Ask Language Log: Long 'i' and 'a' before 'll'

« previous post | next post »

From Dick Margulis:

The first editor of the Encyclopædia Britannica, a Scot named William Smellie, was a distant relative to my wife, whose surname is Smillie, pronounced smiley (the spelling was changed at some point to avoid bad jokes, apparently). I believe William pronounced it smiley as well. John W. Willey (pronounced wily) was the first mayor of Cleveland (the brand new restaurant where my son is a sous chef is named The Willeyville because Willey bought a tract of land on the city's west side and named it that), and there are some towns in England named Willey, although I don't know that they share the same pronunciation. And a shibboleth here in New Haven is the pronunciation of Whalley Avenue, named for the English regicide Edward Whalley and pronounced whale-y, although Our Lady of the Google Navigator, who is Not From Around Here, rhymes it with alley.

I can't blame people who think my wife's name rhymes with Millie or the restaurant is the willie-ville, nor those who have trouble with Whalley, because the way we were all taught to decode a double-ell is that it makes the preceding vowel what we non-linguists call short. But as I find myself at the confluence of these three examples of this unusual (I think) orthographical feature, I'm just curious what the history of it is, if anyone at Language Log Plaza happens to know.

I'll throw this one directly to our readers, before I embarrass myself by speculating that contrastive consonant length might have survived in some dialects through the time of the Great Vowel Shift.  The safe and obvious answer is that the use  of doubled consonants to signal short vowels remained erratic well into the 18th century, and some proper names were frozen in spelling before English spelling was otherwise regularized. But is it a coincidence that these three example all involve 'll'?

I'll just note that the OED's list of variants for ale includes alle, aill, aeill, aell, aiell, aill, all, alle, ayill, ayll, aylle:

Thus

1535 W. Stewart tr. H. Boethius Bk. Cron. Scotl. II. 660   Of wyne and aill takand thame sic ane fill.

And I'll also note this citation from the OED's entry for wily:

1639 J. Clarke Parœmiologia 285   As willy as a foxe.

 



96 Comments

  1. dw said,

    June 20, 2013 @ 9:10 pm

    L-vocalization has surfaced many times during the history of English (words such as all, palm, and folk were once pronounced pretty much as spelled in IPA).

    I wonder whether it is at work here.

  2. Nathan said,

    June 20, 2013 @ 9:48 pm

    Palm and folk certainly have silent ells in my variety of English, but what's up with all? Does someone pronounce that word as just a vowel?

  3. Jerry Friedman said,

    June 20, 2013 @ 9:56 pm

    Robert Burns pronounced all as just a vowel, but I don't think that's what dw is talking about.

    Just out of curiosity, does anyone know what fraction of American native speakers of English now pronounce palm with an /l/? I'd guess it's about half, give or take. (I haven't heard any native speakers pronounce half with an /l/, though.)

    There's another proper name for Dick Margulis's list on the tip of my brain.

  4. Kevin Roust said,

    June 20, 2013 @ 10:23 pm

    More broadly, it seems to me like names are a (the?) key area where (English) pronunciation is nearly arbitrary. Although my last name is pronunced like the identical word, and I am unaware of any other pronunciation for the -oust structure in AmE (oust, joust), that is rarely the pronunciation that is used by people unfamiliar with me. Roast, rust, rost (like lost), roost, and rast (like rasp) all seem more common than roust.

    Definitely true of Americans of many regions, and seems to be true of British and Australians as well (adjusting for the pronunciation of roust). Germans seem to have no problem, though.

    [I can understand oo, folowing Proust, but even that isn't that common of a choice.]

  5. Jerry Friedman said,

    June 20, 2013 @ 10:37 pm

    Namely a common pronunciation of a well-known comet.

  6. maidhc said,

    June 21, 2013 @ 12:32 am

    The name Rollins I've always heard pronounced with a long O like roll or poll. Unlike Collins. Also Tollhouse.

    Isn't Halley's Comet supposed to be pronounced like Hawley? Like Hall or All, I suppose.

    L before a consonant is a different case, I think. Some American dialects have silent L in words like wolf or calculate. And in a famous example Latin falco became French faucon, which was adopted into English until some grammarian insisted on adding the L back in, and then convincing people to pronounce it falcon.

    Some words like walk and talk no one ever pronounces the L but others like palm could go either way. I was taught to pronounce the L in palm, balm, alms, etc.

    On the other side of the original question are short vowels occurring before a single L, as in the name Colin, which seems virtually impossible for Americans to pronounce.

  7. dw said,

    June 21, 2013 @ 3:11 am

    @Nathan Palm and folk certainly have silent ells in my variety of English, but what's up with all? Does someone pronounce that word as just a vowel?

    No one (or no one I know, at least) pronounces "all" with the same vowel as "pal".

  8. Pete said,

    June 21, 2013 @ 4:08 am

    @maidhc: I think some Americans do pronounce the Ls in walk and talk – see the comments (e.g. Roger C) in this LL post from last year: Tawking the tawk, wawking the wawk.

    I presume it's a spelling pronunciation.

  9. AntC said,

    June 21, 2013 @ 5:15 am

    @Nathan what's up with all? Does someone pronounce that word as just a vowel?
    Yes, many dialects around London (either as a vowel, or with a hint of a dark 'l'). What my hyper-correct father stigmatised as London 'drawl' — which us kids parotted back also with a hint of a dark 'l'.

    @Mr Roust: Where does your surname come from? Is it related to Rost (which is German, and I suspect once carried an umlaut). The Rost's are a notable family in New Zealand, with a vowel pronounced variously as schwa or as in roost or roast.

  10. GeorgeW said,

    June 21, 2013 @ 5:41 am

    According to Vivian Cook ("The English Writing System"): "Doubled consonants are chiefly used to show that vowel that precedes them is 'checked' rather than 'free." He uses the terms checked and free for short (lax) and long (tense).

    An example of the -il vs. -ill sequence would be 'pill' vs. 'pile.'

  11. Dick Margulis said,

    June 21, 2013 @ 5:54 am

    @GeorgeW: Yes, that's what I was getting at, the principle (if we dare call it a principle) that these examples run counter to. And these (plus Halley, which I'd forgotten about) all involve the doubled ell. I haven't come up with similar examples involving a different consonant.

  12. mgh said,

    June 21, 2013 @ 5:59 am

    Why is it tall, tell, till, but toll? And why roll and poll, but roly-poly (not rolly-polly)?

  13. GeorgeW said,

    June 21, 2013 @ 6:15 am

    @mgh: While we do have 'poll' and 'pole' we also have 'moll' vs. 'mole' which comply with the rule.

    I suspect there are several reasons for the exceptions as suggested in the opening post. If I recall correctly, sometimes we can tell when a word was borrowed into English by this distinction (pre vs. post Great Vowel Shift).

    I would guess that names are particularly resistant to change.

  14. Dick Margulis said,

    June 21, 2013 @ 6:41 am

    @mgh and GeorgeW: [Manute] Bol, boll [weevil], bowl [of soup or a game]. Pronounced the same everywhere or are there differences in some dialects?

  15. Ø said,

    June 21, 2013 @ 7:22 am

    I seem to recall that the first syllable in the name of the artist Mapplethorpe, once much in the news, had a long "a" at least for some speakers. Or was it just that some people were led astray by the familiar word "maple"?

  16. Ellen K. said,

    June 21, 2013 @ 7:39 am

    No one (or no one I know, at least) pronounces "all" with the same vowel as "pal".

    I was noticing this on another blog, in a discussion about U.S. versus U.K. pronunciation of Spanish words, with regards to the word salsa. Rhyming the first syllable with pal (U.K. pronunciation transferred to my own accent) gives a different L sound than rhyming the first syllable with all (my own U.S. pronunciation). Which I figure perhaps gives justification to one person's claim that the U.K. pronunciation is closer to the Spanish language pronunciation.

  17. Ellen K. said,

    June 21, 2013 @ 7:53 am

    Isn't Halley's Comet supposed to be pronounced like Hawley? Like Hall or All, I suppose.

    That's a new one to me. My understanding is that the correct pronunciation has the first syllable like Hal (rhyming with Al, gal, pal, Sal), but some people incorrectly pronounce it with the vowel of hail (ail/ale, pail/pale, rail, sail, nail). A distinctly front vowel either way, not the back vowel of "all" and "hall". (I can't say "pronounces like" and "rhymes" for the latter group because the L is different, for me at least.)

  18. Lazar said,

    June 21, 2013 @ 7:59 am

    When I first heard of the comet as a child, I thought it was spelled "Haley's comet" and pronounced it that way. When I learned of the correct spelling, I changed my pronunciation to rhyme with "valley". A similar thing happened to me with the actor Jim Carrey – I first heard his name spoken by Mary-merry-marry merged people, so I thought it was "Carey". Later on when I realized how it was spelled, I had to teach myself to say it as "carry" instead.

  19. Morten Jonsson said,

    June 21, 2013 @ 9:15 am

    @Lazar

    Do you try to pronounce every name as it's spelled rather than as other people say it? Good luck with that.

  20. Ellen K. said,

    June 21, 2013 @ 9:30 am

    Well, given that Jim Carrey is Canadian, and thus likely has the Mary-merry-marry merger, simply pronouncing it as Jim Carrey himself does is actually not a good guide for a person without this merger. Or so it seems to me. Barring finding someone with that last name who doesn't have the merger, spelling would be the way to figure out how to pronounce it.

    As for Halley, he's correct. And pronouncing something as it's actually spelled is certainly better than pronouncing something how you incorrectly think it's spelled.

  21. Kevin Roust said,

    June 21, 2013 @ 9:37 am

    @AntC — origins of my family name are slightly obscure, despite being within the last century or so. The story is something to the effect of there being too many Swensons in their area once in the US, so they changed from Swenson to something related to their farm — name (or word for?) a nearby stream (but whether in Minnesota or Sweden I can't say).

    So, while the family is Swedish, the family name wasn't gained until in the US (and then appears to be a place name). Best guess: probably should follow Swedish pronunciation (whatever that would be), but could be based on a German place name (from earlier immigration into Minnesota).

  22. Dennis Paul Himes said,

    June 21, 2013 @ 9:37 am

    Another data point: "Willey" in the White Mountains of New Hampshire (both Mt. Willey and the Willey Range) is pronounced to rhyme with "silly". At least by many people; there is some variation in mountain name pronunciations around there, most famously with "Moosilauke".

  23. Lazar said,

    June 21, 2013 @ 10:23 am

    @Morten Jonsson: What Ellen K. said. It's the same reason why Americans are completely justified in pronouncing the rhotic /r/ in the name of a non-rhotic British celebrity named "Edward" or "Roger" – if one pronunciation in dialect A maps to two or more possible pronunciations in dialect B, in the absence of other evidence, it's reasonable to rely on spelling (even if this doesn't yield the most similar sounding choice).

  24. Morten Jonsson said,

    June 21, 2013 @ 10:36 am

    @Lazar

    The question isn't whether it's correct to pronounce a name the way people in your part of the world do. Of course it is. But there's something a little quixotic about having to teach yourself to do it. And in general, spelling is a very unreliable guide.

    For what it's worth, I don't pronounce my own last name (which isn't Jonsson) the same way my father does; I grew up in a different part of the country than he did. I suppose I could try to reclaim my roots by changing to what he says, but it's not worth it.

  25. Jerry Friedman said,

    June 21, 2013 @ 10:53 am

    On Edmond Halley, a "Science Q&A" from the New York Times, linked from Wikipedia, says, "Contemporary accounts refer to him as Hailey, Haley, Halley, Haly, Hawley, Hawly and Hayley, and presumably pronunciations varied as widely." It adds that modern people with that surname usually pronounce it to rhyme with "alley".

    maidhc: As far as I know, "Colin" normally has the vowel of "collar" in America. (So do "colic" and "solid".) I imagine your comment about "virtually impossible for Americans to pronounce" refers to the fact that in talking about one famous Colin, we generally follow his eccentric pronunciation.

  26. dw said,

    June 21, 2013 @ 11:10 am

    @mgh said,

    Why is it tall, tell, till, but toll? And why roll and poll, but roly-poly (not rolly-polly)?

    "Tall" and "toll" and "roll" and "poll" show the effects of centuries-old L-vocalization. Which is why they don't (in most accents) have the same vowels as "pal" and "doll".

    "Pal" and "doll" themselves are both relatively recent additions to the language (from the 18th and 16th centuries). Hence they missed out on the L-vocalization effect.

  27. Eric P Smith said,

    June 21, 2013 @ 11:30 am

    I don't think I've ever heard 'palm' pronounced with an /l/. (I'm British.) My middle name is 'Palmer', pronounced without an /l/, and it is connected with 'palm' the tree. Unusually (I think) for a Brit, I pronounce 'Palmolive' – the trade name for the soap – with an /l/, by analogy with 'palmitic'.

  28. Rodger C said,

    June 21, 2013 @ 11:56 am

    I've always heard that the pronunciation of "Halley's Comet" was affected for some decades, starting in my parents' day, by the band Bill Haley and the Comets. The original pronunciation with checked vowel was readopted when the comet itself came around again (very disappointingly) in 1986.

  29. Wily Peralta said,

    June 21, 2013 @ 12:01 pm

    How do I pronounce my name?

  30. Mr Punch said,

    June 21, 2013 @ 12:03 pm

    The name of Carlton Willey, a big-league baseball player of the '50s from Cherryfield, Maine, was always pronounced "willy."

  31. Andrew (not the same one) said,

    June 21, 2013 @ 12:38 pm

    I've always heard that the pronunciation of "Halley's Comet" was affected for some decades, starting in my parents' day, by the band Bill Haley and the Comets

    I have heard this as well, but presumably the band had some reason for adopting that name in the first place. Jerry Friedman's source, suggesting that there have been multiple pronunciations from the beginning, looks plausible.

  32. Lazar said,

    June 21, 2013 @ 2:07 pm

    @Eric P. Smith: It's not so common here in the Northeastern US either, but I think it is common in the Midwest – I've heard a lot of people on television pronounce the /l/. For example, I'm pretty sure that David Letterman (host of a late night show) has the /l/, and he's from Indiana.

    @Jerry Friedman: Yeah, Colin Firth, Colin Farrell and Colin Mochrie all seem to be pronounced with the LOT vowel here in the US. Although oddly, I can't think of any Americans with that name besides Colin Powell.

  33. J.W. Brewer said,

    June 21, 2013 @ 2:33 pm

    Note that holly v. holy is a good minimal pair following the "standard" rule, but that the double-l "wholly" is homophonous with the latter rather than the former. Blame it on the suffixation if you like, but then why isn't it "wholely" by analogy to e.g. "solely" (not "solly")?

    As noted above, Whalley Ave. in New Haven is named for Edward Whalley, who arrived in town in the 1660's as a fugitive on the run from the vengeance of Charles II (on account of Whalley's culpability in the beheading of the newly-restored king's father) together with his fellow regicides William Goffe and John Dixwell. Goffe and Dixwell have streets in the same part of town named after them, but the phonology signalled by the double-l in Dixwell is boringly normal, and Goffe is I guess just a proper-nouns-are-weird case because I don't think there are many if any normal English nouns whose standard modern spelling ends with a doubled f followed by a silent e. It would not surprise me if Whalley's surname was not spelled consistently by all of his seventeenth-century contemporaries (indeed, it would not surprise me if it had not been spelled consistently by Whalley himself), but I haven't investigated the case.

  34. dw said,

    June 21, 2013 @ 2:37 pm

    @Jerry Friedman @maidhc — and the "one famous Colin [Powell]" was the son of Jamaican immigrants, so its possible that his unusual pronunciation was due to Jamaican traditions.

  35. Lazar said,

    June 21, 2013 @ 2:47 pm

    @J.W. Brewer: "Wholly" isn't homophonous with "holy" for everyone; it can have an /l.l/ sequence. In my own speech, I contrast:

    holly [ˈhɒːli]
    holy [ˈhoʊɫi]
    wholly [ˈhoʊɫli]
    slowly [ˈslɤʊli]

  36. Ellen K. said,

    June 21, 2013 @ 3:42 pm

    On wholly and holy, how about, homophonous except for possible differences in the consonants (there's also wh vs. h, which are different for some minority of English speakers).

  37. Ellen K. said,

    June 21, 2013 @ 3:44 pm

    Or, maybe not on the why, the w seems to be an addition… (looking it up after posting)

  38. Belial said,

    June 21, 2013 @ 4:10 pm

    @J.W. Brewer: giraffe? gaffe?

  39. LDavidH said,

    June 21, 2013 @ 4:11 pm

    @Kevin Roust: Being Swedish, I can't quite imagine what the Swedish pronunciation would have been. Swedish doesn't have diphthongs, and the letter combination /ou/ would only feature a) in loanwords and b) in compound words where one word ends on -o and the other one begins with u-; they would then be pronounced separately (as in *outtalad* "unpronounced", pronounced with four syllables – sorry folks, I'm not very good at IPA).

    Some similar Swedish words would be *rost* "rust",*röst* "voice", *rusta* "equip", but I don't see how any of them fits the context!

  40. maidhc said,

    June 21, 2013 @ 4:27 pm

    Jerry Friedman: My comment was made on the basis of extensive personal experience.

  41. Xmun said,

    June 21, 2013 @ 4:29 pm

    Perhaps Burns's "Jphn Anderson My Jo" is relevant to the L-vocalization question:

    But blessings on your frosty pow [viz. poll]
    John Anderson, my jo.

  42. Xmun said,

    June 21, 2013 @ 4:30 pm

    Grrrrrrr. Please forgive the mistyping of "John" above.

  43. is said,

    June 21, 2013 @ 5:29 pm

    @Lazar: I think you're probably right that pronouncing the /l/ is common in the Midwest. I'm pretty sure most people around where I was raised (Chicago area) do pronounce the /l/ in palm. I certainly do myself.

  44. Kenny Easwaran said,

    June 21, 2013 @ 6:04 pm

    Not quite the same issue, but the consonant in the word "Aussie" is pronounced in a way that I found surprising the first time I went to Australia, and I think is surprising for most North Americans.

  45. Lazar said,

    June 21, 2013 @ 6:47 pm

    @Kenny Easwaran: Maybe I got more childhood exposure to Australian stuff than other Americans (Steve Irwin perhaps?), but I've known that "Aussie" had a /z/ for as long as I've known the word. And I cringe whenever I hear people say it with an /s/, which, sadly, is most of the time.

  46. Joe Fineman said,

    June 21, 2013 @ 7:23 pm

    From time to time (beginning in the 1950s) I have heard "phallic" pronounced with a "long" a (to rhyme with "Gaelic"). Bizarre.

  47. Rod Johnson said,

    June 21, 2013 @ 7:35 pm

    I'm lost on the Colin thing. What are the alternative pronunciations?

  48. Steve Morrison said,

    June 21, 2013 @ 7:48 pm

    Why the e-with-ogonek in "Encyclopędia"?

  49. Eric P Smith said,

    June 21, 2013 @ 7:57 pm

    @Joe Fineman

    From time to time (beginning in the 1950s) I have heard "phallic" pronounced with a "long" a (to rhyme with "Gaelic").

    That would be Irish Gaelic, /'gelɪk/. Scottish Gaelic speakers get very uppity if you pronounce it as /'gelɪk/: they want /'galɪk/. I've tried arguing that it's the same word, and that it's illogical for me, speaking English, to vary my pronunciation of the word depending on the context, but that seems to touch a raw nerve.

  50. Dick Margulis said,

    June 21, 2013 @ 7:59 pm

    @Steve Morrison: No clue. My original email to Professor Liberman had "Encyclopædia." I'm not sure what glitch transformed it into "Encyclopędia."

    [(myl) I have no clue either -- I just used cut-and-paste to transfer the text.]

  51. Eric P Smith said,

    June 21, 2013 @ 8:00 pm

    @Rod Johnson
    Wikipedia explains: 'Colin' is usually pronounced /ˈkɒlɨn/, although General Colin Powell chooses to use /ˈkoʊlɨn/.

  52. Andrew M. said,

    June 21, 2013 @ 8:05 pm

    @ Lazar: Same here. And I'm American too. I'm not sure where I learned that "Aussie" had a /z/, but I've known it for as long as I can remember. I also cringe when I hear people say it with /s/ (although saying it with /s/ makes more sense given the spelling and given that it's short for "Australian" which of course has /s/, not /z/).

  53. Lazar said,

    June 21, 2013 @ 8:08 pm

    @Eric P Smith: There can also be a bit of a raw nerve if you use the term "Gaelic", rather than "Irish", to refer to the Celtic language spoken in Ireland (even though its endonym, etymologically, is indeed "Gaelic"). My current practice – a bit arbitrary, but not totally without precedent – is to concede the point, using "Irish" for the Irish one and an unqualified "Gaelic" for the Scottish one.

  54. John said,

    June 21, 2013 @ 8:45 pm

    @Dick Margulis: Three different pronunciations for me… bol, boll, bowl.

    I do pronounce the 'l' in palm, too.

    @Lazar: Where I grew up in the US (Western MA) the Irish language was called Erse. Gaelic was what the Scots spoke.

  55. Michael Watts said,

    June 21, 2013 @ 10:20 pm

    Dick Margulis:

    > I haven't come up with similar examples involving a different consonant.

    For me, this is a rule of english writing which is completely unrelated to 'L'. It's related to the phenomenon of "silent e", and is easy to observe with alternative intermediate consonants:

    canning – caning
    ratting – rating
    scrapping – scraping
    betting – completing
    begging – egregious (involving multiple phenomena, including this one)
    shining – shipping
    pining – pinning
    hiding – hitting
    rotten – woken (yes, minimal pairs aren't necessarily easy to come by for every sound)
    humming – exhuming

    As I read English, this rule exists in harmony with some others:
    – A suffix beginning with a vowel, applied to a syllable affected by silent e, overrides (in writing) the silent e: cane -> caning
    – Silent e never applies across a double consonant
    – So, we can indicate that the influence of silent e was *not* present by doubling the consonant, even where it was not originally doubled: can -> canning. We have to do it that way because we can't write the e itself where it should be present (*caneing)

    Then, we can draw a reasonable inference that any doubled consonant indicates that its preceding vowel was not affected by silent e, and must instead have taken the 'short' form.

  56. Ø said,

    June 21, 2013 @ 10:36 pm

    About "Colin": isn't the point that for most USians the vowel that the British use in that name isn't our toolbox? We have to choose between Cawlin and Cahlin, because in our dialects another vowel intermediate between these has been merged with one or the other.

    I would not express this by saying that the name is impossible for Americans to pronounce, though.

    I remember someone once expressing his irritation with Americans for pronouncing the first syllable of "coffee" the same as "cough". He is, I believe, originally an Aussie but coming by way of England.

  57. dw said,

    June 22, 2013 @ 12:47 am

    Ø: About "Colin": isn't the point that for most USians the vowel that the British use in that name isn't our toolbox? We have to choose between Cawlin and Cahlin, because in our dialects another vowel intermediate between these has been merged with one or the other.

    The point the original poster was making concerned the use of the GOAT vowel in Colin Powell's name, which was mistakenly generalized to the whole US. The LOT vowel is generally used for the name in both the US and the UK.

  58. John Walden said,

    June 22, 2013 @ 2:25 am

    Michael Watts' explanation, which I learnt as "The vowel that makes the vowel before it say its name unless it is kept away from it by two consonants" seems to be too shot through with exceptions: "th" seems to get treated as a single letter in "tithe" "swathe" and "bathe" while "x" seems to be its own double letter: it isn't doubled in "boxing" or "boxer".

    Then there's "sliver" "liver" "deliver" and so on and on. Really, it's not that useful except as an often fallible guideline to a best-guess pronunciation (or spelling) of an unknown written (or spoken) word. It's quite a stretch to call it a rule.

  59. GeorgeW said,

    June 22, 2013 @ 6:36 am

    @Kenny Easwaran, Lazar, Andrew M:

    I wonder if the double /ss/ in Aussie signals no intervocalic voicing for Americans. I can think of a few other examples offhand ('Lassie,' 'lessor,' and 'assist') in which the /ss/ is unvoiced.

  60. maidhc said,

    June 22, 2013 @ 7:03 am

    "Colin" is actually my first name, and for the last 40 years or so I have had to endure Americans pronouncing it Coe-lin. In 40 years not a single American has ever pronounced my name correctly. I actually tried to change my legal name to use my middle name, but in America having a name like F. Scott Fitzgerald is now considered close to being a felony. With all the anti-terrorism stuff going on it is not possible to change the format of your name on any official documents. It has got to be FIRSTNAME MIDDLEINITIAL LASTNAME or off you go to Guantanamo. So now I am doomed for the rest of my life that every single US government official I encounter will mispronounce my name, and if I attempt to correct them they will give me this look like "Are you you criticizing me? Because if I feel like it I can have you thrown in jail indefinitely and there's nothing you can do about it."

    Ever since I was two days old I have always gone by my middle name. Like Vice President J. Dan Quayle. I'd like to know how he gets away with it!

  61. J.W. Brewer said,

    June 22, 2013 @ 8:30 am

    Re disagreements on the pronunciation of "Gaelic," I am reminded of the following from the liner notes to Steeleye Span's Below the Salt LP: "So there we was sitting in this dressing room when in comes this guy with glasses and he says, 'Oim afraid youse got de wrong toitles for de toons,' and then he goes and rattles off a whole load of Garlic what we couldn't make out, so we gets the roadies to see him off. Handy things roadies." I assume the joke works better for non-rhotic pronunciations of "Garlic," and I'm not entirely sure what sort of regional BrIsles accent the eye-dialect is supposed to be suggesting. (The track this text relates to is a set of jigs said to have been composed by "Trad." and rightly or wrongly titled "The Bride's Favourite / Tansey's Fancy.")

  62. Akito said,

    June 22, 2013 @ 8:42 am

    Nearly everybody has /z/ in dessert and both /z/ and /s/ in possess.

  63. languagehat said,

    June 22, 2013 @ 8:49 am

    I seem to recall that the first syllable in the name of the artist Mapplethorpe, once much in the news, had a long "a" at least for some speakers.

    It is in fact correctly pronounced with long "a," as you can tell by googling [mapplethorpe pronunciation] or just listening to the audio here. It's a variant of Maplethorpe, which (surprisingly) has nothing to do with maple trees but is derived from the Lincolnshire toponym Mablethorpe (originally "outlying farmstead of a man called Malbert"). I've always been curious as to why some people started writing this variant of the name with a double p.

  64. Ellen K. said,

    June 22, 2013 @ 9:42 am

    Maidhc, I suggest that in a context such as this blog, "virtually impossible for Americans to pronounce" should not be used to mean that Americans don't know how to pronounce it correctly when they see it written.

  65. GeorgeW said,

    June 22, 2013 @ 9:48 am

    @Akito: Good point about voicing the -ss- in dessert and the first -ss- in possess. Isn't -ss word final always (or mostly) pronounced /s/? I think it is intervocalic that Americans tend to voice.

  66. Lazar said,

    June 22, 2013 @ 9:56 am

    @GeorgeW: I don't quite understand what you mean. "ss" almost always represents /s/ in English, except in a few cases (all word-medial, to my knowledge) like "possess", "dessert" and "scissors". Aside from the unique case of "Aussie", I don't think there are any differences among the major national varieties in this regard.

  67. Matt_M said,

    June 22, 2013 @ 10:36 am

    @Lazar:

    "Aussie" is not quite unique. It belongs to a class of clippings in Australian English that also includes "Tassie" /tæzi/ (Tasmania) and "cossie" /kɒzi/ (swimming costume). /mɒzi/ "mosquito", on the other hand, is spelt "mozzie".

    My guess is that the use of "ss" in these words reflects a desire to be faithful to the spelling of the original word (at the expense of breaking the usual convention regarding the link between spelling and pronunciation).

  68. Michael Watts said,

    June 22, 2013 @ 11:29 am

    John Walden:

    The original question relates to how people pronounce written words which are new to them. The fact that a rule has common exceptions isn't really relevant to that question; you'll still go with the rule in the face of the unknown because it's all you have.

    How would you pronounce these?

    Massey
    claper
    frot
    frote
    frottic
    plibbell
    himmer
    vesion
    kessan

    I'll also note that it's not difficult to see why 'th' might appear in the class of "single consonants", or why 'x' might appear in the class of "double consonants", though you do lose the purity of having a rule for writing. But the larger point is that it's not at all necessary for a rule to apply to all, or most, of the words you know, in order for it to have great force on words you don't know. That's the regularization process.

  69. GeorgeW said,

    June 22, 2013 @ 12:57 pm

    @Lazar: 'Lassie,' 'lessor,' and 'assist'

    I was just speculating about -ss- being unvoiced between vowels. But this does not seem to be systematic as a number of exceptions have been posted here.

  70. Lazar said,

    June 22, 2013 @ 1:07 pm

    @GeorgeW: Again, I don't know what you're trying to say about "Lassie", "lessor" and "assist" – those three are all definitely voiceless in accordance with the general rule, as are all verbal forms ending in "-ssing", "-sser" and the like. Much as Michael Watts said, the fact that there are exceptions doesn't mean that there's not a rule – English speakers all over the world would identify an imaginary word like "hesselate" as having /s/ without hesitation. The fact that AusEng has a voiced "ss" in some slang terms is tangential to this question.

  71. GeorgeW said,

    June 22, 2013 @ 2:04 pm

    @Lazar: Given that Australians voice the -ss- in 'Aussie' where Americans doesn't seem to be a "tangential question" to me.

    There is a general rule in American English where obstruents are voiced between vowels. However, this does not occur in Aussie – we retain the voicelessness.

  72. Lazar said,

    June 22, 2013 @ 2:12 pm

    The "Aussie" thing is tangential because it involves slang spellings particular to one country – and slang spellings very often don't accord with the orthographic norms of the standard language. Americans don't pronounce "Aussie" with an /s/ because of an overarching phonological rule, they do it because they're unfamiliar with the norms of Australian slang spelling and therefore misinterpret it.

    There is a general rule in American English where obstruents are voiced between vowels. However, this does not occur in Aussie – we retain the voicelessness.

    I truly have no idea what you're talking about. There is no such rule. AmEng has voicing of /t/ between vowels, but that occurs in AusEng too.

  73. GeorgeW said,

    June 22, 2013 @ 2:57 pm

    @Lazar: "There is no such rule."

    So where is the /z/ in Jesus? Resort? Reason? Positive? Present? Visit?

  74. Ellen K. said,

    June 22, 2013 @ 3:18 pm

    GeorgeW, whatever you are getting at, surely a list of words all spelled with an s representing a /z/ sound is not any kind of evidence for something that applies to all obstruents.

    And your claim about Aussies retaining voicelessness between vowels, where Americans don't, doesn't fit with what has been said about the word Aussie: That Australians pronounce it with a voiced sound, /z/, between the vowels, but Americans, based on spelling, pronounce it with an unvoiced sound, /s/.

  75. GeorgeW said,

    June 22, 2013 @ 3:44 pm

    @Ellen K: I don't have time now to do a more complete study of intervocalic voicing in English. But, how do you explain the examples I have given? Where are the voiced segments coming from? Certainly not the orthography or etymology.

    I have not intended to suggest that Australians retain voicelessness in the word "Aussie." What I have wondered about is why Americans do (retain unvoiced in this word) when our general tendency would be to voice it (see examples that have been given). I speculated that the orthographic double s might be a signal, but that apparently is not the case.

  76. GeorgeW said,

    June 22, 2013 @ 3:48 pm

    P.S. re: intervocalic voicing. It seems reasonable to me that a voiceless obstruent surrounded by vowels (voiced) would be in an environment that could lead to voicing of the obstruent: Jesus (VsV) > Jezus (VzV).

  77. Ellen K. said,

    June 22, 2013 @ 4:03 pm

    I don't know, but none of those are particular to American English. Nor does it suggest any general rule about obstruents.

    attic
    apart
    fashion
    chicken
    narcissism

    Voiceless obstruents between vowels. One even with an /s/. (Okay, Chicken's debatable on whether there's a vowel before the n.)

    As for your list of words, all of them except (possibly) Jesus come from French. That could be a factor. (I add the parenthetical possibly on Jesus because the pronunciation in English may have been influenced by French, even if the dictionaries I have access don't show any intermediaries between Greek and English.)

  78. J.W. Brewer said,

    June 22, 2013 @ 4:04 pm

    Within the set of "Bible names" in English, for a minimal pair one can contrast intervocalic /z/ in Jesus with intervocalic /s/ in Jesse.

  79. Lazar said,

    June 22, 2013 @ 4:10 pm

    But, how do you explain the examples I have given? Where are the voiced segments coming from? Certainly not the orthography or etymology.

    Yes indeed from the orthography and etymology.

    I think you have a misconception about how spelling to sound correspondences came to be in English. The "s" of words like "Jesus", "visit", "positive", "reason", etc. was already voiced in Old French long before they were borrowed into English. In taking on these loanwords we adopted a Romance spelling convention by which single intervocalic "s" tended to represent /z/, and "ss" represented /s/; it isn't as though many instances of English intervocalic /z/ came from /s/ in recent memory.

    What I have wondered about is why Americans do (retain unvoiced in this word) when our general tendency would be to voice it (see examples that have been given). I speculated that the orthographic double s might be a signal, but that apparently is not the case.

    As I've already indicated, the orthographic double "s" is, indeed, the entirety of the issue. The voiceless pronunciation of "ss" is a reliable rule with a legion of examples to support it ("associate", "passing", "sassing", "passage", "hassle", "Cassie", "Lassie", "lessor", "lesser", "massive", "tessellate", "passion"); unsurprising, since this use of "ss" a widespread European orthographic convention that we share with French, Italian, Portuguese and German. Once again, the fact that there are exceptions (the only non-Australian ones being "dessert", "scissors" and "possess" and derivatives) does not mean that there is not a widely applicable rule.

    /s/ and /z/ have been fully differentiated English phonemes for a long time, so you can't speak of some hypothetical word Au/s~z/ie about which Americans had to make a phonological judgment. Americans simply saw an "ss" and pronounced it accordingly. If the word had instead been spelled "Ausie" or "Auzzie", then they would have pronounced it with a /z/.

  80. J.W. Brewer said,

    June 22, 2013 @ 4:29 pm

    Like the overwhelming majority of Christian lexical terms of ultimately Greek origin, "Jesus" would have been mediated through Latin before getting into English (possibly with a separate French stage but probably not). Interestingly enough, however, Anglo-Saxon versions of the Gospels (translated from the Vulgate, not the original Greek) tend not to use the name "Jesus" but render it as "Haelend," which means "healer" and/or "savior." This may be on the authority of Matthew 1:21, and/or intended as a translation of the supposed Hebrew etymology of the original name in lieu of a transliteration. By the time Wycliffe was translating into Middle English, the name was being spelled Ihesus or Jhesus (sometimes without the final -s, depending on case); I don't know if the intervocalic s was pronounced /s/ or /z/. I think of it as /s/ in Latin, but of course the ecclesiastical pronunciation of Latin not only varied over the centuries but varied regionally within Europe, so who knows how it was commonly pronounced in Latin-texted services held in England whether in the 8th century or the 14th.

  81. Jerry Friedman said,

    June 22, 2013 @ 5:06 pm

    Brassiere, dissolve, hussar (for some speakers), hussive (a small sewing-case, usually spelled "housewife"), hussy (for some), Ossington Avenue in Toronto, gossan ("An exposed, oxidized portion of a mineral vein, especially a rust-colored outcrop of iron ore," for some)

    There are still far more words in which intervocalic -ss- is pronounce /s/.

  82. GeorgeW said,

    June 22, 2013 @ 5:32 pm

    @Lazar: " . . . was already voiced in Old French long before they were borrowed into English."

    So what motivated French? Could it be the same phonological process?

    Are you certain that English have borrowed both the orthography and a different pronunciation?

  83. Lazar said,

    June 22, 2013 @ 5:43 pm

    I think it's a development that may have gone back as far as Proto-Romance or Vulgar Latin, but the pattern in most of the Romance languages was that single intervocalic Latin /s/ tended to become /z/, while intervocalic geminate /ss/ remained voiceless and, for many of them French, simplified to /s/. It was definitely already present when English borrowed Latinate words from French.

  84. J.W. Brewer said,

    June 22, 2013 @ 5:46 pm

    The heavy metal singer born John Michael Osborne is now universally known as "Ozzy," tracking the /z/ in the surname as pronounced. One can find scattered early references to him as "Ossie," but that spelling presumably fell out of favor because it didn't reliably cue the pronunciation. Interestingly enough, one can find via google books references to some other people surnamed Osborne who were known by the nickname "Ossie," some of whom appear to be Australian: the AustEng pronunciation of "Aussie" is quite close to the standard AmEng pronunciation of Ozzy, maybe making due allowance for presence/absence of caught/cot merger.

  85. Martha said,

    June 22, 2013 @ 8:09 pm

    I feel like for me, "palm" the tree has a (velarized) l, and the "palm" of my hand has no l at all.

    I know that Australians say "Aussie" with a z, but I wouldn't myself, because it would feel like an affectation, since I'm not Australian.

    maidhc, are you yourself American? (The fact that you mention what Americans can't do makes me think that you are not.) If you aren't, I wonder if what's happening, is like, people see the name and are thinking, "Oh, this person isn't American; surely this name isn't pronounced the regular way." It would be the same as what happens with Kevin Roust's name. Because otherwise it's completely bizarre to me that so many people would be pronouncing it coe-lin.

  86. Eorrfu said,

    June 22, 2013 @ 11:55 pm

    My wife just pointed out that we only pronounce the l in folk in certain contexts. The l is absent when talking about specific individual (my folks, listen up folks) but is pronounced when talking about things like folk art or folk music.

  87. AElfwine said,

    June 23, 2013 @ 2:40 am

    I contacted Dick Margulis, who first raised the question, on another platform. To my ears, "whale-y" does rhyme with "alley". Dick replied that whale rhymes with bail/bale/pail/pale and alley rhymes with Sally/dally. While I recognize that whale/bale, etc. *should* have a long A, it seems that I do not pronounce long A before L. I'm originally from Western New York but have lived away from there for a long time.

  88. Jerry Friedman said,

    June 23, 2013 @ 12:33 pm

    maidhc, if we meet, you'll hear me pronounce "Colin" with the correct vowel.

    I pronounce "folk" and "palm" without an /l/ no matter what they mean. I don't think I've noticed anyone making the distinctions Martha and Eorrfu describe, but I might not have.

    AElfwine's comment reminded me that I've heard people pronounce "sale" in a way that sounds like "cell" to me. (I imagine my pronunciation sounds painfully disyllabic to them.) It also reminded me of a Londoner who said in alt.usage.english that he merges, for instance, "doll" and "dole", and he thinks that's common in and around London. This applies only to the last syllable of the word; he doesn't merge "holly" and "holy".

    Incidentally, I think I have some gemination in "wholly", as Ellen K. implied some people do, so it's not quite the same as "holy".

  89. Jerry Friedman said,

    June 23, 2013 @ 12:36 pm

    The Washington Post says, "Powell's parents, Jamaican immigrants and subjects of the British empire, pronounced their son's name KAH-lin. But Powell's childhood friends in the South Bronx, impressed by the heroic exploits of World War II flyer Colin P. Kelly Jr. (known as KOH-lin), altered the pronunciation and its [sic] been KOH-lin Powell ever since." So that pronunciation does have an American origin—but I've still never heard it in reference to any other American.

  90. J.W. Brewer said,

    June 23, 2013 @ 10:51 pm

    So now we have no idea why Colin Kelly would have had a non-standard pronunciation (or at least why kids in the Bronx would have imputed one to him)?

    The etymology of the given name "Colin" seems to be mixed up with that of the reasonably common surname "Collins," which has that helpful double-l-cueing-short-vowel thing. I have no idea why the orthography settled down differently between the two.

  91. Kira said,

    June 24, 2013 @ 3:35 am

    As another Smellie descendant (changed to Smyllie for similar reasons) I was also told the name was always pronounced "smile-y". I have done a little research into the name, and back in the late 1800s my family's name was spelt "Smeallie" in the census. In my accent (Central Lowlands Scots) I think this comes out as something close to "smile-y".

    I've been told, but have not been able to confirm, that the name originates from the town of Smalley, Derbyshire, UK – and I've no idea if this rhymes with "alley" or "small-ey" (or given the Smellie connection I wondered if it could be "smay-lay"?)

  92. languagehat said,

    June 24, 2013 @ 9:28 am

    I've been told, but have not been able to confirm, that the name originates from the town of Smalley, Derbyshire, UK – and I've no idea if this rhymes with "alley" or "small-ey"

    The usually reliable BBC Pronouncing Dictionary of British Names says it's the latter: [smɔlɪ] (smawli).

  93. David Fried said,

    June 24, 2013 @ 2:42 pm

    Back to Halley's Comet–I recall that at the last appearance of the comet in 1985-86, someone had the bright idea of calling most or all of the "Halleys" in the London phone book to ask how they pronounced the name. The majority said "Hawley."

    The article (I believe it was in Sky and Telescope) agreed that the then usual US pronunciation "Haley" was attributable to "Bill Haley and the Comets."

    Ever since reading the article I have rather self-consciously pronounced "Halley" to rhyme with "pal" but "Haley" still seems more natural.

  94. Joe said,

    June 25, 2013 @ 10:12 am

    @Eric P Smith et al

    I don't know what they say in the Gàidhealtachd, but in the central belt of Scotland I mostly hear /'gelɪk/ (for Irish) and Scots (or Scottish) /'galɪk/. Erse I have heard, but it feels archaic. These days in Scotland, your erse is mostly what you sit on, rather than what you speak,

  95. languagehat said,

    June 26, 2013 @ 8:27 am

    I recall that at the last appearance of the comet in 1985-86, someone had the bright idea of calling most or all of the "Halleys" in the London phone book to ask how they pronounced the name. The majority said "Hawley."

    That's… not a very sensible procedure. God knows what you'd get if you called all the Houstons in the London phone book to ask how they pronounce their name, but it would be irrelevant to the correct pronunciation of Houston Street in New York (which, for those who don't know, starts with the sound of the word "how"). There are only two possible sources of correctness for the pronunciation of Halley's Comet: Halley's own pronunciation, and the pronunciation used by the majority of current speakers (in your area, if there is geographical variation). Since the first is no longer retrievable, the second is what we have left. If pretty much everyone says it with "long a" (which they do), it's pretentious and silly to go around ostentatiously using a different pronunciation for no good reason. (It would still be pretentious even if we could be sure Halley said his name differently, but at least it wouldn't be silly.)

  96. Andrew (not the same one) said,

    June 29, 2013 @ 2:31 pm

    If pretty much everyone says it with “long a” (which they do)

    In Britain nowadays, in my experience, people use a short a, i.e. they pronounce it to rhyme with ‘valley’. (Which implies that the phone book method, if it was indeed used, was unsuccessful.) I was brought up to say ‘Haley’, and went on for a long time thinking of this as ‘right’, and thinking I was being descriptivist in doing so (since this pronunciation can’t be justified by any logical principle, but only on the basis that it’s what people say, dammit). But it looks like I’ve been out-descriptived.

RSS feed for comments on this post