might of, would of, could of, should of

A few years ago, The Telegraph ran an article about Americanisms on the BBC—or rather, an article about complaints about Americanisms on the BBC:
Nick Seaton, Campaign for Real Education, said: “It is not a surprise that a few expressions have crept in but the BBC should be setting an example for people and not indulging any slopping Americanised slang.”
(Tangent: I had to look up slopping, which doesn't seem to be used much as an adjective. Is he using the British slang 'dressing in an informal manner' or the American slang for 'gushing; speaking or writing effusively'? Or is slopping here being used as a euphemistic substitution for another word that ends in -ing?)

But (of course!) half of the 'Americanisms' in their closing list of 'Americanisms that have annoyed BBC listeners' weren't Americanisms. One (face up) was first (to the OED's knowledge) used by Daniel Defoe, the Englishman. Another (a big ask) is an Australianism. But one that really bothered me was this:
  • 'It might of been' instead of 'It might have been'
 Three reasons it bothered me:
  1. Shouldn't it might of been be corrected to it might've been rather than it might have been? That is, of is a misspelling of the similar-sounding 've here. Might've is perfectly good contraction in BrE as well as AmE. Is the complaint that people should say have because they shouldn't be contracting verbs on the BBC, or are they complaining about spelling 've wrong?
  2. We're talking about broadcast television and radio, which are spoken media. You can't see the spelling of what the presenters are saying. So how do they know the presenters said might of and not might've?  Of course, they could have seen it on the (orig. NAmE) closed-captioning/subtitles. But BBC subtitles usually make so little sense that I can't believe anyone would take them as an accurate record of what's been said. (Here's a Daily Mail collection of 'BBC subtitle blunders'.)
  3. I read of instead of 've a lot in my British students' essays. A lot. There's no reason to think they're getting it from American influence, because they'd have to read it and they probably don't get the chance to read a lot of misspel{ed/t} American English. The American books or news they read will have (we hope) been proofread. I suspect that errors like this aren't learn{ed/t}from exposure at all: they are re-invented by people who have misinterpreted what they've heard or who have a phonetic approach to spelling, sounding out the words in their minds as they write.
This particular Telegraph list is one of the things that I mock when I go around giving my How America Saved the English Language talk.  But so far, when I've talked about it, I've just said those three things about it. I have never looked up the numbers for who writes of and who writes 've after a modal verb. I think I've been afraid to, in case it just proved the Telegraph right that it's a very American thing.

I need not have feared! Not only was I right that I see it a lot in the UK, I was also right to feel that I probably see it more in the UK, because —you know what?— the British spell this bit of English worse than Americans.

Here are the numbers from the Corpus of Global Web-Based English. The numbers stand for how many times these variations occur within about 387 million words of text from the open internet.

non-standard of American British
might of 392 672
would of 926 1634
could of 458 821
should of 442 683
standard 've American British
might've 506  277
would've 4921 3121
could've 2379 1502
should've 1685 1140

I've put the higher number in each row in blue bold in my table in order to reflect how it shows up in GloWBE. The blue-bold indicates that those numbers showed up in the darkest blue in the GloWBE search results, like the GB column here:

(The Canadian numbers are distracting—they're not based on as much text as GB and US.)

The darker the blue on GloWBE, the more a phrase is associated with a particular country. So, it's not just that the of versions are found in BrE—it could be said (if we want to be a bit hyperbolic) that they are BrE, as opposed to AmE.

In both countries, the 've version is used more than the misspelling. Nevertheless, the American numbers were darkest blue for these spellings—indicating the correct spellings are more "American" in some way—though note that the British 've versions are just one shade of blue lighter—the difference is not as stark as in the previous table.

The moral of this story  

It looks like the BBC complainers and the Telegraph writer assumed MODAL+of was an Americanism because they disapprove of it. But remember, kids:

Not liking something is not enough to make it an Americanism.

Coulda, shoulda, woulda

When I discovered these facts, I immediately tweeted the would of (etc.) table to the world, and one correspondent asked if the American way of misspelling would've isn't woulda. The answer is: no, not really. Americans might spell it that way if they're trying to mimic a particular accent or very casual speech (I coulda been a contenda!). It's like when people spell God as Gawd—not because they think that's how to spell an almighty name, but because they're trying to represent a certain pronunciation of it. No one accidentally writes theological texts with Gawd in them. But people do write would of in formal text 'accidentally'—because they don't know better, not because they're trying to represent someone's non-standard pronunciation. In the Corpus of Contemporary American English, 75% of the instances of coulda occur in the Fiction sub-corpus; authors use it when they're writing dialog(ue) to make it sound authentic. 

But you do get coulda, shoulda and woulda in an AmE expression, which accounts for about 10% of the coulda data. I think of it as shoulda, coulda, woulda, but there does seem to be some disagreement about the order of the parts:

The phrase can be used to mean something like "I (or you, etc.) could have done it, should have done it, would have done it—but I didn't, so maybe I shouldn't worry about it too much now". (A distant relative of the BrE use of never mind.) Sometimes it's used to accuse someone of not putting in enough effort—all talk, no action. 

The English singer Beverley Knight had a UK top-ten single called Shoulda Woulda Coulda, which  may have had a hand in populari{s/z}ing the phrase in BrE (though it's still primarily used in the US).

Another shoulda that's coming up in the GloWBE data is If you like it then you shoulda put a ring on it. And I can't hear that now without thinking of Stephen Merchant, so on this note, good night!


Postscript, 5 Feb 2016: @49suns pointed out that I haven't weeded out possible noise from things like She could of course play the harmonica. Good point. British people do write could of course (etc) more than Americans do because they use commas less. Americans would be more likely to write could, of course, play the harmonica—and with the commas it wouldn't be caught by the search software. As well as of course, there's of necessity and other things 'noising up' the data.

I'm not going to re-do all the tables because I've posted this now and many have commented on it.  But the good news (for this post) is that the conclusions about of is pretty much the same if we limit the search to modal + of + verb; it's still more frequently British—especially when preceding been, the case that was complained about in The Telegraph. Here's a sample.

An interesting case at the bottom is should of known, which reverses the pattern. This is just because should [have] known—often in should [have] known better—is a much more common phrase in AmE than in BrE. Searching should * known, we get:

Looking more closely at that group, I found that 6 of the 21 American should of knowns were from song lyrics (none of the UK ones were), and one was using it as an example in telling people that they shouldn't write should of

The online interface doesn't like me searching for modal+of+verb, so I've had to search for *ould+of+verb, leaving out might and in the post I also left out must.  But having re-searched those, I can tell you: still dark blue in British, not in American.

The other thing I haven't done, which someone (or someones) else has suggested is what happens after negation. That is a lot more complicated, since there are more variations to consider (since both the n't and the have can be contracted).  I'm really interested in that, so I'm going to write a separate post on it next week. Till then!


  1. 1. I also think of it as "shoulda, coulda, woulda", but given your previous references to regional indicators like Zweigle's hot dogs and salt potatoes, that adds little beyond a reason to wonder if the typical order might vary by region.

    2. I do find "would of" vaguely objectionable, but would never have thought of it as a Britishism (?).

  2. I didn't want to believe it. I kept denying it for years. But reluctantly I've come to regard could of etc as new — non-standard, of course — English constructions. I've just heard too many utterances with the LOT vowel rather than a grunted schwa.

    It means that some speakers have a strange new grammar with new modal auxiliary verbs. These are all PAST TENSE, which is perhaps not surprising since they express PAST TIME modality. One could almost posit PRESENT TENSE forms to express PRESENT TIME: may of, will of, can of, shall of. (No, I'm not serious.)

    I'm not denying the possibility of people saying the Standard MITE-uv while writing might of. But I no longer think that that's what always happens.

    Personally, I almost always write would have etc even when trying to represent casual speech. I'm not fond of spellings like should've. It doesn't quite feel right that there's still a sound there the apostrophe is, even if it's a reduced one. I have the same reluctance to write there're. I write there are even if I want it to rhyme with bearer.

    Maybe some writers share my aversion to written -ve after a consonant and substitute of.

    (Come to think of it, I always feel self-conscious using spellings like rock 'n roll or, even worse rock 'n' roll.)

  3. I don't know if anyone has already mentioned this to you on Twitter, but in other discussions of this point (the links to which I have alas forgotten) it has been pointed out that in British English at least, the "of" in "could of" etc. is sometimes clearly audible, namely when the word is stressed, as it often is in sentence-final position: "well, he didn't score, but he should of." And of course it seems far more likely that a British speaker would say "of" in such instances than an American, because of the extremely common tendency to drop the "h" in British spoken English. This probably explains the greater frequency of the misspelling in BrE: it's because British speakers do regularly hear people saying "could of", "should of" etc..

  4. The spelling I learnt didn't abbreviate to 've after a consonant, so spoken @v was/is represented "have". If others learnt the same, it may account for the lowish frequency of "should've" in BrE texts.

  5. I can hear a difference between "might of" and "might've". Using 'of' instead of 'have' drives me mad! I never thought of it as an American influence. A lot of Brits are using it nowadays, perhaps it's just easier to say. I also hear it from Australians.

  6. "Would of" and the like have long been a pet peeve of mine, but I put them down to ignorance of grammar, not American influence.

    I have a different list of Americanisms creeping into British English that annoy me:-

    "X just got better" instead of "...has just...".

    "Train station" for "railway station" (discussed on another thread).

    Calling clergy "Reverend Smith" when traditional British usage is "Mr Smith".

  7. Our esteemed Lynneguist said, in relation to written versions of "would of": "But people do write would of in formal text 'accidentally'--because they don't know better, not because they're trying to represent someone's non-standard pronunciation."

    To which I say, have you ever read any of Terry Pratchett's Discworld books? It most often occurs in the speech of yer typical working/thieving-class Ankh-Morporkian, who are essentially metaphors for the urban English, innit?

    So TP had it down as an Englishism (rather than an Americanism) and I'd have to agree - I've never really heard it from the US, but loads of times from the poorly educated c*ckwombles in the UK.

    The first time I came across "would of" in print (in a Discworld novel), it was like being tripped up whilst reading - I had to go back and re-read it. It was weird because I couldn't instantly figure out what was wrong.

  8. I would second (or third, etc.) the point made by David and Natch, that you often hear the "'ve" part pronounced /ɒv/, rather than with a schwa, so the two are often distinguishable. I also agree that it seems now to be a standard part of non-standard grammar.

    The complaint about it has been around for decades: it would be interesting to know when such complaints first surfaced.

    One thing that struck me in the figures is that the "of" version is more frequent in BrE than the correct version for "might've" - and it's the only one where this happens. Odd. Could there be a phonetic explanation, because it's after an unvoiced consonant, unlike the other three.

    Finally, wearing a "disgusted of Tunbridge Wells" hat, I'm "shocked" that so many shtoodents make this mistake - presumably Eng Lang or Ling students, to boot!

  9. Lyne's table compares only pairs of written forms might of ~ might've etc. But there's a third possibility: might 'ave. I wonder how common this is. And do people mean it to sound like a TRAP vowel but without aspiration? Do many people say ˈmaɪtæv?

  10. David Crosbie asks: "Do many people say ˈmaɪtæv?" Well, I do, but I could only write it might have.

  11. I agree that the error, or non-standard usage, is more likely reinvented by each user rather than learned by exposure, though the latter presumably occurs too. It's also very old: in my post on would of & co. I cite an example from a 14thC religious text, as well as a wide array of contemporary literary examples. Based on some of the comments, I've come round to the idea that would of etc. may in some cases be dialectal – that the vowel is not always a schwa.

    (Searching corpora for would of produces a few false positives, from strings like "would of course have gone"; I used "would of [v*]" to minimise these.)

  12. The written form is easily checked.

    There are 148 'ave in total, but they're mostly not following modals, and many are main verbs or question-initial, where it's signal(l)ing h-dropping, not contracted have. There are no might 'ave on GloWBE. I can see one might as well 'ave been.

  13. Ach, Stan! I meant to cite your post when I wrote this and then in the late of the night forgot to! Thanks for doing it here.

  14. Yeah, it's definitely possible to hear "might OF" when it's stressed. But, on the other hand, I've also heard (irritating, self-important) people correct "might've" thanks to oversensitively keeping their ears open for "ofs", so...

  15. Yeah, it's definitely possible to hear "might OF" when it's stressed. But, on the other hand, I've also heard (irritating, self-important) people correct "might've" thanks to oversensitively keeping their ears open for "ofs", so...

  16. I can believe that the stressed 'of' happens. The spelling findings suggest that maybe that stressing of it happens more in BrE, and then the pronunciation and spelling would affect/encourage each other. But I can't say much more about that without well-collected phonetic data, which I don't have.

  17. Lynne: Coulda, shoulda, woulda ;-)

  18. For me, it's coulda, woulda, shoulda. (Just Googled this version and discovered on IMDb there's actually a newly made feature-length British comedy/drama with this very title.)

    I can't hear this phrase without thinking of a New Yorker cartoon from some years ago, set (as are so many New Yorker cartoons) in a therapist's office. As usual there's a patient on a couch and the therapist, pen and pad in hand, is doing the talking. The caption is (though I could have the order wrong): "Coulda, woulda, shoulda. Next!"

  19. A couple points:
    1. In my dialect of AmE the pronunciation of have and 've in relevant contexts is identical unless the speaker stresses the "have" on purpose, so one faces a choice when transcribing. In fact, this is true of many contractions (he's vs he is, I'm vs I am, we're vs we are, etc).
    2. I don't think constructions like "would of" are necessarily fueled by ignorance. I personally know perfectly well that these constructions are incorrect. Yet, I've caught myself producing them in typing. There might be some weird phonetic parsing that goes on when people write their thoughts out. I don't know whether that is just a typing thing or includes handwriting too since I haven't done the latter very much lately.

  20. 2. I don't think constructions like "would of" are necessarily fueled by ignorance. I personally know perfectly well that these constructions are incorrect. Yet, I've caught myself producing them in typing. There might be some weird phonetic parsing that goes on when people write their thoughts out. I don't know whether that is just a typing thing or includes handwriting too since I haven't done the latter very much lately.

    I despise admitting it, but I find I make more and more of these boneheaded typing mistakes as I get older. When I go back to review what I've written (as in a comment like this one) I'm all too often appalled by my goofs. It doesn't have to be something as flagrant as could of instead of could have -- sometimes (as happened before I corrected it moments ago) it's typing "something as flagrant as could of instead of should have".


  21. We're talking about broadcast television and radio, which are spoken media. You can't see the spelling of what the presenters are saying. So how do they know the presenters said might of and not might've?

    Yes, this! Moreover, it's not simply that people pronounce "might've" as "might of", but that "of" is very often pronounced "'ve". If you are used to "of" being pronounced "'ve", then when someone says "could've" you might think you are hearing "of", but in fact that isn't what the speaker is saying. Especially if you assume that people who pronounce "of" as "ov" are uneducated idiots and you are disposed to think badly of them. I am sure that people will sometimes hear me saying something that sound like "could of", but what the means isn't that I'm saying "could of" instead of a contraction of "could have". And certainly I say "ov" a lot.

    nineveh_Uk @ LJ

  22. "Could of" infuriates me, but I don't blame America for it!

  23. "Could of" is very definitely a British mistake, not an American one - if I see it on Facebook or in a discussion forum, it is almost a BrE signifier! Americans do use "of" incorrectly, but in other ways: "It would have been better of a plan" and similar convolutions!

    Personally, I know I say "Would of", but ironically. And I wouldn't dream of writing it.

  24. This comment has been removed by the author.

  25. This comment has been removed by the author.

  26. Wow, "shoulda, coulda, woulda" actually sounds wrong to my ears. I automatically go for "coulda, shoulda, woulda." "Coulda, woulda, shoulda" doesn't sound outright wrong in the same way as starting with "shoulda." (AmE - Philly-ish, raised by NY-ers)

  27. Oh and also - I'm not surprised at the high proportion of "might of" since it also appears in grammatical formations ("the might of nations")

  28. I'm in Australia but I'm originally from Northern Ireland.

    I'd just like to add to the chorus of people asserting that they can often hear people very distinctly pronouncing the "of".

    "I. Should. Of. Done. That!"

    Also, round here (and in my native Northern Ireland) the "F" in "of" is usually pronounced hard, as in "off" - at least when it's being pronounced clearly and not run together in joined-up speech. So there's often no mistaking it when you hear it. It's nothing like "should've".

    I really only started to notice it in speech in recent years, to tell you the truth, so I suspect that it's becoming rather more common.

    And the same goes for my family, as we have started saying it ironically to each other - something we never did up until about five to ten years ago years ago when we started hearing people at work and at school saying it.

    My wife and I noticed one friend (Australian) writing it about twenty years ago when we first met him online, and for many years we thought he was the only person who did it!

    It annoys the life out of me, too, so I think I would have noticed it earlier had it been very common. I can't help wondering to myself, "What do they think they're saying?!"

    So I don't really believe that it's (usually) simply a written way of representing the pronunciation of "of" with a neutral vowel and a soft "f".

    Incidentally, I noticed it in a big-production English crime drama recently, but I couldn't tell if they were trying to represent the way an ordinary person would speak, or if the actor actually pronounced it that way in his normal speech! I strongly suspect the former, of course.

    And, I can't remember where, but I recently saw it in written form in a very old document, from at least a hundred years ago. That surprised me as I had thought it was a thoroughly modern mistake, even in written form. I'm not sure if it was a relief or a disappointment to discover that it had been around for that long!

  29. PS: I don't blame the Americans for it either. Americans tend to speak slowly and clearly, with every word clearly audible, which is a good way of avoiding the problem of partly or wholly illiterate people misinterpreting which word is being used.

  30. Incidentally, what about the negative form? "He shouldn't of done that." I'm sure I've heard that construction.

  31. There are 148 'ave in total, but they're mostly not following modals, and many are main verbs or question-initial, where it's signal(l)ing h-dropping, not contracted have.

    Thanks for this Lynne. It seems to conform with what is an instinctive preference for me and a policy Max Wheeler: that have is the preferred spelling for reduced/weak forms of the word after a consonant sound —irrespective of the pronunciation.

    The apostrophe really is a blight on clear thinking. The much vaunted disambiguating force of the possessive apostrophe is only a very moderate enrichment. After all, we get along quite happily without the extra information all the time that we're speaking. Against this minor benefit, the apostrophe creates widespread confusion and anxiety. And sometimes apostrophe placement imposes spurious precision on what is best left ambiguous.

    The apostrophe started as the mark of a suppressed silent letter. And that's the use that still looks right and helpful.

    I really appreciate the spelling differences between
    I'll and ill, he'll and hell, she'll and shell,we'll and well
    I'd and id, she'd and shed, we'd and wed

    The apostrophes provide instant recognition of words that sound very, very different from the other half of the pair. It's the absence of this sort of marking that makes a George Bernard Shaw play-script hard going — at least when you start reading it.

    But all these forms involve a single consonant sound — /l/ or /v/ — following a vowel sound. I'm not so sure it helps the reader if we use use these spellings elsewhere, as in it'll, what'll, that'll or it'd, that'd, that'd. Sure these forms provide information that a 'full spelling' withholds, but just how valuable is this information? I'm sure it doesn't make a text more readable, and I rather suspect to makes it less so.

    (There's also an argument in favour of the -n't spelling as a graphical unit corresponding to a morpheme (word fragment), but that takes us too far from the might of question in hand.)

  32. On reflection, there is a property shared by -n't and -ve. Both are used to represent two sound values: non-syllabic and syllabic. And, of course, both represent grammatical morphemes (word fragments).

    The difference is that it's possible — and for some of us desirable — to use the spelling have for the syllabic version, even if we inwardly hear a schwa (grunt) vowel and no aspiration. There's no equivalent for the negative-forming morpheme; the spelling not implies same the sound as knot.

    I don't think many of us (George Bernard Shaw excepted) fancy spellings like didnt, wasnt. If our medieval ancestors had opted for dident, wasent, I'm pretty sure we'd be happy with them today. But they didn't, so we aren't.

  33. "Could of" is very definitely a British mistake, not an American one

    What Mrs Redboots says three times is true.

    (My only caveat is that for some people it's passed from a British mistake to a British non-standard regularity.)

  34. Apologies for the repeated post! I was having trouble with my browser and didn't realise it had gone through more than once. Will delete two of them!

  35. At the risk of sounding like a "grumpy old man", this is entirely down to the lack of importance given to English grammar in British schools now-a-days.

    As some of your previous commenters have noted using "of" as an auxiliary to "would", "could" and "should" is definitely well on the way to becoming standard English. For example, no matter how many times I corrected my son's speech he is forever using "of" (and that's even when he isn't actively trying to wind me up) he's over thirty now and he tells me that pretty much everyone he knows (and that includes two doctors, a great many chefs [he's one himself], and Jamie Oliver) all use "of" instead of "have", at least in speech if not in the written word.

    I look forward to rotating rapidly in my grave whilst my descendants continue to abuse the language in this way.

  36. Mrs Redboots

    Will delete two of them!

    And spoil my joke?

  37. If you're following this thread, you might want to look back at the post, as I've added a lengthy P.S. about possible noise in the data--and serendipitous findings as I tried to correct for that.

    Enjoying your discussion, as always!

  38. I've given some more thought to written -ve as what is sometimes called an ideograph.

    Chinese characters and Egyptian hieroglyphs, for example, don't represent sounds (well not primarily) but units of meaning. In English, the units of meaning are words and meaningful word fragments known technically as morphemes.

    A few — extremely important — English words and morphemes are grammatical. They don't mean much on their own but are vital in expressing the grammatical meaning of the surrounding words or phrases. Although English spelling is designed to represent speech sounds (like all languages that use an alphabet), at the margins it resembles Chinese and Egyptian in representing some of the grammar words and morphemes.

    An initial analysis
    -'ve represents the grammar word have
    -'ll represents the grammar word will
    -'d represents the grammar word had
    -'s represents the grammar word is
    -'re represents the grammar word are

    This is not quite right
    -'s also represents the grammar word has

    A further refinement
    -'v represents two grammar word with the form have
    1. the form used with a plural subject as opposed to has — technically a finite form
    2. the form used after another auxiliary verb such as will, would, shall, should, can. could, may, might, must, need and also have itself. — technically a non-finite form.

    I suggest that nobody makes mistakes in using -ve spelling to represent the finite have, the one which goes with plural subjects. Number [1] above.

    Mistakes and errors (I could bore you with he difference, but I won't) occur with representation of the non-finite have, the one which forms expressions like might have, must have etc. Number [2] above. A lot of writers don't use -ve because they've unconsciously internalised a rule that it's the sign of the other have (number [1]).

    Oddly, I think we all avoid -ve for non-finite have when it follows a verb combined with to. For example
    dare to have done it, hope to have done it, be said to have done it.

    Although I can say TOOV (ˈdɛə tuv ˌdʌnɪt, ˈhəʊp tuv ˌdʌnɪt etc) I would never write to've and I'm reasonably sure I've never seen it.

    In short, for many British English writers in a hurry, -ve is not an obvious way of representing non-finite have, so they reach for the equivalent graphical unit that we use for different but identical-sounding grammar word, namely of.

  39. I can remember having this use of 'of' corrected at school when I would of/have been about 11, which is over 55 years ago. I was expected to know (and did) why that was wrong. That was in England. I can't remember whether I was criticised for using an Americanism (as mis-spelling a word like colour or theatre would have been) or just being some version of uncouth, but I think the latter.

  40. It seems to me that it's common when a word-final 'd' or 't' is followed by a word-initial 'h' that the 'h' tends to be dropped when the 'h' syllable is unstressed. In rapid speech, when I say phrases like "I paid him off", "I want his book" or "it weighed her down", the 'h' is barely pronounced unless I'm stressing the 'h' word. It's a coincidence that when the 'h' is dropped from the grammatical word "have", the result is pronounced the same as another grammatical word, "of".


  41. Markn

    Word-initial /h/ is lost in rapid speech in all sorts of contexts, not just following word-final /t/ an /d/. But it's surprisingly rare for a writer to show this in the spelling.

    I asked Lynne about the spelling 'ave and she reported that in the data it's largely used to show that the speaker drops his/her aitches — because of accent/dialect, not speed. The same is surely true of 'im and 'er.

    The 've spelling is rarer than the sound it represents — at least among BrE writers. You don't often see What've you done? or Where've you been?

    when the 'h' is dropped from the grammatical word "have", the result is pronounced the same as another grammatical word, "of"

    Not really. if only the 'h' is dropped the result rhymes with chav. This is true if have is
    • a 'lexical' verb What 'ave we 'ere?
    • a grammar word which is a 'auxiliary' verb I 'aven't 'eard that song yet
    • a grammar word which adds PAST meaning to a modal verb I ain't done it. But I could 'ave

    The spelling 've is used
    • when both the /h/ and the vowel are lost We've arrived
    • when the /h/ is lost and the vowel is reduced to a 'schwa' (little grunt) She might've left

  42. As an American( by birth at least) "Might of" is just ignorance of the language.

  43. About "slopping", perhaps he meant "sloppy" (it's common for British peevers to call American English sloppy) and produced "slopping" due to anticipating the -ng in "slang".

  44. @David Crosbie,
    If I understand your two meanings of "have" correctly, I don't think AmE allows the first meaning of "have" to be contracted. For example, I could never write "I've two pieces of bread". If I start a sentence with "I've" the only way to show possession is to follow it with "got", as in "I've got two pieces of bread". Come to think of it, the possessive "have got" sounds very strange when "have" is spelled out.

    So while it's literally true that nobody would make a mistake of confusing the first "'ve" with "of" in the US, it's only because it's not used. On the other hand, despite this distinction, it is pretty easy to orally produce something that *sounds* like "I of two pieces of bread", so your assertion till merits study. I am not so sure that it's true, but don't know how to check.

  45. Boris

    How about these?
    I've no idea
    I've a feeling that...
    I've half a mind to ...
    I've plenty of time

    Jeff Taft's Blues Concordance includes
    I've an old five pound ax: and I'll cut two different ways
    And I cut my little woman : both night and day

    I've an old five pound ax : and I just dropped in your town
    I got women now behind me : just try that old ax on down


    Well well so I've blues in my meal barrel : and the blues on my shelf
    And the blues in my bed : because I'm sleeping by myself

  46. I've no idea
    I've a feeling that...
    I've half a mind to ...
    I've plenty of time

    These are all perfectly good sentences in spoken American English. That said, I'm tempted to suggest -- somewhat in agreement with Boris Zakharin -- that I've is such a brief utterance that for clarity it's likely to be avoided in American English, either by adding got or by refusing to use the contraction at all ...

    I have no idea
    I have a feeling that ...
    I have half a mind to ...
    I have plenty of time

    With respect to the blues lyrics David quotes, it's easy to imagine I've was chosen to line scan well against the melody. Paul McCartney probably had similar reasons for placing got into his lyrics for the song "I've Got a Feeling".

  47. I disagree about the I've sentences being good sentences in AmE. While you can transcribe spoken AmE like that (sometimes), it just isn't done. And I don't think it's just I've. We've you've and they've would be just as unnatural when they represent the possessive have and not helper have.

  48. This comment has been removed by the author.

  49. Dick Hartnell

    With respect to the blues lyrics David quotes, it's easy to imagine I've was chosen to line scan well against the melody.

    That's not how blues lyrics worked back then.

    in the first song, Chaley Patton could easily have sung

    Got an old five pound ax

    And Casey Bill Weldon could just as easily have sung

    Well well well blues in my meal barrel

    Both would have carried the same sense and fitted the melody equally well.

    Of course I wouldn't cite them as typical AmE speakers. They were Black, born in the rural Deep South in 1891 and 1909 respectively. Still it's an indication that I've has been around somewhere in AmE if only as a minority usage — assuming the accuracy of the transcriptions.

  50. This comment has been removed by the author.

  51. This comment has been removed by the author.

  52. It is interesting to note that stressed "could of" tends to be easily distinguishable from "could've" in British speech, because stressed "of" has the rounded LOT vowel, while "'ve" has the unrounded schwa.

    In American speech, stressed "of" generally has the STRUT vowel, which can be quite central. To my British-native ears, at least, it's far less distinct from an unstressed schwa, which can make it quite difficult to distinguish between "could of" and "could've" in spoken language.

    I theorize that this makes Brits of the prescriptive variety especially hostile to "could of", because it is both visually and auditorily salient to them.

  53. I see that my point has already been well made by user Natch above -- apologies for any repetition.

  54. "David Crosbie said...
    Lyne's table compares only pairs of written forms might of ~ might've etc. But there's a third possibility: might 'ave. I wonder how common this is. And do people mean it to sound like a TRAP vowel but without aspiration? Do many people say ˈmaɪtæv?"

    The this is, "'ave" doesn't sound anything like "of"! They have completely different vowel sounds. It seems more like it comes from "'ve" to me. The people I've heard say "might of", seems to be the ones who pronounce "bus" the same as "boss".

  55. This comment has been removed by the author.

  56. Anonymous

    The spellings have, 'ave and 've represent three pronunciations of the same word have.

    Of has two distinct pronunciations, but there is no way of showing this in the spelling.

    I think you may be on to something with STRUT/LOT vowels as a complicating factor. Yes some people say UVV for of. But do any number of speakers have three pronunciations? These would be OVV, UVV and uhv.

    With the spellings of have and of there are a number of questions:

    1. Which word does the writer intend?
    2. Which pronunciation does the writer intend?
    3. Why has he/she chosen this spelling for this pronunciation?

    The complication lies in the third question. Among the possible answers:

    a. because the writer always uses the spelling have
    b. because the writer always uses the spelling have after a consonant
    c. because the writer never uses the spelling 'ave except to hint at accent/dialect ('dropping aitches')
    d. because the writer is tired/careless/in a hurry and confuses the word intended by the pronunciation əv (uhv) — possibly influence by the attitude behind answer [b]
    e. because the writer has changed his/her grammar so that of (in either pronunciation) after might etc is — idiosyncratically — correct

  57. @Anonymous

    The people I've heard say "might of", seems to be the ones who pronounce "bus" the same as "boss".

    Do you mind clarifying what kind of accent you're referring to here? Caribbean or African, possibly?

  58. I'm from the North of England, glorious Yorkshire to be precise. I suspect accent plays a larger part than you suspect. In my telephone voice, wouldn't has 2 syllables so the stress is not on the 've bit in wouldn't've.

    However, in general speaking, up here wouldn't is monosyllabic. Wunt. Shouldn't sounds like shunt and I'll leave you to deduce couldn't for yourselves. The stress then rests equally on the wouldn't and 've components and it sounds identical to me saying wunt ov. And I pronounce of as ov.

    It's not a big leap to assume that the opposite of 'wunt of' is 'would of' with the stress the same. But that doesn't forgive writing it that way!

    Incidentally, my husband corrected an English Lit graduate friend recently who said 'could of'. His response? "I'm talking! You can't criticise my spelling from my speech!"

  59. Carolyn, if your husband's friend is being honest with himself, that means there are four pronunciations of have out there: HAVV, AVV, uhh and OVV.

    Weird, but not totally unbelievable.

  60. David Crosbie, I am as aghast as you and Carolyn at the English Lit graduate's response! And I agree that he (the graduate) is being disingenuous if he actually *does* know that the word is "have" and not "of" in the construction "could have". Although, very sadly, it would not surprise me in the slightest if an English Lit grad in this day and age didn't actually know that!

    But just to be pedantic, which I suppose is acceptable around here: I can immediately think of more than four pronunciations of 'have', and that doesn't seem at all weird to me, given the glorious variations of English usage we are blessed with worldwide.

    I won't even try to enumerate the variations that spring to mind, but the variety depends on at least these two broad factors.

    (a) The speaker's accent - even just among the multitude of British Isles accent there are many variations, and then there's Canadian and US, which is what I hear as "HAYV" (or even "HAYAV" if it's a Southern televangelist), not to mention Australian, New Zealand, South African, Caribbean etc.

    (b) The context. The word may be stressed or unstressed or slipped casually in between two words. "I might've'been" or "I HAVE been at home all day!" Or it may be in what Carolyn describes as a telephone voice or then there's the slouching around at home voice or the speaking to the Queen voice.

    Speaking of telephone voices, when my mother-in-law and other people of her generation from Northern Ireland pick up the phone you would think they were speaking to the Queen! She says "hev" instead of have" until she realises she knows the person at the other end and then it's back to her normal broad Portadown accent.

  61. From all these comments, it sounds to me like there's a pronunciation difference in BrE between "could of" and "could've." In AmE, however, I would pronounce these the same in casual speech. This seemed odd, since I would think that if "could of" and "could've" are pronounced the same, you'd be more likely to see them accidentally switched in writing, and that doesn't appear to be the case.

    So here's a theory: Americans say "could've," and we know we're saying "could've," so as a result we generally write "could've." To British ears, though, it sounds like we're saying "could of." This pronunciation gets occasionally adopted (an Americanism!), but now Brits are writing it "could of" because that's what it sounds like to them.

    And on the "I've..." sentences, I've got to side with Dick Hartzell. It's not that the constructions are wrong, but I'd shy away from them in favor of "I've got" or "I have."

  62. @vp


    The people I've heard say "might of", seems to be the ones who pronounce "bus" the same as "boss".

    Do you mind clarifying what kind of accent you're referring to here? Caribbean or African, possibly?"

    I was thinking or people from the London area and also people from the midlands trying to hide their accent. To me "should've" sounds like "should uv", not "should of"

  63. DC

    You misread me. I wasn't at all 'aghast' at what Carolyn's husband's friend claimed, rather I was intrigued.

    Either he was being disingenuous, which is intriguing in a small way

    OR he really does have a variant OVV (/ɒv/ with a LOT vowel) in his repertoire, which is intriguing in a really big way.

    To be sure, there are a myriad of pronunciations of a word like have as observed by phoneticians calibrating actual sounds. But at the more abstract level of variation within the speech of an individual it's a whole lot simpler.

    At this abstract 'phonological' level, there are only four variants in my speech — and, I believe, in that of most English speakers. There are:

    /hæv/ — with aspiration (H-sound) and a TRAP vowel
    /æv/ — without aspiration ('dropped H) and with a TRAP vowel
    /əv/ — without aspiration and with a commA vowel
    /v/ — without aspiration and without any vowel (as in I've, you've, we've, they've)

    Most dialect variations can be explained as different concrete values to the abstract TRAP vowel.

    • If a New Zealander says HEVV ([hɛv])ɛp for have, he or she also days TREP ([trɛp]) for trap.
    • Ignoring some tricky exceptions, the Northern Ireland TRAP vowel 'deviates' so-to-speak in the opposite direction; it's the same as the PALM vowel. So have and halve sound the same.
    • Your mother and her friends do what's called 'accommodation'. To make a social encounter more comfortable, we unconsciously (well, usually unconsciously) shift our accent towards the other speaker. Some of us accommodate very slightly, some accommodate a lot. Some do it well, some badly.

    In the past, even more than today, people would accommodate toward an accent with more prestige. In your mother's case this might have been English RP or something more Wogan-like. She and her friends accommodated to some hypothetical posh speaker who might be on the line.

    Almost certainly, the accommodation wasn't restricted to specific words like have. Rather, they adjusted their TRAP vowel to sound — successfully or unsuccessfully — like the imagined posher interlocutor.

    In all of these three example, the TRAP vowel is nothing like the LOT vowel of stressed of (/ɒv/) or the commA vowel of unstressed of (/əv/).

  64. Grace

    From all these comments, it sounds to me like there's a pronunciation difference in BrE between "could of" and "could've."

    Not quite, Grace. Most of us are conscious of a single pronunciation, identical for contacted have and unstressed of. For want of a sound recording, I represent this as /ə/ or uhV, or else say that it's pronounced with a commA vowel. This vowel is often called schwa and I sometimes describe it as a little grunt. Alternatively, some people write is as @v.

    There are two problems:

    1. Some people have one pronunciation, but mix up what it represents. For them the spelling might of is just a careless mistake.

    2. Some people have different pronunciations in What have (uhV) you done? and I might have (OVV) done it.

    BrE speakers with one pronunciation (mosts of us) get confused when we hear speakers with two.

  65. I've just heard on BBC Radio Four another variant of reduced have

    I didn't make an immediate note but I think it was MIGHT huhv (ˈmaɪtˌhəv), In any case it was definitely huhv — h've one might write.

    The speaker was a professional broadcaster, so it could have (could've) been a form devised for sounding clear on the air. Or it could have (could've) been a compromise between the man's original accent (Northern Ireland, I think) and something more BBC-ish and RP-like.

  66. Lynne, as a subtitler, I have to take issue with your comment about BBC subtitles! They are, in fact, at least 98% accurate (which means two mistakes in every 100 words). That article is a collection of mistakes from over several years and is not representative of the hundreds of hours of live subtitles we produce for the BBC (and Channel 4 and Sky) every week.

    In any case, I don't think that would be a common subtitling error. We often tend to avoid contractions in live subtitling as they are not always recognised correctly by the voice recognition software we use, so if someone said "should've" I'd be likely to subtitle it as "should have."

    Most of us come from a languages/linguistics background, not to mention we take our jobs seriously, and we try very hard to avoid mistakes such as the ones in the article, but sometimes they are unavoidable!


  67. I admit this is a small sample, but I thought I'd quote it anyway -- it's a comment from Facebook about a baseball team I follow -- to show that phrases like "would of" are alive and kicking in the US of A:

    Also who am I to try to tell a GM who just brought a team to the WS anything,,, I just would of rather seen flores n tejada in a lil platoon at SS with flores getting the majority of the ab's bigger upside then Cabrera [emphasis mine]

    Also note the confusion of "then" with "than" -- another common spelling gaffe that absolutely baffles me. And while I'm at it, the latest spelling howler to hit my radar afflicts Americans -- some of whom are journalists writing for high-profile websites -- who think the past tense of the verb "lead" is also spelled "lead".

    Has anyone else noticed these?

  68. Dick, I'm underwhelmed by your discoveries.

    Consistently accurate, conventional spelling is achieved only in print media, where there are people employed to edit and correct.

    In Applied Linguistics we make a distinction between errors and mistakes. All of your examples are likely to be the result of typos, tiredness, inattention, a wilful spellchecker — in short, mistakes (or 'gaffes' as you call them).

    What is probably — I'm not yet 100% sure — happening in BrE is that would of is developing into a regular, systematic error. An error, that is, for someone trying to speak/write Standard English.

  69. Rosemary

    The other night there was a subtitle sequence on Death in Paradise

    Let's just say that someday once did the same thing for me

    when I was your age.

    And if they hadn't of done it,

    I wouldn't be around today doing the job I love.

    The speaker is supposed to be a policeman on a fictional British Island surrounded by the (mostly) French Antilles. The actor Danny John-Jules (famous as Cat in Red Dwarf) is English and ultimately of Ghanaian descent , but his parents were from Dominica — a real 'British' island with a French past. So the accent he uses could well be a calculated approximation to Dominican — polished somewhat for British and French TV viewers.

    What he says really does sound like hadn't of. And the programme is filmed, so the subtitler had time to consider and choose between hadn't 've, hadn't have and hadn't of.

    Interestingly, the subtitler did go for the 'correct' spelling earlier in the sequence with thing, although the character clearly says ting.

    (Incidentally, the performance was memorable for the fact that the part of the young urchin he was lecturing was being played by daunt's son.)

  70. Oh that spellchecker!

    The part was played by Danny's son.

  71. Dick, I'm underwhelmed by your discoveries.

    Consistently accurate, conventional spelling is achieved only in print media, where there are people employed to edit and correct.

    In Applied Linguistics we make a distinction between errors and mistakes. All of your examples are likely to be the result of typos, tiredness, inattention, a wilful spellchecker — in short, mistakes (or 'gaffes' as you call them).

    Really, David? Have you ever tried googling "then vs than" or "lead vs led"?

    I guess we can chalk up this proliferation of needless explanations to supercilious self-important grammar blowhards who fail to realize that confusing then and than or lead and led are just the mundane result of typos, tiredness, inattention, or a wilful spellchecker.

  72. I suppose we're heading off topic here, but I felt I might respond to your comment, Dick, to the effect that:

    "Paul McCartney probably had similar reasons for placing got into his lyrics for the song "I've Got a Feeling"."

    The thing is... I would almost always say (in speech) "I've got a feeling" or, "I've got a headache", or "I've got an idea". Not "I have an idea" or even "I've an idea".

    It seems so perfectly normal to me that I wonder why you feel that McCartney added the word "got" for musical reasons!

    Maybe it's because I'm from Northern Ireland and he's from Liverpool. Does it sound wrong to Southern English people?

    (Although, as a songwriter myself, I wholeheartedly agree that "I have a feeling" would have sounded terrible - so terrible that McCartney would never have considered it for even a moment! I also couldn't really imagine Ray Charles singing "I have a woman, way 'cross town, she's good to me!"

    On a similar note, I've often noticed that many English (and Australian) people say "I've not heard of that" or "I've not been to the bank today", whereas we N. Irish people would say "I haven't heard of that" or "I haven't been to the bank today".

    Ever since leaving NI I've gone through life wondering what things I'm saying that other people find odd!

  73. If you click on the 'contractions' tag at the bottom of the post, you will find some of these issue already discussed.

  74. So interesting! Completely coincidental lone data point: I just read a 1938 adventure novel by Phoebe Atwood Taylor, one of a series (she appears to have been quite popular) set on Cape Cod, where the opposition between the locals and those from Away is part of the context. Our hero, Asey Mayo, is the consummate local, and Taylor gives him the "might of"/ "could of" construction as an aspect of his characterization.

  75. D C Cardwell:
    Yes, I (English Midlands) would say "I've got a feeling" in informal speech.

    I remember years ago reading a humorous grammar book for children which taught that the word 'got' is a weed which grows in sentences, i.e. you should say "I have a book" rather than "I have got a book".

  76. Fats Waller's song I've Got A Feeling I'm Falling rather knocks on the head the idea of 'got a feeling' as a Britishism. The lyrics are credited to Billy Rose, who was no less American than Fats.

  77. Kate, I can totally get why someone would say it was, like, you know, a weed.

    "Got" is a great rock'n'roll word though - a nice solid, punchy sound that drives the rhythm forward, and to my ears it also does the same for speech!

    And unfortunately (says you) since social networking arrived I've become more and more inclined to write as I speak and not as I write. So to speak.

    But just now, I think for the first time in my life (I'm a bit slow on the uptake) I've properly realised just how ridiculous it is to say it all the time the way we do!

    To say "I've got a feeling" is to say that at some point in the past I acquired a feeling. Whereas in actual everyday common speech the usage suggests a very distinct sense of immediacy, of the present moment.

    I wonder if anyone smarter than me (all of you!) has an idea of whether this slangish use of the word "got" is becoming more acceptable in "correct" written English? Or is it still just plain wrong?

  78. Got was never wrong.

    Even written English has gradations of formality. And only in the most formal of styles is got inappropriate. And even then I'm sure there are acceptable uses.

  79. The OED entry for get is, naturally, endless. Use I. 24.a covers have got as the equivalent of have ('possess') and simultaneously have got to as the equivalent of have to.

    Some impressive quotees

    What a beard hast thou got; thou hast got more haire on thy chinne, then Dobbin my philhorse hase on his tail.
    Fie, th' art a churle, ye haue got a humour there Does not become a man.

    Miss, you have got my Handkerchief; pray, let me have it.

    I have just now got a cough, but it has never yet hindered me from sleeping.

    He has..got C. R. in blue upon his right arm.

    He has..got C. R. in blue upon his right arm.
    The first thing I've got to do..is to grow to my right size again.

    Quite ‘from the heart’—such hearts as the people have got.
    I am very doubtful..whether you have wit enough to understand a word more of what I have got to say this month.

    As a general rule the banker has not got in his possession the money which he owes to his customers.

    .. . and other writers who are not so renowned — at least not to me.

    Clearly, these two related uses have long been a normal feature of written Standard English — albeit in styles the OED classes as 'familiar language'.

    On the other hand, this OED entry has (unsurprisingly) not been fully revised since 1889. I wonder whether the eventual Third Edition will observe was much stylistic restriction.

  80. I searched the blog for have got but there doesn't seem to be a single focus of discussion; it's all over the bog enmeshed in different discussions.

    So this is as good a place as any to report what they say in the Longman Grammat of Spoken and Written English

    has/have as a main verb is much more common in the American Conversation corpus

    AmE 4,900 instances per million words
    BrE 2,600

    has/have got is vastly more common in the BrE CONVERSATION corpus

    AmE 1,000 per million
    BrE 2,600 per million

    has/have gotten is less common than one might expect in AmE and rare to the point of insignificance in BrE

    AmE 1,000 per million
    BrE effectively zero

    Single word got

    AmE 1,800 per million
    BrE 1,800 per million

    The last figures presumably include both sentences like I got it yesterday and sentences like I got rhythm

  81. David Crosbie - you try telling my grandmother, or my English teachers at school, for that matter, that "got was never wrong". We were never allowed to use "got", or "nice" for that matter, and were required to observe the distinction between "can" and "may"!

  82. Annabel, I'll forgive your grandmother. But your teachers were telling you a pack of lies.

  83. The word got may be perfectly serviceable, but what a willingness to use it implies about the speaker is nowhere better illustrated than in George Bernard Shaw's Pygmalion, to wit:

    MRS. HIGGINS [at last, conversationally] Will it rain, do you think?

    LIZA: The shallow depression in the west of these islands is likely to move slowly in an easterly direction. There are no indications of any great change in the barometrical situation.

    FREDDY: Ha! ha! how awfully funny!

    LIZA: What is wrong with that, young man? I bet I got it right.

    Not much doubt that Shaw meant that last line, a little uncouth despite being shorn of Liza's cockney accent, to get a laugh.

    Indeed, the unceasing angst surrounding use of this word can also be found in the comments about it following its definition in the online Merriam-Webster's Dictionary. They include "When did this stupid word hit the dictionary! Why can we not speak correctly?!?!", "You never need to use the word got!", and "GOT should NOT be a word! Which would you prefer? "I got it!" or "I have it!"? I just heard another disc jockey say "Got!" I cringe everytime I hear it!"

    Personally, I cringe every time I see someone who thinks every time is one word. But that's just me.

  84. Well, Dick, I suppose it's of some interest to learn the scale of this mindless prejudice — or is it some fundamentalist superstition? Either way, there's no need to celebrate it.

    As for GBS, the OED quotes him (in his own voice, not that of a character) in one of his letters

    I should muddle at it until I got it right.

    In British English there is virtually no alternative in common use to express the meaning 'calculate the correct arithmetical answer'. And we use it by extension for 'come up with the desired answer or result'.

    When the Irish failed in a referendum to support ratification of the (European Union) Treaty of Lisbon, a second referendum was held with tiny cosmetic changes. Every body, but everybody said

    'The Irish will have to keep on voting until they get it right.'

    There is absolutely no feasible alternative to

    They got it right the second time.

  85. quote: "There is absolutely no feasible alternative to

    They got it right the second time."

    "They were right the second time"
    "They had it right the second time"

  86. Annabel

    "They were right the second time"
    "They had it right the second time"

    Both of these mean something entirely different.

    They got it right the second time was the cynical judgement of those who regard the EU Commission as a bunch of unelected anti-democratic tyrants who will never accept any answer from a referendum other than the one they want.

    Your two sentences express the un-ironic opposite. The people I have in mind would howll either utterance to scorn.

  87. But then, of course, "They got it right the second time" is a different sort of "got" to "I've got a secret" or "I've got a pink umbrella". It was the latter that we were taught not to use. I'm not quite sure what the difference actually is, but I do percieve one!

  88. I've been trying to find a source to help answer this question (which is how I came across your article): can "could of/would of" etc. be considered a colloquialism, or is it just an error? Personally my sentiment is falls with the latter.

  89. That's more a matter of taste, rather than fact.

  90. We'd better be careful how we write about should have from now on.

    Specsavers are reducing their advertising slogan from should've gone to Specsavers to the single word should've.

    So far so reasonable, but Supersavers are in the process of making this spelling copyright!

    See here.

  91. Misinterpreted?


The book!

View by topic



AmE = American English
BrE = British English
OED = Oxford English Dictionary (online)