Wiktionary:Votes/2014-01/Representing the short-a phoneme of Received Pronunciation

Representing the short-a phoneme of Received Pronunciation

 * Voting on: Representing the short-a phoneme (the vowel in trap) of Received Pronunciation: Should we use /a/ or /æ/? We essentially have three options:
 * Option 1: Enforce the use of /a/ for the short-a phoneme in RP.
 * Option 2: Enforce the use of /æ/ for the short-a phoneme in RP.
 * No action: Leave the decision to the editor transcribing the word.


 * Vote starts: 00:01, 24 January 2014 (UTC)
 * Vote ends: 23:59, 22 February 2014 (UTC)


 * Vote created: WikiTiki89 19:48, 17 January 2014 (UTC)
 * Discussion:
 * [[Image:Wikt rei-artur3.svg|20px]] WT:Beer parlour/2014/January
 * [[Image:Wikt rei-artur3.svg|20px]] User talk:Wikitiki89
 * [[Image:Wikt rei-artur3.svg|20px]] '''Wiktionary talk:Votes/2014-01/Representing the short-a phoneme of Received Pronunciation

Support Option 1 - Enforce the use of /a/ for the short-a phoneme in RP

 * 1)  Please read the Beer Parlour discussion.  has now been adopted by the British Library, the OED and Oxford dictionaries, and an increasing number of phonetics textbooks; there are very good reasons for using it. Tying ourselves to an outdated scheme (and one which represents an old-fashioned pronunciation) is not helpful – just look at how many people are already confused by our logo. I see no reason to artificially maintain parity with US transcription given that we already distinguish between such pairs as American  and British  – why is this any different? Ƿidsiþ 08:00, 25 January 2014 (UTC)
 * Note that when a word is pronounced the same way in American and British English, the OED only transcribes it once and uses /æ/ for the short-a. For example, trap is transcribed as and bat as . There is a similar situation with the long-i phoneme, when the pronunciations differ, the OED transcribes it as /ʌɪ/ for British and /aɪ/ for American, but when they are the same the OED only transcribes it once as /aɪ/. Here we have always been transcribing the short-i as /aɪ/ even in RP. --WikiTiki89 15:09, 25 January 2014 (UTC)
 * No, I'm afraid that is not true. The reason you are seeing a difference on the website is that trap: and bat: are both OED2 (i.e. the second edition of 1989). All OED Third edition entries use for UK and  for US. Ƿidsiþ 15:49, 25 January 2014 (UTC)
 * Having looked at the entry for man, I see you are correct. --WikiTiki89 15:54, 25 January 2014 (UTC)
 * 1)  ... for the reasons so clearly set out by Widsith above.  I used to be puzzled by the /æ/ vowel transcription when I first saw it (and had to ask what it was because it was not used at all in the part of northern England where I live).  It is now rare to hear the old RP pronunciation on British radio or TV .  Now that I know what /æ/ means, I mentally convert all /æ/s to /a/s, so I was delighted to see that the OED has made that change for their third edition.  Wiktionary is quick to include changes in meaning, should it not also keep up with changes in pronunciation?    D b f  i  r  s   14:02, 25 January 2014 (UTC)
 * 2) . The users above who favor /a/ are British. The users below who favor /æ/ are American and Russian. I imagine the Brits have more intimate, extensive, and current knowledge of RP, especially modern RP, so I think we should listen to them. And even if US /æ/ and RP /a/ are similar, we already make a distinction between other sounds which are about as similar, e.g. US /oʊ/ and RP /əʊ/. But note my comments in the "abstain" section! - -sche (discuss) 21:18, 9 February 2014 (UTC)

Support Option 2 - Enforce the use of /æ/ for the short-a phoneme in RP

 * 1)  --WikiTiki89 19:07, 24 January 2014 (UTC)
 * Can you please supply your reasoning, like what sources you have considered, what personal experience, and the like? --Dan Polansky (talk) 20:41, 24 January 2014 (UTC)
 * I have already done so at the discussions linked to above. --WikiTiki89 06:56, 25 January 2014 (UTC)
 * Having looked at the discussion, I cannot figure out an answer to the following question: What sources (publications by other people and organizations) have you considered, if any? --Dan Polansky (talk) 08:47, 25 January 2014 (UTC)
 * The second edition of the OED uses exclusively /æ/. The third edition of the OED uses /æ/ when the American and British pronunciations do not otherwise differ. When they do otherwise differ, the third edition of the OED uses /a/ for British and /æ/ for American. In other words, the OED prefers using /æ/ to cover the British pronunciation rather than transcribe the same thing twice. --WikiTiki89 15:21, 25 January 2014 (UTC)
 * Do you have any other sources to mention? Furthermore, do you think the 2nd edition of OED should take precedence over the 3rd edition of OED for the purpose of sourcing? --Dan Polansky (talk) 19:28, 25 January 2014 (UTC)
 * The third edition of the OED has also switched to indicating American /ʌ/ as /ə/ (it's entry for one has: Brit. /wʌn/, U.S. /wən/), which I think is an inaccurate representation of the American pronunciation of the short-u, which does not differ much from the British. For example, the US and UK audio files we currently have a son do not differ at all (except possibly in vowel length):, . Therefore, I think that we should discount the third edition of the OED, as well as any other dictionaries that have followed it, as evidence in this matter. --WikiTiki89 16:32, 29 January 2014 (UTC)
 * ... but the OED has the same pronunciation for son (as in Wiktionary's audio files). I agree that they struggle over the variations in the pronunciation of one, but there are so very many variants of this word on both sides of the pond that it's difficult to represent them all.  I agree with you that the OED can be discounted for some of its opinions on American English (I would go to an American dictionary for an authoritative opinion of that), but it is the standard authority on British English, and I object to "discounting the evidence that you don't like".    D b f  i  r  s   23:07, 29 January 2014 (UTC)
 * , but that is actually because son is a second-edition entry in the OED. The third-edition entries all have separate pronunciations for British and American English, even if they are exactly the same (see met n.1). For example, mutt is also transcribed with the same difference (Brit. /mʌt/, U.S. /mət/). And I am not discounting it because I don't like it, I'm discounting it, because it does not match our practice, so using it to justify our practice is inherently wrong. --WikiTiki89 23:15, 29 January 2014 (UTC)
 * You are correct about the second edition entry. We'll have to wait to see what they do when they revise it.  I thought the vote was all about updating our practice, though we could still retain the combined /ʌ/ (I don't hear as much difference there as I do with /a/ and /æ/).    D b f  i  r  s   23:37, 29 January 2014 (UTC)
 * Well I just gave you some examples of entries that they did update. I don't expect son to be any different. And, that is exactly what I meant by discounting the OED as evidence. If we disagree with the OED about some phonemes, we can no longer use it to justify other phonemes, except maybe as part of a tally of how many dictionaries do what. I don't think anyone has yet named a dictionary other than the OED that does this (except Oxford Dictionaries, but they basically copy the OED in almost everything). --WikiTiki89 00:14, 30 January 2014 (UTC)
 * Yes, I think you are correct that they are going to transcribe the U.S. pronunciation of son as /sən/. I hear very little difference between these vowels, and I hear more difference between regions within each country (especially UK) than between countries.  I assume that the OED has considered an average of pronunciations.  The vote is about /æ/ and /a/, not /ʌ/ and /ə/, so your objection to the OED is a separate issue.  I think everyone agrees that the American /æ/ is almost always higher than the British "/æ/a/".  The question is whether the difference is sufficient to justify using a different symbol.  I hear a big difference; you hear little.  I suspect that, in my case, this is because I am from the north on England (and possibly you have a New England perspective?)  It would be interesting to hear from editors native to other regions of both countries.    D b f  i  r  s   00:35, 30 January 2014 (UTC)
 * My point about the short-u vowel was just to make the point that we cannot say "The OED uses /a/, therefore so should we". And no one has brought up any other dictionaries that use /a/. Yes, I do have a New England perspective on this. In New England, there is a very sharp contrast between two different realizations of /æ/: the normal one and the tense one. The normal one is essentially the basis of my perspective on what the symbol /æ/ represents phonetically (the tense one is realized as [eː(ə̯)] and so is irrelevant here). While the /ɑː/ phoneme's New England realization of [aː], together with the Russian /a/, is my basis for understanding the /a/ symbol. When I hear the British /æ~a/, it sounds very different from the New England [aː] (with the possible exception of some, but not all, Scottish accents). --WikiTiki89 00:51, 30 January 2014 (UTC)
 * Widsith was arguing that latest research suggests that there have been changes, and that the OED, for example, is trying to keep up with changes in pronunciation. No-one was suggesting that we should try to imitate any one dictionary, just that we should not get stuck in the past by keeping outdated conventions.  My point was that we hear things from our own peculiar perspectives (and northern English is closer to Scottish), so a wider view would be useful.  Your New England [aː] appears to be a long vowel.  We distinguish between long and short here in the UK.  For example, my (northern) pronunciation of cart /kaːt/ is just a lengthened version of /kat/ ([käːt] where I live, but [kaːt] in Liverpool) (though if I'm speaking "proper", I would use /kɑːt/).  I do appreciate that those who have learnt a particular interpretation of the symbols will wish to retain them. I've been mentally turning /æ/ into /a/ for many years, and wondering why dictionaries persisted with /æ/ when I seldom hear it on this side of the pond.    D b f  i  r  s   08:35, 30 January 2014 (UTC)
 * I understand that. What I meant was that we can't use the OED as evidence of this sound change. As far as length, I'm aware of all that. The New England [aː] is certainly long usually, but in some cases before voiceless consonants it can be short. But I was only referring to the vowel quality. I've been mentally turning /ɑː/ into [aː], and /ɔː/ into [ɒ(ː)], but that's part of the concept of turning phonemes into realizations. Your [a] is just your realization of the English phoneme /æ/, even if it is becoming more common in Britain. One of the things I hate most about our pronunciation scheme for American English is our failure to represent the difference between /ɑː/ and /ɒ/, and I find myself having to look at the US IPA and the UK IPA, just to find out how to pronounce a word in my dialect. If we had a centralized system of representing phonemes, I wouldn't have to do that. (Side note: I hope you are not confusing /k/ with /c/). --WikiTiki89 17:46, 30 January 2014 (UTC)
 * (Thanks for pointing out my /k/ -> /c/ error, now corrected above. I was concentrating so much on the vowel that I didn't think about the consonant.)  I agree that we can't use the OED as evidence, but current research into pronunciation changes is not particularly common.  Widsith quoted some.  I suppose the problem is that we can never get IPA to work clearly for everyone, but I find it useful to read the difference between "average British" and "average American" pronunciation, even if many regions of both countries have variations.    D b f  i  r  s   18:30, 30 January 2014 (UTC)
 * Also, I have just consulted the Collins Dictionary, the Macmillan Dictionary, and the Cambridge Dictionaries (particularly their entries for cat and ladder). All of them use /æ/ for this vowel even for their British pronunciations. And their British audio files suggest that while, on average, the vowel is lower than the American vowel, it is not what I would personally describe as [a].
 * In the end, I think it is consistency in representing the phonemes that is more important than choosing the symbol that more accurately represents the latest fad in pronunciation. --WikiTiki89 16:32, 29 January 2014 (UTC)
 * I have now taken a closer look at the "huge survey of RP vowel sounds carried out by the British Library" that Widsith mentioned in the Beer Parlour discussion. This survey says that /æ/ is "conservative RP" and that /a/ is "contemporary RP", but listening to the audio sample provided for the "conservative" /æ/, it sounds to me to be higher than what I think of as /æ/. This leads me to believe that /a/ is used to provide a graphical contrast with the conservative pronunciation, without using diacritical marks, rather than to more accurately represent the sound itself (which can't be done without diacritical marks). --WikiTiki89 00:14, 30 January 2014 (UTC)
 * I happened to hear a recording of Dylan Thomas reading "Under Milk Wood" yesterday, and his "conservative" /æ/ sounded higher than what even I think of as /æ/, but it also sounded very, very dated. What would you use /a/ for?  What does [a] sound like if it's not the way I (along with the Scots and the Welsh) say "cat"?    D b f  i  r  s   10:36, 3 February 2014 (UTC)
 * For example, the Polish kat audio file below is what I generally think of as [a]. Also the pronunciation of smart (as [smãˑʔ]) in this 7-second clip from . Those two examples are a bit different, so I might even say that the Polish one is more central ([ä]). --WikiTiki89 18:40, 3 February 2014 (UTC)
 * What about Dutch words with long, like ? 19:00, 3 February 2014 (UTC)
 * That sounds almost exactly (except for the final consonant) like a Scot saying . It's not exactly the same, but it's close to the New England [aː]. --WikiTiki89 19:27, 3 February 2014 (UTC)
 * I perceive a strong difference between the vowel of laat and the vowel of the US pronunciation of hat. I perceive a weaker difference between the vowel of laat and the vowel of the UK pronunciation of hat. - -sche (discuss) 19:43, 3 February 2014 (UTC)
 * Now listen to Gloin say "That's my wee lad, Gimli." in . --WikiTiki89 19:57, 3 February 2014 (UTC)
 * 1)  —Aɴɢʀ (talk) 19:26, 24 January 2014 (UTC)
 * Can you please supply your reasoning, like what sources you have considered, what personal experience, and the like? --Dan Polansky (talk) 20:41, 24 January 2014 (UTC)
 * I said it all in the beer parlor discussion. —Aɴɢʀ (talk) 07:49, 26 January 2014 (UTC)
 * Support That's the way I perceive standard English and the way studied it. I'm not a native speaker of English but the Russian education system was based on British English and all dictionaries and textbooks we used had /æ/ for the short-a phoneme, e.g. /æpl̩/ for "apple". British and American English accents differ but normally, IMO not in this phoneme. I normally hear /æpl̩/ in Australia as well, where pronunciation is much closer to BE than AE but admittedly the sound /a/ for the short-a phoneme also exists in UK but I don't know if this is classified as "RP".  --Anatoli (обсудить/вклад) 03:15, 25 January 2014 (UTC)
 * Changed my vote to "abstain". See below. --Anatoli (обсудить/вклад) 00:08, 18 March 2014 (UTC)

Abstain

 * Abstain for now. I am sitting on the fence for now, but I will probably vote for a particular stance. I may vote "oppose", since this vote seems to be on a factual question, and votes on factual questions are suspect, especially when the voters do not supply their reasoning, sourcing and evidence. As for direct evidence, I bring this: (En-uk-man.ogg),  (en-us-man.ogg);  (En-uk-a cat.ogg),  (En-us-cat.ogg);  (En-uk-a map.ogg),  (en-us-map.ogg). The UK and US items seem very different to me, but I do not know the definitions of the scopes of /a/ and /æ/ to trust myself to properly judge a proper transcription on my own. --Dan Polansky (talk) 09:02, 25 January 2014 (UTC)
 * An idea: here is Polish (Pl-kat.ogg); here is Danish  (Da-kat.ogg) transcribed in kat as ; the Danish transcription escapes me, since the alleged "a" sounds to me almost like the American "a" in "men" or Czech "e". --Dan Polansky (talk) 09:13, 25 January 2014 (UTC)
 * 1) *Thanks for this Dan. I think the audio files of UK/US man:, cat: and map: are a good illustration of why so many British phonologists have moved away from . Ƿidsiþ 11:48, 25 January 2014 (UTC)
 * 2) * Note regarding the US audio file for man and map: The speaker exhibits æ-tensing and therefore pronounces them higher, with /eə/ or /ɛə/. The only example above of the American pure /æ/ is the audio file for cat. --WikiTiki89 15:15, 25 January 2014 (UTC)
 * This just illustrates the problems with regional variations. We each hear a different interpretation of the IPA symbols depending on our own pronunciation.  I don't see how the British "cat" above can be transcribed using /æ/.    D b f  i  r  s   00:04, 30 January 2014 (UTC)
 * Comparing the audio for Polish kat and UK cat, you can tell (in my opinion) that the Polish one is lower than the UK one. --WikiTiki89 00:58, 30 January 2014 (UTC)
 * Yes, I can tell that it is marginally lower, but the difference is very slight to my ear, whereas I hear a big difference between the British and American cats.   D <font color="#00ccff">b <font color="#44ffcc">f  i  r  s   08:04, 30 January 2014 (UTC)
 * , as long as it's done consistently. So technically I oppose both options failing, one should pass, I just don't really care which. 16:11, 25 January 2014 (UTC)
 * Abstain for now. Like CodeCat, I think we need to be consistent. If the RP pronunciation of one word (e.g. hat) were given as /hat/ because the editor who added it preferred /a/, while the RP pronunciation of another word (cat) were given as /kæt/ because the editor who added it preferred /æ/, it would convey to readers that the words had different vowels in RP. If cat and hat in fact have the same vowel in RP (regardless of whether that vowel is the same as the vowel they have in GenAm), we should use the same symbol in transcribing the RP pronunciation of both words. Thus, we should have a rule, either "use /a/ in RP" or "use /æ/ in RP", rather than leaving the transcription up to individual editors. On procedural note: in my view, the status quo of "use /æ/ in RP" continues unless either the "use /a/ in RP" option or the "leave the decision to each editor" option passes, and inconsistency results only if the "leave the decision to each editor" option passes. The structure of this vote almost implies that inconsistency is the status quo (the action of allowing inconsistency is labelled "take no action"), but this is not the case: /æ/ was used everywhere and was prescribed by the relevant appendices until the few /a/-introducing edits that led to this vote, so /æ/ is the status quo that continues unless there is support for changing it. (I want to make my opinion on that explicit, because the vote I am about to cast for /a/ will tie things.) - -sche (discuss) 19:52, 3 February 2014 (UTC) expanded 21:01, 9 February 2014 (UTC)
 * 1)  Changed my vote to "abstain" after a stay in London. /a/ is just too common now. --Anatoli (обсудить/вклад) 00:08, 18 March 2014 (UTC)

Decision
No consensus. Thus, no enforcement one way or the other for now, and the question can be revisited if and when it appears there may be some community consensus. &#x200b;—msh210℠ (talk) 05:26, 25 March 2014 (UTC)