Friday, 6 July 2012


Some people thought I was too cutting yesterday about Classic FM. (I’m spending a lot of time here in hospital listening to FM radio on my mobile.)

So how about this? Yesterday we had Verdi’s “ ˈdiːs ˈɪəreɪ”, i.e. Dies Irae. The announcer pronounced the first word of the title as if it were the German demonstrative rather than the Latin noun, (which I would pronounce in English as ˈdiːeɪz, though I know some would use s rather than my z).

At least he didn’t take it as the English word dies daɪz.


  1. How exactly does one know if a vowel is long in Latin except when it precedes two consonants?

    1. From verse scansion. If you know the intended metre, you know the length of every syllable.

    2. This presupposes, unless the method be circular, that you learn the Latin word first in the context of a poem scanned to you by a Latin teacher who knows somehow which syllable is long and which not.

      How do you English actually scan Latin verses? Do you actually try to mimick the syllable length? In our schools they replaced length with stress, which sounded horrible.

      There is one Vojin Nedeljković of the Univ. of Belgrade who tries to scan Latin verses the way they must have actually sounded, but his 'Viva voce' site seems to be dead.

    3. 1. Sometimes (though not usually) length was indicated in writing by use of the "I longa" or "apex" devices.

      2. From discussions of the ancient grammarians

      3. The development of the Romance languages distinguishes the long and short forms of all monophthongs except /a(:)/

      4. Loanwords in non-Romance languages -- e.g. the borrowing of (via) strata into Germanic as the ancestor of Eng. "street" shows that the first /a/ of "strata" was long.


    4. Wojciech

      How do you English actually scan Latin verses? Do you actually try to mimick the syllable length?

      No we don't. At least my school didn't, and I've never met anybody from any other school who did. And I doubt if you could find many university lecturers who attempt it.

      Latin was a living spoken language here from the reintroduction of Christianity to the Renaissance at least. All that time it shared its phonology with English.

      I seem to remember that Erasmus was amused and saddened that the Latin of England had a different phonology from the Latin of France, the Latin of the Germans etc.

      In modern times Latin was still spoken, albeit less often, sharing the spelling pronunciation of English. Then at the end of the nineteenth century a slow process began: to alter the vowel sounds (and a few of the consonants) of English Latin to something closer to 'authentic' Classical Latin pronunciation.

      But even these changes were in accordance with the phonology of English. The first vowel of dies changed from English PRICE to English FLEECE. Stress (with intonation) is the only device available for voicing prominence. I think it's fair to say that the stress acquired historically determines whether the vowels are long, not vice versa. (Stressed syllables ending in two consonants are long anyway.)

      The picture today is complicated by the Vatican version of 'authentic pronunciation' used until recently in the Catholic Church and still widely used in religious choral works — a fact unknown or ignored at Classic FM. Besides, a few names and phrases continued to be used in the law courts and in Church of England services, where the old pronunciation was initially retained.

      People who speak Latin only occasionally are quite likely to be confused by three conflicting standards: the English old pronunciation, the English 'authentic' pronunciation and the Catholic Italian 'authentic'. You sometimes hear two or even three within the same short utterance.

      Anyway, I don't see the substitution of stress for length in reading verse as a specially English thing. Sure that change happened back in Late Latin?

    5. vp

      4. Loanwords in non-Romance languages

      To be used with caution. You need to be sure that it was the Classical form that was borrowed at the relevant period, not some later development.

    6. David

      are you saying that you chaps (some of you) spoke Latin as your first language in the Middle Ages? That would be news to me. I thought you spoke: English, Norse, later perhaps Old Danish, Norman French and 'frensshe after the skole of Stratford atte Bowe' as your first language, or some combination thereof. If not, what else do you intend to say by 'living language'?

      We (Pl) spoke some Latin too, we relished imagining we were a second incarnation of the Roman Republic and stuff, plus we wrote Latin verses. But as a SECOND language.

      No, the substitution of accent for vowel quantity is not specifically English, I think most modern European languages in 'their' Latins do that, because they are not moraic and stress in their metric plays grosso modo the same role as vowel-length in Greek and Latin. But yet, it sounds terrible.

      As for imitations of moraic patterns, apart from Nedeljkovic there is one Robert Sonkowski of Univ. of Minnesota:

    7. Wojciech

      are you saying that you chaps (some of you) spoke Latin as your first language in the Middle Ages?

      No but England was a trilingual society. Many spoke only English and for a while a minority spoke only French. But a large number of people could converse in English, French and Latin. (In the days when Norse was spoken, there was virtually no French and relatively little Latin.)

      Latin was a living spoken language up to the Renaissance, and a flourishing written language for some time after. This was not a reincarnation of Classical Latin — just the way that clerics spoke much of the time, and wrote even more of the time. Even when Latin was reduced to a medium of written scholarship, there was an awful lot of reading aloud.

      'frensshe after the skole of Stratford atte Bowe'

      I've seen contrasting interpretations of this. The one that makes sense is that the prioress moved in circles of society where French was sometimes spoken, but with an English accent and English peculiarities. By Chaucer's time, the numbers of people in England with French as a first language had fallen below a critical level. The prioress spoke French just as she spoke Latin — with English phonology. That's how we remember former Prime Minister Edward Heath, who delivered speeches in flowing effusive French that sounded exactly like English.

      The Renaissance — helped in parts of Europe by the Reformation — largely killed off living Medieval Latin. But it saw the revival of Classical Latin. This changed the nature of using Latin; people did not learn from what they heard but from the books they read. Bur although the written models new (in the sense that non-Classical texts were largely eliminated) there was no corresponding change in the spoken models.

      I use the term 'living language' for the situation where people learned from native speakers when they were available, but otherwise learned from fluent users who were not native. This would be the case for the French of Chaucer's prioress, and for the Latin of all the clerics and scholars.

      I think I might describe the hobby use of otherwise dead languages such as Manx and Cornish as just about 'living language' — though they don't serve anything like the range of societal functions as Medieval Latin and Anglo-French.

      What Sonsowski does is interesting, but definitely not 'living language'. It's an extreme form of revival, an ultimate acknowledgement that classical Latin is a dead language.

      PS You might enjoy the CD produced by David Crystal's son Ben, with proper actors attempting to do proper drama with revived Shakespearian pronunciation

    8. OK, so Latin, was in medieval England something like English in present-day Netherlands: most 'clerics' speak it, many write it, but as a second language, in this sense 'Engels' is 'a living language' in NL, right?

      Why do you think in pre-Norman England there was relative little Latin?

    9. Wojciech

      Why do you think in pre-Norman England there was relative little Latin?

      Listen to Alfred. OK, he wasn't the first King of all England, but he was vey much the architect of the institutions and culture the pre-conquest English state. Alfred decried the lack of learning in England, and his solution was to institute literacy in English as a pre-condition for literacy in Latin He commissioned (and allegedly contributed to) translations of key Latin works. He circulated copies to centres of education and told them that was the policy.

      Before Alfred, there was little Latin because there was little learning and the Church was not the pan-European force it was to become. The continent was cut off.

      After Alfred, there was a lot more Latin, but written English put it in the shade. Legal documents were in English. Religious texts had to be supplied with English translations. Even the dialogues for teaching Latin were translated — the closest we have to a record of conversational Old English.

      in this sense 'Engels' is 'a living language' in NL, right?

      Yes, but in a different way from Medieval England or Modern Switzerland. In Holland, it's individual bilingualism on a grand scale. It's not the classic societal 'diglossia' where the context and subject matter determines what language (or variety) everybody speaks on a given occasion.

      Besides, I rather doubt that 'Engels' is significantly different from English. And nobody doubts that English is a living language. Latin does live — but largely in the Vatican and occasional pockets within the catholic Church. Elsewhere it's a dead language, or else a revived language — either way with the reproductive future of a mule.

      Early Medieval England had 'triglossia' but unlike Modern Switzerland, only minorities of speakers ever participated in the contexts and subject matter where French or Latin was the appropriate code.

  2. You mean vowels or syllables.

    For aught I know, there are no general synchronic principles for determining the vowel's lenght in Latin, just like in German you have to know that the 'o' is long in 'hoch' but short in 'doch'. Exceptions: before another vowel or a 'h', or before muta-cum-liquida it appears to have always been short.

    Another clue: the historical development of the vowel in the Romance languages, but you have to know them well plus the historical principles.

    See here:

  3. This comment has been removed by the author.

  4. Conversely, the traditional ˈdaɪviːz and Lazarus seems to have given way to ˈdiːveɪz...

  5. Dies Irae is /ˌdiːɛz ˈɪərʌɪ/, according to the BBC's Guide.

  6. A priori = eɪ praɪoraɪ -- that is the pronunciation I picked up from my English mentors.

    Sine tempore = saɪnɪ temporeɪ --- that is how I would pronounce it.

  7. The usual pronunciation is the Italianate Catholic. Those of us who learned Latin at school but have no familiarity with the Catholic mass might well say ˈdi:eɪz ˈi:raɪ or similar. The only authentically English pronunciation would, I suppose, be something like ˈdaɪ‿i;z ˈaɪri:.

    I note that the OED gives 'saɪnɪ ˈdaɪi: as one of the pronunciations of the related sine die. And ˈka:pɪ ˈdaɪɛm is the only pronunciation they give for carpe diem.

    1. Carpe diem

      This answers the Duchesse's question above. It comes at the end of an ode in Asclepidean metre:

      — — — ⏑ ⏑ — |— ⏑ ⏑ — | — ⏑ ⏑ — ⏑ ⏓

      ...........................Dum loquimur, fugerit invida
      aetas: carpe diem, quam minimum credula postero.

      The Wikipedia translation seems as good as any:

      While we speak, envious time will have {already} fled

      So in Classical Latin the first vowel of dies (at the very least in the accusative case form diem) was short. By the time of early Modern English it was stressed and long — as in ˈdaɪi:, ˈdaɪɛm and presumably other forms of dies.

    2. The rule about Latin vowel lengths in the English pronunciation of Latin and Latinate words is that they are used only to determine the stress placement, and are otherwise ignored. Greek words in English likewise get Latin stress placements. Once that's done, the English vowel length (or "length", since the Great Vowel Shift) is established according to the rules of English orthography, so final vowels are "long", vowels before consonant clusters or double consonants are "short", and so on. The Wikipedia article "Traditional English pronunciation of Latin" provides an extremely thorough explanation of the whole convention.

    3. My intuitive pronunciation of "dies" is /ˈdiːeɪs/, by analogy with "res", which I've always pronounced /ɹeɪs/.

  8. How awful! So basically there is no rule, apart from metre, that will tell me what is the length of the vowel. Which then makes Dreimorengesetz almost useless.

    The whole point was not to look in the dictionary every time an unfamiliar word pops up.

    1. Modern publications of Latin texts often indicate length with breves and/or macrons, if that makes you feel better :)

    2. there is not very much by way of rules for contemporary languages with vowel length. OK, in Swedish a long vowel cannot occur before a geminated consonant, but that is pretty much it.

      In many languages there are, however, methods of more or less systematically signalling the lenght or shortness of a vowel in writing. In Latin, apart from modern macros and breves, next to none.

    3. ɪt ˈɾɪəlɪ ˈdʌzn̩t, viː piː, ɪt ˈɾɪəlɪ dʌzn̩t. aɪ θɔːt ɪt wəz ˈpɒsɪbɫ̩ tə nəʊ ʍeə ˈmeɪkɾənz ænd bɾi:vz fɔːɫ wɪðˈaʊt ˈeɪnʃn̩t gɾəˈmeəɾiənz ænd ðæt sɔːt əv stʌf, bət əˈpeəɾəntlɪ nɒt.

    4. ʍ -- I like your 'wh'. How come you use this sound? The Scottish have it, but to judge from above you're anything but a Scottish person.

    5. This comment has been removed by the author.

    6. People sometimes use [ʍ] to sound sophisticated. I admit to occasionally doing so myself :)

    7. As some phoneticians would say, I don't have the WHINE–WINE merger. I know of several people and have heard of several whose surnames start with ⟨wh⟩ and who insist on pronouncing it with ʍ.

      Cruttenden states that “it is often taught as the correct form in verse-speaking” and I believe RADA still teaches it as a distinct sound.

      Plenty of American dialects, mainly in the south-eastern quadrant of the States, also have it. It just comes out naturally to me.

    8. The other way around, WINE–WHINE merger.

    9. I like this sound, and would like to use it myself, but am not sure if I would not make myself ridiculous by so doing.

    10. This comment has been removed by the author.

    11. I think that Johnny Cash had ʍ in his phonetic repertoire. He was from the south-east of the USA. In his version of the song Hurt, listen at 0:55. I think that he uses [ʍ] in "what".

      I've never really understood why this has been considered an RP variant. To me, it seems a clear example of a regional feature. Are there any famous RP speakers who use it?

    12. Brian Sewell does.

      I think I've read somewhere that it comes from a pre-Old English xw phoneme.

      Wojciech, I'm don't think it's such a glaring sound compared to, say, a flapper r. Many people won't even notice.

    13. You're right. He does. The Queen does not though.

      I get the impression that it is stigmatised in the USA. It used to be mocked on the cartoon King of the Hill (set in Texas).

    14. @Ed:

      I think our good host covered this in Accents of English. He said something to the effect that, while RP speakers don't generally learn the wine-whine distinction with their mothers' milk, it is seen as a prestigious feature that is commonly taught in acting and elocution classes.

      When I lived in the UK it would not be uncommon to see RP-like newsreaders or actors using the distinction.

    15. Turns out, I've invented a new sound, a “flapper r”. It was obviously suposed to be a flapped r.

      I've always thought that in the US it is even more unnoticable than it is in the UK and that it just comes out naturally.

    16. @ vp: Yes, I've found it on page 285. I'll pay closer attention to newsreaders from now on to see if any of the English ones use it in this day and age.

    17. You're right. He does. The Queen does not though.

      Does her mother? Did the kings George V or George VI use it? I cannot find any suitable clips.

      I do have to say that though Brian Sewell does sound posh to me his is a very different kind of posh than, say, Elizabeth Howell saying This is direct television from Alexander Palace in 1936.

      It is really funny how people sound posh even when using very non-posh pronunciation features such as, for example, Prince Charles and how there is this constant mingling of working-class speech and posh upper-class talk, what was once posh and sophisticated, now it's dialectal and vice versa.

    18. When I read this comment, I remembered a blog-post from a long time ago that said that no accent stays static over time. I've found it: on 11 July 2007, Professor Wells said that young people never do sound like their grandparents. Brian Sewell is not a young person, of course, but the point is that no one speaks quite like Elizabeth Cowell any more.

      The same is true of other accents. I think that most people would agree that I have a Yorkshire accent, even though I have the FLEECE merger, say -ook words with [ʊk] and pronounce most cases of initial [h]. The "broad Yorkshire" such as in this recording is comprehensible to me, but I would only speak this way for the sake of comedy.

    19. That is a good post because it says which criteria John uses to identify U-RP speakers. Though Brian Sewell doesn't speak like Elizabeth Cowell, and though I do notice a deterioration in his pronunciation over the last few years, he did say he spoke like Vita Sackville-West (here reading from her poem The Land.

      Unfortunately, he obviously doesn't.

      Too bad there are no more recordings of her. Online, at least.

      And Walter De La Mare.

    20. I must say, though, and perhaps I am too harsh, that in a way I bear a grudge that John has never spoken about U-RP, at least not at great lengths. It gets mentioned, but that's about it. I am thus very grateful to Geoff Lyndsey and especially after I've read him saying that the accent Daniel Jones described was a thing of great beauty.

  9. @Ed:

    I get the impression that it is stigmatised in the USA. It used to be mocked on the cartoon King of the Hill (set in Texas).

    In the USA, the wine-whine distinction is commonly found as part of a deep Southern accent. However, it's not the wine-whine distinction itself that's stigmatized, but the broad Southern accent as a whole. When found in a generally non-stigmatized accent, the wine-whine accent tends to be a prestige feature. When I attended an acting class in the US, I was advised to use it for high-status characters.

    1. Not so commonly any more. See this map: the purple areas are those in which /ʍ/ is found in at least half the population. There are no parts of the U.S. in which it is universal any more.

      My wife's from North Carolina, one of the most purple states, and she has it. Then again, she was born in 1943 and left the purple zone in 1969. She thought it rather odd that "in Wales they don't say '/ʍ/ales' for 'whales'".

    2. I'm afraid that in no time we'll make Wojciech sound Laurence Olivier in Richard III.

    3. except that I would not speak with such a piercing, intimidating facial expression... Also, his gait (of an old man) has a strange contrast with his youthful silhouette and face. One feels there is something wrong with him. Not at all a 'glorious summer'...

    4. King of the Hill is mentioned in the Wikipedia article. Apparently, the humour surrounded hypercorrect use of ʍ.

      Is there anyone who uses ʍ in "who" or "whore"?

    5. When my 9th grade English teacher was explaining alliteration to us, she told us quite definitively that "w-" could never be considered to alliterate with "wh-". I had already begun to teach myself about phonetics and phonology, so this really bugged me. If the author uses the same phonemes in those words, the reader uses the same phonemes, and the overwhelming majority of the English-speaking world uses the same phonemes, it's still not alliteration? This was in New England, and to my knowledge she was the only person in the entire school who actually did make the distinction. She didn't sound Southern, so I suspect that she had adopted it.

    6. maybe she ought to have said 'though this be an alliteration to you and a majority of English speakers, it is not an alliteration to all of them' or some such?

    7. Ad Duchesse de Guermantes

      My impression is that sir Laurence

      1. pronounces that word 'lour'd' rhotically,
      2. says a different text altogether from somewhere down the middle onwards
      3. does not look like a paradigmatic example of 'deformity', still less like someone at whose very sight dogs would care to bark etc., which gives this whole monologue a touch of 'coquetterie'.

      Maybe such was the taste of the time (the fifties)?

      His not-just-intervocalic rolled r's are perhaps an attempt at sounding fifteenth century?

    8. I shall have to re-listen, but every time I do, I keep getting stuck at what he says 'instead' of our.

      Do you think he says something to the effect of naʊ ɪz ðə ˈwɪntəɾ əv aɾ dɪskənˈtent? He indeed does 'loose his way' only to find it again after Why, I, in this weak piping time of peace...

      Who is it that said that those r's were used and taught because they were thought of as more dramatic, more impactful and probably a bit more manacing? A sound that can project more from the stage than a mere approximant or a derhoticized pronunciation of a word?

    9. Laurence Olivier was also Henry V. In St. Crispin's Day speech, can you tell me how he pronounces Talbot?

      Funny that he says (in Richard III) ˈbɾuːzed (ɑːmz).

    10. yes, he says:

      naʊ ɪz ðə ˈwɪntəɾ əv aɾ dɪskənˈtent?



      I seem to hear something like 'talldot' or 'talldert'.

      ˈbɾuːzed too has struck me... . There are many striking things in his performance.

      The rolled r's, especially non-intervocalic, are indeed menacing. In England, they used to call this sound a 'dog's sound', I seem to remember from an old book.

    11. On a second hearing, I managed to hear 'talbot', with a 'b', but it's a strange 'b'.

    12. Sounds like ˈtɔːlbət to me (as expected). And what's funny about ˈbɾuːzed?

    13. the 'e', I suppose. Should have been 'bru:zd' or maybe 'bru:zɪd', I think.

    14. Thank you, Wojciech, for going thhrough it all one more time, I'm really grateful. I hope it wasn't too much trouble, but it's really fascinating. To me, at least.

      No wonder the Disney and all Hollywood villains were pretty much all British, don't you think?

    15. This comment has been removed by the author.

    16. This comment has been removed by the author.

    17. I must be deaf, but I keep hearing ˈtɔːlɾet.

    18. Yes, it was fascinating and ... moving. Great literature! And yes, no wonder Hollywood villains British.

      As I said above, sir Laurence's 'b' in 'Talbott' sounds queer, like 'd' or maybe ɾ.. maybe it's the recording...

    19. And while we're at it, how do you pronounce Doerr?

    20. Regarding the alveolar tap... I think John Humphrys of the BBC Radio 4's Today programme uses it quite generously. And he's Welsh.

    21. My surname is pronounced ˈduːə, though I believe other families with the same name say dɔː.

  10. I was taught to use a flapped 'r' between vowels (sorry, horrible), and after 'th'. The former I don't use any more, the latter I still do. I was NOT used to say 'hwat' for 'what' and such-like, though when I familiarised myself with American English (like in the 70ies) the books said 'most Americans' still say 'hwite Christmas' and so on.

    In Scotland they still use the sound, don't they? IN Ireland too, though less commonly, this is at least my impression.

    1. Yes, it is still used in Scotland and Ireland. Glasgow is one area of Scotland where the wine-whine merger is under way. I'm not sure if there are areas of Ireland where it is under way.

      Any book on traditional dialect will tell you that the merger has not occurred in the English county of Northumberland (which borders Scotland). I'm not sure if this is still true. The few Northumberland residents that I've met have had the wine-whine merger.

  11. It would interest me why 'hj' is doing far better than 'hw' in English, that is, 'yooman' and so on are not nearly as common for 'human' as are 'wales' for 'whales' and so on. You guys will certainly know (of) a convincing explanation.

    1. For what it's worth, the English 'hw' sound is usually analyzed as one phoneme, /ʍ/, while 'hy' is always written as /hj/. (Honestly, I'm not sure what the practical difference between [ʍ] and [hw] would be.) Here in North America, /hj/-to-/j/ reduction seems limited to the New York and Philadelphia metroplitan areas. People often make fun of Donald Trump for saying that the winning contestant on his show will receive a "yuge prize".

    2. CEPD writes hw while Wikipedia also uses hw̥, apart from other symbols.

    3. Old English used 'hw' as well:

      Hwaet! Ich swefna cyst secgan wylle
      Thaet me gemaette to midre nihte
      siththen reordberend reset wunedon

      (Dream of the Rood)

      Saying 'Hwaet!' was a favourite way of starting a literary work in English, Beowulf starts too with a 'Hwaet!', methinketh. It was vaguely equivalent to later 'why!' or 'lo!', or modern 'wow!'

      For instance, on the site of RADA one actor recites a sonnet by Shakespeare in which the Swan of Stratford says:

      'why then her breast are dun!'

      in such a tone of voice as if he thought that 'why' is supposed to mean 'for what reason' or 'because of what', whereas in fact it's a 'hwaet!'or 'wow!'.

      Sorry, too much digression.

    4. Wojciech

      For instance, on the site of RADA one actor recites a sonnet by Shakespeare

      The actor is the distinguished Alan Rickman, and he reads beautifully. This seems to be currently the most popular reading — at least among web-users; if you google 'sonnet 130', most of the top hits for video-clips involve a recording of Rickman.

      His reading of why is spot on. Yes, he lends a slight devoicing to stressed white, but it would sound ridiculous for a non-Scottish, non-Irish actor form Britain to say ʍaɪ for why — or, later in the sonnet ʍɛn for unstressed when.

      Google 'sonnet 130 Gielgud' for a British actor of a previous generation and you'll hear almost the same. A more fricative ʍ at the start of white but voiced and unstressed why and when.

      Not a professional actor, but a professional OP expert, David Crystal reads the white, why and when in many ways like John Gielgud: a slightly different PRICE vowel and possibly a tad more friction in white and a tad less voicing in why, but essentially the same.

      His son Ben turns in a more professional reading of Sonnet 116 — both in modern pronunciation and in OP. There's not a great deal of difference in the WH words. In both readings of

      Love is not love

      Which alters when it alteration finds

      There's no stress and a touch of voicing at the start of when. The start of the more prominent which is noticeably voiceless but not fricative in the OP reading. In the 'modern' reading there's a hint of voicelessness.

      The other actors on the Original Pronunciation CD produced by Ben Crystal vary a little, but the general trend is that
      • stressed WH words are more likely to be voiceless and fricative
      when used as a conjunction is extremely likely to be voiced and unstressed.

      in such a tone of voice as if he thought that 'why' is supposed to mean 'for what reason' or 'because of what',

      To my ears the exact opposite is the case. Interrogative Why? is stressed and amenable to devoicing. Therefore stressed devoiced why sounds interrogative — even though the syntax makes that interpretation impossible.

      This why is what the OED lists as

      7. Used interjectionally, before a sentence or clause.

      b. Emphasizing or calling more or less abrupt attention to the statement following (as in the apodosis of a sentence), in opposition to a possible or vaguely apprehended doubt or objection.

      It's not at all like Hwaet!. It can't stand on its own but serves to introduce a clause and the surprising idea it expresses. Phonologically it's a sort of clitic — maybe not entirely resistant to stress but getting on that way.

    5. See, I thought it was Alan Rickman, but haven't had the time to check. How come you didn't recognize him, Wojciech? He isn't overtly posh, but he does give that sort of impression because of the colour of his voice.

      ... it would sound ridiculous for a non-Scottish, non-Irish actor form Britain to say...

      I wouldn't really agree...

      But why is the OP so incorrect, then, if we know how the hw developed over time?

    6. thank you David and Duchesse for your explanation. My comments on whoever read the sonnets were utterly presumptuous, I know, since I don't really know how the British speak (I have had little exposure to living British speech but for work colleagues/bosses) and am aware that their 'inflexions', while short of appearing to me 'hoarse and odd', as T. S. Eliot would have said, do appear as what they are not to me, i. e. as monotonous, because our Polish inflexions are quite different (_they_ would certainly appear 'hoarse and odd' to you guys).

      My long digression on hwaet/why/lo/wow was a joke in part, I am sorry... . But btw, I formerly (when I was young, i.e. 40 years ago) believed that 'why' _qua_ an exclamation was always prononced [waj], even in th'olden times when everybody said [hwaj] for the interrogative pronomem 'wherefore'. But clearly, I now know, I was wrong.

    7. How did the PRICE vowel differ?

      In its starting position. A dialect coach and disciple of David Crystal called Paul Meier summarises thus:

      The PRICE and CHOICE lexical sets.
      This diphthong, too, had a centered onset and started with the schwa, or neutral vowel ,[ə], resulting in [əɪ].

      EXAMPLES: price, tribe, time, Friday, isle, eider, fight, Viola;
      AND choice, point, boil, toy, ahoy, royal

      You can hear him read this and the other lexical set descriptions here. David Crystal's own recordings — some free, some purchasable from this page of his Shakespeare website — sound less exotic (or less Mummerset) because the ə element is shorter relative to the ɪ element when compared with that of his disciple.

    8. I wonder where that claim about CHOICE comes from.

      Both Wells "Accents of English" and Barber, "Early Modern English" claim (IIRC) that today's CHOICE set was divided into two: a set (including most CHOICE words) that was pronounced [ɔɪ], roughly as today, and a smaller subset that derived from [ʊɪ], including words such as "point". This latter subset could easily have been merged with the [əɪ] of PRICE, but not the former.

    9. IIRC, vp, words shifted from one you mention to the other and back a lot. Only a small number of them never had [əɪ] but stuck with [ɔɪ].

    10. Paul Meier's treatment of CHOICE lies at the other end of the spectrum from academic controversy.

      • At one extreme, scholars debate the history of sound changes in individual words — in journals and highly specialised books for academics and their libraries.

      • At a more popular remove, John's Accents of English is accessible to undergraduates and serious amateurs. After outlining some general findings, John observes:

      [the sound of one group of CHOICE words] developed into a variety of [əi] or [ʌi], which led to confusion between these words and PRICE words, so that for instance boil (v.) could be a homophone of bile, or joined rhymed with find. The history of the various diphthongs involved is in fact very complex and the subject of some scholarly do agreement

      • David Crystal did not write for fellow scholars or even for amateur linguists. Rather, he recorded a lot of Shakespeare for actors at the Globe to base their pronunciation on. He made practical decisions on how he read CHOICE words — without confusing the actors with reference to controversy.

      • His son Ben Crystal did more than provide a resource. He directed a group of actors for this CD. Presumably he wasn't even aware of the controversy (not the details anyway), but that didn't matter as his role was to make the important pronunciation decisions. (Minor decisions were left to the actors.)

      • Paul Meier presents David Crystal's scheme to anybody who's interested. Sometimes like Ben Crystal he guides actors personally. But he also publishes. Moreover one of his publications is free on line — see link above. Written for a broad public of strangers, his account is the simplest, least nuanced account on the spectrum. But it has the considerable merit that any old bunch of amateur actors could use it to create a plausible OP (original pronunciation) production.

    11. Paul Meier who, among other things, produces the various accents CDs?

      I visited his website several times and I cannot shake off the feeling that he mispronounces the vowel [ʌ].

    12. @David Crosbie:

      But if Paul Meier just wanted to keep things simple, why not just say that CHOICE words had roughly the same sound they do today?

      That would be even easier to grasp, and, as far as I can tell, closer to what scholars believe (the [ɔɪ] set was larger than the [ʊɪ] set).

      If anyone has counterevidence, I'd be interested to see it.

    13. David Crystal may well discuss the CHOICE vowel in — neither of which I have access to at present. What I have found is this paragraph which appears in his online supplementary notes to his transcriptions of Troilus and Cressida and Romeo and Juliet.

      The Troilus note:
      2.3.5 The remaining diphthongs stay diphthongs, but they have a different quality, and this is perhaps the most noticeable vowel feature of all in EME. In Modern English, lie, now, and joy would be [laɪ], [naʊ], and [jɔɪ], respectively. In EME the first part of the diphthong is articulated in the centre of the mouth, with the sound that we usually hear in the or in the last vowel of sofa: [ləɪ], [nəʊ], [jəɪ], This last one is especially distinctive, when it occurs, and is the source of many puns which weve lost these days. In this play, it is important as the vowel of Troy and Troilus.

      The other note is identical except for the example at the end:

      This last one is especially distinctive, when it occurs, and is the source of many puns which we’ve lost these days (as in loins in the Preface, which would have sounded like lines).

    14. Yes, he says the same thing in his Pronouncing Shakspeare. Basically, he's providing a simplified picture, and consciously so, I think.

    15. I looked at the relevant section in Dobson, "English Pronunciation 1500-1700".

      The mass of detail is almost overwhelming, but a (very simplified) summary is

      1. There is a subset of CHOICE words that had [ɔɪ] in Old French and have consistently retained it in English up to the present. choice, cloister, coy, employ, exploit, joy, Joyce, loyal, noise, poise, quoit, royal, voyage. One native English word, embroider, also belongs in this group.

      2. At the other extreme, there are CHOICE words that derive from Middle English [iː] (i.e. PRICE) -- presumably by analogy after Middle English [ʊɪ] had merged with [əɪ] in some speakers. These words are boil (ulcer), groin, hoist joist.

      3. CHOICE words derived from Old French [oɪ] generally have Middle English [ʊɪ] (outnumbering attestations of [ɔɪ] five-to-one): anoint, boil (vb), cloy, coil (vb), coin, destroy, join, loin, moist, point, poison, purloin, soil, spoil, toil, voice

      4. The remainder of the CHOICE words (derived variously from Old French [o] before a palatal, and from Dutch), seem to have varied between Middle English [ɔɪ] and [ʊɪ], with [ɔɪ] dominant (two-to-one). boy, buoy, foil, oil; loiter

      5. The [ɔɪ] words generally retained their pronunciation.

      6. The [ʊɪ] words followed the development of Middle English [ʊ] (PUT, STRUT), generally becoming unrounded, except after labials. This allowed a pronounciation of [əɪ], homophonous with the PRICE words. Dobson writes that "in good StE [standard] speech, [ʊɪ] was retained until about 1640", but that "in rather less correct speech [əɪ] was in use from at latests about 1640", and "in some [presumably even less correct!] forms of speech ... [əɪ] ocurred as early as the fifteenth century".

    16. Is there a book on English pronunciation from 1800 until 1950?

    17. I'm not familiar with one comparable to Dobson.

      Maybe someone else is.

    18. vp

      If the rhyming and punning evidence is as extensive as Crystal claims, that must surely suggest that of the alternatives that Dobson deduced to have been available, Shakespeare preferred əɪ.

  12. Should this be of interest to any of you: Vojin Nedeljković's moraic recitations of Latin poetry are here:

    I personally prefer them to Sonkowski's or Stroh's, which sound to me as if they wanted to mock Vergil.

    1. ʋɔ̌jin nɛdɛ̌ːʎkɔʋitɕ is also on YouTube. I have no idea why he hasn't made a new web page with the readings.

    2. yes, I have seen that. You like his recitations? I sort of do, especially his Serbian accent. I have heard lots of diverse foreign accents in Latin, but the Serbo-Croatian I like the best (as I also like the sound of the language itself).

    3. I do.

      But you must be one of the rarer people who've publicly stated their love of the sound of Serbo-Croatian, since I can say that, come to think of it, I don't really recall anyone saying that ever. But there must be someone who likes it, on the other hand.

      I've always thought of people thinking of it as a harsh, grating language, with all those affricates etc.

    4. While I personally find, say, Czech (a very soft language, whatever your prejudives) or Bulgarian more beautiful, Serbo-Croat was voted the most beautiful Slavic language at a Slavist conference once. Probably because of the tones.

    5. Did it really? When was that? How delightful.

      I believe Bulgarian and Macedonian sound to Serbs, not all, of course, as an uneducated southern (Serbian) speech, where people often don't know all the cases.

      I'll have to find the exact quote on how Serbs perceive Croatian and Croats Serbian.

    6. they way the speakers of the same language perceive the same language.... . How do Germans perceive Austrians or the French Wallon?

      In fact I quite like all four (or five, depending on your line on the status of Macedonian) southern Slavic languages,but SC most, probably because it hain't got no schwa-like vowels or Roumanian-style i^ or a^-like ones. There is a lot of political projection on such things, since 'we' don't like Serbs we don't like their language either etc....

      Czech is fine, Lipman, I sort of suspect it sounds nicer than my native Polish (though obviously I can't hear it). I understand too much of it to taste it sound, though, while of the SSlavic I understand far less.

      Re cases in Slavic: the farther south in the Slavic area you go, the less complex the case system gets (S.Slavic lang.s are colonial languages, so small wonder) till it disappears completetely in the South Serbian dialects Macedonian, Bulgarian and the (yet existing?) Slavic dialects of Greece.

    7. Duchesse, if you don't mind my asking, what is your L1?

    8. ěŋgleskiː, ali gǒʋoːriːm sr̩̂pskiː. What on Earth does gassalascajape mean and how am I supposed to pronounce that?

    9. Oh, goodness. It's from Ambrose's book, of course.

    10. Nice. Although you got the lengths a bit wrong. Should be:

      ěŋgleːskiː, ali gǒʋoriːm sr̩̂pskiː

      Of course, the lengths in final open syllables are pretty moribund nowadays.

    11. The is intentionally not long since it sounds Bosnian to me.

    12. Ah, but it's quite the opposite - the post-tonic lenghts are best preserved in such environments (following a short rising tone), and you'll find e: in "engleski" many Serbian accents.

      On the other hand the i: in "srpski" and "engleski" does make it sound Bosnian. What I said about the lenghts being moribund in such environments applies primarily to Serbia, but not Bosnia.

    13. I'm not entirely sure that's true. Francuski, engleski and so forth, with long u, second e are entirely Bosnian. Just try lengthening it no end and you'll see. In any case, the final vowel is short in my speech too, but just for 'grammatically correct' reasons I've put them up there. The post-tonic lengths might be preserved, but I wouldn't consider that adjective to be one of those words. In Bosnian I believe that the post-tonic syllable also has a high(er) tone than the one found, say, in the speech of RTS newsreaders.

    14. Maybe in Užice people lengthen that vowel. But not in Belgrade.

    15. True, but then again Belgrade's a wholly different kettle of fish. In the broadest Belgrade vernacular, all of the four tones are neutralized, and the post-tonic lenghts completely absent in most cases.

    16. Like, big cities always simplify (too) complex systems?

    17. Not necessarily - look at say Belfast (what Milroys wrote about it), and New York.

      The allophony of the TRAP vowel in NY vernacular is more complex than any other in North American English.

    18. this is too recondite to me, I am not so well informed about and conversant with the (linguistic and others) intricacies of the Anglophone world. Though by what I know about TRAP in NYC you may well be right.

      I have somewhere read that cases disappeared earlier in Dutch than in German because of the levelling-out effects of the big Dutch-speaking city centres, in Flanders and Randstad.

      Anyway, even though not necessarily, there is a tendency like that, it seems...

    19. Why is it that I have a feeling that gassalascajape teaches English Phonology at the Faculty of Philology (their official English name)?

    20. Well, from his examples, his expertise he keeps revealing in his contributions and such-like...

  13. In "sloppy" or connected speech, daɪz sounds perfectly normal to me. Probably OP daɪiːz with regularly shortened unstressed -> daɪɪz/daɪːz -> daɪz.

    1. "Original Pronunciation" (of Latin in English-speaking areas), as opposed to NP, the New Pronunciation which, rather in vain, aimed at the original pronunciation of Latin of 2000 years ago.

      [EDIT: self-referential captcha: "entaman". I am, man, I am!]

    2. Oh, I see!

      Here is an nice, informative article about th Captcha.

    3. Sorry: Old Pronunciation. (Original Pronunciation would be for Shakespeare plays.)

  14. Wojciech, have you had the chance to introduce yourself to Edith Evans? I think she habitually uses , says ʍɒt ɑː jɔː ˈpɒlɪtɪks (though I'd need better equipment to see whether she says ɔə and something like ˈpʌlɪtɪks).

    Apart from ˈmaɪnɐ ˈmætɐz, ˈgɾouvənɐ, geəl, ɾ is all over the place, ˈʍɒtsoʊevɐ.

    Evans's most widely admired asset was her voice, a highly individual instrument, often imitated but never surpassed. She often professed herself unaware of the extraordinary effect it had on her audiences, but she set great store by clear diction, and in her later years openly criticized the slovenly standards of speech prevalent in the theatre.

    —Oxford Dictionary of National Biography

    1. Yes, quite nice, thank you, I like geəl as the standard vowel of this word displeases me, (I have praised Serbo-Croatian for not having anything like this), but I would stop short of imitating her.

      For one thing, I'm too old for that. For another, well, I once had an English boss, a Phil. Dr. Oxon., who claimed that as a NNS I had no right so use certain words, such as 'to asseverate' since I was (he thought) not supposed to know them at all (actually, I needed that word for practical, not ornamental purposes: to criticise certain people who would solemnly assert but no more that assert certain things). Now I suspect something similar holds for pronunciation, including the 'wh' sound, let alone the rolled intervocalic r's or ɐ's or 'er's (which I, too, like).

    2. I believe that Daniel Jones lists gɪəl, gɛəl, ɟəːl, ɟeəl as standard varieties too, gɚːl, gɛɚl as northern and western varietis, Scottish gɛrl and vulgar gə̞ːl, gæəl, gæl, gɛl, gel.

      What a funny story about your boss! No right to use certain words? LOL!

    3. And actually what she says is something like ˈpɒ̜lɪtɪks, with a less rounded low back vowel, which was very typical.

  15. 'And actually what she says is something like ˈpɒ̜lɪtɪks, with a less rounded low back vowel, which was very typical.'

    Typical for what? Her 'o' in that word is peculiar, but seems different from STRUT, at least from hers.

    She says words in -er with a secondary ascending stress on that syllable, which makes it sound as if she affected French accent. govern'a

    Well, frankly, I am not sure if I'd dream to talking like her... . Even if I was allowed to. But I think my teachers wanted to impart to me something close to Gielgud's (Giełgud's) accent in the scene. All in all, I can't help having the impression that she is caricaturing someone, a type of person defending 'natural ignorance' and such. Is that not so?

    1. For the so-called Empire Speech. That is, the accent codified by Daniel Jones. In his The Pronunciation of English he says for ɔ “open back, with slight lip rounding”. And adds that in many accents “the sound is produced without lip-rounding”.

    2. But isn't she mocking that Empire Speech?

    3. A few remarks/queston:

      that 'gairl' she is referring to is called Gwendolin?

      she has a very high 'a'-vowel: land=lend, she doesn't like 'tempering' with natural ignorance. He was found in a hend-beg.

      plezha (pleasure) again with an ascending French-sounding by-stress on '-a' ('-ure').

      she rolls some non-intervocalic-r's, was that done, or is that the mocking part of her performance? rrequire, cloak-rrroom, prroblably.

      I can't really make out what ever she is called: lady Bradfold, Bradmont, Blackmoore or something...

    4. She is Lady Augusta Bracknell. Her daughter is Gwendolen.

      That realization of the æ phoneme was very frequent and even the Queen uses it nowadays sometimes. She used to use it much more frequently back in her younger years, when she used to refer to her daughter Enn and her son Endrew.

      The Queen also used to say ˈluːsɪ θɾuː ðə bəˈluːn tə suː with a fully back u vowel.

      The alveolar flap was used in a variety of ways, the champion was probably Noel Coward, who used it even after t and d.

      The way her intonation works made some people say (look in the comments) she sounds Indian.

      It wasn't mockery, it was a genuine pronunciation back in the day.

    5. Actually, much of her intonation reminds me of one Ewa Szumańska, a Polish writer and actress (of sorts) who satirised young affectionate 'intelligentsia' ladies in her 'Journal of a young female doctor in the country-side' e.g. here

      all despite the huge difference of the social context etc. etc.

      Many of my studentesses still write unbeknownst to them as if they wanted to emulate 'the young female doctor', here the matter is style and vocabulary, Polish counterparts of 'albeit' and 'whilom' and 'fain' and such-like.

  16. nice posting.. thanks for sharing.

  17. Note that hoist and roil have down-register doublets heist 'a robbery' (from the same metaphor as lift in shoplift, probably), and rile 'anger, annoy'. These reflect borrowings from accents that retained PRICE=CHOICE, though they survived mostly in AmE.

    For what it's worth, my own CHOICE, like my NORTH=FORCE, begins with [o], almost the same as in GOAT except that in CHOICE and to a lesser extent in NORTH=FORCE my lips are actually rounded at the beginning of the diphthong, whereas in GOAT they are merely compressed at the corners, as in my other rounded vowels.

    1. Ah, OK, as writes Lowell in his "Biglow Papers" (Conn.)

      Dear Sir,
      you wished to know my notions
      Bout sartin pints that RILE our land.
      Well, there is nothin' that my nature so shuns
      Ez bein' mum or underhand.

    2. Note that hoist and roil have down-register doublets heist 'a robbery' (from the same metaphor as lift in shoplift, probably), and rile 'anger, annoy'. These reflect borrowings from accents that retained PRICE=CHOICE, though they survived mostly in AmE.

      You mean that the words hoist and roil have a lower-register accent variant pronunciations raɪl, haɪst which are a reflection of the fact that these words were borrowed from accents with equate PRICE and CHOICE words? Such accents only survive in American English?

    3. And, of course, what are these two metaphors regarding heist and lift and shoplift?

    4. Votre Altesse: Mais non.

      I mean that although the distinction between hoist 'lift, elevate' and heist 'steal' was once basically a matter of accent, they are now separate words. Heist and rile are lower register AmE than hoist and especially roil, and people who are not etymologists no longer realize the connection between the pairs.

      A similar doublet derived from another uncompleted sound change, "early loss of /r/" is parcel and passel. The latter is specifically AmE, rarely written, and means '(large) quantity', as in a passel of = a lot of. But both the PRICE=CHOICE merger and the early loss of /r/ happened in England before the settlement of America, so their appearance in AmE is probably a consequence of the general rule of colonial conservatism in language.

    5. The general metaphor is that to steal an object is to elevate it; lift itself, as well as shoplift and heist can also mean 'steal'.

  18. John Cowan

    These reflect borrowings from accents that retained PRICE=CHOICE, though they survived mostly in AmE.

    Record companies last century tried with mixed success to signal the rural nature of songs in the 'Hillbilly' or 'Old Time' catalogues. From a discography of all Country Music recordings prior to 1943 (There are reasons for this date) here are the spellings and number of times used for just one song:

    Boil The Cabbage Down — 1
    Boil Them Cabbage Down — 1
    Boil Dem Cabbage Down — 1
    Bile Them Cabbage Down — 4
    Bile Dem Cabbage Down — 5

    The most popular spelling may look like a 'Negro Minstrel' affectation, but according to another discography, the song wasn't recorded by any Black performer — neither commercially nor for a folk collector.

  19. So, David Crosbie and Lipman, why don't people talk like they used to, e. g., in the 1850s or the 1900s? What happened? What causes the pronunciation changes of an entire language in general and what happened in the particular case of British English?

    Was the accent to artificial and hard to pronounce? What role did politics play? The social changes?

    Why was something that was 'upwards and back' now more open and front?

    1. I don't think anyone has the answer to the question "Why?" in linguistic change. It just happens sometimes.

      My best guess is that it changes as people's lifestyles change. The old stereotype of an upper class person in a country house, waited on by servants and unaware of simple things like buying furniture, is probably something that belongs to the past. Prince Harry decided that he wanted to join the army rather than living such a pampered life, so his speech probably became less upper-class as he worked alongside people with different accents.

      By the same token, working-class accents have changed as people move around more. Most people do not live their whole lives in the same town any more. There used to be a clear difference between the speech of Leeds and York - a mere 20 miles apart. Now there is hardly any difference, and that's probably because 20 miles is not "a long way" any more.

    2. One driver must be the constant presence of in-groups and the occasional elevation of one such group to leadership of fashion.

      The Milroys added a tweak to the model by turning the focus fro groups to networks — the meaning of the term social network before the internet changed it.

      One imagines that every change spread though a network is initiated by some individual innovation. But the converse cannot be true — far from every innovation initiated by an individual comes to be spread through a network.

      The analogy that suggest itself is a flock of birds with many individuals flying briefly in a new direction, only to turn back to following the mass — until one random bird is followed by another, and another and another until the whole flock appears to veer suddenly.

      Where the analogy fail slightly is that not all innovations adopted within groups and network are unconscious. Some are readily adopted and spread because that give individual, groups and network a sense of otherness — not like they parents, not like non-respected authority figures, not like people they find boring, not like people they disrespect.

      And not all innovations stem from an individual. Spelling pronunciations are liable to spring up everywhere when words formerly known only to an educated elite find their way into print that is much widely read. Either that or when education expands greatly — so that the newly educated read words that are never spoken in their circles.

      Moreover not all changes are isolated from each other One changes cans still push or drag another just as the did so spectacular in the Great Vowel Shift.

      I can't see much of a role for politics — except insofar that it influenced social change, causing one social class — or one group or network within a social class — to be more eligible than before to be a leader of fashion.

      Anyway, many changes are observed within closely defined groups not from a social impulse to be different but from a generational urge. I suppose some social groups and networks are more resistant to generational change, but even the Royal family has been observed to change pronunciation over time. The Queen used to sound like her father's generation. Now she sound like her children's generation. And her grandchildren sound nothing like her.