Tuesday 7 February 2012

the quality of SQUARE

I tend to assume that my readers and correspondents have the kind of basic knowledge of linguistics that would involve an understanding of how the term “phoneme” is used (despite the fact that more sophisticated phonologists may well consider that the concept of the phoneme is unsustainable and based on ultimately untenable theoretical assumptions).

But some of the queries I receive show that this is not necessarily the case. Or at least people are not comfortable with the convention that slashes / / are used to enclose symbols for phonemes, but that symbols for speech sounds ~ realizations ~ allophones ~ variants properly go inside square brackets [ ]. (Or you can do as I do in this blog and merely embolden phonetic symbols without reference to their phonological status unless relevant.)

I would like to know your point of view about // and /ɛ:/.

Cruttenden lists [the change from the first to the second] among the "changes almost complete". Also Collins and Mees, in their Practical Phonetics and Phonology, have already opted for the /ɛ:/ symbol since 2003. The open /ɛ:/ is the only one found in Upton's ODP too. […]

I was wondering why both in your LPD and the EPD this alternative pronunciation is not transcribed. I might be wrong, but in my opinion, the /ɛ:/ phoneme could now be considered as part of the phonetic inventory of current BrE.


This question is of course not about phonemes but about the realization and notation of one particular phoneme (the SQUARE vowel).

Here’s what I replied.

I am of course well aware of the monophthongal variant of the SQUARE vowel. Please read what I wrote at http://www.phon.ucl.ac.uk/home/wells/ipa-english-uni.htm .

In LPD I took the decision to be conservative in phonetic notation for RP, sticking for example with Gimson's EPD symbols for the TRAP and GOAT vowels. Upton has made a different decision (including a very regrettable notation for the PRICE vowel).

I personally have a centring diphthong as my usual pronunciation of SQUARE in most phonetic positions. Perhaps that shows my age.

I shall not be changing the transcription in any future edition of LPD, but I agree I ought to insert a note to the effect that many people use a monophthongal quality. This, however, is not a new "phoneme", as you seem to believe. It is an alternative realization of an existing phoneme. There is no possibility of words being distinguished by the choice between the monophthongal and diphthongal variants of SQUARE.

Was I being too pedantic? If so, it’s what comes of a lifetime reading and commenting on students’ essays.

Anyhow, the take-home message is that monophthongal SQUARE is fine. But not obligatory.

27 comments:

  1. "I tend to assume that my readers and correspondents have the kind of basic knowledge of linguistics"...

    Why would you assume that?

    ReplyDelete
    Replies
    1. I don't know if you want a serious answer to that question?

      For the same reason that - say - David Attenborough would assume that his readers/viewers know what an "insect" is, or that a writer on religion would assume that readers know what the word "incarnation" means.

      You can't explain everything as if talking to a five-year-old. You've got to assume some previous familiarity with the subject.

      Delete
    2. 'a writer on religion would assume that readers know what the word "incarnation" means.'

      gosh, you are being exacting. Incarnation! Probably the most difficult theological concept of all, pure nonsense to all religions beside Christianity. (1 Cor. 1:22).

      You picked a poor example, as col. Pickering would have said.

      Delete
    3. I thought it was a good example, precisely because for the more sophisticated reader it is controversial and difficult (just like "phoneme").

      Delete
    4. Yes, this is how I understood your intention Yet, I'd say you did 'des Guten zu viel', as far as controversiality and difficulty is concerned. Anyway, your 'drift' should be clear to everyone, I supposed.

      Delete
  2. "... my readers and correspondents have the kind of basic knowledge of linguistics that would involve an understanding of how the term “phoneme” is used (despite the fact that more sophisticated phonologists may well consider that the concept of the phoneme is unsustainable and based on ultimately untenable theoretical assumptions)."

    An experience which I personally have made (I have never received any formal training in linguistics, I must admit), as well as several other persons known to me with a similar background (academic, but not in linguistics) is this:

    you try to understand 'of how the term “phoneme” is used', and not just behaviouristically, but you try to understand the rationale behind this term and its use, and something like the esssence of the corresponding (theoretical) entity, and then after several attempts you discover precisely that

    'more sophisticated phonologists [not just 'may well consider' but do consider] that the concept of the phoneme is unsustainable'.

    in other words, that linguists do not really know what a phoneme is supposed to be.

    Well---a similar problem bedevils other theoretical terms in all disciplines, you may say... . And this is certainly not quite wrong.

    In the case of the exchange quoted by John, however, the difficulty is located on a much more shallow level, and the person should have informed his- or herself about the use of various brackets, without penetrating in the 'heart of darkness'.

    ReplyDelete
    Replies
    1. I had all my [] changed to // by the publishers for an article that I wrote for the Yorkshire Dialect Society's transactions recently. Somebody didn't understand the distinction.

      Delete
    2. copy-editors are often 'smarter than the radio' as we say here (Pl), or than the author. Maybe a good idea is to append, or rather prepend, a 'Symbol Explanation' sheet (cardboard) to your ms, with all crucial symbols explained in 36 pt type.

      Delete
    3. Don't know that saying -- what is it in Polish? I'm sure I can find a use ;)

      As someone who serves as a copy editor for this kind of stuff from time to time -- I have to say, to keep the balance, that some authors tend to have a hard time deciding what they want. Not to mention being sketchy about the shapes of the symbols in their own manuscripts...

      Delete
    4. mądrzejszy od radia. May be be somewhat 'seasoned', though. Goes back to the (mythical) times when people tended to believe everything they heard on the radio.

      Re authors you are right, too! Some authors copy in a more or less ape-like manner fancy symbols from other authors and are, as a result, not able to explain to the copy-editor just WHAT that squiggle is supposed to be. I know such cases.

      Delete
  3. I'm afraid that this is the price one pays for using an IPA-derived system in a dictionary (or an introductory coursebook etc.). For some reason, the principle of representing specific phonetic realisations using specific symbols seems more compelling (or at least more easily acceptable to the people who happen to find out about it) than the //[] business. Phonemes are abstractions, after all.

    I've seen quite a few discussions stemming from this, including on this very blog, Language Log, etc., etc. One particular group that is inclined towards this kind of thing, and for well-grounded reasons, is language teachers; but I may be being unfair here.

    (And, on the Attenborough-incarnation front, I think a blog is, well, a forum for popular science. What you can assume is an interest in the topic. Knowledge of more advanced abstract stuff -- not necessarily.)

    ReplyDelete
    Replies
    1. Wjarek

      In defence of language teachers, I would claim that most of us have a good-enough instinctive grasp of what a phoneme is — though only those with specific training could articulate it. We engage with dictionaries and course books that use transcription as sources of phonemic information. It's the blooming phoneticians who confuse us!

      I've done more introductory courses in Phonetics than most — three of them specifically for language teachers and two of those three by proper academic phoneticians — so I think I can generalise. What students in teacher-training courses concentrate on learning, whether they realise it or not, is how to articulate a model realisation of each English phoneme — with maybe a very few pairs of allophones. They have no motivation for distinguishing the realisation of a phoneme in the reference accent from the realisation in other accents. The first transcription I learned used a rather than æ and u rather than ʊ. It was perfectly adequate for my purposes as a teacher of English — as long as the materials I worked with used the same symbols.

      But people in general may be interested in accents, language teachers and non-language teachers alike. There used to be a BBC forum where several non-specialists likes to discuss accents. That's when I discovered the nature and value of John's lexical sets. They allowed us to write about sounds on a message board that wouldn't accept IPA. Nobody analysed it, but what we were doing was to exchange hard phonological information and highly impressionistic phonetic information.

      Problems arise when specialists exchange hard phonetic information which is read by amateurs with no formal training and by people like me with serious but circumscribed training. We punters think we understand — indeed, we do understand quite a lot of it, but sometimes less the we think.

      Delete
    2. "In defence of language teachers, I would claim that most of us have a good-enough instinctive grasp of what a phoneme is — though only those with specific training could articulate it. We engage with dictionaries and course books that use transcription as sources of phonemic information. It's the blooming phoneticians who confuse us!

      ...

      We punters think we understand — indeed, we do understand quite a lot of it, but sometimes less than we think."

      The above I emphatically agree with.

      I am even farther from the 'inner circle' in that I teach neither English nor anywhere near a Modern Languages Department, yet I quite often have to either teach _in_ English or _about_ English, or about some other language Modern or Ancient, but _modo obliquo_ as it were.

      For this reason I strongly encourage my students to look up and even purchase John's dictionary and similar dictionaries, all the while bewailing the very low level of foreign language instruction in Poland---and not just there. I also try to raise that level, for as much as my feeble powers go.

      Yet when it comes to using John's dictionary and their brethren, you have to explain to the students the various brackets, IPa alphabet, phonemes, allo- and other -phones and what not--you have to get them to acquire that intuitive grasp David mentioned---within one hour or so.

      If the 'blooming phoneticians' can be of much help here, even if they try very hard, I don't know, I am rather pessimistic.

      Delete
    3. Wojciech

      Students need to think about 'sounds'. These are in fact phonemes, though it's usually not helpful to use that term. As far as students are concerned, sounds should correspond with 'special letters' or even 'phonetic letters' — i.e. IPA symbols as used in their dictionary or whatever.

      The exotic symbols are surely a help rather than a hindrance. Use of ʌ for example is a signal that the 'sound' is nothing like the sound represented by u in Polish, or whatever.

      For students like yours, John's dictionary gives too much information in one particular respect. In order to represent both British RP and General American, he uses a wider range of symbols than would be needed for just one dialect. You could save time by ignoring the standard which your own accent doesn't aspire to.

      I see no reason why students like yours should be bothered with allophones or with square brackets. And if you don't need to distinguish phonetic from phonemic there's no real need for slashes either.

      As for 'intuitive grasp' of what a phoneme is, your students may not know it but they have an intuitive grasp of what Polish phonemes are. That's why I'm confident they can be led into understanding a concept called English sound.

      Where the blooming phoneticians come in is to help us to understand what we will never want to teach our students, for example:
      • the difference between the articulation of phonemes (i.e. realisations of phonemes) in RP and General American
      • the difference between articulations in other accents that we chance to hear and which catch our interest
      • the differences and similarities between speech sounds which are realisations of English phonemes and speech sounds which are realisations of, say, Polish phonemes

      OK you may want to demonstrate differences to your students, but I can't see why this should be anything but a-theoretical.

      I suppose I'm saying that phonemes are all that concern teachers while they're teaching. The blooming phoneticians are of little use then. But when we come home and try to follow the interesting discussions held at places like John's Blog, that's when we need to understand life beyond the phoneme — which is my answer to the question raised in john's OP.

      To be fair to the phoneticians, what they've done very clearly and comprehensively is to describe the articulation of speech sounds in reference accents — insights that can help us to help student of just about any type and competence to improve his or her pronunciation.

      Delete
    4. Ad David,

      thank you. I am in agreement with most of what you're saying except these:

      the phoneticians will enlighten us on why won't ever want to teach our students:

      • the difference between the articulation of phonemes (i.e. realisations of phonemes) in RP and General American

      This I will want to teach my students by all means, since what they hear is GA most of the time, while I more or less approximately and roughly inculcate RP to them, if this be the word for it. Or at least what I know as RP, the 'good ole', with a clear cat-cut difference and all... . In GA it's like 'cayairt' and 'curt', respectively.

      • the differences and similarities between speech sounds which are realisations of English phonemes and speech sounds which are realisations of, say, Polish phonemes

      Oh yes, this is very interesting too, contrastive phonetics, in what way the TRAP-vowel difffers from both the Polish 'e' and the Polish 'a'.

      Maybe I am not bright enough this evening?

      Delete
    5. Wojciech

      • the difference between the articulation of phonemes (i.e. realisations of phonemes) in RP and General American

      I realise now that I was really thinking of vowel sounds. These are very difficult to describe in articulatory terms that make sense to the average student. But surely your own demonstration of contrasting British and American pronunciations would be enough. My attempts at an American accent are pretty pathetic (except, perhaps when singing and copying a native singer) but I would happily dictate word pronunciation from John's dictionary (or any other with American pronunciations). For students' purposes it would be good enough.

      Consonantal differences are more straightforward — except for the tap/flap articulation of /t/ etc.

      contrastive phonetics, in what way the TRAP-vowel difffers from both the Polish 'e' and the Polish 'a'.

      I think of that more as contrastive phonemics. Frankly, I'd simply discourage your students from using a Polish /e/. Their instinct is presumably to use a Polish /a/ because of the association with letter-A. This may sound 'foreign' to some native speakers, but no real bar to comprehension.

      I once worked for an exiled Polish entrepreneur in Italy who made a big deal of æ in his teaching materials — both the sound (He spoke like Daniel Jones) and the symbol (The students learned transcription and did pronunciation homework). I never saw the point and — except when on 'best behaviour' — used my own TRAP vowel, which is closer to Polish /a/.

      Delete
    6. contrastive phonetics, in what way the TRAP-vowel difffers from both the Polish 'e' and the Polish 'a'.

      I think of that more as contrastive phonemics. Frankly, I'd simply discourage your students from using a Polish /e/. Their instinct is presumably to use a Polish /a/ because of the association with letter-A. This may sound 'foreign' to some native speakers, but no real bar to comprehension.


      I'd think they use [a] not because of the association with the Polish letter, but because the contemporary English (England's English) æ sound sometimes more like our a. I discourage them, however, from using either our [e] or our [a] for your [æ], but the exact quality of [æ], and this is phonetics, I believe, escapes them most of the time.

      The Australian/South African [æ] is very much like our [e], the GA [æ] is diverse, but often like our [eja] or similar. But England's English --- well, that's difficult.

      Delete
    7. Wojciech

      because the contemporary English (England's English) æ sound sometimes more like our a

      This is where I believe phonetics (as opposed to phonology) does a disservice to language teaching. If it sounds like [a], then it is [a] and it isn't [æ]. For me there are two useful ways of labelling this phoneme:

      1. the traditional 'short a' — which conveniently doesn't identify the sound, since we say 'ʃɔ:(r)t eɪ'

      2. with reference to John's lexical sets — 'the TRAP vowel'

      I generally avoid speaking of [æ]. if forced, I would say 'the ash sound'. Phoneticians have precise ways of labelling, but I would never use them on specialist students of English, let alone the non-specialists that you teach.

      [John, for example, writes of
      'a front nearly open unrounded vocoid, approximately halfway between cardinals 3 and 4']

      If the problem is one of reconciling what students hear to what's written in reference works, I would use wording like 'where the dictionary has the letter ash' or, if you prefer 'an ash symbol'.

      Thus I would translate

      The Australian/South African [æ] is very much like our [e]

      as

      The short a/the trap vowel/the vowel shown by letter ash in your dictionary [one or more of these labels] in Australian and South African accents sounds like Polish [e]'

      You could add that likewise to many British people it sound like a 'short e'. There was an old joke the definition of sex

      'Sɛks are what they use to deliver coal.'

      It can still be funny if you substitute something else for coal.

      PS A different vowel, but the same sort of raising. A New Zealand physiotherapist, who i was meeting for the first time, spoke the other day about what I heard as strictures. It turned out that she'd said stresses.

      Delete
  4. I am not sure I understand this post correctly. Am I right that Upton in Oxford Dictionary of Pronunciation used /ɛ:/ for SQUARE and something else for NURSE?
    I am thinking of the North of England where stir is homophone of stare. That is another topic then?

    ReplyDelete
    Replies
    1. Yes. For NURSE he writes əː
      (Do not confuse the symbols ɛː and ɜː !)

      Delete
    2. That's only in a few confined areas of the North. Leeds, Bradford, Sheffield, Manchester and the far north all have "stare" and "stir" different.

      Delete
  5. For the record, I don't think you were being at all pedantic. But then we did not take the decision to change the symbol for the phoneme in EPD, either.

    ReplyDelete
    Replies
    1. Ha! That looks like it was changed but someone else took the decision! It's still the same symbol in EPD18.

      Delete
  6. This comment has been removed by the author.

    ReplyDelete
    Replies
    1. Well, I'm an EFL student (native Spanish speaker) and it seems that in your link (http://www.phon.ucl.ac.uk/home/wells/ipa-english-uni.htm) one of the recurring arguments in favour of an "agreed scheme" is not to make EFL pronunciation training more difficult. However I don't think that a common standard is always preferable for the sake of clarity. The "agreed scheme" you talk about will have to adapt eventually to newer, well-established changes (I'm sure you agree with me on this).

      Have a look at the British and American pronunciation examples of "pair" in this online dictionary:
      http://oald8.oxfordlearnersdictionaries.com/dictionary/pair

      Is there anything more confusing than a diphthongal transcription for a monophthong and a monophthongal transcription for a diphthong? You might say that the editors should have chosen oral articulations which more closely resemble the "agreed scheme" of phonemes. But I'd rather lexicographers at least mentioned the possibility of such a widely heard realisation for /eə/ as [ɛː].

      Ten years ago I was struggling to come to grips with phonetics, and I thought I had some type of hearing impairment that prevented me from listening to a centring diphthong in British pronunciations of words like "scarce" or "pierce", while I was able to hear an r-coloured schwa in American speech, which supposedly lacked centring diphthongs. Aware that the British pronunciations I came across could have been allophonic, regional or even socially-stigmatised realisations, and that the phonemic transcriptions in dictionary entries provided little help on this matter, I read the entire introductions of LPD (2nd edition) and EPD (16th Ed.) in search of an answer, but I found nothing. Words like "Europe", "fairy" and "teary" complicated matters futher for me, as I supposed that the presence of a fully pronounced "r" somehow overlapped the schwa I couldn't hear, but I found no evidence of this either in my two much loved pronouncing dictionaries. I started reading some books on English pronunciation and it was Gimson's Pron. of Eng. (revised by Cruttenden) that provided me with many answers, but I only found this book at the age of 23, and the inconsistency between the "agreed scheme" and what I was able to hear persisted for most of my English-learning years.

      I understand your decision not to include the monophthongal variant in every entry, as [eə] and [ɛː] do not contrast. And I don't entirely oppose the idea of conserving the traditional /eə/ to represent this phoneme, given that I'm simply a student and I respect the opinion of linguists. But I'll be really glad to see more detailed notes on pronouncing dictionaries that update and criticise the conservative standard used, as they would have helped me out a great deal during my training years and such observations wouldn't have hurt in previous editions. I'm sure future students will benefit from this inclusion. I can't help but feel empathy for your "phoneme-illiterate" enquirer.

      Cheers.

      Delete
  7. I've posted the same comment twice and it only appeared once I deleted it. Blogspot is pulling my leg!!!!!

    ReplyDelete

Note: only a member of this blog may post a comment.