On American r-lessness

« previous post | next post »

James Fallows has been superintending an interesting discussion at the Atlantic about how strange early twentieth century American announcers sound to us today (There are five articles in the series so far, listed with links here). The comments on his articles suggest that we need make certain distinctions.

There seem to be two things that mainly strike Fallows and the commenters as comprising this way of speaking: its theatrical tone and its pronunciation. The tone issue is easier to parse: announcers on old radio, early television, and in old movies, as well as various newscasters, spoke in an elevated, singsong tone because that tone had been necessary in the era before amplification, when one only experienced announcing in theatrical venues.

In an America new to microphones, the unconscious cultural expectation was that public announcements were couched in a theatrical style – theatrical, as in pitched to be heard unamplified in an auditorium by hundreds or thousands at a time. That tone seemed as natural to people of this time as, for example, recording artists releasing packages of songs of roughly an hour’s length called “albums” seems to us now despite that practice’s roots in a particular physical technology now obsolete, or that pop music is today usually sung in a Southern / black cadence even by whites who don’t speak that way in real life.

Gradually, it dawned on people that microphones allowed a less hotly-pitched, intimate way of communicating: FDR’s Fireside Chats were an example, as was the singing style called crooning. However, the sunny, chirpy delivery persisted as a style well into the 1960s on radio, television and in films simply because of familiarity.

Meanwhile, when today we find the old announcing style peculiar in terms of pronunciation, what is mainly striking us is a particular feature, r-lessness. That is, r’s lost at the end of syllables, such that Carter comes out as “Cah-tuh.” To Americans today, this sounds like Downton Abbey, such that we wonder why people in black and white or on old radio broadcasts sound “British” despite talking about Babe Ruth and Times Square.

As commenters have noted, elocution lessons of the era explicitly taught performers to sound “plummy” (as depicted by Kathleen Freeman’s speech coach in Singin’ in the Rain), but also encouraged r-lessness. However, the Fallows discussion trips up a bit in implying that this was simply a matter of trying to sound British, and that no Americans actually sounded like this in real life.

Britain’s r-lessness was only part of the story of r-lessness in America. American r-lessness has had, basically, three components.

  • One, New England’s – generally traced, indeed, to r-less dialects brought over from Britain by the original settlers.
  • Two, the South – where British influence is often suggested for cities like Charleston and Savannah, but a more convincing explanation in applying to the whole region is that the South was home to slaves from Africa, whose non-native rendition of English elided syllable-final r’s, especially since their native languages almost never had r’s of that kind.
  • Three, the big cities of the eastern seaboard, such as Boston and New York (but not, interestingly, Philadelphia or Baltimore). This is the area most of interest in this discussion. Often it is thought that people in Boston and New York were imitating Brits, but this is only speculation, and what makes it questionable is that working class and poor people in these cities were also r-less. Think of the accent associated with The Honeymooners, The Sopranos, and South Boston – both Ralph Kramden and FDR would say “fee-ah” rather than “fear.” It seems unlikely that humble people in old New York and Boston were much invested in imitating upper-crust Londoners they never met. There was a time when we were to think such people were imitating “their betters” at home, but sociolinguistic theory has dismissed that notion: Ralph Kramden wasn’t trying to sound like FDR. Evidence of humble people’s speech changing in response to wealthy people’s speech is all but non-existent – if anything, the influence tends to go the other way.

There is reason to think, then, that people in New York and Boston were r-less not in imitation of people across the ocean, but for the same reason that (some) Brits had become r-less: natural processes of sound change. R’s are delicate after vowels at the end of syllables, like l’s, and there is no reason that the r-less that developed in Britain could only have happened there – one might even wonder why it would not happen on its own in places elsewhere such as the United States. This explanation, based on how the cookie crumbles here rather than there, would also explain why r-lessness happened in Boston and New York but not Philly or Baltimore.

In any case, with this take on American r-lessness, we are in a position to see two things.

First, Americans most certainly did really talk “that way” in real life – it wasn’t just a pose. That is, elocution schools urging r-lessness were working not just from British Received Pronunciation, but from how actual people spoke in and around New York and Boston, considered cultural beacons for the country.

Here, for example, is Bette Davis, of Lowell, Massachusetts, speaking off the cuff in a 1975 interview. Note the r-lessness – she hadn’t been saying “Peetah! The lettah!” for “Peter! The letter!” in old movies as an affectation. Perhaps for a celebrity like Davis, what began as an affectation became habit? But — accounts are rife of people of this era talking “this way” in real life: Theodore Roosevelt, raised in Manhattan in the 1860s and 1870s, was ridiculed for calling “Mr Speakah! Mr. Speakah!” in the New York State Assembly. Oral histories of countless public figures of this era reveal this kind of r-lessness; it makes almost any New Yorker or Bostonian of the time sound weird to us in recordings.

Second, a person of humble origins like George Gershwin was not talking “like that” in being r-less. Gershwin had the r-lessness that all New Yorkers of his era had – you didn’t need to be trying to sound like FDR, or an announcer, or “cultivated,” to be r-less in the 1930s in New York. In fact, beyond the r-lessness, Gershwin, although clearly having learned to speak “above” the dialect of the streets he grew up on, reveals in the radio shows he appeared on that his native speech is not FDR’s in his pronunciation of er. He has the “oy” pronunciation now associated with old Brooklyn, but which was actually a general demotic New York / New Jersey trait at the time. Note, no one ever actually said “foist” for first – it was more like “fuh-eest,” and Gershwin has that pronunciation here and there on words with er. He is reading from a script, and therefore we know that he would have been suppressing his true vernacular, as sociolinguists would put it. However, in a Gershwin radio appearance available on line, listen at roughly 4:54 for a good old-fashioned New York “fuh-eest.” Gershwin was talking like himself, not “like that.”

As to Fallows’ question as to who the last person to truly talk “that way” was, the truth is that there are still very old New Yorkers and Bostonians living who grew up with the old r-lessness. Any time you read of someone passing away at 100 or so who spent most of their life living on Park Avenue (or somewhere like Bensonhurst, but you’re less likely to read about them) they may well have been r-less in their speech to an extent that would have sounded quaint to us.

Just why r-lessness went out of fashion after World War II rather than earlier or later remains underinvestigated. Soldiers from all across the nation having mixed for four years, perhaps? I doubt that the reason is anything as grimy as one commenter's suggestion that rich people wanted to talk in a way that wasn’t so imitatable, however. Boston cab drivers have never been imitating Boston Brahmins; r-lessness happens by itself.


  1. leoboiko said,

    June 11, 2015 @ 9:18 am

    Incidentally, what was Poe’s /r/ ? A Bostonian in the 19c? When the raven quoth “nevermore”, was it /ˌnevəˈmɔː/ ?

  2. Dick Margulis said,

    June 11, 2015 @ 9:21 am

    I read an article some time ago that associated the r-less feature of the "Brooklyn accent" (toity-toid and toid) with the wave of Irish immigration (Boston and New York being major loci) and the influence of Gaelic phonology. Possibly hogwash that has since been discredited. Or maybe that has nothing to do with the question under discussion. Just bringing it up in case it has some bearing.

  3. Tim Martin said,

    June 11, 2015 @ 10:36 am

    Britain’s r-lessness was only part of the story of r-lessness in America. American r-lessness has had, basically, three components.

    One, New England’s – generally traced, indeed, to r-less dialects brought over from Britain by the original settlers.

    Um… is this correct? British English was still rhotic when America was settled.

    [(myl) My understanding is that New England (or at least Boston and points east) retained enough cultural contact with Olde England to travel along with the development of r-lessness. But I'm not sure what the evidence is for any view of the history.

    Also see Ben Zimmer's comment below…]

  4. mollymooly said,

    June 11, 2015 @ 11:40 am

    John Wells coined "rhotic/nonrhotic" because r-less "r-ful" sounded "awful".

  5. Steve Reilly said,

    June 11, 2015 @ 12:22 pm

    Was an r-less accent also cultivated at prep schools in the US? Gore Vidal grew up in Washington DC, a rhotic city. But he had the r-less accent of FDR. And even the New Yorker George Plimpton had an accent that sounded more British than Manhattan. (When I first watched Ken Burns's documentary on the Civil War, I was puzzled that he got an Englishman to read the diaries of NY stockbroker George Templeton Strong. Then I realized it was Plimpton.)

  6. Peter Shor said,

    June 11, 2015 @ 12:25 pm

    @Tim Martin:

    The London dialect was still rhotic when America was settled. But the Puritans mainly came from East England; do we know whether their dialect was rhotic?

  7. Bean said,

    June 11, 2015 @ 12:32 pm

    So… do linguists do longitudinal studies about how accents change in individuals, over time, as the language around them undergoes natural change? Not just in a second language, but in their native tongue?

    I have recordings from 1981 of my mother speaking to me as a child. She sounds different in them, and not just younger (and annoyed with my young self!). Her accent has changed in the past 35-ish years. But she's just one data point.

    All that to say, the 100-year-old New Yorkers are likely not quite "time capsules" of that accent.

    [(myl) In a word, yes. For example,

    Harrington, Jonathan, Sallyanne Palethorpe, and Catherine I. Watson. "Does the Queen speak the Queen's English?." Nature 408, no. 6815 (2000): 927-928.

    Harrington, Jonathan. "An acoustic analysis of ‘happy-tensing’in the Queen's Christmas broadcasts." Journal of Phonetics 34, no. 4 (2006): 439-457.

    Sankoff, Gillian, and Hélène Blondeau. "Language change across the lifespan:/r/in Montreal French." Language (2007): 560-588.


  8. Sergey said,

    June 11, 2015 @ 12:39 pm

    Perhaps, the literacy and the lack of it has to do with it? Boston and New York were full of new immigrants in the late 1800s. The American "r" sound is not rolling, neither like German/Yiddish nor like Italian/Russian/Irish. It both sounds to the native speakers of these languages as something vaguely vowel-ish and is difficult to pronounce, so it turns into something like "ah" or "oy". And if someone is not particularly literate and is learning English as a spoken language, without much reference to writing, he would just learn it like that. But the literacy had spread in the 20th century, and then perhaps as the "normal" speech spread through the radio, movies and TV, both factors worked towards speaking more like written and homogenizing the local ways of speaking.

  9. Ben Zimmer said,

    June 11, 2015 @ 1:03 pm

    Tim wrote: "British English was still rhotic when America was settled." It's true that the prestige BrE dialect was still rhotic, but the New England settlers brought over non-rhotic dialects that weren't so prestigious at the time. See Richard Bailey's Speaking American: A History of English in the United States, which I wrote about in a Boston Globe column: "A pahticulah way of talking."

  10. David Kidd said,

    June 11, 2015 @ 1:06 pm

    My final paper in Sociolinguistics looked at the way people conceived of themselves as native to Norfolk, Virginia and how that affected rhoticity in their pronunciation of the city name (and only the city name–I assumed that the rest of their dialect was rhotic). There's a least a folk belief, that is somewhat supported by evidence, that natives of the city pronounce it non-rhotically. In interviews, many respondents talked about the "l-lessness" of the second syllable.

  11. Jeff W said,

    June 11, 2015 @ 1:07 pm

    …the Fallows discussion trips up a bit in implying that this was simply a matter of trying to sound British, and that no Americans actually sounded like this in real life

    I think that has to do with James Fallows’s idea that “Mid-Atlantic English” refers to English that is “midway” between North America and Britain, since he says

    To me, “Mid-Atlantic English” is the nom juste for a related but distinct phenomenon (which is also mentioned in Wikipedia). That is the tendency of Americans trying to sound more British, or Brits trying to sound more Yank, to split the difference and speak in an accent whose home ground is no real country but somewhere in the middle of the sea.

    which is not what I always thought the “Mid-Atlantic” in “Mid-Atlantic English” refers to. (In fact, his interpretation never even occurred to me.) Now I’d like to know what it actually means.

  12. J. W. Brewer said,

    June 11, 2015 @ 1:27 pm

    Ben: I think (there have probably been prior LL posts with links on the subject I don't have time to dig up . . .) that some of the studies of Philadelphia English done by Labov and/or his colleagues/disciples have found generational strata, i.e. the prominence vel non of certain distinctive features of pronunciation varies among life-long Philadelphians depending on their decade of birth, suggesting that those whose accents were formed in a particular earlier era do not automatically pick up the subsequent shifts that are forming their children's and grandchildren's accents even though they're living in the same general milieu.

    A few years ago I happened to hear the audio of an interview done with H.L. Mencken (1880-1956) fairly late in his life and was struck by how he had various quirks of pronunciation I strongly associated with my maternal grandfather (1903-2003) and had never consciously noticed in the speech of anyone younger than that. The one I can remember was second-syllable stress on "Broadway" (as in the NYC thoroughfare), but there were others. But neither Mencken nor Granddad sounded particularly close to patrician "MId-Atlantic" in the sense I think is meant, i.e. like Thurston Howell III on Gilligan's Island.

  13. Matt McIrvin said,

    June 11, 2015 @ 2:29 pm

    r-lessness is still quite common in Massachusetts and New Hampshire, though I think it's fading in younger generations. It's not the "announcer accent", though, which sounds to me more like the old upper-class New England Brahmin accent.

    The pop singer ZZ Ward (who is from Pennsylvania, near Philadelphia) sings with a pronounced curl/coil merger on "Put The Gun Down". I'd thought it was nearly gone.

  14. Chris C. said,

    June 11, 2015 @ 3:39 pm

    I'm not so sure the "oy" pronunciation was ever a New Jersey trait. I never recall hearing it. My memory doesn't go back any further than the mid-1960s, but I did spend quite a lot of time in Hudson County, and if it had survived anywhere I'd have thought it would be there.

  15. Jonathan Wright said,

    June 11, 2015 @ 4:24 pm

    Fascinating discussion. Maybe I missed it among all the comments but the process of change needs careful analysis. It's hard to imagine a community of arhotic speakers reverting to rhotic speech without some significant demographic change. I'm British and arhotic and trying to speak rhotically requires massive extra thinking. How did the rhotic children of arhotic parents learn that skill – at school? from other relatives?

  16. Sili said,

    June 11, 2015 @ 5:06 pm

    I had the reverse reaction to Jeff W. I used to think that Mid-Atlantic indeed meant a mishmash of BrE and AmE. I was disappointed to learn that there were Mid-Atlantic states and that the word referred to those. Waste of a good name.

  17. Mark S said,

    June 11, 2015 @ 6:17 pm

    @Jonathan Wright: Indeed. I'm a British-American, but it takes a real effort to pronounce those "r"s. Hugh Laurie, who manages an excellent American accent as Doctor House, said that he has most difficulty with "coronary artery".

  18. Jeff W said,

    June 11, 2015 @ 6:56 pm

    I had the reverse reaction to Jeff W.

    Well, it was easy for me to think that, being from one of those Mid-Atlantic states (New York).

    I’m actually now a bit curious as to how that term “Mid-Atlantic English” arose. I always thought of people like Franklin Roosevelt (obviously from New York) and Katherine Hepburn (from Connecticut) in connection with it but I have no idea if that has any basis in reality.

  19. J.W. Brewer said,

    June 11, 2015 @ 9:05 pm

    "Mid-Atlantic" for that posh/artificial/now-archaic movie star pronunciation = metaphorically somewhere near the Azores. It's not at all the same thing as "Middle Atlantic," for the distinctive pronunciation in the part of the U.S. I grew up in (think as a first approximation of anywhere close enough to either Philadelphia or Baltimore to be able to get decent reception of tv stations from either of those cities). Middle Atlantic varieties of AmEng are rhotic, situated south of the traditionally non-rhotic NYC and New England varieties but north of the traditionally non-rhotic Southern (coastal/Atlantic seaboard subset) varieties.

  20. Viseguy said,

    June 11, 2015 @ 9:41 pm

    Speaking of "foist" (or "fuh-eest") for "first", I have a question: is "tirlet" for "toilet" the other side of the same coin? I never heard the latter growing up in Bensonhurst/Gravesend in the '50s and '60s, except from a maternal uncle who lived in Astoria, Queens — which led the childish me to associate it, probably incorrectly, with Queens. My mother and her brothers grew up in the 19-teens and -twenties on the Upper East Side (in a tenement one block above 96th Street), but my mother wouldn't have been caught dead saying "tirlet". My uncle's twin brother moved to Boston shortly after WWII, but damned if I can remember whether he said "tirlet" or "toilet". They're all gone now, of course — what memories!

  21. Jim Ancona said,

    June 11, 2015 @ 10:35 pm

    My grandmother was born in 1911 to a working class Italian immigrant family in New Jersey. Her non-rhotic accent, which she retained until her death 10 years ago, always sounded "upper class" to me because of the association with movies and old-time radio announcers. On the other hand, my mother pronounces those r's even though she was born in the same area only 21 years later.

  22. Anschel Schaffer-Cohen said,

    June 11, 2015 @ 10:47 pm

    I'd say the non-rhotic New York accent is alive and well, if much less common than it used to be. I went to middle and high school in Hoboken, NJ, and one of my classmates spoke with a non-rhotic, stereotypical New York/Brooklyn/New Jersey accent. I believe he was a second- or third-generation Hobokener, of German, Italian, and possibly Irish descent. And he wasn't alone, I definitely still hear this dialect spoken here and there.

    That said, the fading-out is visible in my own family. All of my grandparents (born in the 20s and 30s) speak/spoke non-rhotically, but my parents (born 1959 and 1960) speak a rhotic dialect that is generally less distinctively New Yorker.

  23. Bloix said,

    June 11, 2015 @ 11:47 pm

    My understanding is that tirlet is a hypercorrection.

  24. Mark F. said,

    June 12, 2015 @ 12:02 am

    The mid-Atlantic accent isn't just General American without the R's, is it? Aren't there other differences? Maybe not; I'm having trouble putting my finger on them. But the vowels just seem a little different.

  25. Roger Lustig said,

    June 12, 2015 @ 12:03 am

    Is this why "er" was used as an interjection in the books of my childhood? Nobody I knew ever said it to rhyme with "fur"; but "uh" could be spelled "er" if RP (or mid-Atlantic) is assumed.

  26. K. Chang said,

    June 12, 2015 @ 1:19 am

    Somehow my mind linked the "r-lessness" with this anecdote shared on Jezebel / Kitchenette blog…

    Full link is at the end, but basically, it's about servers in a restaurant, and this server was extremely confused when the customers, a bunch of Ghanan diplomats, with extremely thick accented English, ordered what sounded like "boof farovin"

    The server checked with the manager, manager had no luck understanding the request either. They brought out the GM, who STILL can't figure out the request either. Finally the server said "let's just give them one small plate each of the appetizers for free, they're cheap". The GM and M reluctantly agreed.

    Their request turned out to be "buffalo wings". ;) As it's the plate that was emptied. :D


  27. Alan Palmer said,

    June 12, 2015 @ 5:05 am

    I'm English and to me a 'mid-Atlantic' accent means a (generally fake) accent that has elements of British and American speech. It's neither one nor t'other. With the arrival of 'pirate' radio stations and more pop music on the radio in the 1960s and 70s the presenters often tried to sound like American DJs, often with unintentionally hilarious results. 'Mid-Atlantic' would also be applied to an American who was trying to sound British.

    We had our own instances of 'announcer-speak' which were similar to the American examples already mentioned. As already suggested, I suspect the tradition of speaking in a theatre, either unamplified or with low-quality equipment produced an orotund, almost florid, delivery. Many of the film stars of the time, notably Celia Johnson and Trevor Howard in Brief Encounter, also spoke in an exaggerated RP accent.

  28. Sidney Wood said,

    June 12, 2015 @ 5:25 am

    There were some questions about when British accents lost rhoticity. The earliest signs were in 15th century spellings in East Anglia and E Midlands (Paston Letters, Lincoln Wills). William Matthews (Cockney Past and Present) has plenty of examples of 16th century nonrhotic spelling variants in London. Mugglestone (Talking Proper) reviews the prescriptive debate for and against rhoticity and concludes that RP was losing its rhoticity from the late 18th century onwards while the debate continued to around 1850.

  29. Bloix said,

    June 12, 2015 @ 7:27 am

    Richard Lustig – yes, e-r spells uh.

  30. Bean said,

    June 12, 2015 @ 8:46 am

    @myl: Cool! (Although I didn't sign up to read the whole article about Montreal French.) I contend that the Queen doesn't count because when you're being recorded you tend to use a "public speaking voice" and more deliberate pronunciation.

  31. Matt McIrvin said,

    June 12, 2015 @ 8:47 am

    I always noticed that Frank Morgan in the 1939 Wizard of Oz (playing several characters, including the Wizard himself) has the curl/coil merger, as in his great line "I can't come back! I don't know how it works!"

    The Wizard claims to be a Kansas man (though his balloon says "State Fair Omaha", which is where he's from in the book); neither origin jibes with his accent, but Frank Morgan was a New Yorker.

  32. Matt McIrvin said,

    June 12, 2015 @ 8:55 am

    @Roger Lustig: I think so, yes, and I never realized this myself until quite recently.

    The curious thing is that sometimes Americans write "er" in print even though nobody ever says "er" as it would be pronounced by most Americans. Presumably they got it from books.

  33. Matt McIrvin said,

    June 12, 2015 @ 8:57 am

    Also, English people often put superfluous Rs in words that they're spelling in comical eye dialect.

  34. Sophie Sofasaurus said,

    June 12, 2015 @ 10:06 am

    "Also, English people often put superfluous Rs in words that they're spelling in comical eye dialect."

    We UK people sometimes use an R to indicate the associated change of a vowel. It isn't really an eye dialect. See "Accents: Which came first 'bath' or 'barth'?" for example: http://www.bbc.co.uk/newsbeat/article/33055518/accents-which-came-first-bath-or-barth

  35. Captain Bringdown said,

    June 12, 2015 @ 11:11 am

    My understanding is that tirlet is a hypercorrection.

    Interesting. Perhaps the same is true for the "peetz-er" pronunciation of "pizza."

  36. Warren Maguire said,

    June 12, 2015 @ 11:49 am

    Dick Margulis: The idea that Irish Gaelic has anything to do with non-rhoticity hardly makes sense given that Irish tolerates some /rC/ clusters (e.g. ard, amharc, ort) despite epenthesis in others, and given that Irish English varieties are almost universally rhotic.

  37. Bloix said,

    June 12, 2015 @ 12:04 pm

    Years ago i worked with a young woman from New Jersey who quit in order to join her boyfriend on his ranch in Montaner.

  38. Coby Lubliner said,

    June 12, 2015 @ 12:07 pm

    Matt McIrvin: "Also, English people often put superfluous Rs in words that they're spelling in comical eye dialect."

    It isn't only in comical eye dialect. The R in Burma (originally spelled Burmah) is supposed to represent Burmese "Bama" [bəmà], and that in Myanmar is for Burmese Myanma [mjəmà] (Wikipedia). Also, A.A. Milne's Eeyore is "is an onomatopoeic representation of the braying sound made by a normal donkey, usually represented as "hee haw" in American English: the spelling with an "r" is explained by the fact that Milne and most of his intended audience spoke a non-rhotic variety of English in which the "r" in "Eeyore" is not pronounced as /r/" (Wikipedia again).

  39. Coby Lubliner said,

    June 12, 2015 @ 12:14 pm

    @Captain Bringdown and Bloix: people sometimes confuse the "intrusive R" of non-rhotic speakers with a superfluous added R, as was the case with JFK's supposed "Cuber"; this R comes in only if the next word begins with a vowel.

  40. Captain Bringdown said,

    June 12, 2015 @ 1:07 pm

    I just took a call asking me to fax him information for the racetracks Belmont and Santer Aniter. His fax number area code is 212.

  41. K. Chang said,

    June 12, 2015 @ 1:12 pm

    @Coby Lubliner — hmmm… the Wikipedia entry on Eeyore also pointed out that Eeyore himself spelled it… "eoR" :)

    I can do a British or Australian accent, but I never thought about what made it special (i.e. rhotic / non-rhotic) and I learned something today! :)

    Going on this drop-syllable thing…

    Military occasionally drops the first syllable when barking orders. "Attention!" is often barked as "ten-SHUN!" and "present arms!" is "'sent HARMS!" Is there a term for this kind of… verbal clipping?

  42. Andrew (not the same one) said,

    June 12, 2015 @ 2:34 pm

    Coby Lubliner: yes, but those R's aren't superfluous; they indicate a modified vowel. 'Burma' sounds nothing like 'Buma' or even 'Bumma', and the 'ur' vowel can't be represented in any way which doesn't involve an R. Likewise 'Eeyore' is not pronounced 'Eeoe'. What I take it Matt McIrvin is referring to is the use of forms like 'barth' in comic representations of the speech of working-class characters; this is superfluous, since in standard southern English 'bath' and 'barth' would be pronounced exactly the same anyway.

  43. Chris C. said,

    June 12, 2015 @ 2:52 pm

    @Viseguy — "Tirlet" is indeed a stereotypical Queens accent, which is why Archie Bunker talked that way.

    @Coby — Similarly, it took me forever to realize that when Terry Pratchett wrote an Igor from Uberwald calling people "marthter", he intended it to sound as if I were to write it "mahthter".

  44. Roger Lustig said,

    June 12, 2015 @ 4:23 pm

    @Bloix: Thanks. Now, what does 'R-o-g-e-r' spell? ;)

  45. David B Solnit said,

    June 12, 2015 @ 4:58 pm

    Coming a little late to this thread, but nobody seems to have pointed out that Gershwin and Vallee don't drop all postvocalic r's. I hear the r loud and clear in Kern and Berlin. Score and work seem to vary, but are rhotic at least as often as non-rhotic. Is it phonological conditioning (e.g. r dropped more often after /a/, less after /ɔ/ and /ə/)? Inconsistent effort to maintain a non-native pronunciation (acrolect)? More research needed!

  46. Pflaumbaum said,

    June 12, 2015 @ 6:44 pm

    @ K Chang

    You're not alone. In my experience, my fellow Englishmen and -women don't actually realise that they don't pronounce their syllable-final, pre-consonantal r's. (Or rather, don't pronounce them a consonants.)

    When doing an American, Scottish or Irish accent they will do the inverse of what you do in your imitations – insert r's (sometimes superfluous 'mozzareller' type ones). But like you, they may not quite know that's what they're doing.

    We also often don't believe we have 'intrusive' r's, though most of us do – 'law and order' is loranawduh(r), and you never hear the runner Jessica Ennis called anything except Jessica Rennis (except by some speakers on the BBC, which may have put out an internal stricture on the matter).

  47. James Kabala said,

    June 12, 2015 @ 7:17 pm

    Jeff W.: To try to make a bit clearer what others have said, "Mid-Atlantic accent" does refer to a hybrid British-American accent (although an American trying to sound British and a Briton trying to sound American will likely sound very different). The separate use of "Mid-Atlantic states" for states halfway down the U.S. East Coast is potentially confusing, but usually should be clear from context. (Although it is true that FDR or Grace Kelly or George Plimpton did indeed come from mid-Atlantic states! Not Katharine Hepburn – while part of Connecticut may be in the New York metro area, I don't think it would ever be called a mid-Atlantic state.)

  48. Jeff W said,

    June 12, 2015 @ 10:55 pm

    Thank you for that clarification, Mr Kabala! Until your comment, I could not tell if my interpretation was right or wrong.

    Although I linked to “mid-Atlantic states” and said I thought the term “referred” to that, I had always thought, in a very inchoate way, that the term arose, really, with reference to “patrician-sounding” people in and around (very loosely speaking) the New York metropolitan area (e.g., Hudson Valley, Oyster Bay, southern Connecticut, parts of Manhattan, maybe Philadelphia) so that it mapped somewhat to “mid-Atlantic states” and maybe derived from it but wasn’t an exact “reference.” I thought that “mid-Atlantic” was a way of referring to that area that (1) avoided the problem of referring to it, in some way as a “New York accent” and (2) evoked, perhaps, the “maritime aspect” of it.

    I was thinking exactly of George Plimpton, Grace Kelly, FDR, Katherine Hepburn, but also Gore Vidal, William F. Buckley. (I never thought Connecticut was a mid-Atlantic state; Katherine Hepburn, who exemplified the accent, was from an area, southern Connecticut, to which, I thought, in my wrong interpretation, the term loosely referred.)

    I’m astonished, sort of embarrassed, to find out that it actually does refer to “halfway between the US and Britain.”

  49. Mark F. said,

    June 12, 2015 @ 11:39 pm

    I don't think most Brits realize how completely flummoxed Americans are by things like Pratchett's "marthter" spellings. Once you start to recognize them, they're everywhere – besides er and erm, there's the Korean name Park (there's no r consonant in the Korean) and the game Parcheesi. I have seen "durr" which I think is just British for "duh". "Dartar" for a pseudo-posh pronunciation of "data."

    The flip side of that is African Americans will sometimes leave out R's intentionally (playfully in informal contexts) to indicate their own pronunciation. I've seen this in basketball players' twitter feeds. This wouldn't make sense in the UK where the white people are non-rhotic too.

  50. Lazar said,

    June 13, 2015 @ 3:38 am

    @Mark F.: I strongly agree about the confusion that that can cause. To be blunt, I've seen many lay* language discussions turn into complete trainwrecks as the result of British "ar" respellings. In one case, a British person asked why Americans pronounce "Parmesan" as "parmejarn". The person's intent was obviously to ask about the widespread American use of [ʒ] in that word, but they were met with nothing but derisive incredulity from all the American participants. I've found that neither side in these discussions tends to have the perspicacity needed to understand the other: the British users of "ar" seem never to understand why that spelling is so nonsensical to Americans, and we Americans, for our part, seem never to understand why that spelling does make sense for Brits. It's absolutely a problem of mutual incomprehension, but I think the onus is on British people to use "ah" and "-a" rather than "ar" and "-er" if they want Americans to understand what they're talking about.

    *By "lay", I mean language discussions among "regular" people – those who don't frequent language forums or blogs.

  51. Mr Punch said,

    June 13, 2015 @ 5:58 am

    I used to know a lot of Americans who spoke rather British-sounding English – In Fallows' Gershwin tape it's Rudy Vallee, a college man from New England, who talks that way. There are fewer every year, but this was the common speech of upper class Bostonians; not an individual affectation, but the way they'd spoken from birth. Fallows has the date of the change about right when he chooses 1961 – many Harvard graduates of the 1940s and a fair number from the 1950s had the accent, very few from the 1960s. One factor, of course, is that prep schools and colleges in the northeast became much more national institutions after WWII.

  52. Keith M Ellis said,

    June 13, 2015 @ 12:43 pm

    I thought that Mid-Atlantic wasn't so much native as it was affected and prestige — like the BBC using RP, it was a standard in American media through the first half of the 20th. It was my impression that its name was a coinage that just reflected a vague sense of it, not that it accurately reflects any deliberate intent to fashion a hybrid of AmEng and BrEng.

    Speaking as a former 80s AM radio disc-jockey, there was (and still is, I'm sure) an informal institutional culture the prescribed certain announcing styles. Elsewhere on the web, I encountered someone (arguing against fry, natch) that all radio announcers and such have some professional standards and training in "elecution" and that, obviously, vocal fry was unprofessional. That's another discussion, but while it's the case that there is formal training in radio and television in some universities here and there, and then of course there's some elecution taught in theater training also in universities, the vast majority of people working in these fields have no such formal training at all and the idea that there's some elocution standard promulgated by formal training is just ignorant. What there is, rather, is informal institutional standards.

    With that in mind, my assumption was that Mid-Atlantic as a media standard evolved partly influenced by some of the regional dialects it's related to, but mostly subculturally and institutionally over time, just as the expectations about tonality and cadence had.

  53. James Kabala said,

    June 13, 2015 @ 7:10 pm

    Remarkably, "mid-Atlantic" seems to be not listed in the Online Etymology Dictionary, so I am unable to learn which of the two very different meanings is the older one. Google and Google Books definitely seem to turn up more results for the "New York New Jersey etc." definition,

  54. Rodger C said,

    June 13, 2015 @ 7:39 pm

    Shouldn't R-lessness mean talking like George R-less?

  55. Mark P said,

    June 16, 2015 @ 10:21 am

    I once heard a British radio game show that (in part) involved identifying homophones. The two words I remember were nor and gnaw.

  56. Matt McIrvin said,

    June 16, 2015 @ 11:29 am


    We also often don't believe we have 'intrusive' r's, though most of us do – 'law and order' is loranawduh(r)

    Intrusive r is a strong feature of non-rhotic Massachusetts accents as well; the subject always reminds me of a local law firm that advertised itself on TV as "the Law(r) Offices of Dane M. Shulman" and assured us that "the law(r) is written to protect you!"

RSS feed for comments on this post