If you think about it
« previous post | next post »
Over the years, we've discussed a number of different sorts of conditionals, including bleached conditionals, concessive conditionals, and baseball conditionals.
But as far as I can recall, we haven't discussed relevance conditionals, as (I think) exemplified in this morning's Stone Soup:
Also known as "conditional speech acts", these are covered in section 5, Factual and relevance conditionals, of Rajesh Bhatt and Roumyana Pancheva's "Conditionals" (chapter 16 in the Blackwell Companion to Syntax). They give these examples:
If I may be honest, you are not looking good.
If you want to know, 4 isn't a prime number.
If you are thirsty, there is beer in the fridge.
They explain that "The if-clause in relevance conditionals specifies the circumstances in which the consequent is discourse-relevant, not the circumstances in which it is true."
Though on reflection, the example in the cartoon is not quite like these others. As the term "conditional speech act" suggests, relevance conditionals can be understood as ordinary hypothetical conditionals in which the consequent includes an unexpressed verb of communicating, something like "If you're thirsty, I'll inform you that there's beer in the fridge".
Val's sentence seems to express a genuine hypothetical in a differently elliptical form. What's left out, it seems, is something about the hearer's uptake, not the speaker's output: "If you think about it, (you'll see that) that's not very long".
Because the "you'll see that" (or "you'll agree that" or whatever) is omitted, there's an opening for Holly's rejoinder, just as in the canonical relevance conditionals ("And if I'm not thirsty, does the beer magically disappear?").
Val's version of the sentence doesn't permit then, which Bhatt & Pancheva give as a characteristic of relevance conditionals:
*If you think about it, then that's not very long.
Adding "you'll see that" or similar material changes this:
If you think about it, then you'll see that that's not very long.
[Update — Kai von Fintel writes:
"Biscuit conditionals" as they're often called (after Austin's example) have been discussed quite a bit in the recent scholarly literature. In addition to the Fulda paper adduced by Joseph, here are some others:
DeRose, Keith and Richard E. Grandy. 1999. Conditional assertions and "biscuit" conditionals. Noûs 33(3): 405–420. doi:10.1111/0029-4624.00161.
Siegel, Muffy E.A. 2006. Biscuit conditionals: Quantification over potential literal acts. Linguistics and Philosophy 29(2): 167–203. doi:10.1007/s10988-006-0003-2.
Franke, Michael. 2007. The pragmatics of biscuit conditionals. ms, Universiteit van Amsterdam.
Predelli, Stefano. 2009. Towards a semantics for biscuit conditionals. Philosophical Studies 142(3): 293–305. doi:10.1007/s11098-007-9187-8.
]
Sam C said,
June 1, 2009 @ 7:47 am
Trivia: deliberately misunderstanding this sort of conditional ('So the beer will disappear if I'm not thirsty?') is a very common form of humour among professional philosophers. I don't know if this reflects training in formal logic, or geeky pedantry, or both…
Ray Girvan said,
June 1, 2009 @ 8:06 am
"… if you think about it …" = the command "think about it".
It seems quite close to what one might call "rhetorical conditionals" that preface a command:
Boss: "If you have a moment, Mr Girvan, could you step into my office to discuss your timekeeping?"
James said,
June 1, 2009 @ 9:08 am
It's got to be relevant that the protasis must be second person.
* If he thinks about it, that's not very long.
The restriction, I believe, is semantic, since there are acceptable first person uses when the speaker is thought of as addressing himself, or addressing a group of which she is a part. For instance, in a lecture, a mathematician could say,
Now if we think about it, the sum must series must converge to 2…
At least, I think that's okay. Is it?
Ray Girvan's suggestion sounds promising to me. Imperatives cannot go in a protasis. Is this construction a way of getting around that syntactic restriction? Maybe the pragmatics go like this: speaker tells listener to think about it; the 'if' is her way of delaying her assertion, the apodosis, until listener has executed the command. That makes this conditional fit into the most general pattern of conditional constructions.
Faldone said,
June 1, 2009 @ 10:40 am
@Sam C:
You've been hanging around with the wrong brand of philosopher. The proper logical response to "If you're thirsty, there's beer in the refrigerator" is "So, if there's no beer in the refrigerator, I'm not thirsty?"
Jonathan Badger said,
June 1, 2009 @ 10:42 am
This basic joke (interpreting IF X THEN Y to imply that not X will yield a different answer) is a staple of computer programmer humor, along with answering questions like "Do you want coffee or tea?" with "yes".
Nico said,
June 1, 2009 @ 11:24 am
Assuming that the consequent in relevance conditionals has a hidden verb of saying just recreates the problem, doesn't it? That is, if "you're thirsty" doesn't hold, then the conditional says nothing about whether or not I'll inform you about the beer.
Of course, I AM informing you about the beer, whether or not you're thirsty, so maybe it's a moot point.
Mary-Ann said,
June 1, 2009 @ 12:14 pm
"If you're thirsty, there's beer in the fridge." I would have thought that this would expand to "If you're thirstly, (you may like to know that) there's beer in the fridge."
Andrew said,
June 1, 2009 @ 12:19 pm
In formal logic, 'If P then Q' does not imply 'If not P then not Q' (though as Faklone points out, it does imply 'If not Q then not P'). So the philosophers in question are clearly relying on conversational implicature to some extent. It seems, however, that they are unable to grasp more complex conversational moves.
Robert Coren said,
June 1, 2009 @ 12:35 pm
@Nico: Well, I guess the point comes back if you interpret it as "If you're thirsty, [you may be interested to know that] there's beer in the refrigerator."
Ray Girvan said,
June 1, 2009 @ 1:23 pm
Jonathan Badger : along with answering questions like "Do you want coffee or tea?" with "yes".
GKP tackled that – in the context of eejits for whom it's not a joke – in Exclusive OR: free dinner and stay out of jail.
Carrie S. said,
June 1, 2009 @ 1:47 pm
along with answering questions like "Do you want coffee or tea?" with "yes".
They get cured of that pretty quickly if you start giving them a randomly-chosen drink–generally, the first time they get the one they don't want. The answer to protests is, "I assumed that since you said 'Yes' that you didn't care what you got."
Though I do know a few people who will answer "Yes" when asked questions such as "Would you like apple or pumpkin pie?", and what they mean is "I'd like a half-width slice of each."
Andrew H said,
June 1, 2009 @ 2:05 pm
James–A little searching on google indicates that forms such as the one you have indicated as ungrammatical, while probably very rare, are not entirely unaccounted for.
If he thinks about it, it's pretty, well, stupid – it's only been a handful of days, but he misses Mello like a physical thing.
If he thinks about it, it's the very rare Scout who attains the rank of Eagle without the involvement, support, and encouragement of others
Also, a type of conditional that has long seemed odd to me is a fairly common sort of phatic expression:
If I don't see you beforehand, have a great vacation!
Perhaps this stems from my overly literal-mindedness, and the fact that (as far as I know) I don't use it. Of course, p=>q doesn't entail q=>p, but conditionals in spoken language are often ambiguous between being conditionals and biconditionals.
Nigel Greenwood said,
June 1, 2009 @ 2:47 pm
@ Jonathan Badger: This basic joke (interpreting IF X THEN Y to imply that not X will yield a different answer) is a staple of computer programmer humor, along with answering questions like "Do you want coffee or tea?" with "yes".
It's worth pointing out that many languages do not share this ambiguity with English. Turkish, for example, uses a construction with the question-tag mi ("Do you want coffee? … tea?"). Chinese uses hai2shi ("Do you want coffee? alternatively tea?"). I'm sure there are other languages that make the distinction: the French ou bien comes to mind.
James said,
June 1, 2009 @ 4:32 pm
Andrew,
Wow, great examples. To me, they simply sound wrong.
Your vacation conditional seems to me to be a biscuit conditional. The protasis is a condition of relevance, not of truth (of course, since the apodosis is an imperative).
Timothy Short said,
June 1, 2009 @ 6:51 pm
I agree that the 'if he thinks about it' construction sounds wrong, but isn't that just because it should be subjunctive?
"if he were to think about it rationally, he would realize how his emotions are over-exaggerated"
"if he were to think about it, it might make him angry"
Though I admit I only have 13 Google hits on this variant and it may only be UK English. But to me it certainly doesn't sound clearly wrong.
Brian said,
June 1, 2009 @ 7:05 pm
Personally, I would interpret the "beer in the fridge" example as:
"If you're thirsty, there's beer in the fridge [and you may have some]."
The example in the comic is harder to pin down, but it seems to me like:
"If you think about it [in a different way than you are], that's not very long."
Flopsy said,
June 1, 2009 @ 7:35 pm
If the beer over falls in the fridge and there's no one around to hear it, does it make a sound?
Joseph said,
June 1, 2009 @ 9:12 pm
See http://dx.doi.org/10.1016/j.pragma.2008.12.010
[(myl) Thanks! That's Joseph Fulda, "Towards a unified theory of if's—The theory of conditional elements: Further evidence from conditionally self-falsifying utterances", Journal of Pragmatics 41(7): 1440-1448, July 2009. The abstract:
"Austinian If's" are those discussed by J.L. Austin in the 1950s, e.g.
I can, if I choose.
There are biscuits on the sideboard if you want them.
I paid you back yesterday, if you remember.
If he's not very original it is at least, a careful piece of work.
Fulda's proposal (I gather) is that "it is … assertion of worries through the medium of warnings which displays the ability to be any of the [various types of] conditional elements." After a quick scan, I'm not sure whether this implies the somewhat grim theory that conditional promises and hopes are merely the superficial guise of worries expressed as warnings. ]
James D said,
June 1, 2009 @ 9:49 pm
@ Sam C: Yes, if you're not thirsty, the beer will disappear. I'll drink it!
Joe Fineman said,
June 1, 2009 @ 11:29 pm
In my time (1950s) at Caltech, if you answered "yes", you got a cup of coffee with a teabag in it, and serve you right.
Some classic impudences of the same kind, where the "if" is suppressed:
"Close the window; it's cold outside."
"So if I close the window, it will be warm outside?"
"This letter is too heavy; it needs another stamp."
"Another stamp will make it lighter?"
Chas Belov said,
June 2, 2009 @ 2:30 am
@Carrie: "Though I do know a few people who will answer "Yes" when asked questions such as 'Would you like apple or pumpkin pie?', and what they mean is 'I'd like a half-width slice of each.'"
In my case, it'd be a full-width slice of each.
@Nigel Greenwood: "Chinese uses hai2shi ('Do you want coffee? alternatively tea?')."
I assume you are referring to Mandarin. I was taught "yìkwaahk" ("or") in Cantonese. The course book (Foreign Service Institute Cantonese Basic Course Volume I) didn't discuss a yes response but implied that to not pick one (that is, to decline both options) would be considered not only rude but a violation of grammar.
Actually, it oddly states "The English possibility of:
A: Do you want coffee or tea?
B: No thanks.
is not covered by yìkwaahk> In Chinese, you would have to rephrase the sentence to say something like 'Would you like something to drink? We have coffee and tea.'"
Unfortunately it does not advise on how B would elicit this replacement phrasing from A were B to require such replacement phrasing.
@Andrew H: I use the construction "If I don't see you, have a happy holiday." If I then feel self-conscious about having said that, I occasionally add (jocularly) "And if I do see you, have a happy holiday."
Noetica said,
June 2, 2009 @ 5:59 am
If you ask me, they're called metalinguistic conditionals.
Faldone said,
June 2, 2009 @ 9:44 am
@ Joe Fineman
I'm having trouble seeing where you could put an if in either of these and even fallaciously draw these conclusions. Now a because works perfectly fine.
Kate Y. said,
June 3, 2009 @ 3:13 am
I grant that in formal logic, "If P, (then) Q" does not imply anything about Not-P.
In real-world communication, however, I think it does indeed imply the possibility of Not-P, (and with it the chance of some alternate Q') or why bring the subject up at all? Considee the subtly different implications of the these two statements:
"If you're going to the store, you can get more eggs."
"Since you're going to the store, you can get more eggs."
The second one is less polite, more of a command, precisely because it removes the possibility of Not-Store. To use "If" is to imply that possibility.
And that is why I flinch at the phrase "If you're just joining me, my guest is…."
Miles Rind said,
June 3, 2009 @ 11:46 am
As the term "conditional speech act" suggests, relevance conditionals can be understood as ordinary hypothetical conditionals in which the consequent includes an unexpressed verb of communicating, something like "If you're thirsty, I'll inform you that there's beer in the fridge".
I can't see that anything along these lines will work at all. If "If you're thirsty, there's beer in the fridge" (call it sentence A) were to be paraphrased as "If you're thirsty, I'll inform you that there's beer in the fridge" (call it sentence B), then it would be appropriate for the addressee of A to say, "Well, I'm thirsty", and for the first speaker to say to that, "Okay, then: there's beer in the fridge"—which would plainly be daft. B, the proposed paraphrase, is structurally parallel to "If you're thirsty, I'll get you a beer"; but in A, the speaker is plainly not offering to tell the addressee something, on a certain condition; he or she is telling the addressee something, which may or may not be relevant. Nor would the proposed analysis fare any better if "I'll inform you" were replaced with "I inform you" (à la "I promise", etc.), as that still entails that the speaker is not informing the recipient of anything unless the recipient is thirsty.
Another way to make the point would be to say that the proposed paraphrase invites a joke-reply parallel to the joke-reply to the original utterance: "And if I'm not thirsty, will you [or, as may be: do you] not inform me that there is beer in the refrigerator?" The possibility of the joke shows that the paraphrase (B) is misunderstood if it is construed as a hypothetical conditional—exactly what was puzzling about the original conditional (A). So the proposed paraphrase does nothing to elucidate the original sentence.
Mary Ann's proposal avoids these problems:
"If you're thirsty, there's beer in the fridge." I would have thought that this would expand to "If you're thirstly, (you may like to know that) there's beer in the fridge."
This is akin to the proposed paraphrase of "If you think about it, that's not very long":
What's left out, it seems, is something about the hearer's uptake, not the speaker's output: "If you think about it, (you'll see that) that's not very long".
I am not sure if this manner of paraphrase does not have problems of its own, but at least it gets off the ground.
Aaron Davies said,
June 8, 2009 @ 12:30 am
"If I don't see you again" seems to me to be eliding a bit of custom/superstition that I find rather hard to paraphrase back into the sentence: "goodbye" and similar things are supposed to be the last thing we say to someone who's going away. The full sentence might be something like "I'd like to wish you a happy vacation on the last occasion of seeing you before you leave, but since I'm not sure whether this is that occasion, I do so conditionally."
mike said,
June 26, 2009 @ 11:52 am
I'm a little surprised that no one has mentioned what is probably the most commonly heard example of this construction (at least, among us earnest liberal types), namely Terry Gross on NPR's Fresh Air and her frequent "If you're just joining us, I'm talking with …". Personally, I never get tired of addressing the car radio with "And if we're not just joining you, who are you talking with?" Easily amused, I am.
Peter said,
February 11, 2010 @ 11:10 pm
There's also the "For those who don't know me, I'm …" sentence some people start speeches with.
My father has started a speech poking fun at this construction. "For those who don't know me, I'm (made-up name). For those who do know me, you'll know my name is actually (real name)"