.g. Wikipedia talk:IPA for Welsh, where the example for /ə/ was changed from "about" to "butter" on the advice of a native speaker. Lfh (talk) 08:57, 24 March 2010 (UTC)

The reason why we can't necessarily discuss these in a central location is because the vowels differ from language to language. Don't let the IPA fool you into thinking languages using the same character for a given sound have the same pronunciation. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 22:14, 24 March 2010 (UTC)

Going back to the original theme of this section, the Spanish [a] sound would be served very well by the word mahi-mahi as an English example in this article. I heard the pronunciation of this word at Forvo, the Pronunciation Dictionary and it is exactly what I expected for the Spanish [a] sound. Mahi-mahi, although derived from Hawaiian, is in the Merriam-Webster dictionary and I think everyone is familiar with it, and with its correct pronunciation. Skol fir (talk) 04:07, 25 March 2010 (UTC)

Then why not take "Hawaii" itself as example? −Woodstone (talk) 15:19, 25 March 2010 (UTC)

Because that has the diphthong /aɪ/. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 18:08, 25 March 2010 (UTC)

I've never heard of Mahi-mahi. (I'm British). Perhaps you can think of another word that has the same vowel for you? How do Canadians pronounce the interjection "ah", as in "ah, I see"? Lfh (talk) 19:25, 25 March 2010 (UTC)

Mahi-mahi is a type of whitefish served in America and elsewhere. Here is one review of the 'ah' sound, from "The 'ah' as in 'father' Sound". Rachel gives several examples, such as the words: father, collar, calm, and a sample sentence: 'The party at the bar was a mob scene.' This selection is also a good example of the "father-bother merger." I think that on the whole, N.Am. English speakers are not very familiar with the true 'ah' sound within a word, apart from the exclamation itself. That is the problem with trying to approximate it "in a word." Skol fir (talk) 20:33, 25 March 2010 (UTC)

For those who would still insist that father is the best example for the Spanish /a/, see this article section in Wikipedia: Father-bother merger. Clearly, father is not a good example if almost all of America pronounces it just like the "o" in bother or lot. This is far from the Spanish /a/ and should be avoided. Therefore, I proposed another example to avoid this mistake, especially for those speakers who are subject to the "father-bother merger." Skol fir (talk) 05:36, 25 March 2010 (UTC)

For most speakers with the merger, it is the closest vowel. Bother is pronounced like father, not the other way around. kwami (talk) 05:39, 25 March 2010 (UTC)

I'm sorry, Kwami. You are right about that. I seem to be fixated on my "Canadian Shift", which would put the merged sound closer to the "o" of bother, in fact bring it to the [ɒ] sound, which is a shift to the back. My reservation about using father should be limited to those who have this back shift from [ɑ(ː)] to [ɒ], which as you pointed out is actually a minority in NA. Let's see if someone has a second example for the "ah" sound besides father, so that the "back-merged" among us are better served.
Skol fir (talk) 20:12, 25 March 2010 (UTC)

I have found another example of the Spanish /a/ sound in English. It is the interjection ta-da, where both "a's" are pronounced as "ah". This could be the second word next to father, which would clarify the pronunciation for most English speakers and avoid the "father-bother merger." If no one objects, I could put it into the table. I am now afraid to do so on my own, as it seems that I have no authority here, while other editors apparently have a carte blanche to single-handedly dismiss any edits without knowing all the facts. :-) Skol fir (talk) 02:39, 28 March 2010 (UTC)

What about bra? A pity proper names are not usable because Obama would qualify. −Woodstone (talk) 03:47, 28 March 2010 (UTC)

Woodstone, the application of bra in this case is a good suggestion. I have seen it used as an example for the "ah" sound at the project Wikipedia:IPA for English, along with "palm, baht and father." I actually like the use of baht because there is no mistaking here what sound should be represented by "ah", although some might argue that this could be construed as a diphthong "ah" rather than as the monophthong "a." I still like ta-da because it is consistently pronounced with the "ah" sound, without being modified by either the "father-bother merger" or the "Canadian Shift," as discussed above. Just my two-bits worth.

As for Obama, I don't see any problem with using proper names, since I have seen them used in other phonetic tables at Wikipedia, and I know of no rules against using them. Then how about Dalai Lama?—His Holiness has two a sounds in his name. The one in "ai" does not count as a monophthong (as in Hawaii) and the final one is more like the [ə] sound. Skol fir (talk) 17:03, 28 March 2010 (UTC)

Agree with Skol fir - I don't see why proper names should be disallowed in principle. I believe Barack Obama has an entry in Longman's Pronunciation Dictionary. Lfh (talk) 10:17, 29 March 2010 (UTC)

At the time Democratic Presidential Nominee, Barack Obama once stated: "Yes, We Can!"—I mean we can add Barack Obama to the English pronunciation of the Spanish /a/ sound, can't we? See the correct pronunciation of this name at Barack Obama and also at How to say: Barack Obama.

Skol fir (talk) 20:34, 29 March 2010 (UTC)

For the Brits among us, we have the word "Yah [jɑː]n Brit informal an affected upper-class person." (from The Free Dictionary online). To me that sounds like a good choice for the Spanish /a/ sound. Skol fir (talk) 18:12, 28 March 2010 (UTC)

We should have words that people recognize. One minor correction about ta-da, a common pronunciation is with the first syllable unstressed. Bra is another good example. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 18:44, 28 March 2010 (UTC)

I concur and say "Ta-ˈda!" Skol fir (talk) 19:58, 28 March 2010 (UTC)

Oh my word - such confusion! Look: "Barack Obama" is a terrible example of how to pronounce anything, because his name is pronounced differently by different people in the USA, much less the rest of the English-speaking world (and we ARE the majority, by the way). I know that it isn't a perfect example but can't we just say that a Spanish "a" is like the "a" in hat? Then most people will get it nearly right. Apart from in New Zealand. Sorry kiwis. —Preceding unsigned comment added by 78.145.204.145 (talk) 21:39, 15 June 2010 (UTC)

It is? What is the variation of Barack Obama? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 05:33, 16 June 2010 (UTC)

/r/ and /ɾ/

I still don't get this guide. I've never heard before as standard a final /-r/. Following this guide, one should pronounce both r in the word alrededor as /r/. I can imagine the English speakers and whoever else who read this struggling to pronounced both r i, n alrededor as trilled r's. Following this rule, all final r's should be /r/. This is not right, all final r's are pronounced as /ɾ/ with some exceptions (Harris also says as /r/, most of the speakers pronounced them as /-ɾ/). Only in some Caribbean dialects are pronounced as /-r/, and in an emphatic pronunciation, same as an emphatic yo can be pronounced different. Words with Latin prefixes, subrogar [suβ.ro'ɰaɾ], subrayar [suβ.ra'ʝaɾ], abrogar [aβ.ro'ɰaɾ], alrededor [al.reðe'ðoɾ], enredo [ẽn'reðo]. Other words without prefixation, honrar [õn'raɾ], israelí [izrae'li], Sri Lanka [(e)zri 'lãnka].

Harris (1991) propone un conjunto de reglas que permiten generar los dos segmentos vibrantes de las representaciones fonéticas, simple /ɾ/ y múltiple /r/, a partir de un único segmento, vibrante simple /ɾ/, de la representación subyacente. Las derivaciones parten de la asignación previa de la estructura silábica. En los contextos donde no existe oposición, las reglas formalizan las realizaciones que hemos descrito, incluyendo la elisión de la vibrante simple ante múltiple:

La representación con /ɾ/ geminada tiene algunas ventajas. Es capaz de explicar hechos fonológicos generales: la realización como vibrante múltiple en el contexto intervocálico a través de reglas que se aplican regularmente. En los siguientes ejemplos mostramos aplicaciones regulares de las reglas que intervienen en la derivación /ɾ/ --> [r]: corre, alrededor, amor romántico.

(a) corre

        |

      /ɾɾ/

        |

     [ɾ - r]      (Regla 1)

        |

     [Ø - r]    (Regla 3)

(b) alrededor (Pay attention, not the final r)

        |

    /C - ɾ /

        |

    [C - r]    (Regla 1)

(c) amor romántico,

            |

         /ɾɾ/

            |

        [ɾ - r]     (Regla 2)

            |

        [Ø - r]     (Regla 3)

En opinión de Harris (1991: 89-90), su propuesta también permite explicar hechos morfofonológicos particulares de determinadas unidades léxicas; así, "la derivación de [r] intervocálica a partir de una geminada subyacente hace posible un tratamiento no ad hoc del tema irregular de futuro que[r]- del verbo cuyo infinitivo es querer. [...] Si derivamos la [r] de cualquier otra fuente —por ejemplo /r/— necesitamos una sustitución fonemática ni documentada ni motivada independientemente, o bien recurrir a un proceso idiosincrásico de suplencia".

• Amor romántico. Final r in amor merges with the trilled r. [a'moɾ ro'mãntiko]-> [a'mo.ro'mãntiko]
• Eterno. [e'teɾno] or [e'terno], r may be pronounced as trilled before nasals.
• Amor [a'moɾ]
• Hablar [a'βlaɾ]
• Ver [beɾ].
• Emphatic may be pronounced, [a'mor], [aβlar], [ber].

Can you change last r in alrededor as it is not trilled? Otherwise one will probably think all final r's are trilled in Spanish. Jaume87 (talk) 23:02, 5 June 2010 (UTC)

That is quite an explicit source, Jaume. For those of you who can't speak Spanish, this is (more or less) what the prose says:

Harris (1969) proposes a set of rules that generates two rhotic segments phonetically--tapped /ɾ/ and trilled /r/--from a single segment (a tapped /ɾ/) from the underlying representation. The derivations are based on the previous assignment of syllable structure. In contexts where there is no opposition, the rules formalize the realizations described above, including the deletion of a tap before a trill:

The representation with geminated /ɾ/ has some advantages. It's possible to explain general phonological facts: the realization as a trill in intervocalic position through rules that are applied regularly. The following examples show regular applications of the rules involved in deriving /ɾ/ -> [r]: corre, alrededor, amor romántico.

According to Harris (1991: 89-90), his proposal also helps explain morphophonological facts particular to certain lexical items: "the derivation of intervocalic [r] from an underlying geminate enables a non-ad hoc treatment of the irregular future tense conjugations of querer...If we derive the [r] from any other source, eg /r/ - we need either a phonemic substitution that is neither documented nor motivated independently, or an idiosyncratic process of substitution."

What we have established, fairly firmly, is that the Spanish rhotics are manifestations of a single underlying phoneme and the limited contrast between them is due to gemination. Everything else is contextual variants.

At dispute here is the realization of this rhotic when it's in the syllable coda. As our article on Spanish phonology states, there is some speaker variation about this. I've opted for the trill, since that seems to be closer to the standard, but Jaume says that this is hyperformal. Remember, it's not a significant contrast in Spanish. It would be nice to have more information about dialectal variation. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 01:51, 6 June 2010 (UTC)

Yes, the Spanish rhotics can manifest with only an r! For sure at the beginning of a word, r-. But also in the syllable coda, which depends on the sentence.

So, how would you transcribe verbs, which end in -r? with the trilled r?!

http://elies.rediris.es/elies4/Cap4.htm (In Spanish)

This is the original source in Spanish, you can check there is an underline below the first r in alreador (not under the last r!).

Please check point 4.3.9. Representación fonológica de las vibrantes.

Una palabra como amo[ɾ - r] se puede realizar con vibrante simple o múltiple final, pero cuando se le añade un sufijo flexivo (amores) o derivativo (amoroso, amorío) que sitúe este segmento en posición de ataque siempre se realiza como vibrante simple. Esto prueba que en la representación subyacente debe constar /ɾ/ y que el proceso fonológico es de refuerzo.

[amoɾ] vs. [amor]. As this source says, both r sounds can be possible in the syllable coda (right at the end of the word, as a following word would change the pronunciation). Most of the Spanish speakers pronounce it with the tapped r, [a'moɾ], [do'loɾ], [ko'loɾ]. If you say, amorrrr! (emphatic), it'd be [amor]. So, what i want to say, MOST of the Spanish speakers pronounce r in the syllable coda with a tapped r, but at the right end of the word when there isn't any other word (no following sound). Whereas in a sentence, I use dolor [do'loɾ] as I could use alrededor [alreðe'ðoɾ]; dolor reumático [do'lo(r) reu̯'matiko], dolor lumbar [do'lor lũm'baɾ], dolor abdominal [do'loɾ aβðomi'nal], [e'terno]. A word itself alrededor without another following word which determine the pronunciation of the last r, would be [alreðe'ðoɾ]. Note if you mention an "emphatic r" in the coda position, you should mention other emphatic consonants in Spanish.

Obstruents p, t, k, d, b, g, can be different as well. Emphatic/reduced consonants.

Either /ɾ/ or /r/. If you emphasize it, it'd be a rolled r, but also if the following word start with an r, or if the following word is any consonant or r (gemination rule). In English, r is always pronounced between vowels in all dialects on its harder rhotic way :D, however in Spanish between vowels it is always a tapped r.

I don't think it is hyperformal or even formal to roll an r in the syllable coda (right end of the word). Ex., comerr, beberr, dormirr. It'd sound too redundant/emphatic; a Caribbean/Andalusian/Argentinian subdialect, or a foreign speaker who doesn't difference between r/ɾ (could be a Russian speaker).

Long ago, rey, reina, words starting with r, as rey, reina were spelled with double r, rrey, rreyna. But never any word was spelled with a double r in the syllable coda; hablarr, dorrmirr, comedorr.

Why does the pronunciation of both r's in alrededor as trilled r's sounds more stantard to you? It is not.

Like I said above, it's my understanding that the trill in the coda is closer to the standard. You say that most speakers pronounce it as a tap, but I'd like to see some sourcing that backs this up. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 05:58, 8 June 2010 (UTC)

http://books.google.com/books?id=gXSveYBqSYUC&pg=PA39&lpg=PA39&dq=pronunciacion+r+final+absoluto&source=bl&ots=6zx5RUS-rM&sig=fvCMRZKDaDKM3KN-Wuqe2AS9EUc&hl=es&ei=C2MOTPjeJOeL4gazptypDA&sa=X&oi=book_result&ct=result&resnum=8&ved=0CCsQ6AEwBzgK#v=onepage&q=fonetica%20espa%C3%B1ola&f=false

Page 39.

Harris says amor can be either [amoɾ] or [amor], but this ambiguity clearly clarifies when there is a following word or with word derivations; [a'moɾes], [amo'ɾios], [amoɾ ĩn'nato] [a'mo(r) ro'mãntiko], [a'mor nup'θjal]. So, in the coda can be both sounds /ɾ/ and /r/, being /ɾ/ more common when there is no following derivation or words, [a'moɾ]. [a'mor] is also possible; emphatic, and depends on the speaker, and dialect.

Jaume87 (talk) 21:16, 8 June 2010 (UTC)

Yes, it depends on speaker. You've said above that you think we should represent this with the tap; your reasining is that most speakers pronounce it that way but I don't think you've established that as a fact yet. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 00:38, 9 June 2010 (UTC)

Well, if you listen to Spanish media, Spanish songs... you will certainly know [a'moɾ] is much more common than [a'mor]. And as Harris says; una palabra como amo[ɾ - r] se puede realizar con vibrante simple o múltiple final.

You should allow both sounds transcription; amo[ɾ - r], al[r]edededo[ɾ - r]

On the Spanish phonology article, the alveolar trill ([r]) and alveolar tap ([ɾ]) contrast intervocalically but are otherwise in complementary distribution: [r] is found after /l/, /n/, and /s/, before consonants, and utterance finally (how you prove this if Harris says both sounds are possible, [r] and [ɾ]).

There are also alternations occurring with suffixation, such as when nouns are pluralized: amor [aˈmor]/allow [aˈmoɾ] is also possible, vs. amores [ aˈmoɾes].

You altered Harris' explanation, amo[ɾ - r]. So utterance final r can be either /ɾ/ or /r/. /r/ occurs in emphatic speech (habla enfática).

Jaume87 (talk) 16:59, 10 June 2010 (UTC)

Spanish phonology already reflects the information that you've provided. I don't think we should have variable pronunciations and I don't think we should change the conventions until sourcing is provided that indicates what you say about the dialectal distribution of these two sounds is true. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 17:21, 10 June 2010 (UTC)