Talk:Near-open front unrounded vowel

Other Languages
Persian:, 'door'. In English, Æ is a 2-step sound: First you open your mouth wide and then you make it a ə (schwa) before closing it. In Persian, the second step is absent.

I am absolutely sure about Persian because I speak Persian. In am also suspicious whether Æ occurs in other languages. The difference is very subtle and most people don't notice it except those who have are bilingual and have studied the phonics of the English language systematically.--02:40, 8 July 2006 (UTC)~

There are two problems with your analysis.

1. In some English dialects, the phoneme /æ/ is phonetically diphthongized but this doesn't mean that [æ] is a diphthong.

2. Being a native speaker doesn't make you an expert in phonology and phonetics and doesn't mean you don't need to cite sources.

Your input is appreciated but you can't say that a sound doesn't occur in a language because it's different than in English. AEuSoes1 06:06, 8 July 2006 (UTC)


 * Pronounce It Perfectly in English (Pronounce It Perfectly in/Book and 3 Audio Cassettes) (Audio Cassette) has a diagram of the mouth showing the two pronuciation stages.  This vowel is a diphthong.  I live in California and everyone pronounces "æ" thus.--06:18, 8 July 2006 (UTC)


 * I believe that it can be diphthongized in English, but that doesn't mean that it's dipthongized in every language. Read the Persian phonology page and you'll see an actual vowel chart taking formants and whatnot into account.  AEuSoes1 06:23, 8 July 2006 (UTC)


 * My understanding of English phonology is pretty systematic. However, that is not the case with Farsi even though I learned it before English.  Nonetheless, I can assure you that my mother does not pronounce "æ" like Californians and she refuses to learn the right way.--06:33, 8 July 2006 (UTC)


 * I hope you didn't put that it was disputed just because you personally disagree with it. You're gonna need some academic disputing to really let that sort of edit fly.  You haven't even indicated what it would be instead.  AEuSoes1 16:09, 9 July 2006 (UTC)


 * It's simple really... if the vowel was a diphthong, why would IPA have one symbol for it? It's common sense. Some Americans say "eæ" especially Californians and people in the KY, OH, IN area. I think that's causing confusion.RedAugust 22:01, 31 August 2007 (UTC)

The phenomenon is known as Californian vowel shift (American version of the Canadian vowel shift).
 * Many Californians pronounce /æ/ as a vowel close to [a]: bad [bad], bag [bag], last [last].

So, should it not be included? Even though the symbol is the same? I don't get it.--Adûnâi (talk) 08:47, 27 November 2017 (UTC)

German /æ/?
Someone has included the pronunciation of the german <ä> here, as in the word Käse. This is actually pronounced with a different vowel (in X-Sampa, it's represented by /E/, I can't get the IPA to work here). /æ/ (X-Sampa /{/) might possibly be present in some German dialects, but not in any "standard" variety.

Russian /æ/?
I'm going to dispute this line:
 * Russian: пять, "five"

At least in my dialect (Kaliningrad region), pronouncing it with /æ/ makes it sound more like (but not quite like) петь - 'to sing'. I definitely say пять with a clear /a/ sound.


 * Check the sources for Russian phonology. Wikipedia's mentions of Russian imply Standard (Moscow) Russian.  Æµ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 00:40, 6 October 2006 (UTC)


 * There's no in Russian, just . It is an allophone of  between palatalized consonants. Peter238 (talk) 06:15, 11 October 2014 (UTC)

English words in German?
Often when English words with this short a sound are borrowed into German, as they don't have the sound per se, it's rendered as /ɛ/, like "bands" being said like "bents" and so on... I'm not sure if this is noteworthy, the right article to put this in, or whether it's original research... so yeah...

86.20.219.123 22:59, 20 April 2007 (UTC)

Received pronunciation
According to the BBC website, the near-open front unrounded vowel (represented by æ) is the sound of the 'a' in 'cat' in received English. This isn't mentioned in the article, which, instead, says that it is the 'e' in 'help'. If you listen to the soundbite on the site, you will see that it does, indeed, sound like the 'a' of received pronunciation, and nothing like any pronunciation of 'e' in 'help'. On this site, it says æ is the American pronunciation of the 'a' in 'cat', which sounds like an American 'a' (which is similar to the 'a' of Received Pronunciation).

The IPA website itself is of little use, but to me, from these other sources, it looks like this page is incorrect. Any ideas?

Is this the a in "ant"?
I'm a United Stateser with a roughly General American accent, and I'm very interested in phonetics. Here's the deal: As far as I can tell, sources seem to agree that the A in words like "ant" and "camp" (at least the way I pronounce it) is an ash. But I can't help but feel that it's actually a vowel in its own right.

I feel that if I wanted to, I could say "man" like /mæn/, with the same internal vowel as the way I say /pæθ/ ("path"), but I think that when I normally say the word, there's something more nasal going on in that A. If I repeat /æ/ sound, I just sound like I'm dully reading a scream ("A-a-a-a"), but if I repeat the vowel in the way I say "ant", it's like the laugh of a nerdy supervillain. I only find this the case for As that precede N or M. Does anyone have the same feeling I do? Is there a word for this change to a vowel, equivalent to the vocalic r? $$\sim$$ Lenoxus " * " 18:37, 4 April 2008 (UTC)
 * Yes, what you're talking about, I think, is nasalization. In English (and many other languages) nasal consonants spread their nasal quality to preceding vowels.  Thus, all vowels are nasalized before nasals.
 * If it seems like more than that, it could be some sort of quality change that is either allophonic or diachronic. In my Californian dialect, front vowels are (diachronically, I think) raised before  so that sing has the same vowel as seen and rang has the same vowel as rain.  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  20:07, 4 April 2008 (UTC)
 * Thanks, aeusoes! That pretty much answers my question. Maybe nasalization needs to be linked more? $$\sim$$ Lenoxus " * " 18:40, 6 April 2008 (UTC)
 * Lenoxus, it is more likely that the phenomenon you are noticing is called æ-tensing. Although nasalization is bound to occur as Aeusoes described, its effects would probably not be overly noticeable to an untrained ear. Whereas the difference between the /æ/ in cat and can is fairly apparent (at least, in many dialects across America). The difference is normally transcribed: [æ] vs [eə]. 96.255.182.118 (talk) 20:18, 16 July 2009 (UTC)

æ in Polish?
I have to disagree with the jajko example, to my ears æ does not exist in Polish, at least not in the standard version of the language. And I have not found any info on this vowel possibly existing in other dialects.. the wiki page on Polish phonology has no mention of it. —Preceding unsigned comment added by 207.112.56.46 (talk) 01:33, 11 April 2008 (UTC)
 * You're right. Examples in Polish should all be sourced.  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  01:51, 11 April 2008 (UTC)
 * The source of this confusion is that Polish is normally central, but it is front  between palatalized consonants. Since the aim of most phoneticians is to avoid diacritics wherever it is possible, a phonetically incorrect symbol  is used for the fronted allophone. If you're looking for extreme precision, it's best that you use canIPA instead (but don't trust Canepari's transcriptions too much, they contain lots of mistakes, such as equating English  with Polish  - that'd work only for Indian English I think). Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 12:31, 27 October 2014 (UTC)

Palatalization in "cat"??
What accent of English is [kʰʲæt] supposed to represent? Source? Grover cleveland (talk) 05:55, 26 March 2009 (UTC)

I've heard some people from my family say it that way, so possible Massachusettian or other New England American English. 75.102.128.131 (talk) 09:17, 3 May 2009 (UTC)

Nasalized version in French?
Isn't the vowel in quinze a nasalized version of this vowel? Should that be mentioned, or is there a separate page for the nasalized vowel? --Trovatore (talk) 05:17, 25 August 2012 (UTC)
 * No, there isn't a separate page for the nasalized vowel. Peter238 (talk) 06:24, 11 October 2014 (UTC)

Not in french
In french, the /æ/ doesn't exist, even in nasalized form. /ɛ̃/ is the correct form. Some phoneticians told that it became /æ̃/ in modern French. But it's not true, it's a normal nasal shift which result of velum opening; it always was pronounced like this. — Preceding unsigned comment added by 109.28.241.116 (talk) 17:50, 1 February 2014 (UTC)
 * You mean that in modern French, is not realized as ? Those are not different phonemes. Besides, what do you mean by "nasal shift"? Peter238 (talk) 06:20, 11 October 2014 (UTC)

Rounded version
Hasn't a rounded version of this vowel been ever attested in any language? — Preceding unsigned comment added by Backinstadiums (talk • contribs) 09:58, 6 June 2014 (UTC)
 * Yes, for instance Danish. See the vowel chart in Peter238 (talk) 06:24, 11 October 2014 (UTC)
 * Another instance are Lehali and Lemerig (Vanuatu), two Oceanic languages featuring a symmetrical vowel inventory contrasting /æ/ and /ɒ̝/. (See pp.194-5 of .) —— Womtelo (talk) 10:43, 11 October 2014 (UTC).
 * Aren't those back vowels though? The Danish one is somewhat retracted front. Peter238 (talk) 12:41, 11 October 2014 (UTC)
 * hi, yes you're right, they are rounded and back (like the higher vowel [ɔ] for ex). I'm not aware of a vowel corresponding to [æ] that would be front and rounded; maybe Danish would be the right example indeed. — Womtelo (talk) 14:00, 11 October 2014 (UTC).
 * Thanks. Yes, Danish is the right example, you can see that vowel chart in Danish phonology. Peter238 (talk) 19:33, 11 October 2014 (UTC)

The sound clip seems to be inaccurate
The sound used in the article seems to sound higher (approaching ɛ) than how it is used in most languages. It seems to reflect the General American pronunciation of the phoneme but sounds too closed for most of the other languages on the list. It also sounds slightly dipthongised (with a [ɘ] following). Could this be changed? — Preceding unsigned comment added by 146.200.25.119 (talk • contribs) 08:51, July 10, 2015‎
 * Agreed, it sounds just like [ɛ] to me 58.136.157.246 (talk) 10:23, 14 July 2016 (UTC)

External links modified (February 2018)
Hello fellow Wikipedians,

I have just modified one external link on Near-open front unrounded vowel. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20120131201021/http://www.latsis-foundation.org/files/Meletes2009/11.report.pdf to http://www.latsis-foundation.org/files/Meletes2009/11.report.pdf

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 05:18, 15 February 2018 (UTC)

[æ] in Turkish
redundant (and you're mistaking allophones for phonemes)

Not. [æ] is in free variation with [a] in many languages (including Azerbaijani, close relative of Turkish). This is not the case in Turkish. Exotic allophonies of phonemes should be noted in respective phonemes as notable facts.Erkinalp9035 (talk) 20:35, 18 September 2018 (UTC)


 * The notation and  (which you used in your edit) is very different to  and . The first pair signifies phonemes and the second one allophones.


 * Your assertion that Azerbaijani varies between  and  isn't supported by Azerbaijani language. What would be the correct wording is that the Azerbaijani phoneme  is phonetically an open front unrounded vowel  (or, you can write it however you want). Again, let's not mistake allophones for phonemes.


 * Why on earth do you feel the need to say that Turkish can't be lower than ? The fact that we don't say that it can is already enough. It's quite obvious that it can't (although this may vary from speaker to speaker, you'd need to provide a source to prove that Turkish  is always near-open.) Plus, let's not forget that it's just an allophone of  in Turkish.


 * An that varies freely with  would be anything but exotic. These vowels are very close to each other on the vowel chart.


 * Also, the source doesn't say that Turkish varies between front and near-front. Kbb2 (ex. Mr KEBAB) (talk) 11:36, 20 September 2018 (UTC)


 * Correction: you wrote instead of the correct  in this edit. Kbb2 (ex. Mr KEBAB) (talk) 11:48, 20 September 2018 (UTC)


 * is allophonous with [a] in most languages. Departures from common practice are notable. Because [a] is understood as /a/ (albeit as mispronunciation, not as allohpone) in Turkish, not /e/. Source: I'm a native speaker of Turkish. Will provide more third party sources on that.


 * An that does not vary freely  is exotic, though.


 * Look above. It is written after citation. Following usual practice of citation, you refer to citation immediately after the fact referenced.


 * Fixed that in the next edit, though. Erkinalp9035 (talk) 20:06, 20 September 2018 (UTC)


 * This doesn't seem to make sense (you're still mixing up phonemes with allophones). Did you mean to say that and  are in free variation in most languages? If so, I strongly disagree. Please prove it.


 * Being a native speaker of Turkish doesn't matter here.


 * No, it's not. I've already said that they are very close to each other on the vowel chart. I'm starting to get a bit tired of these definitive statements said without providing any proof to back them up. We've already had a similar discussion on Help talk:IPA/Hindi and Urdu a few days ago. EDIT: I used we in the broadest possible sense. It wasn't me and you specifically.


 * It wasn't written after the citation but before it. The source for the whole Turkish cell in the Notes column is Göksel & Kerslake (2005:10). I deleted what you put in there because the source doesn't say it. Anyone can delete unsourced parts of articles from Wikipedia.


 * You didn't though. Kbb2 (ex. Mr KEBAB) (talk) 20:24, 20 September 2018 (UTC)

Audio Clip
The audio clip's pronunciation stutters. Can someone provide a clip that doesn't stutter? --Helping end world sexual repression. (talk) 09:55, 20 September 2018 (UTC)
 * The clip is fine, it's probably your browser. Try a different one. Kbb2 (ex. Mr KEBAB) (talk) 11:25, 20 September 2018 (UTC)
 * I tested it on 5 different browsers, and it holds true: https://www.dropbox.com/s/u1p58th9ukevmcs/audio-problem.mp4?dl=0 --Helping end world sexual repression. (talk) 13:20, 20 September 2018 (UTC)
 * Then this is either your operating system's fault (I'm having no issues on Chrome and Windows 7) or you're having problems with the player itself. Try downloading the recording. Kbb2 (ex. Mr KEBAB) (talk) 14:04, 20 September 2018 (UTC)
 * Or try opening the raw file on your browser. I've encountered the same issue and it only happens on MediaWiki's player (TimedMediaHandler) but not on the browser's native player, although it doesn't happen to me with this particular clip now for some reason . It plays just fine when you click the play button for the first time since the page is loaded, but it happens when you play it again—not just with this clip but any clip as far as I can tell (but you don't notice this when a file starts with silence). So this is most likely a issue on MediaWiki's part, and people at WP:VPT may have a solution or at least an answer. Nardog (talk) 14:33, 20 September 2018 (UTC)
 * Thanks Nardog! You're right. That seems to be the case. On a side note, we seem to frequent the same topics online, haha. You're super helpful. Much appreciated! --Helping end world sexual repression. (talk) 08:23, 21 September 2018 (UTC)

Arbitrary removal of content
Kbb2 singlehandedly decided to remove massive amounts of content here without proper justification. S/he apparently mixed two unrelated criteria: "poorly sourced" (what criteria exactly? why not just add a "refnec" tag??) and "languages with few speakers" (what's the threshold?). The latter criterion is totally arbitrary, and lacks any legitimacy in linguistics; also, it aggravates the already strong bias in favor of European languages. How come Bengali, Kurdish, Sinhala get erased?

Incidentally, the criterion of number of speakers is wrongly applied: how did Bengali get removed when it has 260 million speakers, while Limburgish made it to the final list with 1.3 M? Who gets to decide what languages make it to the final list, without even attempting a discussion on the Talk page? I find those undiscussed choices shocking, and arbitrary.

-- Womtelo (talk) 10:05, 6 August 2019 (UTC)
 * Over 30 examples of this vowel is already a lot. Scholarly publications often list just one. Longer tables give an illusion that we're somehow trying to list all of the languages which possess that vowel, which is nearly impossible.
 * I've removed entries without sources irrespectively of how many speakers that language/dialect has. Look at how many sourced entries the table has. We don't have to include any particular language in it, these are *examples* of the near-open front unrounded vowel in world's languages. You can reinstate Bengali and Persian with proper sources per WP:BURDEN. I don't have a problem with that.
 * I've removed Norfolk English and Upper Saxon German because I find those sources unreliable. The first one is at odds with how Trudgill (2004) describes Norfolk English (it mixes up older and contemporary varieties as well as the broad and middle class ones), whereas the vowel chart in Khan & Weise (2013) is also dubious (the vowel space in it looks too compressed, and a contrast between true-mid and close-mid central rounded vowels that are both short and lax to me seems nearly impossible).
 * In close front unrounded vowel I removed languages that have less than 10 million speakers. The threshold could be lower the rarer the vowel is. I've never claimed that this or similar methods are perfect. A list of near 100 examples (now massively reduced) in near-close front unrounded vowel surely is insane, no?
 * The reason I preserved Limburgish was that because it is a sourced entry and it illustrates a near-front (rather than fully front) variety of . Kbb2 (ex. Mr KEBAB) (talk) 09:56, 8 August 2019 (UTC)
 * Correction: it's not that describing Norfolk as near-open is at odds with Trudgill, it's that we should focus on major dialects of English. It's a common sound in English. Trudgill, by the way, describes the older variety of Norfolk  as closer to cardinal  than the contemporary realization, which he describes as "a good deal more open" (so this might mean a vowel that is as open as ). Kbb2 (ex. Mr KEBAB) (talk) 10:11, 8 August 2019 (UTC)


 * hi Kbb2, It's unfortunate that your arguments arrive so late in the debate, when I invited you to come to this discussion much earlier on. You can't just impose your edits the way you did, and crush calls for debates. Discussion allows you to open up to other people's perspectives. Now I'm not interested in Edit-warring or throwing bird names, I'm only interested in science and making WP better; from what I see, deep down you are also driven by a genuine interest in knowledge, so I'll take it from there and focus on the content.
 * The criteria you chose are not consensual, as you yourself rightfully acknowledged. For example, linguists regard as irrelevant the number of speakers of languages when doing comparison: a language with 100 speakers will be as valuable and informative as one with 100 million, with zero reason to privilege the larger languages. Your personal decision to remove languages that have less than 10 million speakers is very problematic, to say the least. Choosing languages with millions of speakers creates a terrible geographical (and genealogical) imbalance in favor of Europe, and automatically erases from the picture thousands of languages from Africa, or from the Pacific – an area with 1500 different languages [!], each spoken by less than 0.5 million. The present entry (on a particular vowel) is also to be used by phonological typologists, who may wish to gain an idea of how widespread in the world such-and-such phoneme is. As you may know, some phonemes are pervasive across all continents, while others are much rarer and found only in the Amazon, or only in Europe. A widespread misconception, for example, even among linguists, is to think that front rounded vowels are only found in Europe, when they are in fact attested in various parts of Asia or the Pacific; so listing such languages in a WP entry is valuable. This is not just my personal preference, but one of typological linguists; it would be a way to avoid the Western bias that is a deep problem on Wikipedia.
 * I share your concern with the number of languages in entries such as this one, and the wish to avoid endless, absurd lists. However, evidently we don't share the same diagnostic. You see too many small languages (& too many unsourced entries), I see geographical imbalance. (And I also acknowledge that the criterion of balance is not the only one that should guide us, it's just one among several.) The list before your edit had 20 languages of Europe (listing altogether 42 varieties), 7 lgs of Asia, 2 of Africa, 1 of the Pacific. Your edits now result in 19 languages from Europe (down from 20), identified through 39 different varieties (down from 42); 2 from Asia (down from 7), 1 from Africa, 0 from the Pacific. Named lg varieties in the table have gone from 81% from Europe, to 93%; from the perspective of the European bias, this is making things not better, but worse.
 * My suggestion would be to keep, in each table, a maximum of 4 or 5 languages per continent, regardless of their number of speakers. (An alternative way would be 3 or 4 lgs per genealogical lg family, but there is then the problem that families vary considerably in their sizes.) I'd rather have Bengali or Persian listed even with a "refnec" tag (knowing that references would be easy to find for such large languages), rather than have 42 dialects from Europe and 3 from the rest of the world. So even if you're into Germanic dialects (which is a fascinating topic indeed), I believe it would create more balance if we could reduce the number of Western varieties listed here, and reinstate lgs from elsewhere in the world, preferably sourced. If we keep a policy of 4 or 5 lgs per continent max in these phoneme entries, we should never have more than 20 or 30 entries in total, so this should be fine for all of us.
 * Best, Womtelo (talk) 19:36, 8 August 2019 (UTC).
 * PS: Following up on our discussion, I reinstated four non-European languages onto the entry, with published sources: Bashkir, Lakon, Persian, Sinhala. I'm not deleting any entries from Europe, but this could also be done to restore some balance. —— Womtelo (talk) 10:50, 9 August 2019 (UTC)
 * More geographical balance won't hurt. The number of speakers can be a secondary criterion if a region (i.e. Europe) gets over-represented. I have added Bambam (c. 30k speakers as of now), representing a small cluster of languages in the interior of Sulawesi which are exceptional in the ISEA area (so far not represented) in having phonemic (that is really realized as, trust my ears). –Austronesier (talk) 17:47, 28 November 2021 (UTC)
 * hi Austronesier, Thanks for adding data about a non-European language! Indeed, geographical (and phylogenetic) is always welcome, and remains badly needed in entries like this one (especially since you're bringing a solid reference too). I had forgotten about this discussion thread, but I confirm that the bias remains heavy in favour of European languages & dialects, and at the expense of other parts of the world. Edits like yours are useful! -- Womtelo (talk) 18:19, 28 November 2021 (UTC).
 * The examples (which were at one time debated somewhere in the Linguistics project talk page) have several functions. On the one hand, we need to address readers who are not familiar with IPA, and therefore have to give them examples they will recognize, thus ideally from the world's major languages. Another purpose is to give a balanced, wide and yet concise range of representative examples for those who already know the basics want to learn more about the prevalence these sounds; this is especially important for less common sounds. And then there's the special case where readers might be familiar with the an IPA symbol, but might have an incorrect idea about its actual value due to their accent: e.g., many people know, , and , but actually equate these symbols with other sounds. So detailed examples from English varieties are in many cases also very helpful (close back rounded vowel does a good job to illustrate that most English varieties don't have cardinal ). It just needs a good (and concise) mix. –Austronesier (talk) 11:29, 29 November 2021 (UTC)
 * I understand, and I agree. In fact, I'm all in favour of including major languages of Europe — as long as this doesn't produce a major imbalance with other parts of the world. best, — Womtelo (talk) 15:44, 29 November 2021 (UTC)

American English diphthong?
In GA, wouldn't this vowel more accurately (and perhaps, necessarily) be described/transcribed as the diphthong [ɛə̯]? Bladesinger46n2 (talk) 02:37, 30 January 2021 (UTC)
 * No, I don't think so. In some American accents, but not GA, the ash has a tense variant, which can actually be used distinctively.  For example, in some New York accents, "can" in the sense of possibility is /kæn/ but "can" in the sense of "receptacle" is /keən/ or some such.  Unless I have that backwards.  Is that what you're thinking of? --Trovatore (talk) 03:59, 30 January 2021 (UTC)
 * By the way, if you're asking for general knowledge rather than because you want to suggest a change to the article, please ask at WP:RD/Language rather than here. --Trovatore (talk) 04:15, 30 January 2021 (UTC)
 * See /æ/ raising. Nardog (talk) 06:16, 30 January 2021 (UTC)

Thank you for your reply. I suppose I have already had a bit of uneasiness concerning this vowel and it's perceived "wedging" within the quadrilateral, which, to me, throws of a certain harmonic balance to the thing. When I actively sound out this vowel, I notice one of two things most regularly: 1) I seem to either slightly glide the vowel to a more central position, or 2) vocalize a long vowel [ɛː]. Bladesinger46n2 (talk) 11:44, 30 January 2021 (UTC)
 * May I ask where you grew up geographically? I don't think that's exactly GA.  Technically we shouldn't be discussing this on the talk page; if you want to reply on my talk page feel free. --Trovatore (talk) 20:10, 30 January 2021 (UTC)