Is this a great photo or what!

« previous post | next post »

That was the start of my heading-comment on a photo of my son Dave. Ensuing back-and-forth on Facebook between me and Andy Rogers (with a relevant interpolation from my son Morriss):

Andy Rogers: Shouldn't it be "or what?"?
Barbara H Partee: I punctuated it as I would pronounce it! Maybe if it was somebody else I might right "or what?".
Andy Rogers: Seems syntactically like a question.
Barbara H Partee: That's true. Well, but how would you punctuate an annoyed "Will you stop that!" It's also a question, but it's pronounced as an imperative. Maybe "Will you stop that?!" Maybe that's what some of those double punctuation marks are for — I've never seen them discussed (but haven't really looked — it's not a category I normally think about.) So maybe we could agree on "or what?!" ?
Barbara H Partee But I have to confess that when I made the original post, a question mark never even entered my head.
Morriss Partee: Would you stop arguing about punctuation or what?!?!??!??!???!!!!?!???!!!??!??!
Morriss Partee: ;)
Andy Rogers: So what IS the relationship among syntactic form, whatever is going on in your ! examples, and punctuation?
Barbara H Partee: ‎(Sorry, Morriss, but wasn't it always like this at the dinner table? Should make you nostalgic!) Andy, I don't know, but somebody must. Maybe I should put a little query-post on Language Log and see what turns up.

So comments are open because I really don’t know! In this domain I’m just a naïve native writer of English, with ordinary education about prescriptive grammar, but they never taught us about what might be called “colloquial punctuation” (maybe it has a name, I don’t know that either.) I wonder if comic strip writers study colloquial punctuation somewhere, or if they just pick it up by paying attention to what other comic strip writers have done. If it’s been studied at all, I’m sure Facebook must be one good corpus-source.


  1. Ross Presser said,

    May 18, 2012 @ 12:01 pm

    My totally unsubstantiated theory: You're not asking a question at all. "or what" has been turned into an emphasis particle. You are very enthusiastic about the quality of the picture, and expect no one to disagree with your assessment; hence the exclamation.

  2. Andrew said,

    May 18, 2012 @ 12:04 pm

    It should have a question mark or an interrobang (which is the name of this thing: ?!, which is not two pieces of punctuation, but a single unit). While its pragmatic function is declarative or even interjective, syntactically, it is an interrogative, and punctuation follows syntactic, rather than pragmatic, cues.

  3. maxh said,

    May 18, 2012 @ 12:10 pm

    I generally use !? (though I've been told it should be ?! (is there nothing too trivial to make rules for)) because a) I don't have an interrobang key, and it wouldn't be recognised and b) I like how it looks.

  4. Dick Margulis said,

    May 18, 2012 @ 12:15 pm

  5. Jonathan Lundell said,

    May 18, 2012 @ 12:19 pm

    A very brief session with the Google suggests that ? predominates, with an occasional ??? and very occasional !. It does seem like the the kind of context ‽ was born for, though.

  6. Jonathon said,

    May 18, 2012 @ 12:24 pm

    There's no firm answer—it's basically just a matter of style. The Chicago Manual of Style 16th edition has this to say:

    6.72 Exclamation rather than question
    A sentence in the form of a direct question can often be styled as an exclamation simply by using an exclamation point rather than a question mark.

    How could you possibly believe that!
    When will I ever learn!

  7. Stan said,

    May 18, 2012 @ 12:26 pm

    Coincidentally, David Crystal wrote about this yesterday.

  8. Ray Girvan said,

    May 18, 2012 @ 12:29 pm

    @Ross Presser: has been turned into an emphasis particle

    I agree. As in "It's a great photo and how!"

  9. John Lawler said,

    May 18, 2012 @ 12:30 pm

    I think Andy's right about the comma, at least. I am one of those who subscribe to the Intonation Theory of comma placement, though I believe Geoff Nunberg disagrees.

    In this case, whether it represents an intonation or not, adding a comma sorts out the tagginess of the or what phrase, and points to the fact that it's an afterthought, not necessarily even part of the sentence, and not subject to the same rules. Including punctuation rules.

    As for the use of interrogation or exclamation point(s), I know of no rules worth following, except "Please Yourself," though Lewis Thomas has more to say on the topic.

  10. Nebu said,

    May 18, 2012 @ 12:40 pm

    Depending on with whom I am conversing, I sometimes adopt an idiolect involving heavy use of rhetorical questions. In such a circumstance, I often punctuate those rhetorical questions with exclamation marks. E.g. "Why didn't you share any with me!" or "How come it's raining!" or "Are you ready yet!"

  11. circadianwolf said,

    May 18, 2012 @ 12:44 pm

    "Maybe if it was somebody else I might right 'or what?'."


    Interrobangs are cool, but I never see the singular character used in practice (though I see plenty of ?!).

  12. Glenn Bingham said,

    May 18, 2012 @ 12:57 pm

    Tempted as I was, applying a style manual to a slang expression seems incongruous. Why shouldn't you do it however it floats your boat!

  13. Ken MacDougall said,

    May 18, 2012 @ 1:03 pm

    Bring back the interrobang.

  14. Ben Zimmer said,

    May 18, 2012 @ 1:03 pm

    Compare the form "How <Adj> is THAT" discussed in my 2006 LL post:

    In a discussion on the American Dialect Society mailing list, Larry Horn noted that the expression tends to have a heavy stress on the final word "that," with a falling intonation at the end. This is best represented as "How <Adj> is THAT" (and not at all well represented by "How <Adj> is that?", which implies a rising intonation at the end of the utterance).

    In the same ADS-L discussion, Seán Fitzpatrick mentioned that such "question-like inversions spoken as statements" were common on the show Friends, giving the example "Could <substantive> BE more <adj>?", e.g., "Could you BE more ridiculous?"

  15. blahedo said,

    May 18, 2012 @ 1:28 pm

    I feel like there are a number of prosodic cues used in spoken English that are difficult to transcribe, because the utterance is syntactically a question but a question mark would normally indicate a very different intonation pattern. There's the low-flat "What" that when it is transcribed in e.g. comics usually gets written as "What." or even omitting the period. There's a descending-pitch exclamation form of "What on earth." (or "What the heck.", etc) that carries a somewhat different connotation than the rising-pitch "What on earth?"

    My general rule is, I use period or exclamation even if it looks funny to do so, if it would look even funnier to use a question mark….

  16. Svafa said,

    May 18, 2012 @ 1:47 pm

    I'm going to agree with Ross, she's not asking a question, despite appearances. I'd say the entire structure "Is this… or what!" fits that description. The "or what" seen here is very likely related to the "or what" seen in questions like "Do you want to go somewhere or what?", but this usage is different as the reader/listener is not being asked to consider anything.

    I'm not sure this deserves an interrobang, though I don't think it would do any harm either. Typically, the interrobang is used when asking a question emphatically or asking a rhetorical question. I'm not sure this example of "or what" qualifies as either (though my "Do you…" example might).

    As to writing the interrobang, ?! seems more common, but !? is still acceptable. I tend to use ?! most often, but use !? if a statement turns into a question partway through (i.e. a declaration made without surety of its validity).

  17. Aaron Toivo said,

    May 18, 2012 @ 1:47 pm

    I think a sensible compromise is: punctuate Standard English according to the usual prescribed rules, and punctuate colloquial English in the manner that best conveys your intended intonation. Intonation is relatively unimportant in the written standard, whereas it's crucial to colloquial language and sometimes even the main thing one wants to convey, and I can't see any good reason to let the rules of formal writing stand in the way of expressiveness in informal writing.

  18. Rod Johnson said,

    May 18, 2012 @ 1:52 pm

    Given that the number of distinct prosodic cues that people use in interpreting speech is probably not 3, it seems unlikely there will be a universally successful mapping to our particular set (. ? !, and I guess people from Céline to the internet have added …).

  19. Adrian said,

    May 18, 2012 @ 1:59 pm

    If it's a question, it has a question mark. If it's *really* a question. Not if it has the *form* of a question. I've had similar discussions about so-called tag questions, for example: "He did really well, didn't he." This isn't a question (unless you really are doubtful), so it doesn't have a question mark.

  20. Tracy said,

    May 18, 2012 @ 2:06 pm

    The single-character interrobang looks clunky to me.

    Is this a widespread problem? Are there any languages with writing systems that have punctuation specifically for a type situation?

  21. Brett said,

    May 18, 2012 @ 2:07 pm

    People (including myself) seem to react more strongly to things that appear structurally to be questions but which (pragmatically speaking) are not interrogatives and are not punctuated as interrogatives. (It seems they are usually punctuated as exclamations instead.) In comparison, sentences that are have the structure of statements but are actually questions and are punctuated as questions (for example, "That is what you are telling me happened?") are completely unremarkable beyond elementary school. The second type shows that word structure is not really sufficient to identify the sentence type unambiguously.

  22. MattF said,

    May 18, 2012 @ 2:15 pm

    If it's a question, then what is the answer? Or should I say "Then what is the answer?!" Or what?!?!

  23. Aaron Toivo said,

    May 18, 2012 @ 2:24 pm

    @ Rod Johnson:

    Most punctuation marks are prosodic cues of some sort, generally intonations and pauses that are normal for the syntactic positions punctuation is used in. For example commas correspond to a lowered pitch followed by a brief pause, and the dash – I don't know how to characterize it, but it's certainly a quite specific effect. A pause that preserves the existing pitch contour, something like that.

  24. Jerry Friedman said,

    May 18, 2012 @ 2:46 pm

    @Tracy: ¿No es suave la forma que se usa en español!

  25. Joe said,

    May 18, 2012 @ 2:53 pm

    I wonder if there are two different ways "or what" can be used, one where it is an interrogative, and one where it is an exclamative. In "is this a great photo, or what," I would say "or what" is an exclamative. It feels more natural for me to think of this as stating "What a great photo this is!" than "Is this a great photo or isn't it?" Exclamatory "or what" follows an interrogative in the same way "and how" follows a declarative. I'm not so sure, however, whether "or what" is an exclamative in a construction like, "Do you want to hear this story, or what"? Here, I think the only available sense is, "Do you want to hear this story, or not?" But I can't figure out a syntactic justification for these analyses. (I'm using a comma before the coordinator because I think of "or what" as a verbless clause).

    I would be inclined to use a question mark only with an interrogative, but it does look a bit odd to me so I imagine a question mark after exclamatory "or what" is not uncommon.

  26. leoboiko said,

    May 18, 2012 @ 2:55 pm

    Everyone interested in this topic should read Nunberg’s The Linguistics of Punctuation. I wish it was more known.

  27. Kyle Rawlins said,

    May 18, 2012 @ 3:33 pm

    It is a question, but an alternative question, which in general end with a falling tone, not a rising tone (cf. Bartels' dissertation etc). In fact I think the intonation on "or what" questions is simply not distinguishable from the general contour on alt Qs. As an added twist, your example is a rhetorical use of an "or what" question, and so acts even less question-like, hence people suggesting that it might be an exclamative etc. (These cases also have a similar scalar component to exclamatives.) I think non-colloquial spelling conventions involve just a "?" on rhetorical questions. Colloquial spelling allows for interrobangs etc. on rhetorical questions, as people have suggested.

    Maria Biezma and I have a recent talk on "or what" questions that substantiates some of the claims I just made, slides can be found at:

  28. Mark F. said,

    May 18, 2012 @ 3:33 pm

    When I was a kid I caught the teacher in an error (or, perhaps, an "error"), as she has punctuated the questions on a test with periods rather than question marks. But, later, when I was writing questions of my own, I could see how it happened. Questions in a list on a test are real questions, but they aren't "prosodically punctuated" as questions, and so it felt somewhat natural not to punctuate them on paper as questions either. (But I think this is a case where a question mark really is what you want, but there are just countervailing tendencies.)

  29. Morriss Partee said,

    May 18, 2012 @ 3:37 pm

    First of all, I want to say "thank you?!" for continuing to use me and my brothers as linguistic guinea pigs, mom.

    However, it is also with great pride that my feeble attempts at interrobang humor are now captured for posterity in Language Log, and no, that comment is not sarcastic.

    Having ten years of experience as a typographer/graphic designer, and priding myself on the use of little-understood punctuation (such as proper quotation marks, apostrophes, ellipses, and ampersands), I am disappointed in myself that I was not familiar with the interrobang as a single symbol. I want to make two points on the interrobang as single mark:

    First, typographically-speaking, this symbol looks horrible at small sizes. The exclamation part of the symbol jutting into the counter space of the question mark is simply obnoxious at anything less than 24 points in size, perhaps a minimum of 64 points should be used. And this makes sense since it seems that the mark began in the advertising world… this mark makes perfect sense for use in a headline, where it will be large, and space is at a premium. If only there were some way to strongly discourage its use in anything less than 18 point type, since the single mark, when displayed at small sizes, looks like an inkblot ruined an otherwise perfectly good question mark. Some fonts provide a more open-looking interrobang, such as this example of Palatino's, in wikipedia, but note that we are looking at a sample which is likely least 144 point in size.

    Secondly: This.

  30. Svafa said,

    May 18, 2012 @ 4:51 pm

    @Kyle Rawlins: I'm not certain the "or what" here is a rhetorical question. Typically, rhetorical questions ask/expect the listener/reader to reflect or consider something. There is no expectation of consideration here. It's more akin to saying "Look at this great photo", and not even asking for a silent affirmation to "Don't you like this great photo?". Maybe that still qualifies as a rhetorical question though, but I've always thought of them as a means of reflection/consideration.

  31. John Lawler said,

    May 18, 2012 @ 5:15 pm

    Depending on the intonation (natch) and also the pragmatic context, or what can represent quite different speech acts.

    Viz, Barbara's sentence, where it clearly is a rhetorical question of some sort, vs

      A: Have this on my desk by tomorrow morning.
      B: Or what?

    where it's an aggressive challenge — not at all the effect Barbara was after.
    And if I can think of this distinction, others can no doubt think of a dozen more.

  32. phosphorious said,

    May 18, 2012 @ 6:49 pm

    Andrew @ 12:04 said:

    "and punctuation follows syntactic, rather than pragmatic, cues."

    Is that a thing? Or are you just creating a rule on the spot. Well played if you are!

    But I would have thought that punctuation is a matter of tone, therefore pragmatics.

  33. John Lawler said,

    May 18, 2012 @ 6:53 pm

    @phosphorious: German punctuation follows syntactic rules, but English punctuation doesn't. Or, more correctly, different punctuation marks in English follow different kinds of rules, many of them contradictory.

  34. Ray Girvan said,

    May 18, 2012 @ 8:46 pm

    @phosphorious: Is that a thing?

    It's a thing, but not an exact one. Whatever the MLA may say, we are still in the transition between rhetorical punctuation ( punctuation representing the rhythms of spoken English, as done by Jane Austen) and punctuation as logical markup.

  35. Rick Sprague said,

    May 18, 2012 @ 9:44 pm


    Not all inverted sentences are interrogatives, don't you know?

    – or –

    Not all inverted sentences are interrogatives, don't you know.

  36. Duncan said,

    May 18, 2012 @ 10:55 pm

    Knowing the context, a parent remarking on a photo of their son (yeah, "their", plural singular, nothing like writing an LL comment to bring out the meta self-consciousness!), I believe an exclamation point is appropriate.

    Certainly in that context, I read that "rhetorical question" as a command, in the form of a question with an implied threat.. You WILL agree with me or I won't be talking to you for awhile (or on a blog, your comment will be deleted, etc). Think of the old folk wisdom about not getting between a mother bear and its cub. That's the (rhetorical, thankfully, at least in a civilized society) level of implied threat and as such, it deserves an exclamation point.

    That's definitely how I'd interpret it in the given parent displaying a picture of their kid, but the same implied threat at a rather lower level applies, in my mind anyway, to, for instance a "tween" gushing about a photo of her latest heartthrob (preferrably with her in it) : "Is this a great photo or what!" Addressed to a peer, that really /is/ a threat of "or what" meaning I won't be talking to you for a week (or will be de-friending you on FB or whatever) if you disagree.

    That's what an exclamation conveys in this context, to me as a reader. A question mark wouldn't have the same implied threat, with the contrast implying a level of indifference — maybe what the friend who really doesn't care but knows they'll get de-friended if they don't agree might put it, if they echo the photo and caption.

    The ?! combo (either order) would be acceptable in either case as well, and would be what I'd be inclined to use. The implication there is a somewhat more subdued "polite" threat, accepting that it is a question and of society's rules for questions, thus by implication, an acceptance of possible disagreement, if only to be polite about it, or elsewise, agreeing with a friend, to be polite about it, thus this is more ambiguous.

    The order I'd use would depend on which I wanted to stress, the question or the exclamation bit, with the "important" one first. If I wanted to express enthusiasm without the implied "you BETTER agree!", I'd put the question first, conversely if I wanted the implied threat.

    But the combined form "interrobang"? This discussion is the first I've seen of it AFAIK. It /does/ make sense, but I agree with MP's opinion of the aesthetics at "ordinary" print sizes and thus would be unlikely to use it myself even if it /was/ easily found on my keyboard. Plus it doesn't have the additional flexibility of being able to choose which one comes first. (Of course, me being a Linux geek, you'd be more likely to impress /me/ with a "shebang", #!/bin/sh, or a reference to "sudo", or the like. See for a classic comic commentary on the second. That's a true geek's form of the same level of implied threat, only to a geek that makes it explicit! =:^)

  37. Ray Dillinger said,

    May 18, 2012 @ 11:54 pm

    In modern English I have always regarded punctuation as encoding intonation and rhythm. I would read the sentence as being intoned differently with different punctuation marks.

    Trailing "or what" after a syntactic question is rarely intoned as a question. In cases where it is, it should get a question mark. In the usual case however, an exclamation point or a period is indicated.

    Some people who read my short stories are puzzled by my writing dialog where some characters use serial (Oxford) commas and some do not, whereas outside of quoted dialog I use them uniformly. It's the same kind of principle at work. I'm just writing the characters as having different topolects or speech rhythms. And doing so successfully I think, because more people identify the characters correctly as "belonging" to a particular region or group, without necessarily even noticing exactly why, than are puzzled by the "inconsistent" punctuation.

    When writing I take every device I can get to deepen characterization. I stop short of using "eye dialect" however, because whether or not people speak using those pronunciations, they do not write using those spellings. I don't want to produce something that will appear alien, and possibly offensive, to the groups my characters belong to.

  38. C Thornett said,

    May 19, 2012 @ 1:17 am

    Should 'or what' be classed as an informal question tag? If so, that would suggest a question mark, but as Adrian suggests, none of the question tags (isn't it / didn't you) are always spoken or intended as questions. This loose group of expressions seems to arise from speech, and it seems natural to punctuate accordingly, pace all the ESL/EFL textbooks which indicate a question mark. Written contexts for these tags seem to be predominately informal, so going with intonation and meaning seems acceptable.

  39. DrSAR said,

    May 19, 2012 @ 1:45 am

    I find it interesting the that the order of ? and ! in the interrobang when used in chess notation matters:
    !? an interesting move (that may not be best)
    ?! a dubious move (one which may turn out to be bad)
    (source wikipedia and some distant memories)

  40. John Walden said,

    May 19, 2012 @ 2:42 am

    “What ho!" I said.

    "What ho!" said Motty.

    "What ho! What ho!"

    "What ho! What ho! What ho!"

    After that it seemed rather difficult to go on with the conversation.”
    ― P.G. Wodehouse

  41. Adrian said,

    May 19, 2012 @ 4:31 am

    Ray: "we are still in the transition between rhetorical punctuation ( punctuation representing the rhythms of spoken English, as done by Jane Austen) and punctuation as logical markup"

    If this is a trend, it should be resisted. Especially if it's a trend being promoted by teachers of English as a foreign language, many of whom are clearly more interested in making their job easier than in understanding/teaching reality.

  42. languagehat said,

    May 19, 2012 @ 7:46 am

    It is a question, but an alternative question, which in general end with a falling tone, not a rising tone […]. In fact I think the intonation on "or what" questions is simply not distinguishable from the general contour on alt Qs.

    Thank you! I was reading down the thread, horrified that no commenters on a post at Language Log, of all places, seemed to be aware that wh– questions do not take "question intonation." There is no difference between the intonations of "Is this a great photo or what?" and "What is this?" and therefore I see no reason on earth why a question mark would be problematic.

  43. Chris Cooper said,

    May 19, 2012 @ 7:54 am

    The title of the BBC-TV comedy news quiz "Have I Got News For You" is officially punctuated that way on BBC websites. This distresses me greatly: for me it just has to be "Have I Got News For You!" But more distressing is to read "Have I Got News For You?", which I've seen.

  44. Rod Johnson said,

    May 19, 2012 @ 9:14 am

    @languagehat: thanks for pointing that out. It's one of the things I was getting at with my (overly terse) comment way back there.

    What this discussion suggests to me is that "question" isn't really a semantic/pragmatic category but a morphosyntactic one. The set of speech acts expressed by sentences with question structure is so heterogeneous, and the divisions between them are so subtle and vague, that it makes me think that "question" is one of those naive folk categories that doesn't really stand up to analysis, except at the level of subject-verb inversion.

    It's tempting to imagine that in some Garden of Eden state the question construction meant some well-defined set of things, and it has since accrued a lot of pragmatic baggage like Barbara's original example, but it seems more likely things have always been this murky. At any rate, it shouldn't be surprising that punctuation is applied in a pretty slapdash way. The idea that punctuation could ever be truly "logical" seems like a pretty naive hope to me unless we have a pretty impoverished idea of the semantics underlying that.

  45. Barbara Partee said,

    May 19, 2012 @ 9:30 am

    Thanks, everybody (and don't stop) — I'm learning a lot. Probably we all are. When I'm back where I have higher-speed internet access I'm going to check out the suggested references, like Kyle Rawlins' and Maria Biezma's talk, and I'm certainly going to get Geoff Nunberg's book — I'm pretty sure I had heard of it before but I had never managed to get interested in punctuation before!
    And @Language Hat, yes, I know that wh-questions don't have rising intonation, but I still feel a difference between the way I pronounce the exclamation I had in mind and the way I would pronounce the same string as a question (as in a question like 'Did he get fired or what?'). When I wrote that exclamation, a question mark never even occurred to me. I think this is all really interesting and more subtle than I had anticipated — phonology, syntax, semantics, pragmatics all involved, plus the differences between formal writing and writing to represent colloquial spoken language.

  46. Ed said,

    May 19, 2012 @ 10:03 am

    i'm writing my dissertation on imperatives and i've had to stipulate up front that only sentences which are syntactically imperative get punctuated with ! in my examples. for quoting examples from other sources, who are more liberal with their exclamation points, it gets messy. my best solution for those currently is (!)

  47. Ed said,

    May 19, 2012 @ 10:36 am

    sorry to double-post, but i wanted to comment on the notion that these might be exclamatives. to me, Barbara's example doesn't seem equivalent to an unadorned exclamative in a language that has more robust syntactic encoding of them, such as Italian Che bella foto! "What a nice picture!" (An exclamative with a tag question might come close — Che bella foto, no? — but that's a non-native judgment.)

    a quick skim through some recent papers on exclamatives by Paul Portner and Raffaella Zanuttini didn't turn up any discussion of the "or what?" construction, but they would be the people to ask for a definitive answer.

  48. Joe (same as above) said,

    May 19, 2012 @ 12:52 pm

    I'm just thinking aloud, and I need to read the link, but I'm just not sure if I would consider "Is this a great photo or what" to be an alternative question. I can see how it could be used rhetorically in a different construction. Something like "Are you my father, or what?" (said to the speaker's father) could be rephrased as "Are you my father or aren't you, withe the presuppostion that my father isn't behaving in a manner consistent with being a father. But to me, at least, Is this a great photo or not" would not be rephrased as "Is this a great photo or isn't it." it could be rephrased as "Is this a great photo or is this a great photo." but I wouldn't consider that to be a question, rhetorical or otherwise. It might be seeking agreement, but I think that agreement with the speaker's attitude, not the statement concerning the photo (whereas the question about the father does question the properties of my father). Or perhaps not. I need to read the links and think more carefully about the issue.

  49. Barbara Partee said,

    May 20, 2012 @ 3:19 am

    @Rod Johnson,
    I was with you until you started sounding as if you thought the heterogeneity of uses was a bad thing. 'murky' etc. I wouldn't say these uses won't stand up to analysis, but rather that the analysis is going to be pretty non-trivial, and will require insights from a lot of fields from phonetics to pragmatics to sociolinguistics to orthography. Realizing that in the course of this post is why I suddenly consider punctuation an interesting field, not just something where prescriptivists and descriptivists throw darts at each other, and I'm eager to get hold of Geoff Nunberg's book. Sounds like it could well be one of those things you can offer as a project topic to your brightest students in Linguistics 101, the kind to whom I used to suggest Gleitman and Landau "Language of the Blind Child" and I forget what-all else, books that are simultaneously accessible, enjoyable, and a turn-on to linguistics.

  50. Rod Johnson said,

    May 20, 2012 @ 9:09 am

    Barbara: Bad thing? Not at all. By "murky" I just meant more complex than our naive analytical apparatus of "question, statement, command, exclamation, etc." can describe. Sorry for not articulating that better.

  51. Kyle Rawlins said,

    May 20, 2012 @ 12:52 pm

    I finally had a few moments to respond to some of the responses to my comment. Sorry for the length, but these things are complicated and interesting! We are planning a paper on this topic that will be more readable than just slides, but alas it is not close to being ready.

    @svafa "Typically, rhetorical questions ask/expect the listener/reader to reflect or consider something."

    I'm not familiar with the notion of rhetorical questions that you presuppose (perhaps it is from language teaching, or literature?), but I think you may have in mind a subclass of what the linguistics literature (sometimes) calls rhetorical questions; in particular, the class that Searle dubbed 'exam questions'. (This is not a particularly great name but it is the only semi-standard one I'm aware of.) These are questions where the speaker already knows the answer, and so isn't genuinely seeking information, but wants to find out if the hearer knows the answer, or provoke some other similar response. In linguistics, though it is isn't really a settled matter exactly what a rhetorical question is, this is not really considered a prototypical type of rhetorical Q. Rather, rhetorical questions are taken to be questions where basically, the question itself implies what the speaker thinks the answer is or might be. The particular theory of rhetorical questions we were adopting (due to Caponigro and Sprouse, in a 2007 Sinn und Bedeutung paper "Rhetorical questions as questions"), in treating "or what" questions is one that I think more or less sums up the current line of thinking in this literature. The basic idea is pretty simple — "A question is interpreted as a rhetorical question when its answer is known to the Speaker and the Addressee". They work this out in somewhat more detail. In terms of the pragmatic "force" of rhetorical questions, on this kind of account they act very much like assertions or even exclamations. As I said, this isn't the last word on rhetorical Qs, and C&S' account has some problems, but I think it is a very clear statement of where the field is at. (See our slides / their paper for some other citations if you are interested.)

    C&S are dealing with examples like "I don’t think we should have Onavi on our short list. (After all,) what does he know about semantics?" (their ex. 1), where the question implies that the speaker thinks Onavi doesn't know anything about semantics. This has in common with exam questions that the question is not information-seeking, but not in common that it inherently implies what the speaker thinks the answer is.

    Non information-seeking uses of "or what" (e.g. "is he an idiot or what?"), you can probably see, fit this definition very well. The puzzle we were trying to deal with is why — since they seem to be syntactically indistinguishable from more information seeking "or what" questions, e.g. "Are you coming to dinner or what?"

    @Joe "I'm just not sure if I would consider "Is this a great photo or what" to be an alternative question."

    You are right that the argument needs to be made that these are alternative questions, and this is part of what we were doing in the talk. Here is a sketch of the argument. Basically, we looked at the characteristic properties of alternative interrogatives/questions, and tested "or what" questions with these properties. Aside from the "what", "or what" interrogatives seem to be intonationally and syntactically indistinguishable from alternative interrogatives. They are marked with "whether" when embedded, trigger raising of an auxiliary in root clauses, allow arbitrary numbers of disjuncts before the last one, and take the characteristic alternative question intonation. An attested example (from COCA) of an embedded "or what" question is: "At first, Miina couldn't tell whether the boy was playing a trick on her, or was drunk, or what." I didn't spell out in my previous comment what pitch contour I had in mind, but what this involves is a pitch accent on non-final disjuncts, and a falling tone on the final disjunct. A similar contour is found in lists in general. In information-seeking uses they can have very similar, though not identical, pragmatics and uses to "or not" alternative questions, and alt Qs in general. In fact, we couldn't even come up with any hypothetical arguments that would demonstrate that the _weren't_ syntactically and semantically alternative questions, though I would welcome any arguments to that effect. (Note that arguments from the _pragmatics_ are not really actually very effective in understanding the structure. This is because questions, and in fact all speech acts, are incredibly flexible, and so the point that construction X in context Y has a use that does not match its apparent pragmatic force, is not in and of itself an argument about the syntax and semantics of that constructions, but rather something that needs to be analyzed given the syntax and semantics.)

    Also, in the [very small] sample of languages we have looked at, there is a question type that closely matches the English case, and adopts an alternative question-like structure. The point of variation, in this sample (mostly romance languages), actually seems to be in the correspondent of "what".

    Our analysis is that they are alternative questions used when the speaker wants to leave the full set of options open, where normal alt Qs usually close them off. Consequently the null hypothesis is that they have the syntax and semantics of alternative questions — the linguistic puzzle being how to analyze the differences that show up in their use.

    You mention redundant alt Qs like "Is this a great photo or is this a great photo." These are also clearly alternative interrogatives in every discernable respect, but they are rhetorical in the sense that I defined above — they imply their own answer (but not necessarily rhetorical by other notions of that concept that people have been assuming in this thread). In fact, our proposal is that rhetorical uses of "or what" questions amount to asking a redundant alt Q, for various reasons that I won't go into here. We then gave an analysis of redundant alt Qs that explains why they act rhetorical in the linguistic sense, and how this follows from the semantics of an alternative question.

  52. Kenny Easwaran said,

    May 20, 2012 @ 1:35 pm

    I've encountered the opposite confusion sometimes. I wonder if this is common? Sentences beginning with "I wonder" often seem to get punctuated with a question mark, even though syntactically they're not questions. (This was a pet peeve of Eszter Hargittai on Crooked Timber a few years ago:

  53. Kyle Rawlins said,

    May 20, 2012 @ 2:16 pm

    Also, on the punctuation issue, which is very interesting. As far as I can tell there are something like four main candidates for what should determine a choice of '?' vs '!' vs '.', and I'm actually getting kind of interested from this thread in throwing up some simple mechanical turk experiments to see what people's intuitions are, in a more systematic way.

    1. the syntactic/semantic properties of the main clause, i.e. is it an exclamative, interrogative, or declarative. My understanding of convention is that this is actually how standard formal written English works, more or less. But interestingly from the discussion in this thread, it is clear that people's intuitions don't match this consistently.
    2. the pragmatics of a sentence in context — how it is used, and what "force" does it have. I.e. a syntactic question with a force similar to an assertion might receive a ".", or a declarative sentence used questioningly might get a "?"
    3. the intonational pattern. The basic intuition in some responses here is that rising intonation = "?". This is probably right in some cases, but as @languagehat clearly articulated, it breaks down for "wh"-interrogatives (which have flat or falling intonation), and it also breaks down for alternative questions (which have falling intonation).
    4. the emotional/expressive content of a sentence in context. I think this leads to the addition of extra "!"s in sentences that are otherwise questioning. For example, I've done some work on "what the hell" questions which in written texts often can receive a "?!" even though they are very clearly questioning.

    It isn't clear that these factors can be fully disentangled, in terms of predicting people's intuition, which is part of what makes all this interesting. For example, Christine Gunlogson's work discusses examples of declarative sentences with falling tones, that are still used in a questioning way in the right context. These clearly (to my intuitions) get a ".", in this case mainly because of the pitch contour. But as I noted above, the pitch contour doesn't determine the punctuation in other cases.

  54. Joe said,

    May 20, 2012 @ 3:58 pm


    Thanks for your long response. That was very enlightening, and I really have to take a look at the longer work when I get a chance. I expressed myself a bit sloppily: I agree with your claim that the syntax is that of an alternative question. I was thinking more about the the influence of illocutionary force on punctuation. For me at least, "Is this a great photo or what?" is used as indirectly equivalent to an exclamatory statement, something like, "What a great photo this is, isn't it!" with falling intonation on the tag. The problem, as you and Language Hat point out, is that the final element of alternative question would have the same falling intonation. But I still think the questioning component is backgrounded relative to the emotive exclamatory component in "Is this a great photo or what," "Is he awesome or what," etc., which leads to the use of (!) by some writers. So my intuition goes with #2, as you probably would have guessed.

    (I wouldn't, however, say that the questioning component is backgrounded in "Did he fall off a bicycle or what?" or "What is he exactly? A liberal, moderate, a centrist or what?" or other examples you cite. So I guess for me the backgrounding really only occurs when the alternative are identical).

    Thanks again for the comments and the link. Look forward to reading it.

  55. Rod Johnson said,

    May 20, 2012 @ 4:01 pm

    Very interesting, Kyle. I can see that this is just the tip of a very large iceberg. Regarding 3, the reverse issue is also relevant: statements with rising intonation (uptalk).

    I'm curious whether you can even define what a question is, in a way that works crosslinguistically. (Was that one?)

  56. Chad Nilep said,

    May 20, 2012 @ 8:44 pm

    I was reading the comment @Andrew? about how punctuation always follows syntax? And then several people mentioned prosody? And I was thinking about up-talk? and how that's usually indicated? And I just don't know.

  57. Jeremy Wheeler said,

    May 21, 2012 @ 1:42 am

    Languagehat says: -no commenters on a post at Language Log, of all places, seemed to be aware that wh- questions do not take "question intonation."-

    Assuming that I've understood what he means by "question intonation", wouldn't it be more accurate to say: – frequently do not take "question intonation."-?

    I am thinking here of the one word question, used to express surprise, or to question what was just said, "What?", where many speakers I know would most certainly give it a question intonation. Then there are questions such as "what did you say?", where the intonation changes, depending on which words take emphasis: "WHAT did you say?"; "What did YOU say?"; and so on. I also wonder whether placing the "what" at the end of the sentence, for example in "You said WHAT?", affects the intonation.

  58. Nick said,

    May 22, 2012 @ 3:02 am

    Question mark should come after the exclamation mark as it looks neater. So for example "wtf!?"

  59. Victoria Simmons said,

    May 22, 2012 @ 4:15 pm

    I'm learning a lot from this analytical discussion. Speaking as a folklorist, though, I think folk speech within a folk group sharing a frame of reference for written communication should be punctuated consistently with the context. So Barbara had it right: it was an exclamation (with a bit of implied threat).

    This sort of thing comes up in transcription fairly often. As a transcriber of speech texts and oral histories, I've punctuated in this way frequently. You never want to imply an upswing where it didn't happen (or wouldn't happen) in speech.

  60. Victoria Simmons said,

    May 22, 2012 @ 4:55 pm

    Note that, although this sort of exclamation anticipates no reply beyond a general affirmation of awesomeness, the idiom has generated a tradition of folk sarcasm:

    "Was that a great concert or what!"
    "Or what."

  61. Kelt Locke said,

    May 23, 2012 @ 3:15 pm

    About 4 decades ago I read an article in Time Magazine about the interrobang, which was easy to make with a typewriter. It would be nice if the Windows character map – which gives us ౠౠ and ♰ and ¶ – could include this wonderful punction mark.

    – Kelt Locke

    [(myl) The interrobang is there in unicode at code point U+203D. You can enter it in HMTL as &#8253; = ‽ No need to appeal to Windows.]

  62. Johan Palme said,

    May 24, 2012 @ 12:59 pm

    I assume with time the "or what" can almost lose its character of a quiestion entirely, like with certain versions of the dialectal british "innit":

  63. Rod Johnson said,

    May 25, 2012 @ 5:18 pm

    In case anyone's still reading down here… I just posted this forum message: "Now why was javascript.options.strict set to true, I wonder."

    I waffled about whether to end this with a question mark or not, but finally decided the tag made it not a question syntactically. But really, neither option felt right.

  64. Glenn Bingham said,

    May 25, 2012 @ 11:00 pm

    To whatever extent "or what" displays questiontude, it is not a request for an alternation (yes-no) nor an exam question–it doesn't involve data, but projected yet unknown personal feelings–nor a request for new information. It seeks confirmation of a statement. If asked of a man, a lack of confrontation (silence) confirms the statement. If asked of a woman, a "yes, it is/was" or "mmmm" or "you're so right" is appropriate to confirm the sentiment. Right?

RSS feed for comments on this post