"Didn't use(d) to be"

« previous post | next post »

Tim Leonard sent along the Nov. 2 User Friendly strip, with a question:

Tim wrote:

For me, "didn't used to be" is obligatory, but awkward. I Googled "didn't use to be" and found this page in which several people assert that only "didn't use to be" is correct, with a supporting analysis (but no statistical data).

So I counted Googlehits and got:

"didn't used to be"/"didn't use to be" (with region = UK): 1220000/96400 = ~13
"didn't used to be"/"didn't use to be" (with region = US): 16300000/2340000 = ~7

That was a relief, since only the first choice is grammatical for me. But the second choice is much too common to be simply an error. Do you have any light to shed on it?

The forum assertions that Tim links to seem to be suggesting that use should be treated like e.g. try in

… tried to be …
… didn't try to be …

The analogy seems plausible, but it's clear that a large majority of English speakers and writers don't see it that way.

The key insight here, I think, is that "used to" has been sort of part-way re-analyzed as an aspectual auxiliary, usually pronounced [ˈjus.tə] and sometimes written "useta".  The "to" part has been incorporated into this new verb, so that "used to be" is no longer "used [to be]" but rather "[used to] be", where "used to" is just the way to render the word [ˈjus.tə] in standard spelling, just as "want to" and "going to" are the ways to write wanna and gonna.

One factor that may be involved in this process is that [ˈjus.tə] can be a reduced pronunciation for both "used to" and "use to". So when someone wants to render "It didn't [ˈjus.tə] be" in standard spelling, they may be tempted to spell the verb as "used to", even if their internal grammar represents the form as "use to".

However, I can testify that my internal grammar doesn't have any analysis involving a verb use in any of these forms. The first piece of evidence is from pronunciation. The final consonant of the genuine verb use is voiced, so that a careful pronunciation of "… (didn't) use to …" should be [ˈjuz.tu], and similarly [ˈjuzd.tu] for "… used to …". But both of these are out of the question for me.

This might just show that I've re-lexicalized this usage as a different verb use with the pronunciation [ˈjus] (like the noun). But this "verb" would be seriously defective:

*He uses to like spinach.
*He may use to like spinach.
*Could he use to like spinach?

And it wouldn't allow any immediately following adverbs (or even parentheticals or filled pauses):

*She used clearly to like spinach.
*She used, I think, to like spinach.
*She used uh to like spinach.

In other words, at least for me, "used to" (in its aspectual sense) is just a word. Specifically, a verb with very restricted distribution.

What about the 10% or so of English speakers who prefer to write "… didn't use to …"?

There are two obvious possibilities. One is that their internal analysis is different, and "used to" really does involve the preterite form of a verb use. The other is that their internal grammar is the same as mine (and Tim's), but their attempts to make sense of the conventional spelling (and historical source) for "used to" lead them to the artificial spelling "didn't use to".

My money would be on option two.


  1. Pomplemoose said,

    November 5, 2010 @ 5:50 am

    I've always said (and written) "He didn't use (pronounced 'yoose') to do that. Never seemed strange to me.

  2. Twitter Trackbacks for Language Log » “Didn’t use(d) to be” [upenn.edu] on Topsy.com said,

    November 5, 2010 @ 5:58 am

    […] Language Log » “Didn’t use(d) to be” languagelog.ldc.upenn.edu/nll/?p=2756 – view page – cached November 5, 2010 @ 5:36 am · Filed by Mark Liberman under Linguistics in the Tweets about this link […]

  3. Aaron Toivo said,

    November 5, 2010 @ 6:06 am

    My money would be on a mixture of both option one and option two… depending on how old the [justə] auxiliary is. If it's only really gelled in the last half-century or so, there may well be people whose internal grammars haven't caught up yet. Or I could just be suffering from the recency illusion.

  4. James said,

    November 5, 2010 @ 6:17 am

    I would prefer [ˈjusənt] in speech to either of the alternatives offered, which both sound clunky to me. I've no idea how I would write it, though. In fact, I'd probably try to avoid the situation entirely with some periphrasis like 'never used to'.

  5. Benvenuto said,

    November 5, 2010 @ 6:25 am

    I wondered whether "used not to be" is effectively dead. Google returns 30400 UK, 298000 US; ie. 1/40 and 1/55 of "didn't used to be".

  6. Scott said,

    November 5, 2010 @ 6:25 am

    I usually (usedually?) go for "He never used to verb." Seems to work alright for me. Except in interrogatives… Didn't we used to say "Used he not to verb?"?

    My head explodes.

  7. GeorgeW said,

    November 5, 2010 @ 6:40 am

    I did a brief check in my personal email corpus. I am Floridian born and raised and those with whom I correspond are similarly over represented (a little).

    I found frequent occurrences of 'used to' with several meanings and 'use to' with only one.

    1. Past habitual (I think the past is what motivates the /-ed/ spelling). My wife, an English 2nd-language speaker, has acquired this without the habitual meaning. A couple of times in the past is good enough for her 'used to' where I need many more.

    2. Something to which one is accustomed: "She is used to eating lunch at 12."

    3. Instrumental (past): "She used X to help her Y." This was the only usage that I found for 'use to': 'She uses X to . . ."

    I have no problem with "didn't used to," but in the cursory search didn't find any.

  8. outeast said,

    November 5, 2010 @ 6:53 am

    I only use 'didn't use[d] to' in speech; if I did use it in writing, it would only be in writing down colloquial speech. (And in that context I think I'd hesitate, then opt for 'didn't use to'; 'didn't used to' feels really horrible to me, though I'd use it if I was really sure that that was what the speaker was actually saying of course.) In writing I exclusively use 'used not [verb]'.

  9. Mykyta said,

    November 5, 2010 @ 6:53 am

    That's strange. Most of the grammar aids I use (predominantly OUP and Cambridge) for my ESL teaching insist on "didn't use to" and "did [subject] use to" for questions. It never occured to me that "didn't used to" is not only valid, but quite a frequent form as well.

  10. C Thornett said,

    November 5, 2010 @ 6:56 am

    The current BrEng EFL/ESL textbooks that I am familiar with (a reasonable number but not all by any means) teach 'didn't use to' with *'didn't used to' as an error, on the analogy of 'tried to/didn't try to'. I should say that these textbooks all make a point of using idiomatic language. In quick speech, it's very difficult to hear a difference, so the concern is with written language, and especially with examinations of writing skills. I notice that the SOED prefers 'didn't use to' with 'didn't used to' marked as informal.

    Would you ask 'Did you use to?' or 'Did you used to?' (Just curious, not prescribing.)

    [(myl) In speech, of course, it's [ˈjus.tə] or [ˈjus.tu]. In writing, both the forms you quote seem wrong, and so I'd avoid writing either one.]

    The shift seems to be developing with the loss of the meaning 'habit, custom, usual activity' from the present tense of use. Not *'she uses to like spinach' but something more like 'she uses spinach a lot' meaning 'eats it a lot'. From the SOED: T. Bridges Use War abroad, at home use Peace. S. Williams He uses no exercise. Present tense 'use' is now almost exclusively 'utilise, make use of'.

    It would be interesting to know whether 'didn't use/used to' have been running as alternatives or regional variants for a long time or whether 'didn't used to' is replacing an 'older didn't use to'.

  11. rone said,

    November 5, 2010 @ 7:03 am

    Spanish has a verb for this form of 'to use', "soler".

    "didn't used to" is ugly so i avoid it. "didn't use to" is less ugly.

  12. Astrid said,

    November 5, 2010 @ 7:08 am

    Another ESL teacher here. I remember when I first taught "used to" and remember coming across this question in my own head. I did hesitate, but then was sure that "didn't use to" was correct. This is indeed the way it is taught. Generally though it is taught as a separate unit, and I even drill the "usedta" pronunciation. Students don't usually have an issue with the form, but the meaning is something that doesn't always come across clearly the first time (a past state or habit that isn't true anymore).

  13. Tim Leonard said,

    November 5, 2010 @ 7:30 am

    So that's interesting. EFL/ESL teaching is deprecating the form that ~90% of English speakers use.

  14. Mark F. said,

    November 5, 2010 @ 7:43 am

    I once came across a New York Times article that used "did not use to be." It was in the text of the article, i.e., the reporter's voice, not reported speech, and it was a hard news article, not a feature or a column. It was striking.

    I after that I assumed that there was an established written convention and that was it. I see that's not so.

  15. Ellen K. said,

    November 5, 2010 @ 7:44 am

    For me, "used to" is pronounced [ˈjuz.tu] and "use to" is pronounced "[ˈjus.tu]. That is, the spelling difference represents a pronunciation difference. At least when I read them. Now I'm curious if that's true for any one else.

  16. Jens said,

    November 5, 2010 @ 7:50 am

    If you don't consider "used to" to be the preterite form of "use", then where does the preterite "didn't" come from?

    I really don't see how you can put the preterite conjugation into "didn't" without changing the original predicate verb of the sentence from its preterite conjugate form to the infinitive. If as you claim "used to" is not a preterite verb, then where is your original sentence's predicate?

  17. C Thornett said,

    November 5, 2010 @ 7:53 am

    Re Tim Leonard: 90% of all English speakers world wide? Written and spoken English? Is there a reliable statistic somewhere?

  18. GeorgeW said,

    November 5, 2010 @ 8:00 am

    @Ellen K: Not me (Floridian), I use (juz) a voiceless (jus.tə) for both.

  19. Bobbie said,

    November 5, 2010 @ 8:10 am

    The one that bothers me is : I shouldn't OF worried about this one… (instead of shouldn't HAVE worried)

  20. Pflaumbaum said,

    November 5, 2010 @ 8:18 am

    This was discussed on here quite recently:


    after which I looked it up in CGEL, and have now forgotten what it said. All these sound fine to me:

    'It didn't use to matter'
    'It didn't used to matter'
    'It used not to matter'
    'It used to not matter'

    – though my phonetic antennae aren't sensitive enough to tell me which of the first two I use.

    'It usedn't (usen't?) to matter' sounds more old-fashioned but still okay.

  21. BobH said,

    November 5, 2010 @ 8:20 am

    FWIW, internally, I use "useta" in much the way you describe here. But "didn't used to" feels wrong to me; "didn't use to" looks much better. I have no reasoning behind this feeling.

  22. Colin John said,

    November 5, 2010 @ 8:31 am

    I find both alternatives slightly awkward and would habitually use 'I used not to'. (BrE, in my 50s)

  23. Mike M said,

    November 5, 2010 @ 8:38 am

    I never thought about this before and had assumed that 'didn't use to be' was prescriptively correct (though I definitely use 'useta' as one word in speech. I think my preference is determined mostly by the semantics of did vs. was….namely that did takes the bare form of the verb while was takes the past participle – hence, since *'wasn't used to be' is clearly not acceptable, 'didn't use to be' seems like the correct alternative.

  24. Brian said,

    November 5, 2010 @ 8:45 am

    I personally can't believe that "didn't used to" is so popular. To me, saying "It didn't used to be" would be equivalent to saying "They didn't liked to walk." I'm floored to learn that I'm strongly in the minority.

  25. John Cowan said,

    November 5, 2010 @ 8:51 am

    I remember the first time I encountered I was used to 'I was accustomed to' in a book, something I could never say either then (as a child) or now, and really felt the difference between the author's English and my own.

  26. Aoife said,

    November 5, 2010 @ 9:34 am

    It usen't be/ usen't to be. Anyone else use that?

    Might be Hiberno-English, but don't have much time to check right now.

  27. iching said,

    November 5, 2010 @ 9:38 am

    Another fascinating example of a construction I use all the time but to which have never given a moment's thought as to the how or why. An Australian English speaker, I am in the didn't use to (apparently minority) camp.
    @C Thornett: I think Tim was taking the 90% figure from Mark's original post, which was apparently based on Googlehits. When I tried repeating the experiment for Australian sites, I got a similar ratio in favour of didn't used to be so I must be out of step with the consensus here.
    I did a search on the Corpus of Contemporary American English (COCA) from 1990-2010. For written sources there were 42 instances of didn't used to be and 5 instances of didn't use to be , while spoken sources favoured didn't used to be by 15 to 0, so 57 to 5 for all sources. When the search was for didn't used to vs didn't use to the ratio narrowed to 113 to 30. An example for the latter had both forms:
    "You didn't use to call me Sir, you used to call me Stan"
    I am sure I have never been formally taught about this, so I must have picked up my didn't use to idiolect from the ether, somehow resisting the majority. How does that work?
    I think I can shed some light on my internal grammar. I definitely feel I am influenced by the analogy with tried to be and didn't try to be but I don't feel that "use", pronounced [jus] is a normal verb (in distinction with "use", pronounced [juz]).
    As Mark says it is a kind of aspectual auxiliary. It doesn't have a tense of its own but makes the following verb into a past tense. But "use to" (or "used to") is not an absolutely indivisible entity for me. I think Mark has hit on the key to this when he says that for him
    She used clearly to like spinach.
    She used, I think, to like spinach.
    She used uh to like spinach
    are not allowed.
    For me it is crystal clear that they are all absolutely fine. Even She used, a long time ago, to like spinach. is OK as far as I am concerned, although I have to admit that She didn't use, a long time ago, to like spinach does sound weird.
    Now that this has been pointed out to me, I think the "used to", indivisible unit aspectual auxiliary idea would make a lot of sense if I could only retrain my brain to think that way. I could then just stick a "didn't" in front of it without a worry.
    Thanks for the interesting post, Mark.

  28. The Ridger said,

    November 5, 2010 @ 10:00 am

    Bobbie, yours is a spelling complaint, which is radically different from the topic of this post. It only looks similar because it too involves a very reduced auxiliary verb (have, pronounced of by virtually everyone).

    I know I pronounce "didn't use to" the same way I would pronounce "didn't used to", and it's very likely that I only spell it "use" because of my age (over 50). If we spelled it as most of us say it – when it's a semi-modal and not a lexical verb – we'd spell it "usta" and there wouldn't be any worry over whether we had 2 "past tense" forms.

  29. Bill said,

    November 5, 2010 @ 10:02 am

    I think I'd say "didn't use to". I would certainly say "did not use to" rather than "did not used to", which for me indicates that there is not grammatical reason for "didn't used to".

  30. ENKI-][ said,

    November 5, 2010 @ 10:05 am

    I find both these constructions vomit-worthy, and had I discovered myself beginning to say them, I would quickly backtrack and say "he once wasn't". But, aesthetics have little to do with grammar.

  31. Ryan Denzer-King said,

    November 5, 2010 @ 10:07 am

    Maybe a little OT, but I've always interpreted "used to" as some sort of modal, perhaps partly because people around me (in Georgia) used it in double modal-like constructions, e.g., "I used to could do it".

    [(myl) This one may be more widespread. I grew up in rural eastern Connecticut, and in my family, things like "might could" were a sort of dialect joke; but "used to could" seems idiomatic to me, though of course informal.]

  32. Spell Me Jeff said,

    November 5, 2010 @ 10:10 am

    I don't think the way personally I use or parse the phrase is germain to this topic. I am grossly over educated, and language is my business.

    I can say that after 25+ years teaching freshman comp, I consistently run into the phrase being used "incorrectly." No numbers available, but seriously, almost no one gets it right, or if they do on occasion, they'll blow it the next time.

    Considering the way the phrase seems to be used by large numbers of people, I am inclined to analyze it with Mark, as a single word that functions as an auxiliary, specifically to mark the habitual past.

    If indeed this is a problem, I suspect part of it arises from what I perceive as a difficulty in analyzing the semantics of the phrase. As an idiom, it makes perfect sense. Immeasurably few native speakers, I'm sure, misunderstand the phrase. But ask the man on the street to explain what the word "use" is doing there, and you're likely to get a blank stare. Difficult to inflect a verb whose function is mysterious. Not so difficult to inflect a modal, since in English most modals inflect the same way for all persons, even if they are further modalized:

    I had to go to the store
    John had to go to the store

    I didn't have to go to the store
    Suzy didn't have to go to the store

    I useta go
    John useta go

    I didn't useta go
    didn't didn't useta go

    The case for analyzing useta as a one-word auxiliary, conventionally spelled (like so many other words) so as to indicate its etymology, is compelling.

  33. Spell Me Jeff said,

    November 5, 2010 @ 10:11 am

    "didn't didn't" ? Nice . . .

  34. I.D. Mercer said,

    November 5, 2010 @ 10:25 am

    I agree with "outeast" above: Neither "didn't used to" or "didn't use to" is something that I would use in formal writing. I tend to think of it as a speech-only construction. In formal writing, I would rephrase. So I tend to think of it as something that doesn't really have a "correct" version in Standard English. (I am Canadian, aged 36.)

    I pronounce "I used to" with a voiceless S. The S is very close to the T, time-wise, so I cannot tell whether I am saying "didn't used to" or "didn't use to", but I guess it sounds closer to the former since "use" as a verb would have a voiced Z sound if one were to pause slightly between words (which one probably doesn't in this situation).

    Now that I think of it, the very fact that the S in "used to" is pronounced like an S, as opposed to the ordinary past tense "used" where it's pronounced as a Z, is evidence that "used to" is a separate modal verb (as far as the spoken language is concerned).

  35. Alexander said,

    November 5, 2010 @ 10:38 am

    Mark, you wrote: "but their attempts to make sense of the conventional spelling (and historical source) for "used to" lead them to the artificial spelling "didn't use to"." Do you mean by this that they are choosing "use" for the spelling, because the NOUN "use" is pronounced [jus]? That, in any case would be one guess of what's going on.

  36. nicole said,

    November 5, 2010 @ 10:43 am

    I do think I have a different analysis of this from you and Tim. I definitely voice the "s" in all contexts, and something like "She used, I think, to like spinach" is completely grammatical for me. I could also say "used never to" and, of course, "used not to."

    I am a bit surprised to see the Google stats lean so heavily the other way; I have always considered it an "error" well on the way to standardization but I didn't realize I was so much in the minority.

  37. Jerry Friedman said,

    November 5, 2010 @ 10:49 am

    @Ellen K.: I feel sure the people who write didn't used to and the people who write didn't use to pronounce them the same way. At least, I don't remember ever hearing didn't [ˈjuz.tu] VERB. All we're talking about is a spelling difference.

    I tried checking Google Books, but mentions in grammar books seem to greatly outnumber uses "in the wild".

  38. Ran Ari-Gur said,

    November 5, 2010 @ 11:05 am

    Saying that "used to" is a word with a restricted distribution doesn't really rule out the possibility that some speakers replace it with "use to" when the context is already preterite, rather like how most speakers replace "never" with "ever" when the context is negative.

    By the way, some speakers already write "use to" even when there's no "didn't"; see e.g. http://books.google.com/books?q=%22I+use+to+think%22. I guess this is halfway between "used to" and "useta".

  39. Nijma said,

    November 5, 2010 @ 11:33 am

    ESL grammarians Betty Azar (American) and Raymond Murphy (British, but I have the North American version) both say "used to" for habitual past, "didn't use to" for the negative, and "did you use to" for the question form.

    But what really bothers me is to mouse over the cartoon and see "click to embiggen".

  40. dfan said,

    November 5, 2010 @ 12:04 pm

    Funny, I was just thinking about this yesterday and resolved to look up the Language Log post on the subject that I assumed must have been made years ago.

    I write "didn't use to" for your reason #1: to me, "use" is a verb, and the phrase is analogous with "didn't try to" or "didn't want to".

    I realize that it makes "use" a pretty weird verb, but there are plenty of other weird verbs that are hard to do general things with, like "may".

  41. Jerry Friedman said,

    November 5, 2010 @ 12:14 pm

    Google News search 2009–2009 (number checked by looking at the last page of hits):

    "didn't use to": 340
    "did not use to": 90

    "didn't used to": 924
    "did not used to": 79

    "never used to": 906

    There are too many false positives for used not to and used to not for me to count the relevant hits now.

    I'm astonished that the incorrect (sorry) spelling outnumbers the correct one in edited text. Of course, a slight modification of spellcheckers would drastically change the numbers in either direction.

  42. mollymooly said,

    November 5, 2010 @ 12:18 pm

    Hiberno-English can treat "used" as fully modal, with no to-particle:
    –He used never smoke, used he?
    –Oh, he used, all right, but his brother usedn't.
    That's a rather contrived example. In practice people vacillate. I find Bertie Ahern saying "the Order of Business used to take 15 minutes" and "The Taoiseach used never take questions on Thursdays" in consecutive utterances in the Dáil; and later "They have been very helpful in opening up inspections and giving us reports and data, which they used not do." — a doubly Irish construction, with its unreduced "not".

    "Usedn't" is often spelt "usen't" to reflect the pronunciation, just as "mustn't" is often misspelt "musn't". (While standard forms have spellings and misspellings, nonstandard forms have only spellings, never misspellings.)

    [(myl) Interesting! Now I wonder what role Irish influence played in the history of these constructions in America.]

  43. Iulus said,

    November 5, 2010 @ 12:28 pm

    Some observations on my own English (PNW):

    /jusd tu/ and /jus tu/ are neutralized to both be /us tu/ or /us tə/.
    On the one hand, this is natural since terminal alveolar stops get neutralizing before a following alveolar stops, regardless of voicing. But on the other hand, this is weird since outside of the "use(d) to" construction I consistently voice the /s/. I want to say this is neutralization followed by voicing harmony, but there are many examples where this isn't the case. I always says things like [kips duing] or [she's doing] or even [bɛs du] for /bɛst do/.
    So I'd have to say I'm treating /jusd tu/ as a special case. And while I think /jusd tu/, this is likely influenced by writing, since I don't understand what /jus///juz/ would need to be past with the past modal /dɪd/, which in all other cases requires a present verb.

  44. Iulus said,

    November 5, 2010 @ 12:30 pm

    Heh, obviously I should have wrote [ʃis duɪŋ] instead of [she's doing]. and /bɛst du/. Sorry Language Loggers, I'll blame the morning on this.

  45. Rasselas said,

    November 5, 2010 @ 12:35 pm

    Read the following sentence out loud:

    Ask Liz which travel agent she used to book her cruise to the Bahamas this year.

    If you pronounced "used to" as a unit with voiceless sibilant, you probably reached the end of the sentence without finding a place for the gap linked to which travel agent. Once you realize that the gap comes after the preterite of the lexical verb use, you will probably voice the final consonants, as Liz did when she answered:

    I used Abel Baker to book my cruise this year, although I used to book cruises with Charlie Delta.

  46. áine ní dhonnchadha said,

    November 5, 2010 @ 12:37 pm

    The verb is [juwz] when it means "employ" but [jus.tə] as habitual for me. I hear it all the time and use it without anyone blinking.

    I didn't [jus.tə] go out at night, but now I prefer it.

    The spelling of [jus.tə] is the issue. I think I use "used to".

  47. Rasselas said,

    November 5, 2010 @ 12:40 pm

    My point in the foregoing comment is that modal "used to" is quite distinct from lexical "used" followed by a "to" phrase.

    [(myl) Indeed. That was also my point in the body of the post.]

  48. Erik Zyman Carrasco said,

    November 5, 2010 @ 12:41 pm

    "One [possibility] is that their internal analysis is different, and used to really does involve the preterite form of a verb use."

    I think this is actually the case for me (a nineteen-year-old Manhattanite who's majoring in linguistics). Unlike MYL, I can pronounce use to as [ˈjuz.tə] and used to as [ˈjuzd.tə] in careful (or even not-so-careful) speech. (Unsurprisingly, I have the written alternation too, so I never write things like Didn't she used to come here a lot?) That (if my analysis is correct) my mental representation does involve a verb use and is hence analogous to he tried to… ~ didn't he try to…? may be at least partly a result of extensive exposure to writing, since the standard/prestige spellings of both forms favor this internal analysis. Of course this means that, as MYL pointed out, this use would be "seriously defective," but I think that's the right analysis for me.

  49. John F said,

    November 5, 2010 @ 12:49 pm

    I would take it to be simply the 'insert do/did rule', which I think still holds true for 'use to', which is kind of still a verb

    present: I use/I do use; I do not use=I don't use
    past: I used/I did use; I did not use=I didn't use

    I'd make the further analogy:
    I _drove_ to work habitually. (to drive is in the past tense)
    I _didn't drive_ to work after the accident. (I don't know the technical term, but the act of driving was in the past, while the verb form is the same as the present tense.)

    Also, "must have" and "must of" are very different. Though one could stretch it to fit, since 'to have' can mean 'to possess' and 'of' indicates the genitive. Even so, "that must have been scary" still beats "that must of been scary" because even though the contraction is "must've", the 'have' goes with the 'been', so 'of' won't work in that case.

  50. Mr Punch said,

    November 5, 2010 @ 1:05 pm

    I'm in the "used to" (but informal only) camp. Never analyzed it before, but I see "used to" as a marker of continuous/habitual action in the past, rather than as an actual verb, and therefore it's uninflected.

  51. Barrie England said,

    November 5, 2010 @ 1:22 pm

    In Twelfth Night, Act II, Scene iv, line 44-46, Orsino says

    ‘The spinsters, and the knitters in the sun,
    And the free maids that weave their thread with bones,
    Do use to chant it.’

    What if Orsino were saying what the spinsters and others did in the past? Wouldn’t he say ‘Did use to chant it’?

  52. groki said,

    November 5, 2010 @ 1:41 pm

    another 10%er here who used to reject "didn't used to be." eventually I may be used to it, but I certainly didn't use to be.

  53. Qov said,

    November 5, 2010 @ 1:46 pm

    I remember running up against this as a kid and after trying and rejecting various formes, concluded that "used to be" was a construction that could only be used in the present tense. I then tried to find other constructions in English that were similarly constrained, but the grownups I asked to help me didn't understand the question. A round of drinks to you all for being grown ups who understand my child self.

  54. Ellen K. said,

    November 5, 2010 @ 2:09 pm

    From the original post: just as "want to" and "going to" are the ways to write wanna and gonna.

    I see "want to" and "going to" as separate constructions from wanna and gonna, used in different registers of English. So, when I write "want to", I don't see it as representing the spoken "wanna". If, when writing, I have in mind the spoken English "wanna" (or "gonna"), I will write it that way, rather than using the standard written English "want to" (or "going to").

  55. John said,

    November 5, 2010 @ 2:13 pm

    I think if I pronounce "used to" in careful speech, not only is the s voiceless, but there's a word boundary gemination effect with the (unvoiced) -d and the following t. Do others produce/notice this distinction between "used to" and "use to"?

  56. John Thayer Jensen said,

    November 5, 2010 @ 2:17 pm


    I wondered whether "used not to be" is effectively dead

    Dunno – it is certainly what I would normally say, but I must confess I have spent a great deal of time reading 19th Century letters, so maybe I have been contaminated. I am inclined to think that when I was young (born in 1942) I would more likely have said "didn't use to be."

  57. dw said,

    November 5, 2010 @ 2:42 pm

    I suppose everyone is aware of the history behind "used to":

    The verb "use" once had a lexical meaning "to be in the habit of". One could say things like "I use to take a walk every day", meaning that I am in the habit of taking a daily stroll. This is exemplified in the Shakespeare quotation above.

    Over time this sense was lost except in the past tense. People no longer saw "used" as the past tense of a verb "to use". This is reflected in the change of pronunciation. /just/ for this sense as opposed to /juzd/ for the lexical verb, as in "he used a screwdriver".

    The traditional negation is " used not", as in "he used not to be great". That is still the form I normally use. However, this makes "used" an unusual verb because most non-auxiliary verbs can only be negated by adding an auxiliary form of " to do". Thus we get "didn't use(d) to". I suppose that "didn't use to" is more logical, but language generally isn't very logical. The fact that "used to" and "use to" (in this sense) are homophonous for most speakers as /justU/ or /just@/ means that there is inconsistency between "didn't used to" and "didn't use to". The fact that until recently, these forms were largely confined to spoken language adds to the inconsistency.

    In writing I would still find "didn't used to" somewhat jarring. This is supported by a Google Books search on books published since 2000:

    "used not to" 14,800
    "didn't used to" 8,010 hits
    "didn't use to" 4,780 hits

  58. Adrian Bailey said,

    November 5, 2010 @ 3:00 pm

    Spell Me Jeff said: "I don't think the way personally I use or parse the phrase is germain to this topic. I am grossly over educated, and language is my business."


    Meanwhile, let me add my inconsiderable weight behind the 90% camp. "Didn't used to" is correct. When I see "didn't use to" I see a vain attempt to conjugate the item "used to" as though it was part of the verb "to use". As has already been pointed out, the fact that the s is pronounced differently is a good clue that we're talking apples and oranges here.

    Compare: 1. the hammer I didn't use to fix that shelf
    2. The hammer didn't used to be on that shelf. (Or even: usedn't to be)

  59. Ben Hemmens said,

    November 5, 2010 @ 3:04 pm

    English speakers who prefer to write "… didn't use to …"?

    I'm one of them, and I've always thought that was the right version, but I've no idea why. I can't remember any conscious effort to choose it rather than "didn't used to"; I don't think the latter was used either in speech or in writing where I grew up, which was Dublin, Ireland. I think the first time I encountered "didn't used to" was a cockney character in a novel by Nevil Shute, which I remember because it seemed strange to me (at the age of 10 or 11) and I filed it next to not pronouncing your H's. Maybe it's an Irish thing.

  60. GeorgeW said,

    November 5, 2010 @ 3:05 pm

    @Ellen K.: "I see "want to" and "going to" as separate constructions from wanna and gonna, used in different registers of English."

    Yes, there are register differences, but there are also syntactic differences:

    1. She is going to go home.
    2. She is gonna go home.
    3. She is going home.
    4. *She is gonna home.

  61. Ben Hemmens said,

    November 5, 2010 @ 3:15 pm

    –He used never smoke, used he?
    –Oh, he used, all right, but his brother usedn't.

    I can make that a bit more specific and say from my childhood that I'd associate that quite strongly with northern Ireland. My da said things like that (he grew up in Dundalk) and so did a friend and his family who were from Belfast. Maybe some other non-Dublin people said it too, but I didn't be knowing too many of them back in them days ;-)

  62. Joe said,

    November 5, 2010 @ 3:42 pm

    I tried to submit this before and I didn't work, so I apologize in advance there's a double post.

    Anyways, I agree with most of what Mark says, but I don't quite get the second possibility he raises. Isn't it possible that those who say "didn't use to" aren't trying to make sense of the conventional spelling and historical origins but unaware of it. For such speakers, the word [ˈjus.tə] is spelled "use to" as Mark points out. It is highly defective, since it can't be used with present time meaning (sorta the reverse of "must" and "HAVE got," which can't be used for past time meaning) and doesn't exhibit subject-verb agreement (again, like "must"). If these speakers would say, "he use to . .." it would be perfectly normal for them to form the negative by saying "he didn't use to."

    One thing I don't quite understand, however. Why call [ˈjus.tə] in "he didn't use(d) to . . ." an auxiliary? Wouldn't the fact that it requires do-support in negatives mean that it would have to be lexical aspectual verb? (although, as I indicated above, it has a lot in common with modal auxiliaries in ways other than the NICE properties).

  63. Jerry Friedman said,

    November 5, 2010 @ 3:46 pm

    @dw: Of the first page of GB hits on used not to that I got in repeating your search, not one was relevant. They were all from grammar books or constructions such as "Caution should be used not to reintroduce pre-Vatican II practices into the post-Vatican II liturgy." Of the first page for didn't use to, only two were relevant, and of the first page didn't used to, six were. I suspect that your numbers for the latter two aren't too misleading but those for "used not to" are way off.

    I see I was wrong in saying that people who write didn't use[d] to wouldn't distinguish the two forms in pronunciation.

  64. GeorgeW said,

    November 5, 2010 @ 3:52 pm

    @Joe: "Why call [ˈjus.tə] in "he didn't use(d) to . . ." an auxiliary?"

    Because it cannot be used as a full verb. It can only be used with another full verb –

    'She ˈjus.tə use the library more,' or
    'She used the library more,' but not
    *'She ˈjus.tə the library more.'

  65. dw said,

    November 5, 2010 @ 4:28 pm

    @Jerry Friedman:

    We must be doing different searches. On mine all but two of the visible results on the first page for "used not to" are relevant.

  66. John Walden said,

    November 5, 2010 @ 4:34 pm

    EFL/ESOL teachers have bickered about that 'd' on more than one ocasion.



    Here's Pepys on the subject, unwittingly: "a fat man, whom by face I know, as one that uses to sit in our church" ; "I did this night give the waterman who uses to carry me 10s" ; "But this I know, that it did not use to bee soe".

    My own view is that the 'd' shouldn't be there. Marking for the past twice has no precedent that I can think of, English doesn't usually do anything twice. But there are far too many people who think the 'd' is there for a descriptivist to argue any different.

  67. nicole said,

    November 5, 2010 @ 4:57 pm

    As has already been pointed out, the fact that the s is pronounced differently is a good clue that we're talking apples and oranges here.

    Since at least one person other than me has noted she does not pronounce the "s" differently, I'm not sure how dispositive this is. Perhaps notably, I grew up in the same region (not Manhattan, but southwestern CT) and also majored in linguistics, and may be contaminated by having read an awful lot my whole life.

  68. John Lawler said,

    November 5, 2010 @ 5:11 pm

    There's really no good solution, barring written relexification that matches the spoken varieties; there are after all, three different 'used to's.
    First, and most relevant here, for many speakers of modern English, used to — pronounced /justə/ (or /justu/ in careful speech, but never with /zd/ instead of /st/) — is a past frequentative auxiliary verb requiring an infinitive complement, and signalling frequent activity in the past and no activity in the present.

    This is in contrast to the idiomatic verb be used to — with the same pronunciation of the used to part — which is a full predicate (not an auxiliary verb) requiring an NP object, allowing gerund complements with optional Equi (but not infinitives of any kind), and meaning only "accustomed to, habituated to".

    Both of these are idioms, as the special meanings and pronunciation oddities show. There is also the simple construction they both descend from: v:Used + pp:to [NP].

    He used to live/*living on the West side.
    He's used to *live/living on the East side now.
    A shovel is used to dig with.

  69. Peter G. Howland said,

    November 5, 2010 @ 5:26 pm

    As an aside @Nijma, when I first came across the "Click to Embiggen" note, I thought it was beer-out-the-nose funny and I still crack up when I see it. I'm not sure, but I think the intent is humorous.

  70. Ran Ari-Gur said,

    November 5, 2010 @ 5:38 pm

    @John Lawler, "or /justu/ in careful speech": That seems pretty rare to me. In my experience it's usually /justu/ before a pause and often /justu/ before a vowel, but it would have to be very careful speech to make me pronounce it /justu/ immediately before a consonant.

    (But I guess this is also true of "to" in other senses, and also of "the" and partly "a". So "used to" is mostly only exceptional, in this regard, in that it's disyllabic.)

  71. John said,

    November 5, 2010 @ 5:55 pm

    John Walden: "English doesn't usually do anything twice."

    I see what you done did there! (And I really, really mean it.)

  72. Joe said,

    November 5, 2010 @ 6:11 pm

    @George W,

    I agreee that [ˈjus.tə] is like a modal auxiliary in that it selects a bare infinitival complement. So, as I said, it is an unusual lexical verb.
    But I'm not sure if we can say it is an auxiliary because it requires a "full verb" to follow it. At the very least, you have to expand the class of complements to include infinitival and gerund-participles. Otherwise, you would have to exclude "ought" and "is" as auxiliary verbs, and I don't think anyone wants to do that. But not every verb that requires an infinitival complement is an auxiliary verb. Take a verb like "tends" as in "he tends to get angry," for example. Is that an auxiliary? It requires an to-infinitival complement: *he tends angry. But it can't be used without do-support in interrogatives (*Tends he to get angry?) negatives (*he tendn't to get angry), etc. So I guess I would say that [ˈjus.tə] (as well as gonna and wanna) are lexical verbs, but they are usual ones, and I was wondering whether Mark agrees with that classification, or whether he is proposing a different means of categorizing the difference between lexical and auxiliary verbs than diagnostics like do-support in interrogatives, negation, etc.

  73. Jerry Friedman said,

    November 5, 2010 @ 6:20 pm

    @dw: Six of the books on the first page you got are books on grammar or usage. I think they're irrelevant to the frequency of used not to because they're talking about it, not using it. (However, one of them does have some comments on these forms, including that the Early English Prose Fiction database has three instances of used not to, three of did not use to, and one of never used to.)

    One is not the "aspectual" construction we're talking about.

    Two are from textbooks (of French and Turkish) for English speakers that use "used to" as a formulaic translation of the imperfect. This strikes me as on the border between use and mention, but if you look at the texts you'll notice many signs that they're a good deal older than 2000 (thou usedst not to write, no illustrations, and the like).

    And one is a concordance to the Shakir translation of the Qur'an, which was written in the early 20th century by somebody who was probably not a native speaker of English.

    So I don't think any of them tell us anything about the frequency of use of didn't used to in contemporary English, which I take it is what a Google count for the last ten years is supposed to tell us. Some of the discussions in grammar books show that like you (and me and other commenters), some people find didn't used to jarring.

  74. Jerry Friedman said,

    November 5, 2010 @ 6:26 pm

    (Sorry, there seems to be controversy over who wrote the Shakir translation and when, and maybe native speakers of English were involved, but I still don't see it as relevant to contemporary usage.)

  75. Ellen K. said,

    November 5, 2010 @ 6:31 pm

    George: You're number 3 in your reply to me isn't a "going to" — there's no "to".

    Yes, "going" is used in other ways too. But that's really beside the point.

  76. Pflaumbaum said,

    November 5, 2010 @ 6:35 pm

    John Walden said: "Marking for the past twice has no precedent that I can think of…"

    The past perfect marks for past time twice, no?

    Or how about the (admittedly somewhat bizarre) conditional construction with had have? As in

    Had I have been there…

    If he'd've lived to see this…i>

    Not sure if that qualifies… I have no idea what's going on with it at all.

  77. Julie said,

    November 5, 2010 @ 6:59 pm

    Well, I do double things now and then, although not in formal writing. But my internal grammar doesn't allow "didn't used to," and I automatically make it "didn't use to." I analyze it as a match for auxiliary formations like "want to, need to, etc."

    "I wanted to go." "I didn't want to go."
    "I used to go." "I didn't use to go."

  78. Craig Russell said,

    November 5, 2010 @ 7:00 pm

    I couldn't imagine writing anything other than "didn't use to" myself, and would never have suspected that "didn't used to" was anything other than an isolated mistake. But I suppose I would have been wrong.

    However, now that I have given the matter some thought, given the overwhelming majority of people who (on websites indexed by Google) write "didn't used to", I wonder if there is a third explanation to add to MYL's two for "didn't use to." (Forgive me if I am repeating something that's been said in the comments; I didn't see this but it seems too obvious to have been missed).

    My third explanation is that (as I suspect anyone else who has graded a good number of student papers will attest) often people write "use to" for "used to" in situations that don't involve "didn't". E.g. you'll often see "when I was a kid I use to go to the movies all the time." The explanation for this is obviously the [ˈjus.tə] pronunciation that any rendering of the word represents. So for people who invariably write "use to" in these "normal" situations, there would be nothing unremarkable about "didn't use to".

    In other words (from the perspective of people like me who, I suppose, write "didn't use to" for MYL's reason 2), part of the 10% of "didn't use to"-ers may be doing it right for the wrong reason.

  79. Andre Gallo said,

    November 5, 2010 @ 7:11 pm

    Maybe a valid analogy here would be one with "have", which when followed by the infinitive has a similar use to that of modal auxiliaries, but which doesn't follow the same structural rules as those. Thus:

    I had to work late on Tuesday, but I didn't have to get up early on Wednesday. (rather than …I didn't had to get up early…)

    Also, the pronunciation of "have to", indicating obligation, is different than that of "have" in other uses.

    There's a question about "used to" that I'd like to ask: when you say "I used to love listening to this song when I was a kid", does that mean that now you don't love listening to it, or are you only talking about the past, and not giving any information about the present?


  80. Craig Russell said,

    November 5, 2010 @ 7:17 pm


    I think the extra "have" in "Had I have been there" is something different. Has LL covered this before? I am under the impression that sentences like "I couldn't have done it" have been reanalyzed to "I couldn't of done it" (a commonly seen (mis)spelling), and that "of" has taken on a new extra meaning as a particle marking a verb as past contrary to fact.

    The new "of" particle became so separate from the word "have" that was originally contracted into "'ve" that it gets re-added to sentences that already have auxiliary "have", e.g. "If this hadn't of happened, I would be a lot better off."

    I would personally consider all of this non-standard, along with your "Had I have been there," in which I take your "have" as an instance of the same "of" particle that I am describing. But I feel that I hear it commonly enough that it's more proper to call it a (relatively) new feature of English than a mistake.

  81. Ellen K. said,

    November 5, 2010 @ 7:33 pm

    Craig Russell, I think there's something to what you say. For me "Had I've been there" (my spelling of what you call the "of" particle) is unremarkable, but, yet, "Had I have been there" I can't even parse… I can't make sense of it. So clearly the "'of' particle" of I've is not simply a reduction of the word "have".

  82. peterv said,

    November 5, 2010 @ 7:35 pm

    James Thurber's advertising slogan for a brewery fallen on hard times:

    "We still brew good, like we used to could."

  83. Ethan said,

    November 5, 2010 @ 7:41 pm

    Take a verb like "tends" as in "he tends to get angry," for example. Is that an auxiliary? It requires an to-infinitival complement: *he tends angry.

    At least in US political jargon, this does not hold.
    "The West Coast tends Democratic, while the Midwest tends Republican".

  84. Pflaumbaum said,

    November 5, 2010 @ 7:48 pm

    I don't understand how you guys can be so sure you're saying 'didn't use to' or 'didn't used to'. I can't really hear the difference when I say them naturally, and when I say them emphatically it kind of defeats the purpose of working out which is natural.

    @ Craig –

    The CGEL does note that the of spelling is commonly seen for the reduced 've in the have had construction. It also confirms that the construction is non-standard, though common.

    The of spelling is also common elsewhere, though, of course. And I think people really are analysing it as of, not just mis-spelling it, because you hear it in emphatic speech:

    Well you damn well should of

    Otto Jesperson remarks, I recall, that of is a rapacious cannibal of a word, having departed from its original meaning, 'away from', and taken over great reams of English syntax. So it wouldn't be that surprising if it carried on usurping the perfect.

  85. John Walden said,

    November 5, 2010 @ 7:59 pm

    Pflaumbaum. The tendency of English not to use belt and braces unless it's for something is right there: the past perfect marks the past twice precisely because it is the past of the past.

  86. Craig Russell said,

    November 5, 2010 @ 8:08 pm


    I agree with you (as I said above) that the 've is being analyzed as "of" and that as such the word has taken on a life of its own, being used in places where logically "have" has not place. (e.g. "If you hadn't of done that…") I guess I would personally be more inclined to classify this as a new word, separate from the preposition "of" but spelled and pronounced the same way, rather than a new meaning that the old word "of" has taken on. I'm not sure how you would test the different though.

    But I don't see (at least from how it sounds in my own idiolect) how "You damn well should of" with emphasis on the final two words is proof that this is different from "You damn well should've," which is what (I think) I would be thinking in my head if I said this.

  87. Benjamin Lukoff said,

    November 5, 2010 @ 8:10 pm

    I'm not sure which it is in my case: different internal analysis or (unconscious) attempt to make sense of the conventional spelling, but as long as I *have* been spelling, I've written "didn't use to be," and never understood why so many people did otherwise. If writing "ice tea" for "iced tea" is so common, and since "ice cream" has been preferred over "ice cream" for decades, you'd think that "didn't used to" would have been replaced by "didn't use to" by now.

    If my internal grammar is different and "used to" is a preterite, the verb is obviously defective. I wonder which one is a more likely explanation.

  88. dw said,

    November 5, 2010 @ 8:27 pm

    @Andre Gallo

    Also, the pronunciation of "have to", indicating obligation, is different than that of "have" in other uses.

    Is it? How?

  89. GeorgeW said,

    November 5, 2010 @ 8:41 pm

    @Joe: "The Oxford Concise Dictionary of Linguists" defines auxiliary as "A verb belonging to a small class which syntactically accompanies other verbs . . . Auxiliaries typically mark modality, tense or aspect . . ."

    I would say that [ˈjus.tə] is an auxiliary because it cannot be stand alone as a main verb and it marks aspect (past habitual). Although not a necessary criterion, the lack of susceptibility to inflection, I think, suggests this as well.

  90. GeorgeW said,

    November 5, 2010 @ 8:52 pm

    @Ellen K.: Good point. I was just trying to demonstrate (but not very well) that 'gonna' cannot stand on its own like the verb 'to go.'

    One can be going to school, but cannot *'be gonna to school,' or *'gonna to school.'

    So, 'going to' and 'gonna' are not always syntactically interchangeable.

  91. Pflaumbaum said,

    November 5, 2010 @ 9:04 pm

    @ Craig

    Yeah, in my idiolect you can hear the /ɒ/ in emphasised of.

    Why would you classify it as a new word, though? If we'd done that every time of had taken on a new grammatical role, we'd have an awful lot of homonyms by now!

    It's true that you might be hard pressed to find a language that formed its perfect tense with a preposition and a participle like that, but a) the perfect is only arguably a tense anyway and b) it's not particularly 'logical' to use have as the auxiliary anyway (though admittedly a lot of languages do it).

  92. Filius Lunae said,

    November 5, 2010 @ 9:13 pm

    The first thing that comes to mind when I read about this construction is something Miss Cleo used to say. She was a pseudo-physic in the US, some people may remember her, claiming to be of Jamaican descent. In any case, she used to say:
    -Is there a gentleman in your life?
    -Did he use(d) to play football when he was in high school?

    Like others who have commented, I too thought for a while that that "did he use(d) to" was something to be relegated to speech only, and informally, at best. In writing, I would've probably changed this by: a)writing it without the "did", "He used to play football?" or b)dropping the "used to", "Did he play football in high school?".

  93. Andre Gallo said,

    November 5, 2010 @ 9:14 pm

    Do you really have to go? ['hæftə'goʊ]
    Have a nice day. ['hævənais]

  94. groki said,

    November 5, 2010 @ 9:56 pm

    @Pflaumbaum: The past perfect marks for past time twice, no?

    actually, there are two different pasts being marked in past perfect, so it's not quite the "didn't used to" issue.

    the perfect is a verb that took place prior to some reference point, and the past is the reference point being prior to now (when the utterance happens).

    I had eaten = at some time in the past, there was an even earlier time that eating happened.

  95. groki said,

    November 5, 2010 @ 10:52 pm

    @John Walden:

    and by the time I commented, you had made the point almost 2 hours earlier!

    (past perfect win, but refresh fail.)

  96. wren ng thornton said,

    November 5, 2010 @ 11:49 pm

    Just to add another datum. In my idiolect (Maryland/DC raised, with a generous dose of Portland Oregon since then):

    For the positive case "I used to" [jus.stu] or [jus.stə] is the only option for past habitual meaning, whereas "I use (it) to" [juːz (ɪt) tu] can only mean current utilization. In the indicative the instrument (often pronominal) is requisite between "use" and "to", so it's only in imperatives and the like that the two words would be adjacent. Though the "it" is often weakened by devoicing the vowel or cliticizing onto "to".

    For the negative case, "I didn't use to" is the only correct spelling, with the [jus.stu] or [jus.stə] pronunciation for past habitual and the [juːz tu] pronunciation and indicative-obligatory instrument for the utilization reading. The reasoning is that the past is already marked by "did", so duplicating the past with "used" would be incorrect and sounds godawful— we don't say "I didn't wanted the present", "I didn't gave a damn", "I didn't had to do it", "I didn't went to the store", nor any other such atrocity. These are starkly ungrammatical for me (exactly like "didn't used to"), whereas things like "use(d) to could" are quaintly non-standard but acceptable (though I wouldn't use it except for amusement). The utilization reading also allows "don't use to". The past habitual reading disallows "don't use to" because there's no past marking, and "don't used to" is forbidden because the past tense has skipped over the first verb in the cluster ("do"), which is forbidden generally.

    And finally, "was used to" would be the correct past habitual for the pluperfect, where the "used" is the VBN form not VBD. My pronunciation gives a stronger emphasis to the palatal onset or to the final [st] cluster depending on prosody, something like [wʌz.juːst tu] vs [wʌz jːus.stu], and weakening of [tu] to [tə] doesn't sound right here though it can become a [t-] clitic if the prosody breaks between "was used" and "to X". There's definitely some weird cliticization stuff going on here that doesn't happen at all with the verb "use" [juz], though I doubt I have the records to figure it out impartially for my own speech.

    I'll grant that my cliticization/pronunciations may be strange, since I seem to have an uncharacteristically agglutinative take on English (contractions like "-'ll" can follow anything, not just pronouns; multiple contractions like "-'d've" and "-'d'n't" sound perfectly fine; etc), but I'm surprised that these aren't the ubiquitous grammaticality judgements in other American English dialects since I've never noticed anyone using "didn't used to" anywhere I've lived (western Maryland, DC, Kansas, Portland, Baltimore, southern Indiana).

  97. wren ng thornton said,

    November 6, 2010 @ 12:05 am

    Edit: The obligatory instrument of the utilization verb can also be provided by a passive subject, naturally. So passive indicatives and deontic modals can also cause adjacency between "use" and "to" ("cups are used to drink from", "this should be used to catch water",…). So I should've said it's mainly the active indicative which forces non-adjacency.

  98. Antariksh Bothale said,

    November 6, 2010 @ 12:26 am

    To me (raised in India), didn't used sounds just plain awkward. In fact, before reading this post, I won't have had second thoughts about declaring it ungrammatical. It's only now that I realize that didn't used is fairly commonly (and probably more often) used.

  99. James Wilson said,

    November 6, 2010 @ 1:52 am

    Sorry, can't duplicate. Googling "didn't use to" turns up 8.9 million results, while "didn't used to" turns up 7.9 million. Moreover all the prescriptive grammar sites say it's "didn't use to". So why is this not just a common misspelling? It wouldn't be the first instance of metanalysis in the history of the world.

  100. Leslie said,

    November 6, 2010 @ 1:53 am

    I also remember being a bit taken aback when I first started teaching EFL with BrE books (Headway series, mostly) ten years ago, and they taught "didn't USE to". For me, AmE, from central Texas, 'used to not' is probably most natural, but 'didn't USED to' is also grammatical.

  101. dw said,

    November 6, 2010 @ 2:57 am

    @Andre Gallo:

    Are you sure this isn't just progressive assimilation of the /v/ of "have" to the voiceless /t/ of "to"?


    "I'll have to make some soup"

    different from

    "I'll have tomato soup"

    with [hæf] in the former and [hæv] in the latter?

    If so, do you mind if I ask what variety of English you speak? I've never heard of this distinction before.


  102. dw said,

    November 6, 2010 @ 3:17 am

    *She used clearly to like spinach.
    *She used, I think, to like spinach.
    *She used uh to like spinach.

    These all sound grammatical to me.

  103. Julie said,

    November 6, 2010 @ 3:58 am

    Even fully emphasized, "Well you damn well should have" ends in "shoulda." In fact, "shoulda" might even be more emphatic than a more precise pronunciation.

    @Andre Gallo: "I used to love listening to this song when I was a kid," has a nostalgic feel to it, and usually means "I haven't heard it in a long time, and I'd forgotten how much I love it."

    "Had I have been there…." No, that can't be another "have." Must be "of." I'm pretty sure I use that construction in speech and pronounce it "uh." "Had I-a been there…" In writing, it's "Had I been there…."

    I suspect that it's modeled (falsely) on sentences with modals and perfect tenses. "Could I have been there?" "Would that I had been there…."

  104. Pflaumbaum said,

    November 6, 2010 @ 8:37 am

    @ Julie – I assure you it is another have, cf CGEL p151, where it's called the 'double perfect'. The authors remark:

    "It appears to be increasing in frequency, and though it is not yet established as a standard form, it is used by many who in general speak standard English."

    As regards should of, there are different pronunciations in different dialects of course. When I emphasise it, being a BrE speaker I don't say shoulda; most of the stress goes on should, but I'd be likely to give it the full have, though maybe with the /h/ dropped. But I do also hear a fully articulated of from some speakers, especially young ones, and I've probably done it myself on occasion if speaking carelessly.

    @ groki – definitely, I wasn't claiming that it was functionally similar to didn't used to, I was just taking up John's challenge to find another construction that marks the past twice.

    Also, the past perfect doesn't even have to mark past time at all, as in the 'doubly remote' construction:

    If they had been alive now they would have been horrified.

  105. Ellen K. said,

    November 6, 2010 @ 8:53 am

    Craig Russell, I know I don't personally see 've/"of" as having any connection with the preposition "of", and yet, as I noted in my previous post, I find it acceptable where a full "have" would be unparsable. So, using it has something different from a full "have" does NOT have to mean associating it with the preposition "of".

  106. Bob Ladd said,

    November 6, 2010 @ 9:39 am

    I was going to say the same thing that Andre Gallo said about have (to), and I'm surprised that dw is surprised that Andre wants to compare them. To me, it seems clear that there are plenty of analogues to the distinction that Rasselas illustrates above with respect to used to. Like

    What kind of opportunities does he have to go skiing? with [v]


    Why does he have to go skiing? with [f]

    On the other hand, dw's reaction shows clearly that English speakers are happy to identify the first syllable of hafta with the verb have, whereas what the whole topic of the thread shows is that lots of speakers have lost the connection between the first syllable of used to (or usta, if you prefer) and the verb use. That's why they're unhappy about writing didn't use to in I didn't use to go skiing, even though it's phonetically analogous to writing didn't have to for the "hafta" pronunciation of I didn't have to go skiing.

    As for dw's example I'll have to make some soup and I'll have tomato soup, the pronunciation of have in those two certainly seems very different to me.

  107. greg said,

    November 6, 2010 @ 12:20 pm

    It's interesting that you mention "try" as an analogy; I've recently become fascinated by the way some people use "trying to" as a single word, often spelled "tryna."

    It's interesting to me because this word actually prevents one from marking a common distinction by the placement of "not." That is, people who use "tryna" can't say

    (1) I'm trying not to do that.

    as opposed to

    (2) I'm not trying to do that.

    I would sometimes use these to mark a distinction between between (1) making an effort to avoid doing X and (2) making no effort to do X.

    See, e.g., this Positive K song (http://www.youtube.com/watch?v=VvYIpa1Ulvw) that contains the lyric: "I'm not tryna hear that, see," where he clearly means to express that he's making an effort to ignore what she's saying. This struck me as odd when I first heard the song, and over time I realized the explanation is that "tryna" is a single word and so blocks the "trying not to."

    I have since begun using "not tryna" as I would have used "trying not to." I get a kick out of it. And of course once I became aware of the phenomenon I see it everywhere; it's surprisingly common but I hadn't noticed it before I heard the song.

    Sorry, not being a linguist I'm not sure the best way to express what's going on here, but I hope I've gotten the idea across.

  108. Dw said,

    November 6, 2010 @ 1:02 pm

    @Bob Ladd;

    Would you, or anyone else with the voiced/voiceless distinction between the two senses of "have", tell me where this distinction is current? I've lived in England and California and have never perceived such a distinction in my own or others' speech. For comparison, I do have the voiced/voiceless distinction in "used to"

  109. Nijma said,

    November 6, 2010 @ 2:35 pm

    Peter G. Howland,when I first came across the "Click to Embiggen" note, I thought it was beer-out-the-nose funny

    Part of why it's funny is you expect to see a mouseover message something like the NYT uses, also because you understand it immediately without having to resort to Urban Dictionary, but still I find it slightly louche in a way I would not find "enbiggen", for reasons I can't explain.

  110. Jerry Friedman said,

    November 6, 2010 @ 3:24 pm

    You can get few enough results to check individually by specifying the verb and searching newspapers to avoid usage discussions (though I did eliminate one).

    "used not to be" 70

    "used to not be": 25

    "didn't use to be": 171
    "did not use to be": 45

    "didn't used to be": at least 800, but the search engine is rebelling.
    "did not used to be": 50

    As far as written text is concerned, I got the impression that things aren't as bad for used not to be as they look, since a larger proportion of the others were from quotations of spoken comments. But didn't used to be is still the clear winner.

    I tend not to split infinitives with not (or never), but I think I'm more likely to say used to not [verb] than used not to [verb], which may indicate that used to is one word somewhere in my brain.

  111. wren ng thornton said,

    November 6, 2010 @ 5:23 pm


    Why can't they say "I'm tryna not do that" (perhaps contracted as "trynnot")? Those seem to be the natural way of distinguishing vs "not tryna", and it works fine for other analogues like "I wanna not get caught in traffic" or "I'm gonna not see that movie". I agree that the "not Xna" versions are the unmarked negation, but the alternative is grammatical for me (for "gonna", "wanna"; I don't have "tryna").

    Also I think the "not tryna" for 'trying not to' is related to other dialect considerations in AAVE. That is, when emphasized or enunciated I think you'll find "not trying to" with the same meaning as "not tryna". In my experience it's a lexical difference in the meaning of "try".

  112. greg said,

    November 6, 2010 @ 7:16 pm

    @ wren

    Hmm, I suppose one could say "tryna not do that." That does involve splitting the infinitive, but nothing wrong with that. It definitely seems wrong to me, but I'm really not the right person to ask, since as I said to the extent that I "have" "tryna" it's something I've consciously adopted even though it wasn't part of my dialect. Maybe it's just that "-na not" is awkward to say, and the further contraction to "trynnot" is a bridge to far?

    The "gonna" and "wanna" sentences you mention do strike me as grammatical, but they're also not what I would say. I'd prefer:

    I don't wanna get caught…
    I'm not gonna see…

    Interestingly, though, "I'm going not to see that movie," strikes me as definitely ungrammatical, and I disprefer "I want not to get caught…" although I wouldn't say it's ungrammatical. In either case, there isn't the same logical distinction to be made as with "trying to," so there's nothing for the placement of "not" to distinguish.

  113. greg said,

    November 6, 2010 @ 7:20 pm

    Actually as I think about it, I think that "I'm NOT tryna X" is better for emphasizing the negation than "I'm tryna not X." Try saying both with a heavy emphasis on "not."

  114. Craig Russell said,

    November 6, 2010 @ 7:54 pm

    @ Ellen K.

    Yes, this is why I argued above that I would prefer to consider it a separate word from the preposition 'of' and not an extension of the preposition's use. I offer the spelling 'of' simply because that is how people commonly represent it in writing, and because it is a homophone with 'of'. I don't consider it the same word though because it is not being used as a preposition and because it is clearly not etymologically related to "of" (since it is obviously derived from a contracted 'have', although as you say it has moved on to be used in situations where 'have' won't go). Pflaumbaum seems to be arguing to consider it an extension of the preposition, though, and maybe that is the understanding of some users of this word.

    I call it a 'particle' above because it reminds me a bit of the ancient Greek word ἄν (an) which is generally referred to as a particle. This word isn't directly translated, but (among other uses) marks a past tense verb as contrafactual:

    τόδ᾽ ἐποίησας (tod' epoiesas) — "You did this."
    τόδ᾽ ἐποίησας ἄν (tod' epoiesas an) — "You would have done this."

    While this is not exactly the same as this new use of "of" (or 've or whatever you prefer to call it) it seems somewhat analogous.

  115. Ellen K. said,

    November 6, 2010 @ 11:42 pm

    Craig Russell, as I read what you wrote before, I see you saying that this "of" particle, as you call it, is being seen by people as the same word as the preposition "of". That is what I was replying to.

  116. Jason Eisner said,

    November 7, 2010 @ 12:05 am

    Mark (MYL), are you suggesting that "didn't use to" is a kind of compromise or analytical error on the part of those of us who prefer it?

    I don't really understand your analysis. I certainly share your judgments about the spoken forms, and am happy to agree that "useta" is a verb that subcategorizes for a bare infinitive.

    But doesn't that make it exactly the same as "hafta," "wanna," and "tryta"? The analogy seems strong. Each of these verbs is written as a simpler verb combined with "to," often with a special pronunciation, and as you say, this combination is uninterruptible in the case of "useta" and "hafta" (at least for us — though not for iching, nicole, or dw above).

    Following that analogy would support the "didn't use to" spelling:

    Each of these combined verbs shows tense and agreement fully analogous to that of the simpler verb. The inflectional pattern is exactly the same (e.g., "hafta" is irregular in exactly the same way as "have"). Paralleling this, the standard written form of the combined verb simply combines the inflected form of the simpler verb with "to" (even though the pronunciation changes, e.g., for "hafta," "hadta," "useta"):

    You had to / You didn't have to / Did you have to / Do you have to / I have to / She has to / They have had to / You would have to / We seem to have to / They made him have to / You've been having to
    You wanted to / You didn't want to / Did you want to / Do you want to / I want to / She wants to / They have wanted to / You would want to / We seem to want to / They made him want to / You've been wanting to
    You tried to / You didn't try to / Did you try to / Do you try to / I try to / She tries to / You have tried to / You would try to / We seem to try to / They made him try to / You've been trying to

    Compared to "hafta," "wanna", and "tryta," the only peculiar thing about "useta" is that it can only appear in the past tense. (At least in modern usage: dw above says it could formerly be used in present tense.) I don't see why this calls for a different spelling convention than is used for the other combined verbs! So I would write:

    You used to / You didn't use to / Did you use to / *Do you use to / *I use to / *She uses to / *You have used to / *You would use to / *We seem to use to / *They made him use to / *You've been using to

    It's not clear to me yet why "useta" only appears in the past tense in contemporary English (related to its contemporary semantics? or just an arbitrary restriction?). But unlike some commenters above, I would not conclude from this past-tense restriction that "useta" doesn't take tense at all, since syntactically it does — however you choose to spell "Did you useta," the form of "useta" there is presumably the bare infinitive rather than the past, because the bare infinitive is what is used with do-support in English. Nor would I conclude that "useta" or "use" is an auxiliary, since (at least in my dialect) it can't move like other English auxiliaries:
    *Use you to smoke?
    *No, I usen't to smoke.
    *Useta you smoke?
    *No, I useta not smoke.
    (with the meaning "It is not true that I useta smoke")

  117. Ethan said,

    November 7, 2010 @ 1:03 am

    (at least in my dialect) it can't move like other English auxiliaries:
    *Use you to smoke?
    *No, I usen't to smoke.
    *Useta you smoke?
    *No, I useta not smoke.

    If you look up-thread a bit, you'll see that other people consider both (2) and (4) acceptable, and in some dialects more natural than "didn't useta smoke".

  118. Joe said,

    November 7, 2010 @ 3:34 am

    I think Jason is raising the same issue that I did unthread. The question isn't whether "use" (in the aspectual sense) is not an auxiliary for some speakers. If someone says, "Used he to smoke?" it is. Something like "he used not to be" is ambivalent depending upon whether the verb or the infinitival clause is being negated). The question is why "use" in "he didn't use(d) to" would be considered one. The standard argument would be that, the form "usedn't" would be auxiliary, "didn't use(d) would be lexical ( because it requires do-support), and the diachronic process would be the historical trend in which fewer and fewer verbs behave as auxiliaries (have is undergoing a similar process: although there are still speakers who say, "have I to sign the form?" (auxiliary), it is more common to say, "do I have to sign the form (lexical).
    I'm still not sure why Mark's analysis wouldn't apply to Tim.

  119. John Walden said,

    November 7, 2010 @ 5:15 am

    The rise of do-support has been resisted, optionally, by verbs that have some modality about them (I daren't, I hope not, I suppose not, I believe not etc) as well as most successfully by the nine or ten 'German Modals' (can could will would etc). Then there are cases like 'need' where the lexical form is more factual and less subjective than the modal auxiliary form. 'To-support' also implies less modality, both in meaning and form: 'ought to' is less modal than 'should' in every sense, while still not needing do-support in Standard English (*You didn't ought to have done that)

    On that basis "he used not to smoke" probably should have a shade of difference in meaning from "he didn't use(d) to smoke": The lexical form should present the information as fact while the auxiliary form should have some modality. Which it doesn't seem to on paper, as far as I can see.

    I'm going to make another wild assertion: that language rarely has two different ways of doing exactly the same thing. It's probably relevant that in the auxiliary form the emphasis falls by default (for me) on the strong "not" and in the lexical structure on the main verb:

    He used NOT to smoke

    He didn't USE(d) to smoke

    Is usedn't in use anywhere?

  120. Jon Lennox said,

    November 7, 2010 @ 5:15 am

    @dw: I definitely have the voiced/voiceless distinction in "have". Raised Western Massachusetts, currently live in the New York City area.

  121. GeorgeW said,

    November 7, 2010 @ 7:28 am

    @Jason Eisner: "Nor would I conclude that "useta" or "use" is an auxiliary, since (at least in my dialect) it can't move like other English auxiliaries"

    What would you say about 'gonna?' It cannot, like 'justa,' be fronted except with ellipsis?

    Maybe 'justa' and 'gonna' are both on a similar grammaticalization glide path toward full modal status but not quite there yet.

  122. Pflaumbaum said,

    November 7, 2010 @ 9:08 am

    @ John Walden

    My Yorkshire grandparents said usedn't to, and it's possible my mother (about 60) occasionally does too. I'll listen out for it…

  123. Joe said,

    November 7, 2010 @ 10:26 am

    @John Walden,
    I would analyze most of your examples differently. "Not" in "hope not," "believe not," etc doesn't negate the verb. "not" is more a "pro-form," that negates an antecedent in an interrogative: Q: do you think it will rain? A: I hope not (I.e., I hope it will not rain). You can see this clearly by comparing it with "I hope/believe/suppose so." Likewise, if you say "I used NOT to smoke," it could be argued that you are negating "to smoke" not "used" (cf. I tried not to laugh, where it is the laughing, not the trying that isnegated) In such cases, it is hard to tell whether "use" is an auxiliary or lexical verb, which is why I gave "usedn't" as an unambiguous auxiliary (and the fact that that form is rare is precisely my point).
    I wonder whether present tense aspectual "use" fell out because present tense can convey habitual events: the present time equivalent of " I used to cycle to work" is "I cycle to work." (I haven't look up any examples in the OED to see if present tense aspectual "use" had a difference in meaning, however).

  124. Jason Eisner said,

    November 7, 2010 @ 10:48 am

    @myl: Now that I've answered (I hope) your curiosity about why some of us write "didn't use to," I'd like to turn it around and ask why some of you write "didn't used to." You say that "useta" is "just a word" for you. How does that lead you to a spelling? Do you feel that any spelling is as good as any other, and "used to" is less likely than "use to" to be misread and mispronounced as involving the lexical verb "use"? (I do worry about misreading when I write "use to" — whereas everyone knows how to read the more frequent bigram "used to.") Or is it that you feel that this is some kind of fixed construction that is truly invariant across "He V smoke" and "Did he V smoke," in contrast to all other verbs of English?

    @Joe, @Ethan: Thanks for clarifying, Joe — yes, that is what I meant. When I say "auxiliary," I am only talking about the feature on an English verb that says it can move rather than taking do-support. Which verbs take this feature varies across time and space.

    @speakers of dialects where "use" is an auxiliary: Since you can say "He usedn't to smoke" and "Used he to smoke?," apparently "useta" is interruptible for you. Can you also interrupt it in other ways — that is, are Mark's examples below okay for you?
    *She used clearly to like spinach.
    *She used, I think, to like spinach.
    *She used uh to like spinach.
    I don't suppose you can say "To smoke is what I used, but to drink is what I want"?

    @GeorgeW: The construction you're asking about is "be gonna," not "gonna." It seems to pattern fine with "hafta" and the others. The "gonna" part is phonologically reduced and uninterruptible. The construction itself is inflected just like "be," and the "be" part retains the feature (sometimes called "auxiliary") that allows it to move: "Are you going to…?"

    Or are you talking about dialects where "gonna" doesn't require the "be," as in "They gonna dance the night away"? I have that construction in speech, and for me at least, I think it is part of a more general pattern of being able to elide present-tense "be" under certain circumstances. Here are two pieces of evidence that it is only elision. In tenses other than the present, the "be" is overt: "They were gonna dance the night away." And the elision account correctly predicts that I'd ask "You gonna dance the night away?" (as short for "Are you gonna dance the night away?"), rather than "*Did you gonna dance the night away?" or "*Gonna you dance the night away?" Perhaps other commenters have more data about their own dialects.

    @dw and @Jon Lennox: I grew up in New Jersey. My pronunciations are
    have tomato soup ['hæv.tə] have to make a soup ['hæf.tə]
    has tomato soup ['hæz.tə] has to make a soup ['hæs.tə]
    had tomato soup ['hæd.tə] had to make a soup ['hæd.tə]
    In other words, the final consonant of "have" or "has" (but not "had") is devoiced in the "have to" (= must) construction. I've never noticed anyone using other pronunciations and I think they'd sound deeply odd to me. dw, where are you from?

  125. Joe said,

    November 7, 2010 @ 11:12 am

    For me, "use" is a lexical verb, so I was only referring to those for whom it would be an auxiliary (and I think they are very much a minority). What I think do (and I would be curious if others do so as well) is distinguish "Did he use(d) to smoke? ( where "use" is compounded with "to") and "What did he used to do?" ( where I don't believe "use" is compounded with "to"). But I haven't thought about this enough.

  126. GeorgeW said,

    November 7, 2010 @ 11:42 am

    @Jason Eisner: "Or are you talking about dialects where "gonna" doesn't require the "be," as in "They gonna dance the night away"?"

    No, I was referring to the grammatical as opposed to lexical role of 'gonna': She's gonna go home. Here, 'gonna' indicates future tense.

    I realize that 'justa' doesn't require a 'be' auxiliary. However, it also has a grammatical role (habitual) rather than lexical in sentences like:

    1. She justa go to school every day.
    2. *She justa go to school once.

  127. Jason Eisner said,

    November 7, 2010 @ 12:33 pm


    No, I was referring to the grammatical as opposed to lexical role of 'gonna': She's gonna go home. Here, 'gonna' indicates future tense.

    Okay: then we are all on the same page, so I trust you accept my answer about how the "be" in "be gonna" (not the "go") is the part that inherits its tense, agreement, and movement patterns from "be." This is analogous to how "hafta" inherits those patterns from "have," and — at least in my analysis! — "useta" inherits them from "use."

    You are being careful in emphasizing that you are not talking about the lexical meaning of "go," but I think that is already implicit in the whole discussion. We've all been talking only about the constructions like "be gonna," "hafta," and "useta" ("justa"). As you say, these have aspectual or modal meanings that are quite unrelated to the meanings of "be going," "have," and "use." They also have altered pronunciations that are not available for the lexical "go," "have," and "use." (For example, you can't say "*She is gonna the fair," "*I think I'll haftamato soup," or "*I usetamato soup to wash out the skunk smell." You can't even do it in cases where the "to" really is the infinitive marker: "*She is gonna sell her cow" (to mean "She is going [in order] to sell her cow"), "*Diamonds are what I hafta slake my need for material goods," "*Some people useta impress their friends" (where "use" = use drugs).)

    My claim is that despite these differences of pronunciation and meaning, the constructions "be gonna," "hafta," and "useta" nonetheless inherit their morphology and syntax from lexical "be," "have," and "use." And they also conventionally inherit spelling in the case of "be" and "have" — so why not also in the case of "use"?

  128. GeorgeW said,

    November 7, 2010 @ 12:53 pm

    @Jason Eisner: "And they also conventionally inherit spelling in the case of "be" and "have" — so why not also in the case of "use"?"

    I think the operative word is "inherit." I would suggest that over time they lose their compositional origin (except among historical linguists) and are analyzed as single units with a grammatical (vs. lexical) meaning. I suggest that the variation in negating, agreeing, etc. are because these are in a transition stage. I would guess that at some future time this would become standardized.

    FWIW, I couldn't quickly find 'justa' cited as an example of grammaticalization. But, 'gonna' is frequently cited including in "Historical Linguistics" (Cambell, MIT Press, 1998).

  129. Dw said,

    November 7, 2010 @ 1:46 pm

    @Jason Eisner:

    I grew up in England and live in California. I checked wih my wife, who grew up in Texas, and she doesn't seem to have the voiced/voiceless distinction in "have" either. Is it possible that it's an East Coast thing?

  130. Ellen K. said,

    November 7, 2010 @ 2:07 pm

    I've spent my life in Missouri and Eastern Kansas, and I have a voiced/voiceless distinction in "have". It's possible "have" is partially devoiced in something like "have tomatoes", but, still, it certain remains a v sound, and, in particular, if I were to pause between "have" and "tomatoes", then "have" would get a solid v sound and would never sound like an f. (If I say it with an f, it sounds like I'm talking about cutting tomatoes in half.) Whereas, if I put in a break between "have" and "to" in something like "I have to go", the final sound in "have" would probably be voiceless.

  131. GeorgeW said,

    November 7, 2010 @ 4:41 pm

    Isn't the 'have' /v ~f/ distinction one of possession vs. compulsion (used with to)?

    "I hav tomatoes" vs. "I haf to go."

  132. boynamedsue said,

    November 7, 2010 @ 5:32 pm

    In EFL teaching, students are consistently taught that "didn't used to" is incorrect, and all examining boards in the UK agree that it is an error. I personally don't use it, but I don't know whether that's because I learned to teach it like that or because I've never said "didn't used to".

    It's obviously something that was unacceptable in writing in the early 20th century, but which is evolving into the dominant form.

  133. GeorgeW said,

    November 7, 2010 @ 5:34 pm

    As I think about it, I also have the same vd-vl distinction for the 3rd person based on possession vs. compulsion.

    "She haz tomatoes" vs. "She has to go."

  134. kenny said,

    November 7, 2010 @ 6:57 pm

    I agree with everything in this article except one thing: I find "She used, I think, to like spinach." totally acceptable.

  135. James said,

    November 8, 2010 @ 1:31 pm

    Within the first page or so of Catcher in the Rye, Salinger uses both "used to" and "use to" correctly.

  136. Mira said,

    November 8, 2010 @ 4:10 pm

    I'm an EFL teacher, and I confirm what boynamedsue said. We're supposed to teach "used to" and "didn't use to", unless, apparently, it's a question, in which case we can say "Did you used to…?" This all seems completely artificial and divorced from the way actual English is spoken. I've never made this distinction myself, and both "didn't use to be" and "didn't used to be" look sort of wrong to me — if I had to use that phrase, I'd probably say "used to not be", which would be dismissed by my boss and coworkers, I'm sure, as ungrammatical and American.

    EFL English can be a little bizarre sometimes. Lots of materials I've come across written by Czechs have said that "Used you to like spinach?" and "I usedn't to like spinach" are correct. This makes my head hurt.

  137. Alces said,

    November 8, 2010 @ 6:24 pm

    I'm from Merseyside, England and I've never been sure how to write 'didn't use(d) to'; I think 'didn't used to' is what I say, although it's hard to tell since the extra [t] in [dɪdənʔjusttʰə] can easily be dropped in quick speech.

    To me, 'used to' seems to be a single syntactic unit (though not a phonological one, what with the gemination and aspirated /t/), with little to do with besides etymology with the verb 'use'. Pausing after the 'used' would sound strange. It'd then make sense that it should stay 'used to', and not change to 'use to' for this one usage.

    @dw: I definitely have 'have to' realised as [haftʰə] (and 'has to' as [hastʰə]), distinct from normal 'have' [hav]. I'd never put an [f] in 'I have tomatoes'; that would sound quite strange to me although I've read that it's a feature of Yorkshire dialect.

    I also have 'used to' realised as [justtʰə] rather than [juːzdtʰə]; I've never heard this latter usage in fact.

  138. Jonathan said,

    November 11, 2010 @ 3:54 pm


    It's true that "used to" is pronounced with /s/, while the real "use" is pronounced with /z/, but that by itself doesn't mean they aren't the same word.

    For me "have to" is pronounced with /f/, while real "have" is pronounced with /v/. Even in careful speech or with a pause that "ve" is pronounced /f/. Yet the "have" in "have to" is clearly the verb "have", and is conjugated the same: "he has to" (with /s/) "he had to".

  139. Jonathan said,

    November 11, 2010 @ 4:00 pm

    This raises a question for me: is the "to" glued more strongly to the preceding word with some verbs than with others?

    It seems like "not" has to be placed before "to" in some cases, after it in others.

    I'll try not to. (Not "I'll try to not." "Not" is glued to the following (absent) word.)
    He used not to OR He used to not [like them]. (I think both are acceptable? "Not" is floating between the words.)
    He has to not [touch the line with his toe and he'll win a prize]. (Better than "he has not to", but both are weird. "Not" is glued to the preceding word.)

    Does this show anything?

  140. Jonathan said,

    November 11, 2010 @ 4:06 pm

    Another weird thing about the gluing of to.

    In New York taxicabs, there are ads that say "Lotos knows which files to download and which to not." This sounds really bizarre to me: "Lotos knows which files to download and which not to" would be more like the English I speak.

RSS feed for comments on this post