Wednesday, 28 April 2010

ən ɪnˈɔrgjərəl ˈlɛktʃər?

For years I had been teaching people that speakers of rhotic accents pronounce an r sound wherever there is a letter r in the spelling, and not where there isn’t. Non-rhotic speakers on the other hand — which includes me and most of the students I lectured to at UCL — don’t, since we have lost /r/ except before vowel sounds.
I would illustrate this with pairs such as court–caught, larva–lava, sore–saw, career–Korea, tuber–tuba, manner–manna, western–Weston — and dear–idea.
Then I started to notice that sometimes I would hear rhotic speakers who nevertheless seemed to be pronouncing r at the end of idea. How could this be? Anyhow, I had to start expressing myself rather less sweepingly, more circumspectly. Some words (not only colonel) seemed to be exceptions. They had no <r> but they did have [r] for some rhotic speakers. (See also my blog for 11 October 2007, and Accents of English vol. 2 p. 343.)

(We non-rhotics do sometimes say aɪˈdɪər, of course, but only before vowels: that’s our “intrusive r”, ðə ˈveri aɪˈdɪər əv ɪt. But that’s a quite different matter.)

Now the Scottish (and correspondingly rhotic) phonetician Jim Scobbie has revealed the shocking facts. He reports that he has an “extra” r in the items Chicago, unauthorised, theatre, idea and, as he has just noticed, in inaugural.

There is no historical r at the relevant point in these words. The explanation is presumably false analogy in words first heard pronounced by nonrhotic speakers. If a Scot hears an English person say ˈɔːɡən organ and knows that it corresponds to Scottish ˈɔrɡən, it’s entirely understandable that he would infer that for English English ɪnˈɔːɡjʊrəl the corresponding Scottish form must be ɪnˈɔrɡjərəl. Spelling doesn’t come into it.

In rhotic accents outside Britain, where the influence of England’s non-rhoticity is weaker — in America, in Canada, even in Ireland — you would expect this sort of thing to be less likely to happen.


  1. I'm positive that I've also heard inaugural with an extra r before. I would expect inauguration is somewhat less likely to acquire the extra r, though.

    Here's a piece on 'intrusive intrusive r':

  2. Religious recordings by rural Southern Black speakers frequently feature 'Nora' -- i.e. Noah with an intrusive r.

    Is this a general feature? Or does it result from performance style?

  3. Nice spot by Jim - I too have an extra R in 'inaugural' (but no [j]), but not in 'inauguration', as Jongseong predicts. Many (rhotic) speakers from Tyrone in Northern Ireland, myself included, also have R in 'khaki' and 'lager', and I've sometimes heard it there in 'Chicago', 'sofa', 'idea' and even 'pillow' (pronounced as 'pillar').

  4. Jack Windsor Lewis28 April 2010 at 12:50

    John’s generalisation is made when he has on his teacher’s hat giving a rule-of-thum’ to students who don’t have English as their mothertongue. When he substitutes his analyst’s hat he’ll be saying something different. Jones in EPD1 in 1917 acknowledged the existence of variants of words like barren as /`bӕrn̩/ and quarrel as /`kwɒrl̩/. In 1979 I drew attention to the large numbers of words like American, Arabic, miracle, terrrible and borrowing which have pre-consonantal /r/ in their lexical forms for very many speakers. See Section 3.4.

  5. Well, there is also the rhotic pronunciation of Goethe as /ˈgɝ.tə/ and Gödel as /ˈgɝ.dl/, where mid front rounded vowels in German names are approximated with the rhoticized HERD vowel. It was quite surprising for me when I first heard my American logic professor use such a pronunciation for Gödel.

  6. I've heard a handful of pronunciations like this from speakers of English rhotic accents (West Country type). Isn't a rhotic pronunciation of "last" actually recorded in the Survey of English Dialects?

    As mentioned in the comments to the article Jongseong mentioned, there's also the "warsh" phenomenon found in some parts of the US.

  7. @Jongseong:

    …and don't forget about Schoenberg [ʃɝnbɚg]

  8. In Broad New York and Boston, intrusive-intrusive r (thanks for the term, Jongseong!) hyper-corrects final schwa to schwar in COMMA words (Cuba > Cuber) and diphthongs that end in schwa (idea > idear). Like plain old intrusive r, this is a socially stigmatized pronunciation, even though it can be heard pretty far up the class scale.

    Wash > warsh doesn't occur in urban New England (including New York metro area) - at least, I haven't come across it myself - but it does in many parts of the south, midwest, and even points further west. I think that phenomenon is traceable to early Scotch-Irish settlement and then migration patterns from there, but I'm not sure about that.

    I didn't know it could be heard across the pond. You learn something new every day!

    As a matter of purely incidental interest, I have a client with a ripe Noo Yawk accent that - among other delights - turns California to ˌkælɪˈfɔnᴊɚ and joint to ʤɝnt.

    I once had a client for whom idea > ideal. As he was an Italian-American from northern New Jersey, this wasn't traceable to ancestors from Bristol. Anyone else come across this in the US?

  9. I should clarify, perhaps: New York/Boston is a historically non-rhotic area, but the parts of the US where wash > warsh are, so far as I know, all rhotic.

  10. I wonder if hyper-rhoticity has something to do with the derivation of the term 'tater' for potato, which I think is US-specific (and nowadays only common in the term 'tater tots').

  11. Robert Plant of Led Zeppelin often uses pre-vocalic intrusive /r/ at the end of PALM, THOUGHT and COMMA words, but once in a while he uses an /r/ in START words even if a consonant follows.

  12. For me, career-Korea and tuber-tuba are differentiated by the quality of the unstressed vowel as well as the /r/; Korea has unreduced o and tuba has unreduced a.

    In Chicago, where is the intrusive /r/ placed? Are we dealing with Chicargo or Chicager? I can imagine either. In Chicago itself, the pronunciation of the word is a shibboleth: the "respectable" classes use THOUGHT for the stressed vowel, and PALM is the mark of the lower class, AAVE-speakers (often the same thing), and outsiders.

    Listening to Kennedy's Bay of Pigs speech, I hear only linking /r/ in Cuba, though there are probably people with intrusive r in this word. Idear, though, is a different story: it's a separate phonological shape that some rhotics idiosyncratically use, notably Eugene McCarthy, the 1968 Democratic candidate for President, who was always talking about "new idears" in his Minnesota accent. I don't know of any whole dialect region that uses it, though.

    Jongseong: My mother was born in Germany and moved to Detroit, where she ended up teaching German at Wayne University. While riding the streetcar (tram) home and back, she made a collection of the different conductors' pronunciation of Goethe Street, a regularly announced stop. The most common form was /ˈgiti/; the most exotic, /go.ˈiθ/.

    Amy: I think it's correct that warsh (and Warshington, etc.) follow the Scots-Irish. But are you sure your client actually has /ɚ/ in joint? The (now recessive) CHOICE-NURSE merger historically had a diphthong, roughly [ɜɪ], causing people lacking the merger to hear CHOICE in NURSE words ("Thoid Street") and NURSE in CHOICE words ("eating ersters"). I suppose this could have changed with the rhotacization of NURSE in recent years.

  13. In a much-belated reply to Nigel Greenwood, Gwoyeu Romatzyh was invented by a non-rhotic speaker of BrE, though not a native one: Yuen Ren Chao, the great Chinese philologist. It consistently uses postvocalic r as a tone marker only: the spelling used for the final retroflexion of official Putonghua, spelled r in Hanyu Pinyin, is l.

  14. One advantage (I suppose) of a non-rhotic accent is that the NURSE vowel is more easily available as a native equivalent to foreign ø or œ.

    In words like Schoenberg I hear many different AmE realizations (I see that LPD (1990) gives the GOAT vowel for AmE "Schoenberg", the FACE vowel for AmE "Goethe", and the FOOT vowel for AmE "Goebbels"!).

    In the first syllable of "meuniere" LPD 1990 gives the STRUT vowel for the first syllable in AmE.

    In more nativized words like "entrepreneur" LPD gives the NURSE vowel for AmE, but I only ever hear CURE -- which I presume is a spelling pronunciation.

    In proper names there is even more variety. Prominent Republican politician John Boehner's surname is pronounced with the FACE vowel, which always feels odd to me. Accidental senator George LeMieux uses the GOOSE vowel.

    I am fairly confident that any BrE speaker would use NURSE for every one of these. Am I wrong?

  15. @Jongseong: I don't know the origin of the rhoticity in "Tater Tots," but final schwar where most English and a few American accents have schwa is certainly not uncommon in the US. I think it follows roughly the same Scotch-Irish lines that gives us warsh for wash. We generally read "feller" and "winder" with rhoticity, and so the non-rhotic pronunciations of such spellings would have to be rendered as "fella" and "winda" (or "winduh") to be properly understood. (A.A. Milne's "Eeyore" is a joke lost on most Americans.)

    Yet another alternative pronunciation of "window" is "windy" - which I think must trace its origin to Scotland.

    @John Cowan: Yes, of course I'm sure; I heard him do it only last Friday. This client is in his sixties, hails from Far Rockaway, and has a classic CHOICE-NURSE merger, which he has taught himself to suppress with only partial success. So while many of his NURSE words no longer have ɜɪ most of the time, but instead have a sort of ɝ made with a flattened tongue on the r-coloration (that I haven't yet figured out how to transcribe), and he rarely substitutes ɝ for ɔɪ, those features are still part of his idiolect, and he is liable to produce them occasionally when he is either very pressured, or extremely relaxed.

    I think I have a recording of him telling me how, when he was a kid, he and his friends regularly made fun of a girl who moved to the neighborhood, because she didn't pronounce "toilet" the normal way; that is, as "terlet."

  16. BTW, am working on a breakdown for rural Somerset accents right now, and one of the marvelous samples from SED has intrusive r in warsh, warshing. I wonder whether our director will let us use this feature in our production; and if so, whether I will encounter any resistance from the cast.

  17. Here in Canada, I've received report from a Newfoundlander that "warsh" is dialectal for "wash" as well. This is a feature of Eastern Canadian speech, apparently also reported in Prince Edward Island, so I presume this is part of a general New England phenomenon, north and south of the border.

    I've heard absolutely *no* native-born Manitobans say this. (Us Central Canadians have enough speech anomalies of our own to deal with.) And... I admit... we may even pee ourselves laughing at anyone who does insert misplaced r's.

    All in all though, this ahistorical /ɹ/ is by far an exception rather than norm so there's no need to confuse sheltered non-rhotics and helpless ESL students.

  18. "taters" occurs in Sam Gamgee's Mummerset in "Lord of the Rings".

    This Wikipedia list is relevant, if short.

    I certainly have heard "Chicargo" and "sarcrament" in southern Ireland. The first could be by analogy with "cargo", but the second..."sarcastic"? Introspectively, the PALM vowel in my "sacrament" is definitely closer to the START vowel than most words in the class. I don't know why that is, but maybe for some speakers with similar accents it strays so far it tips over into the START class and acquires a /r/ accordingly.

  19. (A.A. Milne's "Eeyore" is a joke lost on most Americans.)

    Or Winnie-ther-Pooh, for that matter.

    Concerning names with oe, there's The Simpsons' Matt Gr[FACE]ning, which I had assumed to be Matt Gr[GOAT]ning before I knew better.

  20. The SED shows us that 'warsh', 'taters', 'winders', etc. are/were found in southwest English dialects (which often have/had extensive hyper-rhoticity). Not at all surprising they turn up in Eastern Canada.

    Since hyper-rhoticity is only found in a few words in Ulster dialects, especially new(ish) and uncommon words such as 'sofa', 'lager' and 'khaki' (partly because a distinct PALM vowel is almost non-existent there), I suspect that 'Scots-Irish' isn't the source of forms such as 'warsh', 'taters', 'winders', etc. in North America (although 'kharki' turns up in Canada too, according to AoE: 491).

    But it might well be the source of 'windy', which is a pronunciation you can still hear in Tyrone today.

  21. Typically, Scottish has the same vowel for TRAP, PALM and START, and is rhotic, but has a back vowel allophone before tautosyllabic /r/. People vary a lot, and I suspect tend to lose these "incorrect" forms for coda /r/. I don't have one in lager or palaver, but contrariwise for years had no final /r/ in Gibraltar.

    Vowel quality variation without any effect on the rhotic is possible before intervocalic non-coda /r/ in the stressed syllable of a trochee, e.g. safari, but not in Barry, Harry. Variation is also seen on suffixation, so either TRAP or START is possible in starry, where I would expect a correlation between the frequency of the derived /ar#e/ and the likelihood of START: high freq or opaque derived forms (carry) will have TRAP, starry seems about halfway, and a nonce word indicating how "R-like" something is... i.e. "R"-y... *has* to be START for me.

    One related innovation by Scottish rhotic speakers is use of what for them is an allomorph -ry in forms like vanillary and bananary. Furthermore, the /r/ appears in spelling for both rhotic and non-rhotic speakers as far as I know, which is unusual for intrusive /r/.

  22. It's pretty common to render the German oe with the FACE vowel. The old song is "Danke shane," isn't it? There's a "Gatey" park in Sacramento. (Named for the donor, not the poet.) Took me a long time to figure that out. Using NURSE to pronounce the German makes perfect sense to me, but it's much closer to the German if I try to pronounce NURSE as a non-rhotic speaker would.

    In my native accent, ɔ does not exist without an r or l following. Accordingly, I might possibly infer an r or l where none exists, particularly if the speaker is non-rhotic and I'm unfamiliar with the word.

  23. You can definitely hear "warsh" and "Warshington" in my hometown in the Midwest U.S, although I've never actually pronounced them that way and I think those pronunciations might be receding. I have no idea why people pronounce them that way. I've never heard a decent explanation.

    We also occasionally go in the other direction and "drop r's" in words like "surprise" and "governor" (the first "r" in both; "governor" is also commonly two syllables but that's a different story). Yes I know this is also a feature of southern Irish English and that you cover it in Accents of English. This has actually caused spelling problems for me and I thought "surprise" was spelled "suprise" up until just a few years ago, because I never did (and still don't) pronounce the first "r" in that word.

    You can hear "karki" for "khaki" in Canada as someone mentioned. I heard this might come from Canadian soldiers imitating the pronunciation of English troops during some war, though I can't remember which one. It's understandable that they heard it that way because of the PALM-START merger. Although it makes me wonder why Canadians didn't just use their own FATHER (PALM)- BOTHER (LOT) vowel. Is there really some sort of subtle phonetic difference between PALM and START in RP (and accents like it) that Canadians could somehow hear?

    I've also heard "idear" for "idea" from completely rhotic northern Vermont speakers, even when the next word did not begin with a vowel (even at the end of statements, etc.). So obviously this can't be an "intrusive r", at least not in the non-rhotic sense of the term.

    On a somewhat related note (related to the "karki" thing anyway), I just thought of the term BARF-BATH merger. Are Southern English kids going to "the barf room" these days? I don't know anymore. Nice blog (thread, post, whatever).

  24. I'm American and I usually hear "Goethe" as /ˈgɝ.tə/ (or maybe even rhyming with "Earth" though I'm not sure about that) and "Goebbels" rhyming with "gerbils". They sound perfectly normal to me. But then again I don't hear those words/names every single day.

    Matt Gr[FACE]ning sounds normal to me because that's how I've heard it pronounced the times I've heard it pronounced. I've also heard the last name "Moechnig" pronounced /ˈmɛknɪg/ by people who have it.

  25. Is there really some sort of subtle phonetic difference between PALM and START in RP (and accents like it) that Canadians could somehow hear?

    Not in RP. In Mid-Atlantic accents as taught to Hollywood actors in the 30s and 40s I believe that there was supposed to be a distinction: PALM = /ɑ:/ and START = /ɑə/.

    I don't know whether any native accent has, or had, such a contrast. Perhaps some East Coast US?

  26. "I heard this might come from Canadian soldiers imitating the pronunciation of English troops during some war, though I can't remember which one."

    It was World War II. I should have known that.

  27. The use of the FACE and DRESS vowels for German ö (and, for that matter, the FLEECE and KIT vowels for German ü) is not phoneme replacement on the part of English speakers, but reflects rather the fact that for many German speakers ö and ü have been unrounded to e and i. Some examples from around Austin and other parts of Central Texas, where I grew up: the town of Gruene is pronounced just like "green"; Robert Mueller Municipal Airport was pronounced "Miller" (pronouncing it "Mewler" identified you as an outsider); Koenig Lane is pronounced "Kaynig" (again, pronouncing it "Kernig" or "Koanig" identifies you as an outsider), and so forth. All of these pronunciations reflect the pronunciation of the German settlers themselves, who had already unrounded these vowels back in Germany.

  28. That's right. In fact, most German (and all Yiddish) dialects used to have ö and ü unrounded, until the advance of the standard language reversed this to a degree even in dialects and certainly in regional variants of the standard.

  29. Anon: I just thought of the term BARF-BATH merger. Are Southern English kids going to "the barf room" these days?
    1. because the American word "barf" is not used in Britain: we say instead "to be sick (all over the floor)";
    2. because the bathroom for us is the room with the bath in it. The place you "go to" in Britain is the toilet (or lavatory, loo, bog, etc).

  30. I went to uni in Newcastle Upon Tyne in the 1990s with a Cockney 'geezer' who became famous for having 'barfed' in the 'barf' (even though the toilet was right next to it) after drinking too much...

  31. I also notice that various English accents insert an "r" sound into words containing the spelling "aw" not only as in saw ("I saw-r-a man"), but also in the middle of words and phrase such as "drawing" ("draw-ring") and "awe-inspiring" ("ore-inspiring"). Is this becoming a general rule or are those two words exceptions?

  32. To continue my previous post, am I right in thinking that even in the accents with the most intrusive-r occurrences the word "clawing" would not be pronounced as "clawring"?

  33. How do you Americans pronounce the ö in Eyjafj... <g,r&dVF!>

  34. "1. because the American word "barf" is not used in Britain: we say instead "to be sick (all over the floor)"

    Well yeah. Obviously we could say that too. Buy it's not as specific or descriptive.

  35. Gadi: am I right in thinking that even in the accents with the most intrusive-r occurrences the word "clawing" would not be pronounced as "clawring"?
    No, you are not right. See the relevant entry in LPD, where an optional r is shown in "clawing". There's nothing special about that word, and the rule is a very general one.

  36. "No, you are not right. See the relevant entry in LPD, where an optional r is shown in "clawing". There's nothing special about that word, and the rule is a very general one."

    I guess it's that time of month again.

  37. @Gadi:
    "intrusive r" even extends to "withdrawral", where the suffix is morphological rather than inflectional.

  38. About "khaki" in Canada.

    If you had a general idea of what a (southern) British accent was and how it corresponded to your own, and you heard a Briton say "khaki," you might think one of two things - either that it was the BATH vowel, in which case you'd use your own BATH vowel; or that it was the START vowel, in which case you'd use your own START vowel.

    And in fact, both pronunciations, "cacky" and "carky" can be heard in Canada. The pronunciations "cocky" and "cahky" (which are actually different for those Canadians who pronounce "foreign a's" in a special way) seem likely to be due to American influence.

  39. Hardly American influence, Anonymous, since Americans use TRAP in khaki.

  40. Reading this is making me feel Orwkward.