Wednesday, 11 November 2009

diphthong, schmiphthong

A customer wrote to the publishers of the Longman Dictionary of Contemporary English to point out what she thought was an error in the list of diphthongs given in the Pronunciation Table on the inside front cover. The last two entries there read
She objected that these
ne sont pas des diphtongues (voyelle se modifiant en cours d'émission, prononcée en 1 seule syllabe) mais des voyelles non phonémiques /i/ et /u/ suivies d'un schwa /ə/, appartenant à 2 syllabes contiguës mais distinctes, malgré l'absence de consonne intermédiaire.
[are not diphthongs (a vowel modified in the course of production, pronounced in one single syllable), but non-phonemic vowels /i/ and /u/ followed by a schwa /ə/, belonging to two adjacent but distinct syllables, despite the absence of any consonant between them.]
Longman asked me to draft a reply.
In LPD I have a whole-page panel discussing what I call Compression. You could also call it Varisyllabicity. The point is that some English words are varisyllabic: the number of syllables they contain is variable. Thus actual can be pronounced as two syllables or as three; peculiar can be pronounced as three syllables or as four.
The sounds represented as and here can be, and frequently are, pronounced in a single syllable. Alternatively, particularly in slow, careful, overarticulated speech, they can be pronounced in two syllables.
If "compressed" into a single syllable, they are diphthongs. Daniel Jones treated them as a particular type of diphthong, rising diphthongs. I have sometimes called them crescendo diphthongs. (It may be difficult or impossible to distinguish them auditorily from the corresponding sequences of semivowel plus schwa, or .)
I would think you get the same alternation in French, for example in the name Louis, which is usually monosyllabic [lwi] but can also be said disyllabically as [lui].


  1. I should think you were a bit exasperated that Longman thought that merited a reply! "des voyelles non phonémiques /i/ et /u/"?? In solidi??

    You say Daniel Jones treated them as a particular type of diphthong, rising diphthongs, but as you say, only for wə or jə realizations. He conceived of them primarily as centring diphthongs, surely.

    Do you remember a competition in the Listener to write a short story with nothing but monosyllables? They disallowed 'real', with an appeal to the effect "Does anyone really think that's a monosyllable?" (if my memory is anything to go by). Obviously they had only etymological ears, and heard not!

  2. Rising is about where the vocalic and the consonantal parts are, I thought. Centring diphthongs come falling and rising.

  3. Yes Lipman, but canonically these centring diphthongs are falling ones.

  4. Most diphthongs are schmiphthongs, actually. Anyone remember who it was that said, "The syllable is not a phonological unit"?

  5. Probably misunderstood you - I thought you implied an opposition between centering diphthongs and rising (or falling) diphthongs.

  6. Some confusion here. ɪə as in near and as in sure are falling diphthongs for DJ, his numbers 20 and 21. But tautosyllabic (which he wrote ĭə), as in peculiar, and tautosyllabic (DJ's ŭə) as in actual, are his rising diphthongs. They are numbers 22 and 23 in the 8th edition of An Outline of English Phonetics. It is these latter two that the French lady is concerned about.
    All are centring diphthongs, which is not the issue.

  7. Furthermore, the writer is a francophone, for whom lune has two syllables (after all, it's CVCV, no?), and voyelles three. Their ears are entirely etymological.

  8. Anything that isn't centralising is very disturbing to the French mind.

  9. Dunno what she meant by "non phonémique /i/", but don't minimal pairs such as studded ~ studied show that /i/ *is* a distinct phoneme than /I/ after all?

  10. And what about triphthongs? There are speakers for whom ours is one syllable and hours is two, but for others higher and hire are homophones despite the morpheme boundary in the former...

  11. Are you talking about the second vowel in what's written happy and pronounced in MRP as huppee, sometimes regarded to be different from both /i(ː)/ and /ɪ/? Otherwise beat and bit or the like might be handier minimal pairs.

    I'm not sure any variety of RP that doesn't merge unstressed /ɪ/ with schwa makes a difference between studded and studied.

  12. What's the structural difference between ours and hours?

  13. John, it looked to me like a trivial misunderstanding between Lipman and me. I don't really see any confusion.
    I see no evidence that we don't all agree that ɪə as in near and ʊə as in sure are falling diphthongs for DJ, his numbers 20 and 21, and falling diphthongs of the centring variety at that. "Tautosyllabic" is a bit of a problematic word, if you sympathize at all with my above interpretation of your word "schmiphthongs", but I see no evidence that we don't also agree that his ĭə and his ŭə, numbers 22 and 23 are rising diphthongs, and rising diphthongs also of the centring variety.
    Moreover it looks obvious that he assigned them different numbers and wrote them like that because he still conceived of them as rising diphthongs as distinct from jə and wə, which I don't think any of us would want to call diphthongs at all. So please forgive the shorthand in my first post on this: I should have said "You say Daniel Jones treated them as a particular type of diphthong, rising diphthongs, but as you say, only for wə or jə -*type* realizations of centring diphthongs" (though we may just have to agree to disagree on whether DJ conceived of 20-23 inc. primarily as centring diphthongs which come in falling and rising varieties.) But you had made it quite clear in yesterday's piece that you were not talking about what I referred to above as the canonically falling centring diphthongs, but precisely those wə or jə -type realizations, i.e. DJ's notionally tautosyllabic dipthongs 22 and 23 and therefore in this case notionally both rising and centring, which perceptually shade into the corresponding sequences of semivowel plus schwa: wə or jə. In acknowledging that it may be impossible to distinguish them auditorily from those sequences, you may perhaps be saying that the only way they can be distinct at all at such a low level of distinctive realization is in terms of phonatory intention! I can't think what other criterion you might have for implying there is anything to distinguish at all in such cases.
    I suppose the most charitable assumption would be that it is your symbols for DJ's numbers 22 and 23 that the French lady is concerned about. (Why had I felt so sure it was a lady?) But who knows what *was* the issue for her? You say she was reacting to uə as in 'actual' and iə as in 'peculiar' in the list of diphthongs. We have been familiar with similar broad transcriptions of what are typically falling centring diphthongs for longer than I care to think, but I believe your motivation for using these particular transcriptions in LPD for these phonetically mercurial contexts was the same as for using i to represent the similarly moot final vowel of 'city'. But this lady certainly does *not* agree that they are centring diphthongs, or even diphthongs of any sort, as she says "appartenant à 2 syllabes contiguës mais distinctes", and I assume you agree that any attempt to make sense of "voyelles non phonémiques /i/ et /u/" is hopeless. In fact she makes so little sense in this quote that it is also moot whether she is aware of the concepts of centring diphthongs whether rising or falling at all, or whether she is as etymologically hidebound as the Listener judges I mentioned above.

    Lipman, like you I'm sceptical that any variety of RP that doesn't merge unstressed /ɪ/ with schwa makes a difference between studded and studied, and even more so that there are speakers for whom ours is one syllable and hours is two. Unless sporadically!

  14. Lipman, what do you mean by "MRP"? If it means "modern RP" or something, wouldn't speakers of it who have [i] for HAPPY distinguish "studded" and "studied"?

    Also, are we actually talking about merging unstressed /ɪ/ with schwa, or just about using schwa in certain endings? Speaking for myself (not an RP speaker though) I do have unstressed /ɪ/ in some places, just not in "studded" etc.

  15. MRP - mainstream RP (or mockneyfied, if you like).

    And that was exactly my question. But the person above might have been under the impression that it's actually studd[ɪ]d ~ studd[iː]d.

  16. Hello again, Lipman. By 'the person above' I take it you were referring to army1987, whom you asked "Are you talking about the second vowel in what's written happy and pronounced in MRP as huppee, sometimes regarded to be different from both /i(ː)/ and /ɪ/?" And I think that that second vowel in 'happy' was a good guess, and that to make a minimal pair of studded~studied it would take someone who has the vowel there for which JW has proposed /i/ as distinct from both /iː/ and /ɪ/, especially if they level-mix their morphology and phonology! Perhaps that will meet JHJ's point about your use of MRP, and that when you say "I'm not sure any variety of RP that doesn't merge unstressed /ɪ/ with schwa makes a difference between studded and studied" you're just talking about that sort of context.

    Systems with demotic modifications like that were at first called Advanced RP, and then I think Modified RP as they got increasingly modified. The young Royals were often cited as yardsticks. I was born aged and surrounded by ageds, and I have not got any younger, so I am used to mainstream RP meaning something quite different, but I don't think I would quite have had the temerity to call the MRP you are talking about mockneyfied. A nice mnemonic, though. My trouble is that I am so out of touch that I don't know how mockneyfied.

    I get the impression that there is a whole spectrum of mockneyfication. A good example for me (whose mockneyfication has not been far ahead of the Queen's) would be taxis/taxes/taxis/taxies, between which I make absolutely no distinction, probably because I have /ɪ/ in taxi, and plurals as well as pasts. I would be at a loss to say which combinations of those pop up as minimal pairs at various points on the continuum.

  17. Mallamb - if you're interested in how those four words are pronounced by a man in his 20s who self-identifies as an RP speaker, I could transcribe how I say them for you.

  18. I am agog, Leo! What a lucky strike to come here just now!

  19. Here are the raw data:

    taxis - ˈtæksiːz (plural of taxi)

    taxes - ˈtæksɪz

    taxis - ˈtæksɪs (as a Latin-looking word)

    taxies - ˈtæksiːz (what a plane does)

  20. Aargh! I was thinking "four?" I thought I had three. I remember I changed the order to make the sequence more intuitive, and I must have copied instead of cut! Perhaps I shouldn't have told you that before you came up with 4. I wonder what they would have been. If the singular structural taxis, Im sure it shouldnt be there for either of us, so may I have the plural of that, which at least looks as if it belongs there, and may belong with taxies for some people, and say that for me the opposition is between that and the other three?

  21. I see I hadn't refreshed the page. So of course you did include the singular structural taxis.

    Sorry, but what about the plural of it?

  22. Hang on, yes, I thought singular "taxis" (a word I hypothesised by analogy with "praxis"!) would probably end in /s/ for both of us. But if I pluralised this singular "taxis" to "taxes" I would get ˈtæksiːz, which is the same for me as the plural of "taxi" and also as the verb for what a plane does ("taxies").

    For me, then, those three are in opposition to "taxes" - ˈtæksɪz - the present indicative of "tax" and also the plural of "tax".

    What a tangle.

  23. No tangle at all, dear boy. (Why did I think all along that you were one? You could have been a Leonora. Observe how I asked of the onlie begetter of this discussion, "Why had I felt s o sure it was a lady?" What a sexist! Must get closer down to bedrock in the hunt for my flak jacket.)

    What it means to me is that you can reasonably be expected to have the same difference between studded and studied yourself (grabs tanzen, fortunately the one with traditional cotton padding against our ecologically recalibrated central heating), but although it appears we do agree on the plural of taxis, and that it at least looks as if it belongs there, and may belong with taxies for some people, once again I must ask "Are you sure?" You can never take too many reality checks. Sure the /iː/s are the same length, are you? Sure Gk taxes and noun/verb taxies are not a minimal pair?

  24. You were sure she was a lady because Professor Wells refers to her as "she" in his opening paragraph. I hope that puts your mind at ease!

    I certainly distinguish studded and studied. Remind me, who was it who had them as homophones?

    I think it would take a phonetics lab to tell my Greek "taxes" from my regular English ones, if indeed there is any difference.

  25. Mind duly at ease. Thanx. Would that such angst were always so observably misplaced. But think ladies~Hades. And then keep thinking. I don't think you'll need a phonetics lab.

  26. David Marjanović13 November 2009 at 23:39

    singular "taxis" (a word I hypothesised by analogy with "praxis"!)

    It really exists, at least in composites: google for phototaxis and chemotaxis for instance.

  27. Yes, I looked it up afterwards - a relief to find out it does exist.

    Mallamb - you're trying to influence my answers, I'm not falling for that!

  28. So didn't either of you trust me when I called it the singular structural 'taxis'?

    Trying to influence your answers, Leo? Moi? Just trying to save you the sweat of hot-footing it to the nearest phonetics lab.

    I did say "think ladies~Hades. And then keep thinking." But I would never put you at risk of monitoring yourself until you go into a synaptic loop and succumb to Unrealitätsgefühl! Can't you just think of people you know who might have different phonotactic structures for pairs like that?

  29. I think I say taxes (pl. of taxis as [ˈtɛksiːz] with a long i only when I want to avoid ambiguity. Or I would; can't seem to remember the last time that was necessary.

    Talking about Leo - the other day, I noticed I pronounced it the same as Lear. I think I use quite a variety from [liːɵʊ̯] to [lɪə], and the latter probably rather rarely. Nevertheless, prof. Wells, is this a normal smoothing phenomenon? This specific combination isn't common, with l[iːə]tard and l[e]pard.

  30. plural of "tax" [taksəz]
    plural of "taxi" [taksɪz]
    plural of "taxis" [taksiːz]
    (similarly "ladies" with [ɪ] and "Hades" with [iː])

    (Northern near-RP, born 1970s)

  31. Thanx Lipman and JHJ.

    "taxes (pl. of taxis as [ˈtɛksiːz] with a long i only when I want to avoid ambiguity"

    Ambiguity with what? You say with a long i, implying you are talking about ambiguity with the pl. of tax or taxi or both, but if you are with Leo on this you may mean only ambiguity with the pl. of taxi, in which case you may mean a longer i. Where are you from, that you say [ˈtɛksiːz] with ɛ and leopard with [e]? Your provenance may be germane to this.

    As for Leo, we need to know which one. The poster here says he's male, so Leo may well be his unabbreviated name, and I would find it hard to imagine this being indistinguishable from Lear, though a tendency for [əʊ] to be smoothed to [əː] was noted in the earliest reports of Advanced RP which I mentioned here somewhere, and it was even suggested at that point that this would lead to a linking r in e.g. 'slow up'. You may know him, and know that he is a Leopold or a Leofranc or something, but even those I would still expect to be abbreviated to [liəʊ] etc. The same applied to the name Leonora that I mentioned above.

    The only relevance I can see for 'leopard' is that you are telling us you have raising for both æ and ɛ.

    So JHJ your pls of taxis and taxi are the same as mine, so sure enough you have "ladies" with [ɪ] and "Hades" with [iː].

    "Northern near-RP" doesn't really click with anything particular in my experience. Is it from your family or your schooling? I have seen JW described as a Northern RP speaker, but the bracketing of that is obviously Northern  (RP speaker), and only Northern in origin at that. Nothing seems to have made any inroads on his RP!

  32. Ambiguity between sg and pl.

  33. I'm using "near RP" in the sense used by JW in Accents of English: roughly, an accent which shows enough regional features to be outside the usual meaning of "RP" (e.g. no trap/bath split) but which isn't perceived as strongly regional.

  34. "Ambiguity between sg and pl."

    If you say so, Lippo (or should that be Lipper?), but it doesn't seem very likely to be the only reason, what with the singular having final [s] an' all. If anything I hear more people these days giving these Gk and Latin pls not only a long vowel, but a secondary accent. Singulars too, like [daɪəˈbiːtiːz] increasingly have the etymologically correct long vowel, but I still plug away with [daɪəˈbiːtɪz]. I don't see how it can be influenced by my plurals, especially now I have so much more occasion to be influenced by those of the etymological persuasion fresh out of medical school (the doctors are all getting younger, like the policemen): I never in my life thought it was plural, and I have always had the long vowel, and probably the secondary accent, in Gk and Latin pls.

    JHJ, yes I thought that was what you meant. But you see the ambiguities. The nearest it came to clicking with anything particular in my experience was the old Manchester Grammar School. I don't know about now.

  35. "You may know him, and know that he is a Leopold or a Leofranc or something"

    Leofranc! Thanks for giving me a chuckle. But no, it's plain Leo, and it ends with [əʊ]. Nobody's ever called me [lɪə] or [liə] or [li:ə], all of which I guess I might hear as either "Leah" or "Lear" depending on context. In fact, I think my own "Lear" varies between [liə] and something like [lɪ:], whereas my "Leah" is always [liə]. Though I expect "Leah" and "Lear" must be consistently homophonous for at least some people.

    Plural of taxi vs. plural of taxis (or, ladies/Hades) - now that I think about it, a longer [i:] for the latter definitely seems intuitively plausible at least for somebody, though whether I make this distinction with any consistency I really can't tell.

  36. - Leo: I'm quite sure I don't usually pronounce it entirely smoothed 'n reduced, most of the time it's probably something like [lɪəʊ̯], but I "caught myself" saying [lɪə] the other day, so I was - and am - curious if this is very much out of the ordinary for smoothers or not.

    - [ɛ] and [e] - never mind that, it's probably more a matter of IPA interpretation.

    - relevance of leopard - to show that the specific combination of sounds is rare, and ME /ēō/ has other modern equivalents, too. (Could have written libbard, for that matter.)

    - Die a bee tease? Well, untreated diabetes certainly increases the level of sugar in your blood, but I'm not sure that's enough to attract those insects. Otherwise, English isn't Latin, and it's a bit strange to heed the "correct" lenghths when you're pronouncing ē as i anyway.

    - taxis, pl -es: I'd find it very awkward to make the distinction only through s versus z, sg [tæksɪs], pl [tæksɪz], even with the usual lengthening before -z. Brings us back to the question of some weeks ago, why the Greek and Latin -s is pronounced voiced in -es = [iːz].

  37. Do I detect some consciousness-raising going on?

    I think my own "Lear" varies between [liə] and something like [lɪ:], whereas my "Leah" is always [liə].

    [lɪː] is jolly interesting, isn't it? I first heard it from Australians and then it spread like wildfire. At least that's what I think happened. I certainly wasn't aware of it before the Australians. And perhaps 'wildfire' is wrong too. The influence of the Australian mafia on the media is more like some other irresistible force of Nature. And the melting pots of the big cities have plenty of firebreaks between them. If it takes over completely in BrE we shall have a true correlation of length with /ɪ/, and /i/ will have to find a new home.

    But if you get so far as [ɪː] are you sure you start from [liə]? Of course if your "Leah" is always that, it's the etymological thing I mentioned in the first post, isn't it? Only in this case it's justified.

    So these people for whom you expect "Leah" and "Lear" must be consistently homophonous, do you think they would have [lɪə] for both, or [liə]?

  38. Assuming they do exist, I would guess [ˈli.ə], as a disyllable. That is how I say Leah. And "near" would be [ˈni.ə] for these people of course.

  39. Sorry I missed your post, Lippers (for you did not enlighten me).

    - [ɛ] and [e] - never mind that, it's probably more a matter of IPA interpretation.

    FE – I've seen them assigned like that on and off over the years.

    Right, I did think it was a bit strange to heed the "correct" lengths in diabetes etc., but that's what these medics (actually of all ages) are doing. But it doesn’t seem to mean there has been a Classical revival! (The report on an attempted one here can't even get its subheading into Latin: and it has been claimed that the Rooneys were torch-bearers for another in calling their son Kai! Anyway these medics are as impervious to my pronunciations as I am to theirs: they say 'amaurosis few jacks' and I doggedly say 'amaurosis fugax'.

    - taxis, pl -es: I'd find it very awkward to make the distinction only through s versus z, sg [tæksɪs], pl [tæksɪz], even with the usual lengthening before -z.

    Well I certainly couldn’t see how you would be making it without any such lengthening, which is why I was puzzled. I'm still not quite quite sure what you're saying. Would you rather say [tæksiːs] or something for the plural? And then have say [beɪsiːs] for the plural of basis and [beɪsɪz] or [beɪsəz] for the pl of base?

  40. Most people pronounce it [fɪlɪp].

    I have
    bases (pl of base): [beɪsɪz]
    bases (pl of basis): [beɪsɪz], unless there's some ambiguity, then [beɪsiːz.]

    That so strange?!

  41. Nope, nil humani alienum a me puto. Indeed for once I may have caught on sooner rather than later. But I did choose that example because I thought it was by way of having "some ambiguity"! It's the old upper limit of distinctive realization again, isn't it? And you have one.

    I consider myself duly enlightened, and privileged to call you Philip. But you don't say whether you are Phillip. People are, you know. Pro tem, I will just consider myself provisionally privileged.

  42. ll. (Parents shouldn't name their sons like that and go on to send them to a school where they teach you Greek. Sooner or later your classmates will find out you're a fat aficionado rather than a lover of horses.)

  43. I've got an unetymological l too, and I'm a worse case than you because it is not the second but the fourth l: Llewellyn. This is because it's a surname (not hyphenated, so masquerading as a given name), and it's surnames that this seems to happen to, isn't it? So a back-formation from Phillips may appeal. Or you could take up Fat Liberation and fight the thinnist mafia.

  44. BTW I meant to dispute that I say "die a" anything. It's dire Beatties, like some dreadful Scottish family.