Monday, 12 March 2012

the Lindsey system

Some of you will have read a thought-provoking blog posting by Geoff Lindsey (trading as English Speech Services), in which he argues for a radically revised transcription system for the vowels of the English of England.

ɑj bɪlɪjv ɪts ɪmpootnt nɔʔ dʒəs tə dɪskəs transkrɪpʃn sɪstəmz əz ə θɪjərɛtɪkl (θɪɪrɛtɪkl) ɛksəsɑjz bət oolsəw tə trɑɪ ðəm awt ən tɵw ɪnvɛstɪɡɛjt haw ðɛj wəək ɪn praktɪs. səw hɪjə (hɪɪ) ɡəwz.

wɔʔ dɵw ɑj fɪjl (fɪɪl) əbawt ɪt ɪn dʒɛnrəl?

əpɑɑʔ frəm ɪʔs dɪskənsəətɪŋ ənfəmɪlɪjarətɪj, ɑj fɑjm mɑjsɛlf rɪjəlɪj (rɪːlɪj) lɔŋɪŋ tə bɪj əlawd sɪmblz fə tɵw fəwnəlɔdʒɪkl ɛntətɪjz ðət ɑjm əkəstəmd tɵw.

• wən ɪz ʌ. sɪns STRUT əŋ commA ə nɔʔ məədʒd fə mɪj, ɑjd lɑjk tə bɪj ɛjbl tə kəntɪnjɵw tə jɵwz ə sɛprəʔ sɪmbl fə ðə foomə. ɑjd bɪj hapɪj tə mɛjk ɪt ɐ rɑɑðə ðən ʌ. bəʔ dʒɛf dɪsmɪsɪz STRUT əz ə məəkjɵwərɪjəl (məəkjɵɵrɪɪl) ən fəŋkʃən lɑjʔ ɡrɛjl (fə wɪtʃ hɪj wɪnz tədɛjz mɪkst mɛtəfə prɑjz).

• ðɪj əðər ɪz ðə happY/schwee sɪmbl i. fə happy, wɪtʃ dʒɛf wɵd rɑjt hapɪj, ɑjd lɑjk tə bɪj əlawd tə rɑjt hapi. (ɑj ɪmadʒɪn ðət “hapɪ”, ðəw fənɛtɪklɪj akjərəʔ fə mɪj, wɵb bɪj sɪjn əz ɪnsəfɪʃnʔli mɔdn əm mɛjnstrɪjm.)

• dʒədʒɪŋ bɑj rɪjakʃənz ɑj tɪpɪklɪj ɡɛʔ frəm nɛjtɪv spɪjkəz, ɑj θɪŋk mɛnɪj pɪjpl wɵd wɔnt ə spɛʃl ɔw fə ðə vawəl (vaal) əv cold, soul, fənɛtɪklɪj vɛrɪj dɪfrənʔ frəm ðɛɛr əw əv coat, so.

ðə stɛjtəs əv smɵwðd vawəlz ɪz nɔʔ klɪjə (klɪɪ) ɪn dʒɛfs sɪstəm. ɑjm θɪŋkɪŋ nɔt əwnli əv ðɪj ɑjə → ɑɑ əv science, lion ən ðə awə → aa (?) əv power, sour bət oolsəw əv ðə NEAR ən CURE vawəlz (vaalz). am ɑj səpəws tə rɑjt here əz hɪjə oor əz hɪɪ? fə jury (ɪf ɪʔ rɑjmz nɑjðə wɪð story noo wɪð furry) ʃal ɑj rɑjt dʒɵwərɪj oo dʒɵɵrɪj? oo dəznt ɪʔ matə? səw fɑɑr ɑjv ɡɪvn jɵw bəwθ pɔsəbɪlətɪjz. bət ɪf wɪjə (wɪɪ) ɡənə jɵwz ðɪs sɪstəm ɪn tɛksbɵks ən dɪkʃənrɪjz wɪj nɪjd tə mɛjk əp ɑɑ mɑjnz wɪtʃ wən wɪj prɪskrɑjb əz ðə noom. ɑjd lɑjk tɵw ɔpt fə ɪɪ ən ɵɵ, bət ɑj səpəwz ðɛj wɵɡ kɔnstɪtʃɵwt ən ənnɛsəsrɪj kɔmplɪkɛjʃn fər ɪj ɛf ɛl pəəpəsɪz.

ɑj dəwnʔ θɪŋʔ ðə sɪmbl ɵw ɪz wɛl tʃəwzn. frəm ə pjɵwəlɪj (pjɵɵlɪj) praktɪkl pojnt əv vjɵw, ɑjm əfrɛjd ðə lɛs ɛkspəət jɵwzəz əv fənɛtɪk sɪmblz wɵd fɑjnd ɪʔ dɪfɪklʔ tə kɪjp ɵw dɪstɪŋkt from əw, oor ɪndɪjd bəwθ əv ðɪjz dɪstɪŋkt frəm aw. ɛnɪjhaw, ðə dɪfrəns bɪtwɪjn GOAT ən GOOSE ɪz moo ðən dʒəs lɪp rawndɪŋ ɔn ðə fəəst ɛlɪmənt əv ðə dɪfθɔŋz. ðəz ə dɪfrəns ɪn vawəl (vaal) hɑjt. səw ɑj θɪŋk ə bɛtə sɪmbl fə GOOSE wɵb bɪj ɨw.

wɔʔ dʒɵw θɪŋk?

kɔmɛnts (ɪn transkrɪpʃn, plɪjz, ɪf jɵw kən manɪdʒ ɪt) wɛlkəm.

63 comments:

  1. Bət mɛjbɪj ðɪs ɪz ə mɪzəndəstandɪŋ,

    Eh.

    Bt mɛebɪ ðɪs ɪz ə mɪzn̩dəstæˑndɪŋ - it's not just about the transcription but also about the system, and you don't speak the accent it represents. Of course, for happY and cold, your arguments are very valid in my view.

    (Sorry I didn't write this in transcription, but it would take too much time.)

    ReplyDelete
  2. Ɑɪ θɪŋk ðæt ʃwiː (ᵻ) ænd ʃwuː (ᵿ) ʃʊd biː ədɔptᵻd ɪn ðiː Ɑɪ Piː Eɪ. Bʌt Ɑɪ doʊnt əˈɡriː ðæt ɪt meɪks sɛns tu rəɪt "t" æz ə ɡlɔtəl, ðoʊ.

    ReplyDelete
  3. Not only would əw and ɵw be difficult to distinguish in the handwriting of 'less expert users of phonetic symbols', but even in this printed text I found myself constantly misreading one for the other.

    ReplyDelete
  4. These are different issues. concerning the t, the question is what Mr Lindsey wants to describe, and if the accent he has in mind has a glottal stop where other accents have a t or a k, it makes perfect sense.

    ReplyDelete
  5. ɑj əgrɪi ðəʔ ɵw ɪz badlɪi tʃəwzn, fər ənəðə rɪjzən: ɪf ðə fəəst ɛləmənt ɪz sɛntəd ən əwpənd frəm trədɪʃənəl ʊ to ɵ (wɪtʃ mɛjks sɛns tə mɑj mɑjnd), wɑj ɪz ðə sɛkənd ɛləmənt stɪl w (kɔrəspɔndɪŋ tə trədɪʃənəl u)? ʃoolɪj səmθɪŋ lɑjk ɵɥ wəb bɪj moo kənsɪstənt.

    oolsəw, wɑj haz ðə lɛŋθ dɑjəkrɪtɪk bɪjn rɪplɛjst wið dəbəl vɑwəlz? ɑj dəwʔ sɪj ðə pɔjnʔ.

    ReplyDelete
    Replies
    1. fə joo lɑɑst kwɛstʃn, jɵw kɵd sɪj ðə ðɪs vɪdɪjəw http://www.youtube.com/watch?v=kN9ywVglcCk

      Delete
  6. [aɪm ˈvɛɹi ˈhæpi tə ˈsɪi ðəʔ GOOSE ɱ̩ FLEECE ɹ̩ ˈfaɪnɫ̩i ˈtɹɛənskɹaɪbd əz ˈdɪfˌθɔːŋz, ˌɹæðɹ̩ ðn̩ ˈlɔːŋ ˈmɑnəfˌθɔːŋz. wʌɾ aɪ ˈdoʊn ˈlaɪk ɪz ðə jʉus əv j/w ɪn ˈdɪfˌθɔːŋz.
    aɪ əˈgɹɪi ðət STRUT n̩ commA ʃəb bɪi ˈkɛpt əˈpɑɹt, sɪns ˈfjʉu ˈpɪipɫ̩ ˈmɹ̩ːʤ ðəm kəmˈplɪiʔli ɪn ə ˈsɪikwəns laɪk son of, bət bəˈfoːɹ ə ˈpɔːz (e.g. Rosa, BrE teacher), ˈmɛni ˈpɪipɫ̩ ˈjʉuz ðɹ̩ STRUT ˈsaondz.]

    ReplyDelete
  7. fər ɪj ɛf ɛl, ɪts ə kmplɪjt nɔnstɑɑtə. ðɪj əwnlɪj wɛj ðət ənjɵwʒəwəl sɪmbəlz aktʃɵwəlɪj hɛlp ləənəz ɪz ɪf ðɛj droo ətɛnʃən tɵw ən ənɛkspɛktɪd sawnd.

    wɛn ɑj stɑɑrtɪd ɪn ðɪj ɪj ɛf ɛl trɛjd, ðɪj əwnlɪj ɛgzɔtɪk sɪmbəlz rɛgjələlɪj jɵwzd ɪn tɪjtʃɪŋ mətɪjrɪjəlz wə ʌ ənd ə. ðə kənvɛntʃn əv mɑɑkɪŋ vawl lɛŋθ təw ɪmplɑj vawl kwɔlɪtɪj mɛj nɔt əv bɪjn dɪskrɪptɪvlɪy akjɵwrət, bət ɪt tʃɑjmd wɪð ðɪj ɛkspɛktɛjʃnz prədʒɵwst bɑj ðɪj ooθɔgrəfɪj əv ɪŋlɪʃ — and, ɪn mɛnɪj kɛjsɪz, ðə ləənəz ɛl wən. ən ɛksɛptʃn wəz ɔftən mɛjd fə ðə trap vawl — æ~ a:, nɔt a~ a:. ðɪs wəz tə rɪmɑjnd ləənəz nɔt tɵw sawnd tɵw kɔmen — ɑj ɪj nooðən.

    aksənts əðə ðən ɑɑ pɪj dɪdnt bɪlɔŋ ɪn ðat jɵwnɪvəəs.

    əpɑɑt frəm əbandənɪŋ ðɪj ɛdwoodɪjən ɛɪdʒ, ɪj ɛf ɛl həz oolsɵw — fə ðə məwst paat — dɪtʃt fənɛtɪk transcrɪptʃn. ðə rɑjz əv tʃɪjp pəəsənəl aksɛs tɵw rikoodɪd sɑʊnd wəz lɑɑdʒlɪj rɪspɔnsɪbəl.

    ReplyDelete
  8. I think I'd prefer a compromise between the two systems.

    Specifically, I'd definitely adopt ɔ for LOT, o: for THOUGHT, ɑɪ for PRICE, and ɛ: for SQUARE.

    On the other hand, I'm not too keen on the j/w in diphthongs. And I agree with others, Geoff should do something about GOOS and GOAT being too similar in his system.

    ReplyDelete
    Replies
    1. Would you adopt oɪ for CHOICE? As I said on Geoff's blog, I've not noticed any change in the CHOICE vowel, but he says that its first element has moved in the same direction as THOUGHT.

      Delete
  9. ðɪs sɪstəm oolsəw pɵts ɪŋglɪʃ ɪn ðə tɑjpəlɔdʒɪklɪj rɛɛ pəzɪʃən əv havɪŋ ɪ wɪðawt i, wɪtʃ sɪjmz ɑɑtɪfɪʃəl sɪns FLEECE rɛndʒɪz frəm ɪjii.

    ReplyDelete
  10. If I understand correctly, Geoff's blog is about a transcription system for EFL learners (and, I suppose, for their teachers (if not more so)), transcribing the modern common British (exemplified by his choosing sound clips of actors in their 20s and 30s) to aim for as an EFL speaker, not to transcribe accurately, or even being able to describe accurately, the speech of the entire British population. So I must say I find John's objections a bit moot. I do share though the objection to the ɵ symbol as being to similar to ə, and perhaps ʉ is a better choice.

    ReplyDelete
  11. ðə trəbl wɪ ðə GOOSE vawl ɪz ða? ðɛɛr ɪs naw ə ðʒɛndə splɪt. ɑj ðɪŋk ða? ɨw ɪz əwnlɪ sɵwtəbl foo jəŋ fɪjmɛjlz. ɪ? mɑj? bɪj bɛtə tə jɵws ʉw foo mɛn.

    ReplyDelete
  12. (Sorry, don't have time for IPA).

    If ə is to be used for both unstressed schwa and the stressed STRUT vowel, then it is important to indicate stress location. I notice that neither JW nor anyone else commenting so far has done this, which is one reason (among many) why I find them difficult to read.

    As a concrete example, dɪskəs could equally be "discuss" or "discus".

    Admittedly, the same objection applies to ɪ used for stressed KIT and unstressed rabBIT. But the STRUT-schwa merger makes the ambiguity significantly worse in practice.

    (For me, like JW, schwa and STRUT are not allophones, and I hear a voice inside my head screaming "nooooo!" whenever ə is used for STRUT, but I realize that I'm probably in a pretty small minority in this respect).

    ReplyDelete
    Replies
    1. If ə is to be used for both unstressed schwa and the stressed STRUT vowel, then it is important to indicate stress location. I notice that neither JW nor anyone else commenting so far has done this, which is one reason (among many) why I find them difficult to read.

      ᵻndɪjd.

      Admittedly, the same objection applies to ɪ used for stressed KIT and unstressed rabBIT. But the STRUT-schwa merger makes the ambiguity significantly worse in practice.

      ɑj lɑjk ᵻ, az mɑjkəl ɛvəsən sɛz.

      Delete
    2. I think this STRUT-schwa merger is the result of an EFL approach, because hardly any native speakers pronounce their STRUT and word-internal commA sounds alike. On the other hand, a lot of people pronounce the final sound of comma as unstressed STRUT before a pause, which makes commA — frankly — quite an unfortunate choice to represent the schwa. (Perhaps something like campUs would have been better.)

      So it's not that people pronounce their STRUT vowels as schwa, but the other way round: /ə/ gets realized (before (and for some ever after) a pause) just the way their STRUT is. That's one less reason to favor ə as the outcome of the merger, which isn't even a real merger then, just a contextual neutralization.

      Delete
    3. hardly any native speakers pronounce their STRUT and word-internal commA sounds alike

      You mean ‘any native speakers in England’? Even in Wales such a merger is normal AFAIK.

      Delete
    4. No, I meant any native speakers in the world. Wales is an exception, indeed, and as far as I know even their merged sound isn't [ə], but something like [ɜ]. Still, the number of such speakers is almost negligible when compared to all the native speakers of English in the world.

      Delete
    5. Really? I was under the impression that a lot of America has the same for STRUT and COMMA, even phonetically, not just in terms of positional allophones.

      Delete
    6. @teardrop:

      [ə] for STRUT is not uncommon in the USA.

      Still, the number of such speakers is almost negligible when compared to all the native speakers of English in the world.

      I think not.

      Delete
    7. No. Word-internal /ə/ is for most AmE speakers [ə], a true mid-central vowel. STRUT on the other hand is usually more retracted and/or more open than [ə], somewhere between [ɐ] and [ʌ], or [ɤ] and [ʌ]. For some speakers, it is closer to [ɜ], which is still more open than [ə] is, though the audible difference is not so huge between them, it's realitvely easy to misidentify them.
      In a word like fungus the two vowels are qualitatively different. I really doubt any speaker pronounces the first vowel as close as [ə] in spontaneous speech.
      In a word like umbra before a pause, the two vowels are most likely identical: both are STRUT, neither [ə].

      Delete
    8. John W, can you tell us?

      Delete
    9. There are plenty of speakers in the English midlands and in Wales who have identical qualities in the two syllables of fungus.

      Delete
  13. ɵw is the most confusing part of this system, since phonetically this suggests something not very different from əw, but in reality it's probably more like ʉw, starting with a close central vowel, for most southern-english speakers that I know.

    Also, I kept getting annoyed reading his article and it using the terminology "Standard British" instead of something like "modern RP". That's the picky scot in me, I guess.

    ReplyDelete
  14. dʒɛf sɛz ða? NEAR ənd CURE kən bɪj sɪjn az FLEECE + ə ənd az GOOSE +ə. ðɪs sədʒɛsts [ni:ə] ənd [kju:ə] tə mɪj, bə? nɪjðə iz nooməlɪ sɪjn az RP. hɪj adz ða? ðɛj kan bɪj smɵwðd tə [nɪ:] ənd [kjɵ:]. iz ðis nɔt ə tʃɛjndʒ ɪn aksɛnt rɑɑðə ðən smɵwðɪŋ?

    (I'm pretending that I speak in the sort of accent that Geoff's describing rather than applying his symbols to my own speech)

    ReplyDelete
    Replies
    1. That's an empirical question: do people distinguish dimorphemic seer from monomorphemic near, at least potentially? (Bet the answer is “Some do and some don't.”)

      Delete
    2. I don't distinguish near and seer, but I would distinguish cure from sewer. I think that [kju:ə] sounds old-fashioned. [ni:ə] is still alive and well, but I don't think that it would be used in the south of England.

      Delete
  15. aj əˈgri wɪθ ˈkɪliən ðət wi kən ˈownli dʒɐdʒ ə trænˈskrɪpʃən ɪf wi now wɐɾ ɪts ˈpɚpəs ɪz. ɑr wi trajɪŋ tə ˈrɛndɚ wɐɾ ə spɪˈsɪfɪk ɪndɪˈvɪdʒuəl sɛz ɑn ə pɪˈtɪkjəlɚ əˈkejʒən, æz ajm brɔdli duɪŋ ˈhiɚ, or ɑr wi ˈdʒɛnrəlajzɪŋ ˈovɚ ˈiðɚ or bowθ dɪˈmɛnʃənz? ɑr sɪmˈbɑlədʒi wɪl ənˈɛvəɾəbli bi ˈdɪfrənt diˈpɛndɪŋ ɑn haw mɐtʃ ˈditejl wi nid tə blɚ.

    ɪn ˈɛni kejs, aj priˈfɚ ˈjuzɪŋ ˈsɛmivawəl ˈlɛtɚz (or ɛls ði nɑn-ˈsɪlæbɪk dajəˈkrɪɾɪk) tə dʒɐst ˈrajɾɪŋ tu ˈvawəlz ənd ˈlivɪŋ ɪɾ ɐp tə ðə ˈridɚ tə now ɚ gɛs wɪtʃ ɪz ðə glajd -- vɛri ˈɪrətejtəŋ! aj ˈɔlsow θɪŋk ɪts ˈvɛri əmˈportənt te juz ðə ˈprajmɛri strɛs mark; ɪts ə ˈfownim lajk ɛni ɐðɚ, ənd aj fajnd ɪt ˈdifəkɐlt tu rid ˈlɛŋθi trænˈskrɪpʃənz wɪðˈawɾ ɪt. For ˈtɛknəkəl ˈrizənz ajv əˈvojdəd dajəˈkrɪɾɪks ɪn ˈrajɾəŋ ðɪs, ðow ajd lajkt tu əv juzd ðə ˈsɪlæbɪk dajəˈkrɪɾɪk ɪnˈstɛd əv ən and əl ɑn sɛvrəl əkejʒənz.

    ReplyDelete
    Replies
    1. When I see two consecutive vowels in a transcription, I always assume the second one is non-syllabic unless told otherwise. FWIW, the examples on official IPA chart don't mark the second half of the diphthong in phonetician specially, and explicitly mark the syllable break in react.

      Delete
  16. It would interest me if you chaps agree that the FOOT-vowel (the short 'oo') has really become something like the French e-muet in 'le' etc., as Lindsey pretends, in modern RP. Just curious---I don't hear modern RP (or for that matter ancient RP, other than from my ancient 'English-for-beginners', the '50ies, LPs) very often.

    His system is a bit 'gewoehnungsbeduerftig' as the German says---you have to get used to it, but marginally more than any more traditional transcription, methinks. Besides, it's sort of funny to read one's English as he really speaks it, that's a very unEnglish phenomenon. And I also don't like the identification of STRUT with the schwa. But one gets used to everything, as Sappho says.

    ReplyDelete
  17. wɛn ɑj sej wɔt dʒɛf əz retən əbɑwt ðə KIT vɑwəl, ɑj əm wɔndəreŋ wɛðər jɵ kɵdənt rɛjz ə semələ kwɛstʃən əbɑwt ɪ lɑjk ðə 2012-03-06 pəwst rɛjzd əbɑwt ɱ. ɑɑ ðɛɛ rejəle lanɡwədʒez ðət hav ə desteŋktev /ɪ/? ət lejst en ðə fjɵw endəwjərəpejən dɑjəlɛkts ɑj nəw, ðə trənskrepʃən əv ðə vɑwəl betwejn /i/ ənd /ɛ/ sejmz tə bej ənrelɛjted tə fənɛteks. enstɛd, e ez tʃəwzən ef ðɛɛr ez ə trədeʃənəl əsəwsejɛjʃən weð ðə lɛtər ⟨e⟩, ənd ɪ ez tʃəwzən ef ðɛɛr ez ə trədeʃənəl əsəwsejɛjʃən weð ðə lɛtər ⟨i⟩.

    əf koos, dʒɛf ez rɑjt ðət jɵwzeŋ e ez pətɛntʃəle meslejdeŋ. səw hej tʃɵwzez praktekəl (oo trədeʃənəl) əwvə fənɛtek rejzənz. fraŋkle, ɑj θeŋk ðə sɛjm pojnt kan bej mɛjd əz wɛl əbɑwt ðe əðə sɑjnz he z tʃəwzən: ðɛj ɑɑ pətɛntʃəle meslejdeŋ ənd ə moo trədeʃənəl trənskepʃən wɵd bej hɛlpfəl, oor ejvən ə sestǝm lɑjk trɛjɡər ənd blɔk's (ɑj həwp ðat's hɑw ðɛj ɑɑ prənɑwnst). ɑj lɑjk ðət et's ə radekəl prəpəwzəl, bət en mɑj əpenjən, et's nɔt praktekəl ət ool.

    ReplyDelete
    Replies
    1. I'm not sure what you mean with "indoeuropean dialects" (languages?), but it seems to me that it is generally accepted that most Germanic languages at least distinguish between /i/, /ɪ/ and /ɛ/. Perhaps you don't know many Germanic languages except for English, but in my native language of Dutch, and if we can trust Wikipedia in German and the Scandinavian languages as well, there is a clear difference.

      Delete
    2. Kilian,

      in older styles, a let us call it this way 'noble mannerism' was in currency to call languages of a certain family 'dialects', for instance 'West Germanic dialects' (English, Frisian, Dutch, German, roughly), Uralic dialects (Finnish, Hungarian, Estonian, Udmurt and so on), Romance dialects (French, Spanish, Italian...) and so on. It's out-dated and out-moded nowadays, so small wonder you have problems with it.

      Concerning the gist of what j. mach wust is saying (sɛjeŋ), I must confess I too am sort of hard put to make much sense of it. Maybe the Indo-European dialects he knows are really 'few', as he calls them himself. But 'though this be a madness there is method in't', as we know, and it first springs to mind that what he means is the distinction between /i/, /ɪ/ and /e/, and not (as in your reply) /i/, /ɪ/ and /ɛ/. Second, one Germanic language I know (of) which doesn't seem to have a real /ɪ/-/e/ distinction is Danish; that is, for aught I know, all words with a /ɪ/ in related words in other Teutonic languages have either /i/ or (most often) /e/ in Danish. But the Danish /e/ is extremely high and narrow, and unless you have a very well Danish-attuned trained ear you can hardly distinguish it from what you from other languages know as /ɪ/. So j.mach wust has a point, probably.

      Delete
    3. ɑj wɔz rɑɑðər θeŋkeŋ əv laŋgwedʒez sətʃ əz eŋɡleʃ oo dʒəəmən. ɑj nəw ðət ðɛɛr ɑɑ trənskrepʃənəl sestəmz əv eŋɡleʃ ðət əpəwz /ɪ/ (foo ðə KIT vɑwəl) ənd plɛjn /e/ (foo ðə FACE vɑwəl). hɑwɛvə, ɑj hav səm dɑwts wɛðə sətʃ ə desteŋkʃən ez rejəle bɛjzd ɔn fənɛteks oo wɛðər et ez dʒəst kapetəlɑjzeŋ ɔn rɛdele əvɛləbəl sembəlz. ðə KIT sɛt bejeŋ ə tʃǝkt vɑwəl ənd ðə FACE sɛt ə frej vɑwəl, ɑj səspɛkt ðɛɛr ɑɑr əðə fənɛtek kjɵwz sətʃ əz lɛŋθ ðət defərɛnʃejɛjt ðə tɵw sɛts moo klejəle ðən vɑwəl hɑjt, ejvən ðoo ðejz əðə kjɵwz wɵd nəw lɔŋɡər əlɑw ə wən-sɑjn trənskrepʃən. ðej entənaʃənəl fənɛtek alfəbɛt ez nɔt əbɑwt ǝlɑweŋ wən-sɑjn trənskrepʃǝnz, bət əbɑwt rɛprezɛnteŋ fəwnejmz.

      ɑj kənfɛs ɑj havənt dən mɑj həwmwəəks lɵkeŋ əp wɛðə ðɛɛr ɑɑr ɛne laŋɡwedʒez ðət fejtʃər ən ɔpəzeʃən ǝv /ɪ/ ənd /e/ ðǝt ez nət ǝkɔmpəned bɑj əðə fənɛtek defrənsez.

      dʒəəmən ez oolsəw trədeʃǝnǝle transkrɑjbd weð /e/ ənd /ɪ/, bət əz en ðǝ kɛjs əv enɡleʃ, ðə desteŋkʃən ez rejəle ekɔmpəned bɑj ə tʃɛkt-frej desteŋkʃən ðət manefɛsts en desteŋktev lɛŋθ, ənd ɑj dəwnt dɑwt ɑj kɵd fɑjnd əgɛn ðə foomənt ənalezez ðət ʃəwd /e/ tɵ bej ət lejst əz kləwzd əz /ɪ/.

      en dɛjneʃ, əkoodeŋ tɵ wekepejdejə, ðej əlɛdʒd desteŋkʃən əv /e/ ənd /ɪ/ ez əkɔmpəned bɑj ə desteŋkʃən əv strɛs. səw ǝgɛn, ðɛɛ kən bej nəw trɵw ɔpəzeʃən bǝtwejn ðǝ tɵw sɑwndz.

      Delete
    4. I find it extremely difficult to plow through such texts, especially if you feature various idiosyncracies, such as the one you are championing.

      You must be really enamoured of phonetics to write and, what's more, read such-like, but you folks surely are.

      But in Danish there simply ain't no /ɪ/ adall, this is at least the official theory I have been acquainted with for like 30 years or maybe 35, doesn't matter. I must add that I do not really have much Danish, just 'knowledge by acquaitance'. Anyway, where from your Average Standard Germanic you would expect an /ɪ/, for instance 'spille', play, they have an /e/, /sbelə/. This sprang to my mind as I was reading your first post. But this /e/ of theirs sounds to my ears extremely high, very much like your /ɪ/. All their front vowels are as it were lifted upwards or whichever way you put it, for instance their /a/ sounds like SAE /ɛ/ (it used to be an ash-vowel, but is no longer).

      Delete
    5. I know seem to understand why mr. j. mach wust referred to Indoeuropean languages as 'dialects': he is a germanophone Swiss and thus for him language is by definition dialect. The Germanophone Swiss speak various dialects, a dialect, for instance St. Galler Swiss-German, is their first language, so small wonder that they have a tendency to see all languages as 'dialects'. I must add that I personally quite like Swiss German 'dialects'. Other than that, a language is just a dialect with an army and a navy, as a Jewish-Russian I believe linguist once put it... . Doesn't quite work for Luxemburghish, for they don't have a navy (or do they?)

      I am saying this of course a bit tongue in cheek, but not quite...

      Delete
    6. @j. mach wust: I find it rather silly to say that /I/ and /e/ aren't distinguished in German because there's more to the phonemes than height alone. It is very clear that they are separate phonemes, and there's plenty of good reason to ascribe them different phonetic symbols, especially because /I/ pairs with /i/, and /e/ with /E/, not /e/ with /I/.

      Delete
    7. (Sorry, I don't have my IPA keyboard layout available at the moment.)

      Why, I don't doubt that the German phonemes traditionally transcribed as /ɪ/ and /eː/ are distinct phonemes. I am just saying that there is no phonetic reason for using two distinct signs -- you might as well write /e/ and /eː/. Indeed, I think I have seen formant analysis that proved German /ɪ/ to be at least as open or even slightly more open than /eː/ -- exactly the opposite of what the traditional suggests.

      Of course, there is more to transcription than just phonetics, but also the association of phoneme pairs that Kilian has pointed out. However, this does not justify the choice of two distinct signs. What are the other of "plenty of good reasons", Kilian? I can imagine tradition, the mere availability of the [ɪ] sign, or the wish for having one-sign transcriptions for all phonemes (without [ː]). I agree that all of these are valid reasons, but none of them is a phonetic reason that would justify a distinct [ɪ] sign.

      Imagine the IPA had no [ɪ] sign (only [i] and [e]) and German were not a well-known Middle European language, but an exotic Middle New Guinea language. Would the suggestion to introduce a new IPA sign [ɪ] for the description of that language stand any chance? Of course not, since people would rightly point out that there is no need for a new IPA sign. The existing signs [i] and [e] along with [ː] are perfectly sufficient for transcribing the exotic Middle New Guinea language known as German: /iː/ /e/ /eː/ (even though that obscures some consonant pairs, but hey, exotic vowel systems have exotic pairs). You might as well repeat the same story with English or other European languages. That's the IPA's traditional bias towards European for you.

      I have now done the homeworks and looked through Ladefoged and Maddieson's (1996) The Sounds of the World's Languages. They mention no distinctive /ɪ/, but only systems with distinctions of /i/ /e/ /ɛ/ (or /i/ /ɪ/ /ɛ/). While they mention one possible language with a five height vowel system (without confirmation), even that language has only the distinction between /i/ /e/ /ɛ/ (p. 289 s., Amstetten Bavarian dialect).

      Delete
    8. I'm not sure this is correct on a purely phonetic level, but anyway: what do you make of short /e/ in learned words and loans? Genom is different from *ginnom and *gehnom.

      Delete
    9. Short answer: That's an easy one: I would doubt that! ;-)

      Long answer: The status of unstressed vowels in learned or loaned words is marginal at best. I suspect that unstressed syllables in such words may have a complementary distribution of /eː/ (shortened to [e]) and /ɪ/ (using the traditional transcription without phonetic base): /eː/ never before consonant clusters, /ɪ/ only before consonant clusters. At least, a word like *ginnom feels unlikely to me and I can think of no likely example.

      Perhaps Million vs. *Melion? To me it feels like if there is any distinction between the two words, then it is rather a distinction of syllable cut, that is, with the /l/ in Million being closely tied to a preceding /ɪ/, giving rise to something like /mɪlˈjoːn/ [melˈjoːn] – again a consonant cluster –, while /l/ *Melion would have a loose tie to the preceding sound, thus giving rise to a trisyllabic pronunciation /meːliːˈoːn/ [meliˈjoːn]. So the example is no good because of the possible [lj] cluster. Any suggestions for an unstressed /ɪ/ not followed by a cluster?

      Delete
    10. The short close phonemes are marginal, of course, but where they exist, /e/ is different from /ɪ/, unless you're speaking a Ripuarian dialect of the "Setzen Se sech" variety. How about Melone and *Millone?

      Delete
    11. (Ɑjm əɡən at mɑj ɑjpejɛj kejood!)

      Ən ənstrɛst /ɪ/ fɔləwd bɑj ə seŋɡəl kɔnsənənt – əz en Millone – stel fejlz ənlɑjkle tɵ mej. Mɑj səspeʃən ez ðat /ɪ/ ez nɛsəsrele fɔləwd bɑj ə kləstə (Miltone, Miljone etsɛtrə) ənd ðət befoor ə seŋɡəl kɔnsənənt, əwnle frej vɑwəlz kan əkəə (Milone, Melone). Ef ðat səspeʃən ez trɵw, nəw ɔpəzeʃən wɵd bej pɔsebəl.

      Əf koos, ə trɵw ɔpəzeʃən betwejn /ɪ/ ənd /e/ dəzənt hav tɵ bej en dʒəəmən. Ɑjd bej vɛre hape tɵ sej eɡzampəlz frɔm əðə laŋɡwedʒez (ef Ladefəwɡed ənd Madezən prɔvɑjd nəw sətʃ eɡzampləz, et dəzənt ðɛj dəwnt eɡzest).

      Delete
  18. I notice several commenters are using the question mark symbol ? for the glottal stop instead of the actual glottal stop symbol ʔ. This makes it harder to read, because every time we encounter a question mark, we have to figure out if it's a question mark or a glottal stop. Using the correct symbol would make reading punctuated transcriptions much easier.

    Notice the question at the end of the original post (and other questions within it)

    wɔʔ dʒɵw θɪŋk?

    That would be much more problematic to read had he used the same symbol for the glottal stop and for the question mark.

    ReplyDelete
    Replies
    1. Ellen,

      I share this concern of yours. I also can't help feeling these glottal stops are somewhat overused. Do the respective authors really speak like that all the time? As an American, you must feel a bit strange with so much glottal-stopping around.

      I also sort of miss the distinction between 'light' and 'dark' l in all (?) of the above transcribed texts. Maybe it's gone in modern RP, as it seems to be in modern General American? Well, in Estuary English they famously say 'peopw' and so on... . Or maybe the reply is that that distinction is somehow irrelevant? Yet, for the benefit of a foreign learner it should be maintained, I think...

      It also perplexes me that John is writing: wɪjə (wɪɪ) ɡənə (we are going to), I used to think this t-deletion (innernational airport) is something American. Maybe to make American readers like you feel more at home?

      Delete
    2. Wojciech

      1. On the one hand, we have the opposition between two 'substantial' allophones — i.e. abstractions that find concrete realisation in articulated sounds. On the other hand we have the opposition between a substantial allophone and a multi-purpose marker of style and/or syllable boundary.

      A dark l in English has no status as anything other than an allophone of /l/. Nor, indeed, does a clear l.

      A glottal stop performs many different functions in English — and not consistently for all speakers of all accents. For different speakers it may be various things, including
      • a marker of an initial vowel
      • a reinforcement of a consonant
      • a substitute for a consonant
      And it may perform the function in various ways, such as
      • in few or many speaking styles
      • at word boundary or syllable boundary

      In short, the occurrence of a glottal stop is not predictable. Knowing the speaker's accent and the current style of speaking makes it more predictable, but you're still likely to be surprised occasionally.

      By contrast, you only have to listen for a little while to a speaker to know when he or she uses a clear l and when a dark. The norm is for clear in leaf and dark in feel and speakers with a different distribution soon stand out as exceptions — quite often predictable from what you know of their accents. After a while, there are no surprises.

      2. Taking Ellen's example (and losing the Lindsay notation)

      • omitting the ʔ
      wɔdʒu: θɪŋk or even wɔddʒu: θɪŋk would suggest a rather unusual pronunciation

      • keeping the t
      wɔtdʒu: θɪŋk is pretty well impossible
      wɔtʃu: θɪŋk or perhaps rather wɔttʃu: θɪŋk is authentic, but a different pronunciation

      Of course, we non-Americans can retain the t, but then do you stays as two syllables — at least that's how it works in my speech.

      3. To cite wɪjə (wɪɪ) ɡənə (we are going to) is a little misleading. John was actually saying — and listening to himself saying — if we're going to use this system in textbooks and dictionaries. The sheer length of the clause, and the relative unimportance of the information at the start of it, place more pressure to reduce to gənə. I bet there are contexts and styles in which John keeps the t. In fact, I bet there are plenty of Americans who 'reinsert' the t in certain contexts or styles.

      Delete
    3. ad glottal stop

      I used to think that glottal stop is somehow 'marked' in the sense that the more thereof you have (as a Londoner, Glasgovian, and such) the more 'popular', maybe (I am using this term with much uncertainty as I am not really initiated to such issues) 'working-class' your pronunciation sounds. Is that quite wrong?

      Re dark l and clear l. I know. But I was saying: for the benefit of a foreign learner of English (much of the whole fuss with transcriptions and stuff is for his/her/its benefit, isn't it?) I remember in the middle of a discussion on the existence or otherwise of vowel length in Italian (I am nearly obsessed with this topic, but have good reasons to be so) I said 'you can sound perfectly correct in Italian without drawling your second 'a' in 'italiano'' and someone sneered at me saying (writing) 'yeah, and you can sound perfectly correct in English with a clear 'l' in your 'well''. Well ... (dark l)... you can't. At least if you want to sound standard RP or standard General American. The other way round: darkening all your l's in RP (in GA it's the default option)---I cannot say how you then sound. There is a certain l-darkening tendency in some variants of spoken Polish and I can tell you that it sounds a. very 'folksy' b. --- de gustibus ...--- I would say ugly.

      Re 'wɪjə (wɪɪ) ɡənə', yes, you're right, I suppose. But then the question arises, at least to this commenter: shall we use phonetic transcriptions in such a way as to give account of the most subtle variations, due to 'relative unimportance of the information' and other such --- which will make life hard to our readers? Or only the most 'basic' ones, whatever they be? 'How shall we these dayes wryte, 'eggys' or 'eyren'?' as Caxton would have said.

      Delete
    4. Wojciech

      I suspect that all the uses of glottal stop have spread in recent decades both geographically and up the social spectrum. This may have been amplified as sensitivity to hearing them has also spread.

      I asked my wife whether she thought it would help Russian students to mark the light and clear l's in transcription. She thinks not.

      'yeah, and you can sound perfectly correct in English with a clear 'l' in your 'well''. Well ... (dark l)... you can't.

      Here's a famous/infamous TV interview of a politician who consistently uses clear l almost everywhere. (I think I detect a darker l in the word bawl). Listen to Mail, whole, dismissal In another clip I heard Michael Howard say Well with an extremely short clear l — virtually a stop. But I posted the link to this interview because it's considered a classic.

      And listen to this irish comedian saying Bible, rules, still, gospel, while, fall etc. he does say Well once — but the l is sort-of intervocalic in wɛliˈtʰru: for Well, he threw that together.

      There are times when reduction such as gənə are the actual teaching objective, not just a peripheral complication. Reduction of going to wouldn't be very high on my list of priorities, but after has, have, had, are, was, were, can, could, should it probably deserves a mention.

      Delete
    5. 'I asked my wife whether she thought it would help Russian students to mark the light and clear l's in transcription. She thinks not.'

      in Russian all 'l', unless post-iotised (if I may coin a word) are dark. Hence small wonder. In Polish by contrast, at least in 'good' Polish, all of them are clear. Our people say 'well' with a clear l, deplorably. To them, a marked distinction in transcription would be helpful, I'd insist.

      Re gənə, OK, I've got your point.

      Delete
    6. David

      I have seen/heard part of the Howard interview by Paxman and I can only say: Heavens beware us from learning/teaching English pronunciation from such flickers. Howard's face is a 'dead pan', one almost thinks he suffers from an ailment or such. The overall impression is very very off-putting, at least on me, to put it mildly.

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

    ReplyDelete
  20. The most disconcerting aspects seems to hang on

    • the starting point of the GOAT vowel
    • younger Southern English speakers' pronunciation of the STRUT vowel

    My belief is that neither is of any great interest or importance to foreign learners. Yes, the old use of o: when I started teaching had the drawback of encouraging some students to use a o sound from their own language. But that's not the most serious of errors — provided, of course that they make a distinction with the vowels of LOT and THOUGHT.

    The sound that it's most useful to symbolise is the commA vowel — because we really want students to master reduced forms of has, was etc etc. It may be a bit if a fiction that ə is exclusively a short vowel in English, but it's a very useful myth.

    Moreover, a distinct symbol for the STRUT vowel is highly useful to remind students not to use any sound reminiscent of orthographic u in their own language. Using a different symbol from the commA vowel accords with the way many of us speak and, more importantly, it signal a stressed vowel.

    The ʌ~ə distinction may not be as real as we pretend, but it helps to remind students to avoid GOOSE and FOOT and to assign stress or not as appropriate.

    A distinct symbol for the TRAP vowel was important once upon a time, but I agree with Linsey that it's no longer vital to remind students to dissociate the sound from that of orthographic a in their own language.

    The problem with ɑj, ɛj, and especially ɪj is that they don't look like anything in any language — and oj is unhelpful to most. The same goes for ɑɑ, ɛɛ, əə — except possibly for students with double vowels in their native language, such as Finns. And oo is a digraph in English, which is surely bound to confuse.

    Yes, the symbols will work if students can be persuaded to learn them thoroughly — but that's true of any system. What worries me is the confusion they'll cause while students are still trying to learn the system — assuming that they'll ever co-operate.

    ReplyDelete
  21. CORRRECTION

    It may be a bit if a fiction that ə is exclusively a short vowel in English, but it's a very useful myth.

    Sorry! What I meant to say is exclusively an unstressed vowel.

    ReplyDelete
  22. 'The ʌ~ə distinction may not be as real as we pretend, but it helps to remind students to avoid GOOSE and FOOT and to assign stress or not as appropriate'\


    much as the Lindseyan collapsing of STRUT into schwa displeases me, I'd observe here that our people (Pl) tend to pronounce the latter vowel as /a/, ('tend' is an understatement here), thereby merging it with TRAP, sadly, the culprit being in part the symbol 'ʌ', which looks like a funny 'a' to them. Maybe someone should invent another symbol, not reminiscent of either the above or the schwa symbol.

    In our spelling we have quite a few 'aj's and 'ej's and 'oj's. Swedes too employ 'ej', for instance in 'mej' (me), 'dej', (thee), 'tjej', (girl), 'tejp' (tape, esp. adhesive tape) and a few others. ɑj, ɛj are different, I know, but they are close enough to suggest the right pronunciation. The Dutch are famous for their 'ij's (/ɛj/, actually). They too have double vowels, as do Germans, albeit to a smaller degree. The Lindseyan cause is not yet lost.

    ReplyDelete
    Replies
    1. our people (Pl) tend to pronounce the latter vowel as /a/, ('tend' is an understatement here)

      That's precisely why I'd like them to have ʌ before them as a constant reminder that it should be kept free from the memory of anything related to anything in the orthography.

      Actually an a pronunciation isn't such a disaster — provided that students don't confuse it with TRAP, PALM, BATH or commA. Confusion with FOOT wouldn't matter — if it wasn't something stigmatised (by some) in native speakers.

      The transcription can't teach a suitable pronunciation — at least not in the sort of classrooms that I've experienced. But it can remind students of what the teacher has taught them.

      In our spelling we have quite a few 'aj's and 'ej's

      But not ɑjs and ɛjs, that was my point.

      The Dutch are famous for their 'ij's (/ɛj/, actually).

      Again, not ɪj. And the ɛj value is a distinct hindrance, not a help at all.

      They too have double vowels, as do Germans, albeit to a smaller degree.

      I've forgotten what Dutch or Afrkaans I ever knew. I'm familiar with German aa (and, of course, the word Afrikaans) and German ee. Again they don't use ɑ or ɛ, and in the latter case the spelling is reminiscent of the wrong sort of e-sound. Plus the fact that ee, like oo, is a digraph in English spelling with unhelpful sound-correspondences.

      i:, u:, ɔ: may not be ideal, and a:, e:, o: are actually misleading to some extent. Nevertheless, the colon convention is less distracting than a new symbol to be learned and the symbols are only similar to spellings, in no case identical.

      Using ɑ: rather than a: is an improvement — closer to what phoneticians do — without being too much of a distraction to the student as he or she gets a better grasp of phonetics. the same goes for using ɑ, ɛ, ɪ in diphthongs. The same goes for using ʊ rather than u.

      The NEAR and SQUARE vowels call out for a ə element for when we represent the intervocalic (including linking-r) value. I used to be happy with and , but the adjusted vowel symbols are more 'accurate' and don't get in the way of recognition by students.

      Frankly, any attempt to represent the GOAT vowel is going to place a learning load on students. Some things are difficult because they're difficult, which I think is the case here. So I suppose it's best to seize the nettle — get the difficulty over quickly by learning the symbol early, whatever it is. Either that or stick with the misleading o:.

      An extra bonus is that any representation with : or two letters points unequivocally to a stressed value. It would be nice if a similar set of symbols pointed to unstressed vowels. This is only possible with ə— all the more reason to reserve the symbol for the commA (or lettER) vowel.

      Delete
    2. David

      'Actually an a pronunciation isn't such a disaster — provided that students don't confuse it with TRAP, PALM, BATH or commA.'

      actually, they do confuse it [STRUT] with TRAP, that's the problem.

      I agree that the Lindseyan transcription is sort of weird, and I too am in favour of the colon. I just tried to do justice to Mr. Lindsey, pointing out that his method is not _that_ crazy as it might seem.

      Delete
    3. Wojciech

      actually, they do confuse it [STRUT] with TRAP, that's the problem.

      And the only way i can see transcription being of any help is if:

      FIRST
      Teachers demonstrate the different sounds and give students plenty of practice.

      SECOND
      Students are frequently reminded of what has been taught by the symbol ʌ for STRUT words. It doesn't matter what symbol is used for TRAP words, providing that it is different. For your students there seems to be a case for persisting with æ.

      I'm not saying that Lindsey's system is 'crazy'. It's an elegant solution to the wrong problem.

      Delete
    4. David

      Polish students make an /a/ out of STRUT, whatever little or much they hear of modern RP makes them make an /a/ out of TRAP as well, and given that they hear General American most of the time they also make an /a/ out of LOT. Cat, cut, cot = /kat/. In this sense I agree with your FIRST and SECOND above. The symbol /ʌ/ has deplorably come to signify (if anything, then) an /a/ to them too, it also looks like a funny 'A' without the horizontal bar. For this reason, the Lindseyan merger of STRUT and schwa, tho' false in and by itself, even in America, could have (had) beneficial didactic uses, like for instance making them aware that STRUT is not really an /a/.

      Delete
    5. Maybe someone should invent another symbol, not reminiscent of either the above or the schwa symbol.

      WTH would be wrong with /ɜ/ for STRUT (provided we don't use it for NURSE too, but WTH would be wrong with /əː/ for NURSE?)?

      Delete
    6. Wojciech

      As I keep insisting, phonetic transcription can't help students who haven't already mastered the sound in question. (Students of a language, that is. It's completely different for students of phonetics.)

      I strongly suspect that there are many students who will never derive any help from it.

      Delete
    7. Army1987

      the trouble is, we do use it for NURSE most of the time.


      David

      OK, I understand and least of all am I inclined to think that phonetic transciption by itself can teach anything. I am privately quite happy with the traditional transcription. But as I have written earlier on, part of the blame for the wrong pronunciation of STRUT by most Poles is carried by the traditional symbol for that vowel, which is perhaps all too suggestive of /a/.

      Delete
    8. @ Wojciech: The merger of STRUT with schwa may be confined to only a few areas (e.g. Wales), but it would not cause any misunderstandings across Britain (and probably not anywhere in the English-speaking world). However, using [a] for STRUT does cause misunderstandings (especially in the north of England) and should be avoided.

      In addition, I don't think that you've realised how far the glottal stop has spread in Britain. The transcriptions in the comments are not using it especially often. It is very widespread for /t/ when not at the start of a syllable and there is no sign of the trend's being reversed. I remember reading in Petyt (1985) how the glottal stop had become very widespread in Huddersfield, where it had not previously existed, within just a generation.

      Delete
    9. Ad Ed

      Ad primum: yes, that is exactly why the Polish habit (and not just Polish) of equating STRUT with /a/ has always infuriated me, and why I have been opposing it all my lifetime.


      Ad secundum: no I have not realised (how far the glottal stop has spread in Britain), I am going by my impressions based on scanty materials. I have never had much exposure to living British speech). But I believe you of course.

      Delete
    10. @Army1987:

      If you wanted to distinguish STRUT and schwa but otherwise stay within the spirit of the Lindsey system, then /ɜ/ for STRUT, /ɜɜ/ for NURSE and /ɜw/ for GOAT might be a reasonable solution. This way the /ə/ symbol really would only occur in unstressed syllables, with occasional exceptions such as one pronunciation of "Ms" or where a weak form gets stressed, which I do sometimes.

      For my own accent, which I think is close to but not really part of (too Northern, though I don't have FOOT=STRUT) the family of accents he's talking about, I do want a schwa/STRUT distinction.

      I do think that if ɵ and ə are confusingly similar then that is a fault of the IPA not of Geoff Lindsey. He does seem to be describing a vowel very much in the area of the IPA symbol ɵ.

      Delete