Lady Bracknell strands even adjunct prepositions

« previous post | next post »

Lady Bracknell, in Oscar Wilde's The Importance of Being Earnest, is one of the most terrifyingly pedantic and correctly spoken characters in all of English theater ("a monster," Jack Worthing says of her, "without being a myth, which is rather unfair"). And I have mentioned her usage in lectures on numerous occasions to point out, when talking about preposition stranding, that she does strand prepositions. But as I watched Mark Thomson's wonderful production of the play at Edinburgh's Lyceum Theatre last Friday night (get tickets now, readers in eastern Scotland), I suddenly noticed something new about what she says when Jack Worthing gives his age:

LADY BRACKNELL: … How old are you?
JACK: Twenty-nine.
LADY BRACKNELL: A very good age to be married at.

A preposition phrase (PP) like at the age of 29 is very clearly an temporal adjunct, not a complement. So Lady Bracknell is prepared to strand a preposition even in a temporal adjunct PP!

The thing is that this is usually treated by syntacticians as ungrammatical, not in the sense of being forbidden by some grammarianly edict that might or might not match the facts of usage, but in the sense of being instinctively avoided by native speakers and felt as unacceptable. Lots of cases of stranded prepositions that are heads of locative, directional, or temporal adjuncts seem clearly unacceptable to many speakers. Some examples (with the stranded preposition underlined):

Who did you talk to today? *Who did you talk near today?
How many mayors have you written to this week? *How many towns have you cycled towards this week?
What do you laugh at most often? *What do you laugh during most often?
Tell me something you are happier about, will you? *Tell me something you are happier since, will you?
The things I had been working for were now unimportant.     *The things I had been working despite were now unimportant.

I am not of course suggesting that there is a clear-cut true generalization here about stranding adjunct prepositions being always disallowed. Quite the reverse: the example from Lady Bracknell itself is solid evidence against such a generalization. I take her utterance to be entirely well formed in Standard English, and thoroughly idiomatic; so the generalization that prepositions can be stranded only when they head complements, not when they head adjuncts, is (despite the foregoing examples) not accurate.

What exactly makes the examples in the red column above seem less felicitous than the ones in the green column has yet to be identified (obviously, it is likely to have something to do with plausibility and naturalness of the situation depicted, and frequency of the habit of stranding in that particular phrase; pragmatic concerns rather than syntactic ones, in other words).

But the issue of the "preposition at the end of the sentence" (a misnomer: not one of my examples above has the stranded preposition at the end of the sentence) should not concern sensible people. It is not kind of grammatical sin. Lady Bracknell, an archetypal Standard English speaker if ever there was one, strands not only complement PP heads but also adjunct PP heads, whenever she damned well pleases.

[Would-be commenters: Please note that Language Log now has a penalty of death for perpetrating the two most common jokey remarks about the stranding of prepositions; so before you submit a comment, think about whether you want a hit man on your trail. Thank you.]


  1. Yerushalmi said,

    November 3, 2010 @ 8:16 am

    Personally I don't have a problem with the top two items in the red column. The third one is borderline; the bottom two feel instinctively incorrect to me.

  2. Dierk said,

    November 3, 2010 @ 8:51 am

    I cannot rephrase the actual Bracknell utterance to not have the preposition stranded. Are those lambasting such use telling us to scrap complete sentences in favour of completely different sentences to avoid their pet peeves? Or am I just dumb again?

    Might also be that Oscar Wilde couldn't write … much less in style.

  3. John Atkinson said,

    November 3, 2010 @ 8:52 am

    I do have a problem with those top two items, Y. But it's the same problem as I have with "I talked near Jack today" and "I cycled towards five towns this week" — they're not the sort of thing anyone would say — not meaningless, but not far off it.

    I agree with you on the other three.

  4. Yerushalmi said,

    November 3, 2010 @ 8:57 am

    The situation in which those top two items would realistically be said would have to be a very bizarre one ("Okay, we're going to have a contest to see how many people you can have a conversation near!"), but so long as the situation could be envisioned, so could the sentence.

    The other three, however, would be phrased very differently in my head even if a situation naturally leading to the sentence were to arise.

  5. Don Monroe said,

    November 3, 2010 @ 9:03 am

    My gut feeling agrees with Yerushalmi's about the five red examples.

    My guess is that the differences lie in whether the combination of verb and preposition is familiar or idiomatic: "work for" is, "work despite" is not. When the combination has a recognizable meaning of its own, it feels more natural to keep the two words togeteher. It seems akin to the unseparable prefixes of German verbs, which would often be prepositions on their own but do not float freely like true preopositions.

  6. Ray Girvan said,

    November 3, 2010 @ 9:04 am

    I think a difficulty with those discussed above is that there's an interaction between the semantic and the grammatical: "Who did you talk near today?" is weird not because it's ungrammatical but because it's a spacial connection no-one makes. The actual construct is fine if used with a verb where such connection is normally noted: for instance, "Who did you sit near today?"

  7. Jens Fiederer said,

    November 3, 2010 @ 9:09 am

    A HIT MAN on your trail? I thought it was a Linguistics scholar, or has the professor gone professional in an only weakly related field?

    [We delegate, OK? —GKP]

  8. L said,

    November 3, 2010 @ 9:47 am

    I'm glad the top two jokey comments are off-limits, because it makes room for more obscure stranded preposition jokes.

    Why, just the other day I brought a book to my daughter's upstairs bedroom for a bedtime story. When she saw it, she exclaimed with dismay, "Daddy, what did you bring that book that I don't want to be read to out of up for?"

    [Enzo has been dispatched to deal with this commenter. You will not be hearing from him again. —GKP]

  9. Vance Maverick said,

    November 3, 2010 @ 10:18 am

    Dierk, try "A very good age at which to be married," which is to my ear reasonably idiomatic, though not as good as the original. Wilde could definitely write (as well as being a serious classicist).

  10. Pflaumbaum said,

    November 3, 2010 @ 10:40 am

    @ Ray Girvan – I agree. These, for instance, seem fine to me:

    (1) Who did you stand near today?
    (2) What goals have you worked towards this week?

    Maybe I've missed the point in recasting them though?

    @ L – when Pinker gave that example the book he had the book be on the subject of Australia – "…to be read to out of about Down Under up for?"

  11. Jonathan Mayhew said,

    November 3, 2010 @ 10:51 am

    That's a very good reason for a person to have a hit man after for.

    [Enzo: Deal with Jonathan too. —GKP]

  12. Mr Fnortner said,

    November 3, 2010 @ 11:09 am

    The pink adjuncts are not so ungrammatical as unexpected, and thereby jarring. With a bit of a setup, the sentences are easy to digest, and unremarkable.

  13. outeast said,

    November 3, 2010 @ 11:15 am

    @Dierk,Vance – '…a very good age for marriage' would work too.

    [No, outeast, you're not following the drift here. Everyone agrees that if you rephrase completely you can avoid facing the decision about whether to strand a preposition (…to be married at) or prepose it with the wh-word (…at which to be married). The issue under discussion here is whether stranding is grammatical. (It is, of course; though millions of Americans seem to believe otherwise.) Don't just tell me you can cop out and not face the issue. —GKP]

  14. John Roth said,

    November 3, 2010 @ 11:18 am

    The first three sound grammatical to me. At least, I can contrive a situation where I'd say that, and where I'd understand it if someone else said it. I can also imagine a situation for number five, although the phrasing is a bit contrived.

    The exception is number four, which I'd flag as something a native speaker of German who learned English as a second language might say.

  15. richard howland-bolton said,

    November 3, 2010 @ 11:34 am

    "There are spies everywhere. Who did you talk near today? Your loose lips have already sunk several ships!"


    [Exactly right. What seems to be the key is that here you have created a context where having talked near someone can be of real relevance, whereas in most contexts that couldn't conceivably be a topic of relevance to anything. This is what I meant about pragmatic concerns. —GKP]

  16. Ray Girvan said,

    November 3, 2010 @ 11:48 am

    @ Pflaumbaum:

    Actually, with a contrived context, it makes perfect sense.

    A: "How did you get that black eye?"
    B: "I got thrown out of the theatre."
    A: "Why?"
    B: "Well, ever since the Duke of Edinburgh spoilt my enjoyment of Hamlet by talking all the way through, I've made it my project to annoy the rich and famous by sitting near a celebrity in the theatre and talking non-stop; see how they like it."
    A: "So, who did you talk near today?"

    [Same point as the previous comment, but 15 minutes later. Pragmatics is the key. —GKP]

  17. David L said,

    November 3, 2010 @ 12:24 pm

    I have this friend, a keen cyclist, who loves to take long rides out on the open road, in the fresh air, but gets nervous whenever he rides close to any sort of built-up area. The traffic, the noise, the people — it all induces a kind of phobia in him. But this fear meant he often had to go long distances out of his way to get home. So he started weekly sessions with a therapist, who encouraged him to ride a little closer to conurbations than he normally would, to push against the limits of his anxiety. My friend is a diligent sort, and worked hard at this exercise.

    Naturally, his therapist began every session by asking "How many towns have you cycled towards this week?"

    [Yep, pragmatic forcing once again. You guys are getting the hang of how to push apparently unacceptable sentences toward sounding much more acceptable. —GKP]

  18. Mark F. said,

    November 3, 2010 @ 12:58 pm

    Back in the 80's I learned about a model of Universal Grammar in which a large list of properties that a language might or might not have were somehow stored in the brain at birth, with a set of default on-off values. Learning language would then just be a matter of learning which parameters to change away from their defaults.

    It seems to me that the kind of fluidity of grammaticality judgments that we see here tends to argue against theories like that. At least for the constructions where there are degrees of felicity, it seems unlikely that the brain is wired to expect one rule or another.

    My recollection of the theory I describe may be wrong, and it may by now already be thoroughly be discredited anyway. But still.

  19. Barrie England said,

    November 3, 2010 @ 1:00 pm

    'What exactly makes the examples in the green column above seem less felicitous than the ones in the red column' Wrong way round?

  20. Mr Fnortner said,

    November 3, 2010 @ 1:08 pm

    On some level, the prepositions sound like inept choices, as in "She did it on accident," or "He is suffering of the flu." (I know neither preposition is stranded.) I found that I stumbled with each sentence until I resolved the word choice error in favor of an alternate meaning. As I was doing so, I felt I was reading a non-native speaker's words as another explanation of the unexpected preposition. With the last three sentences, I find it very difficult to keep the alternative meanings in my mind. As soon as I look away, they vanish and must be reconstituted when I reread the sentences.

  21. KevinM said,

    November 3, 2010 @ 1:18 pm

    I think you're just intimidated by Lady Bracknell. (So am I.)

  22. Coby Lubliner said,

    November 3, 2010 @ 1:37 pm

    What struck me about the two columns was the nature of the prepositions themselves, more than the context. The ones on the left (to, at, about, for) are typically given a reduced pronunciation in fluent speech (as are others such as on, with etc.), while the ones on the right (near, towards, during, since, despite) are usually pronounced fully, with their own stress. Is that a coincidence?

  23. Neal Goldfarb said,

    November 3, 2010 @ 1:46 pm

    Hmm. So the bedtime-story story is one the two most common jokey remarks about preposition stranding? That's not what I would have figured. I know that No. 1 is the one that Churchill didn't say. (Actually, one of the many ones that he didn't say.) But I figured the Princeton/asshole joke for the other one.

  24. Charles Gaulke said,

    November 3, 2010 @ 1:47 pm

    I obviously have no idea what a stranded preposition actually is, because I can't for the life of me see the difference between the red and green sentences apart from some of the red ones being odd questions to ask someone in most cases. The forth red sentence actually seems much more natural and likely to me than its green counterpart (though note, Mr. Roth, that I am a native speaker of English who knows practically no German). Time to dig into the archives I guess.

  25. Army1987 said,

    November 3, 2010 @ 1:48 pm

    *Now* I get why "What value of Z does this reach a maximum for?" looked so weird to me this morning!
    BTW, for some reason, "A very good age to be married at" sounds OK to me, but none of the example in the red column of the table does.

  26. GeorgeW said,

    November 3, 2010 @ 1:48 pm

    "How many towns have you cycled towards this week?"

    I think the difficulty with this is that 'towards' implies approached by did not arrive where 'to' implies arrived.

    1. She drove towards New York City but didn't get there because of car trouble.

    2. ?She drove towards NYC and got there quite late.

    3. She drove to NYC and got there quite late.

  27. Chandra said,

    November 3, 2010 @ 1:54 pm

    I think you can even build a pragmatic case for the fourth one in the pink list:

    Person A: I'm lonelier since my wife left me. I'm more miserable since I lost my house. I'm more desperate since my business went bankrupt. And to top it all off, I'm sadder since my dog died.

    Person B: Tell me something you are HAPPIER since, will you?

  28. groki said,

    November 3, 2010 @ 2:17 pm

    What exactly makes the examples in the green column above seem less felicitous than the ones in the red column has yet to be identified

    I've seen several purported "identifications" out there–some even in comment streams here at LL. so it seems more fitting (though admittedly riskier, Enzoically speaking) to suggest instead:

    there's no majority on by which one explanation has yet to be agreed.

  29. Xmun said,

    November 3, 2010 @ 2:32 pm

    @Neal Goldfarb. All right. I, for one, have never heard the Princeton/asshole joke. What is it?

  30. Peter Taylor said,

    November 3, 2010 @ 2:43 pm

    The thing which strikes me most about "A very good age to be married at." is the verb. I would interpret "to be married" as a state rather than an action – for the latter "to get married" is the main candidate, and "to marry" is plausible. Is this a change in usage or difference in dialect?

  31. groki said,

    November 3, 2010 @ 3:34 pm

    @Xmun: perhaps Neal Goldfarb is referring to a version of this joke.

  32. Xmun said,

    November 3, 2010 @ 3:48 pm

    Thank you. I had read it before, but had forgotten it.

  33. blahedo said,

    November 3, 2010 @ 3:54 pm

    Looking over the pink examples, I have little problem with them (assuming pragmatics are taken care of), but I have even less of a problem if the stranded preposition *actually ends the sentence*. For example, the third one is much better as simply "What do you laugh during?" and the fifth in a frame more like "He discovered the difficulties I had been working despite."

    I'm not sure why, though. It may have some sort of adjacency problem, in that if one of these stranded prepositions is followed by anything I make a first pass at parsing it as unstranded, with the following thing as object, whereas if it actually ends the sentence there's no such possibility.

    What I love about this is the possibility that the simplistic way of phrasing the rule might in fact be exactly wrong—that stranding (some kinds of) prepositions might in fact be *only* acceptable if they *do* end the sentence.

    (PS: gkp's post about the new death penalty policy was the funniest thing I'd read on LL in ages and I'm all giggles again just from rereading it just now. Fantastic!)

  34. Jens Fiederer said,

    November 3, 2010 @ 5:12 pm

    Not related to stranded prepositions, but this whole "pragmatic forcing" think reminds me of the one thing so far I have found objectionable in The Cambridge Grammar of the English Language: when I read "* The train has been arriving for two minutes now." I wanted to use white-out on that "*", because I immediately got the picture of some railroad employee who does not realize that a train on the outskirts of a station is having mechanical difficulties telling somebody 'The train is arriving this very minute.' with "The train has been arriving for two minutes now." being the irritated rejoinder.

    To me it feels like there is a very real difference between a phrase that can been "pragmatically forced" and one that is jarring under any imaginable circumstance, even if Google search counts can not detect any difference.

    I wonder if there are any experimental linguists that take putative "*" sentences and ask subjects to interpret them.

  35. Jens Fiederer said,

    November 3, 2010 @ 5:16 pm

    Brr. "think" vs. "thing" could be almost passed off as intentional, but my "can been" instead of "can be" could not possibly be pragmatically forced into acceptability!

  36. dazeystarr said,

    November 3, 2010 @ 6:14 pm

    Straying just a bit from the topic, and regarding Mr Fnortner's comment above, is it fair to characterize "on accident" as an "inept choice"? I'm legitimately wondering here.

    I've always said "by accident", and I assumed that was standard, but I feel like I hear "on accident" more and more frequently. I've been wondering if it's an accepted dialectal, perhaps regional, variant. (I grew up on the East Coast of the US and now live on the West Coast.)

    I'll admit it rubs me the wrong way, but I have a friend who argues that it makes more sense, as it parallels its usual opposite, "on purpose". He has me half-convinced.

  37. Jerry Friedman said,

    November 3, 2010 @ 6:14 pm

    I have no problem with How many towns have you cycled to this week? Or through.

    I object less to The things I had been working in spite of were now unimportant. than to the example sentence with despite. Maybe because I'm more used to seeing of stranded than the more formal despite? Using I'd helps even more.

  38. Mathis said,

    November 3, 2010 @ 6:34 pm

    Regarding German verb prefixes, it should be noted that stranding verb parts also occurs in English. The following two sentences have the same word order (and meaning):

    Ich habe einen grünen Hut auf. / I have a green hat on.

    Here, auf is a detached prefix, while English on is more of a suffix, as infinitive constructions show:

    einen grünen Hut aufhaben / to have on a green hat

    While probably unrelated to the matter discussed here, is this kind of preposition stranding also being denounced as ungrammatical?

    PS: I’m neither a native speaker of English nor a professional linguist, so I hope that the examples I’ve given are valid.

  39. The Ridger said,

    November 3, 2010 @ 7:13 pm

    @Peter Taylor – I believe the "get passive" is later than Wilde, or perhaps just more American. At any rate, nowadays you're right: the 'be married' is a state while the 'get married' is the change of state.

  40. marie-lucie said,

    November 3, 2010 @ 7:58 pm

    But I have heard some old ladies say "I was married in 1934", for instance (meaning that's when the wedding took place).

  41. marie-lucie said,

    November 3, 2010 @ 8:11 pm

    I find the near examples problematic, but I would have no problem with next to or close to: Who were you sitting next to at the concert?, for instance. In addition to to being much more frequent than near, both generally and in a "stranded" position, Who were you sitting near? a) sounds like a request to repeat the name of the person near whom you were sitting, and b) implies that you sat in someone's vicinity but not in the next seat, something which would normally be of much less interest to a questioner than the person sitting right next to you, whom both you and the questioner might know, or whose identity might cause the questioner to gain information about your social life.

  42. Richard Sabey said,

    November 4, 2010 @ 2:32 am

    Peter Sarstedt didn't mind stranding a preposition that heads a directional adjunct.

  43. Pashtie said,

    November 4, 2010 @ 3:02 am

    A beautiful lesson from 30 Rock:

    Ex-Wall Street banker: We'd be honoured if you'd come with?!
    Tracy Jordan: You shouldn't end a sentence with a preposition, at.

  44. Robert said,

    November 4, 2010 @ 3:23 am

    I'd interpret 'Who were you sitting near?' as asking for a list of people, which would be relevant when tracing an infection, or a rumour. Who were the people sitting close enough for you to have caught measles from? Who was sitting close enough to have overhead your amusing anecdotes about what the managing director and her lackey have been up to?

  45. Plegmund said,

    November 4, 2010 @ 5:38 am

    Four out of the five red cases give vague information when there are obvious alternatives that would be clearer: talk to, cycle through, laugh at (if that's what happened in each case), Does this offend against Grice's Maxim of Quantity? They have that flavour to me.

    I can't really see that the fifth has a problem of this kind, however.

  46. Peter G. Howland said,

    November 4, 2010 @ 7:12 am

    I don’t have a big problem with Stranded Prepositions, either as a prescriptivist’s rule against them or descriptivist’s acceptance of them. They either work well in the effort to communicate clearly, or they don’t. The example of Lady Bracknell saying, “A very good age to be married at.” works just fine for me if Oscar Wilde intended an emphasis placed on the word “at”. This invites a comparative connotation.

    But in the other examples (pink) shown, I first got the impression that they were composed by someone thoroughly unfamiliar with standard American English. And even when I attempted to superimpose my limited concepts of British English construction on them, they simply made no sense and did not translate into the “acceptable” (green) versions shown. “Why would anyone talk like that? What the hell does that mean?” erupted involuntarily from my normally silent method of reading and attempting to comprehend. It’s possible that “WTF?” with all its vowels and consonants intact was also uttered by me with enthusiastic curiosity.

    Then! Some very creative commenters above presented plausible scenarios in which such phrases might obtain. Even though they appear to be against the “rules”, these constructions could work. It takes a smattering of weirdness to get it done, but hey…it could happen!

    Disregarding a stranded preposition’s acceptance, or lack thereof, there comes a point at which certain phrase structures are instinctively avoided, and I believe this grows out of grammatical maturity. An ideal example might be:

    Child: “I don’t know where it’s at.”
    Adult: “I don’t know where it is.”

    Unless used to emulate or participate in the speech of youth or the under-educated, Stranded Preposition Structures not having other pragmatic sensibilities can easily be avoided by those who care to speak well.

    [You are truly, depressingly, failing to get the point. There is nothing grammatically immature about sentences like What were you laughing at? or It's something we should talk about, and the Lady Bracknell examples underline that. What is wrong with you, still thinking that this is about whether or not one might "care to speak well"? Speaking well is compatible with preposition stranding. The notion that there is something wrong with stranding is a myth. One tries a hundred times to point this out, but it seems to be like talking to a wall. —GKP]

  47. GeorgeW said,

    November 4, 2010 @ 8:20 am

    @Plegmund: "Does this offend against Grice's Maxim of Quantity?"

    Some seem so to me as well, particularly the second (cycled towards). The fact that they require some very specific and unusual contexts suggests this.

  48. John F said,

    November 4, 2010 @ 9:06 am

    Beavis and Butthead Do America gave a lot to laugh at.

  49. Pflaumbaum said,

    November 4, 2010 @ 11:05 am

    This is a different issue, to do with missing pronouns rather than the stranded prepositions as such… but what are people's opinions on

    (1) Are you coming with?
    meaning Are you coming with me/us?

    To me it's completely natural, but when I went to university away from London I discovered that a lot of people found this almost incomprehensible. Strangely, a sample of my friends found that with one exception, only the Jewish ones found it grammatical – though this might be an artifact of the small sample size. They also felt

    (2) Do you want? (when offering something) and

    (3) <Did you enjoy?

    to be grammatical, which I don't. Are these influenced by Yiddish? How natural are they in AmE?

  50. John Roth said,

    November 4, 2010 @ 11:37 am

    I think (1) Are you coming with, and (2) Do you want? are perfectly acceptable in spoken English — as long as the missing complement is obvious in the immediate context. I've used and encountered both of them, and I haven't noticed any association with Jews or Yiddish (my surname to the contrary, I'm not Jewish and I don't know Yiddish). On the other hand, I probably wouldn't notice that kind of association, so it doesn't mean it isn't there.

    However (3) Did you enjoy? is less so. I'm not sure why.

  51. GeorgeW said,

    November 4, 2010 @ 11:49 am


    1 & 2 don't work in my dialect of AmE. 3 would be unusual even in a clear context.

  52. Pflaumbaum said,

    November 4, 2010 @ 11:49 am

    Thanks John. Are you an AmE speaker?

  53. Xmun said,

    November 4, 2010 @ 11:49 am

    (1) is a calque of "Kommst du mit?" or "Kommen Sie mit?"

    (2) I can't imagine any good context for. I'd expect "Do you want to?" or "Do you want [+ direct object]?"

    (3) reminds me of that irritating expression used by waiters when serving food: "Enjoy!", to which I always want to respond: "Don't you mean 'Enjoy it!'?"

  54. John Cowan said,

    November 4, 2010 @ 1:41 pm

    Is near really a preposition, anyway? Prepositions don't normally inflect for the comparative and the superlative. Maybe it's just an oddball adjective that governs a NP.

    (Historically, it was the comparative member of the set nigh, near, next, but these three words have gone their separate ways.)

    [There's very good evidence that it is indeed (sometimes) a preposition: [1] it allows right as a modifier (right near the window), and [2] it allows preposing with a wh-phrase (the window near which I placed it), and [3] it fills the second complement role for put (Put it near me), [4] it acts a non-predicative adjunct (Near the town, there was nothing), and so on. Plus some evidence that it's an adjective. I think it might have to be dually categorized. —GKP]

  55. Xmun said,

    November 4, 2010 @ 2:32 pm

    The CGEL says of "near" that it is "highly exceptional in its syntax, combining a number of adjectival properties with those of a preposition" (p. 609). In modern English its comparative is "nearer" and its superlative "nearest". The nigh-near-next graded set is long obsolete.

  56. Ellen K. said,

    November 4, 2010 @ 2:47 pm

    "Are you coming with?" seems ordinary and unremarkable to me. Me or us is clearly implied.

    Xmun, I believe "Do you want?" would mean either "Do you want this?" or "Do you want it?". With either a non-verbal showing or previous conversation indicating what they are talking about. I can especially imagine this being used when a physical object is shown, with the "this" dropped.

  57. marie-lucie said,

    November 4, 2010 @ 2:49 pm

    I have heard "Are you coming with?" from someone from a part of Canada which has many people of German origin. It is a translation of "Kommst du mit?" (from the verb "mitkommen"), and the person in question did not know that it was not used in other parts of the country.

    JC: The nigh-near-next graded set: I had always thought that there was some ancient relationship between those words, but had never made the morphological connection. Thank you for pointing it out.

  58. GeorgeW said,

    November 4, 2010 @ 5:04 pm

    I think the strangeness of 'are you coming with' relates to the constraints on ellipsis. Radford ("Transformational Grammar," Cambridge University Press, 1988) states the following: "Only VPs (Verb Phrases) can undergo Ellipsis (under appropriate discourse conditions)."

    So, according to Radford's principle, ellipsis of the complement of the preposition would be ungrammatical.

  59. Rubrick said,

    November 4, 2010 @ 5:26 pm

    I felt deeply compelled to post a stranded preposition joke, but on seeing what happened to other wags I decided it was against my best interests. Once again, it's all about pragmatics.

  60. Rasselas said,

    November 4, 2010 @ 6:03 pm

    Granted that your five good and five bad examples are as good and as bad as you say, I still don't think the explanation can lie in the distinction between adjuncts and complements. I find it easy to come up with examples of adjuncts that strand with the same felicity as Lady Bracknell's "good age to be married at".
    Who did you go to the game with? I went to the game with Tim.
    When do you need this by? I need it by Tuesday.

    These preposition phrases are adjuncts, but the prepositions are central prepositions: short, commonly clitic, with no obvious derivation from words of other categories. Such prepositions, I submit, are the easiest to strand.

  61. GeorgeW said,

    November 4, 2010 @ 8:19 pm

    @Rasselas: I think in your examples of Wh-questions, the complement is just fronted to form a question:

    You did go to the game with C(who) > C(who) did you go to the game with

    You do need this by C(when) > C(when) do you need this by

  62. Jangari said,

    November 5, 2010 @ 12:19 am

    'Ungrammatical' is defined in the original post as:

    …being instinctively avoided by native speakers and felt as unacceptable.

    [That was not given as a definition, but only as a rough indication of which concept of grammaticality I meant — the scientific one, not the prescriptive one. —GKP]

    I fully endorse this definition, but it immediately entails that grammaticality is a continuum, both in the sense that people will rarely agree on where the boundary between grammatical and ungrammatical is (as shown in this comment threat) but also in the sense that individuals will judge things as more or less grammatical as a result of being more or less instinctively averse (respectively) to certain forms.

    I think the forms shown in red are all fine under the right pragmatic conditioning. This is not to say they have the same grammatical status as the green ones; likely as a function of high frequency collocation (as someone pointed out early in this thread) they are less likely to invoke instinctive aversion and (under the above definition) would therefore be more grammatical. In the examples that have been pragmatically forced, the reader had been primed to the low-frequency collocation by the context. That is, in the example of the theatre, the section immediately before the 'sit near' set up the reader to the possibiity. But I guess this is exactly what we mean by pragmatic forcing.

    I'm trying hard to think of an example of a stranded preposition that is unequivocally ungrammatical, but I can't.

    [You haven't thought hard enough. Try
    *We should ask about what the hotel is closed because of. This happens to be a case where you can't prepose either: *We should ask about because of what the hotel is closed is disastrous, and *We should ask about of what the hotel is closed because is even worse than that. Sometimes there's no way out! —GKP]

  63. John Cowan said,

    November 5, 2010 @ 1:36 am

    Marie-Lucie: Playing "spot the hidden superlative" in the Indo-European languages is a very enjoyable game. For example, the ordinals first and last are both old superlatives whose positives survive as for(e) and late respectively. The well-known suppletive forms better/best, worse/worst also once had positives, the first of which shows no trace in any Germanic language, and the second survives only in North Germanic. More/most are the comparative and superlative of the lost adjective mo, preserved as late as Shakespeare's day. Erst(while) is the superlative of ere, and little/less/least also fit here. On the other hand, against, amongst, betwixt, whilst are old genitives that owe their excrescent t to association with true superlatives.

    In Latin, we similarly have prae/prior/primus, which shows that the notion of 'first' as a superlative goes back to PIE, and ultra/ulterior/ultimus. More obscurely, there is bru:ma 'winter' < brevis. Auxerre in France probably goes back to an old superlative of whatever the Gaulish version of altus was (though the Latin name was different), as it sits on top of a hill. But the most peculiar use of the superlative ending is in German, where the large ordinal numbers are formally superlatives: as fleissig 'diligent' is to fleissigste 'most diligent', so dreissig 'thirty' is to dreissigste 'thirtieth'. Apparently this is a calque on Gallo-Romance, where the superlative ending -issimus fell together with the pseudo-ending -e:simus in vice:simus, tre:simus as *-iesme.

  64. Cecily said,

    November 5, 2010 @ 5:25 am

    @Pflaumbaum, November 4, 2010 @ 11:05 am: I presume you're not talking about usage in London, England?

    Your examples "Are you coming with?", "Do you want?" and "Did you enjoy?" are not remotely idiomatic to my ears (born, raised and still living within 25 miles of London). I thought they were common in AmE, but GeorgeW says otherwise. I wonder where they do come from?

  65. Pflaumbaum said,

    November 5, 2010 @ 7:54 am

    @ Cecily – Yes, London, England.

    Since are you coming with? is completely, unequivocally natural to me – I've said it a thousand times – and yet seemingly unacceptable to most English people, I'm trying to work out which dialect I'm getting it from. Is it Cockney-influenced North London, my northern parents, or the Jewish aspect, which the Kommst du mit connection suggests might be it – though my memory is that the phrase was common in the eighties at primary school, where there were barely any Jews.

    Do you want? and Did you enjoy, though, both sound odd to me.

  66. Cecily said,

    November 5, 2010 @ 10:21 am

    Intriguing. According to this survey, it is quite common in the US, but not especially popular there:

    Now we need a similar UK survey…

  67. Ellen K. said,

    November 5, 2010 @ 10:45 am

    "We should ask about what the hotel is closed because of" is fine for me, though I'd be more likely to say "We should ask why the hotel is closed".

  68. groki said,

    November 5, 2010 @ 12:49 pm

    We should ask about what the hotel is closed because of.
    ? We should ask about because of what the hotel is closed.
    * We should ask about of what the hotel is closed because.

    for me, (1) is grammatical, (2) is questionably grammatical yet charming to my mouth/ear, and (3) is ungrammatical and ugly.

    but in honor of GKP's * on (1), a dialog:

    –"ask not about what the bell tolls because of."
    –"since it's me?"
    –"nah: it's just ungrammatical."

  69. groki said,

    November 5, 2010 @ 12:52 pm

    @John Cowan: thanks for the superlative reveal of hidden superlatives. I take it that mo meant "some"?

  70. Peter G. Howland said,

    November 5, 2010 @ 1:19 pm

    I “get the point” Dr. Pullum. As I stated in my opening sentences, I don’t have a problem with accepting a preposition stranding structure as long as it communicates clearly. As your examples succinctly show, it is both natural and functional in many instances and I have no interest in perpetuating the “myth” that there is something inherently wrong with the structure. You seem to have ignored a key phrase, “…not having other pragmatic sensibilities…” in my closing paragraph. My point, as depressingly wall-like as it may seem to you, is simply that “immature” structures can easily be avoided, and it is these with which I have an issue.

  71. John Cowan said,

    November 5, 2010 @ 2:27 pm

    groki: Perhaps originally, but it always seems to mean 'more' in even the earliest English. The OED connects it with the obsolete English adjective mere 'famous, beautiful, excellent', but 'big, large' in other languages. This has nothing to do with the modern adjective mere, which is from Latin.

  72. L'Esprit de l'escalier said,

    November 5, 2010 @ 5:05 pm

    Elizabeth Bennet said

    For what do we live, but to make sport for our neighbors, and laugh at them in our turn?

    but in Present-day English the idiom "what … for" always strands the preposition. In informal speech it may be the commonest way to ask for an adjunct of reason or purpose. So I ask:
    What is Geoff Pullum questioning preposition stranding in adjuncts for?

  73. Peter G. Howland said,

    November 5, 2010 @ 6:49 pm

    @Pflaumbaum, et. al
    In a wide variety of settings, from boardrooms to barrooms and beyond throughout the western U.S., I have heard the “You want to come with?” phrase for several decades. As the structure is not natural to me (and please, I have no problem with the preposition with which it ends!) I don’t use it. But I have tried to observe carefully to determine the settings and circumstances in which it is uttered, and they seem to be typified by the following example:

    “Me and her are going. You wanna come with?”

    This is said by those who know perfectly well that “Me are going” makes no grammatical sense, but who don’t want to sound “snooty” (or whatever reason their trenchant aversion to proper speech is based upon) by saying “She and I” or “Carol and I”.

    It also seems that this clipped structure is most often used by (Oh my! You folks are going to want to stone me to death for this one!) adequately educated females in the 16 to 36 age group. I have also noted it in speech by modestly educated males of any age group who might on occasion say “You wanna go with?”, but this is silently followed by the implicit “them guys?”

    It appears that the phrase is used most often by those who would prefer to effect an apathetic construction in order to sound cutely casual and cool. It’s simply a culturally cultivated habit who’s perpetuation is one of the ways in which language usage evolves. This is, by the way, something I can readily go along with! (Howland said, casually stranding his naked preposition for all the cool people to see.)

  74. Xmun said,

    November 5, 2010 @ 7:56 pm

    Does anyone still use the word "withal"? If so, how?

  75. groki said,

    November 5, 2010 @ 8:31 pm

    @John Cowan:

    so I guess mo / more / most meant "more" / "even more" / "as much more as possible."

    also (willfully ignoring some years of temporal displacement), mere (obs.) and mere (mod.) are another kind of cleave-like auto-antonym pair.

    cool, thanks.

  76. Pflaumbaum said,

    November 5, 2010 @ 8:54 pm

    @ Peter G. Howland

    Me are going indeed makes no grammatical sense, but me and her are going is a different matter. That's because the so-called accusative is the default form of the pronoun in English and is natural in all situations except when it is the non-coordinated subject of a finite verb. There's plenty of proof of this, see CGEL for examples. It's true that CGEL regards it as non-standard in co-ordination, but see for instance Thomas Grano's Stanford thesis which argues that this is to a large extent the result of prescriptive forces.

    As for Are you coming with, for this speaker (though I'm Southern Standard BrE – apart from in this respect it seems!) there's no element of trying to be cute, it's the normal way of saying it.

  77. Army1987 said,

    November 6, 2010 @ 8:31 am

    I think there are a preposition near and an adjective near (and maybe an adverb, as well?) with the same meaning, and in some cases they are impossible to tell apart (much like it's impossible to say whether the fast in She is going fast is an adverb modifying the verb or an adjective acting as a predicative). For example, near can take a complement (near the valley) or inflect for comparative (nearer) but not both (*nearer the valley).

  78. Army1987 said,

    November 6, 2010 @ 8:33 am

    To expand on the last post: note that in none of the four examples in GKP's comment to Cowan's post you could use nearer or nearest.

  79. Pflaumbaum said,

    November 6, 2010 @ 8:41 am

    @ Army1987

    What's wrong with nearer the valley?

    "They decided to pitch camp nearer the valley."

    Sounds fine to me (see post above for dialect).

  80. Army1987 said,

    November 6, 2010 @ 8:47 am

    @Peter G. Howland:
    So, in Where are you from?, does the stranded preposition have "other pragmatic sensibilities", is it "immature", or what?

  81. Jens Fiederer said,

    November 7, 2010 @ 10:43 am

    I think a rather interesting point of Howland's has been lost in the midst of the usual brouhaha.

    I would be willing to bet that "maturity" really DOES make a difference in what is grammatically accepted and/or produced (not in the pejorative sense of referring to an adult usage as "immature", but in the intended sense of a child's learning the language), quite possibly including the stranding of prepositions, with an adult's set of possibilities considerably more restricted than a child's.

    This might be old hat to professional linguists, but us regular joes haven't read much about that sort of thing outside of "How your baby is growing up" books.

  82. James Wimberley said,

    November 7, 2010 @ 7:44 pm

    GKP: ¨What exactly makes the examples in the green column above seem less felicitous than the ones in the red column has yet to be identified …¨
    What if some of us find the green column more felicitous? Enzo will be very busy.

  83. Dionne said,

    November 8, 2010 @ 9:03 am

    "Are you coming with?" is a very South African phrase, and it could be that you are surrounded by ex-pat Saffas in London. I always thought it came from Afrikaans people directly translating "Kom jy saam" which is perfectly grammatical, but now, maybe not. Same thing for "Do you want?".

  84. J said,

    November 9, 2010 @ 1:26 pm

    Regarding hidden superlatives, the positive form of better/best does show up here and there. In older High German, you see the adverb "baz", which pretty much just means "better". There is also the noun "boot" as in "to boot", from the same root, albeit with a different ablaut grade. It also shows up in older German (MHG "buoz" and interestingly enough in the expression "ze buoz" = "as a remedy", IIRC).

RSS feed for comments on this post