It went past me on the radio as I was going to sleep, so I didn’t get the details of either form or context, but the crux of the matter was the possibility of either can or can’t in
I’ll see if we can/can’t [do something or other]
Huge numbers of both on the net. Compare these two:
I’ll see if we can’t do something for you in the next version. (link)
But I’ll see if we can do something for you so you can try it out. (link)
At first glance, it looks like this is a case of simple negation indifference (as Chris Potts labeled it in 2004): adding or removing a negation without change of meaning.
There are (vaguely) parallel cases that Potts inventories (and that I’ll look at in a moment), but this one has its own assemblage of features, three different factors. And, I’ll argue, the variants are semantically close but nevertheless distinct.
The Potts 2004 list had five items:
(1) vulgar minimizers like squat in
Eddie knows / doesn’t know squat about phrenology.
(from Paul Postal)
(2) (that / I) will teach, as in
That’ll teach you to / teach you not to tease the alligators.
(some discussion here)
(3) the endlessly denigrated (though incredibly popular) could care less, as in
I could / couldn’t care less about monster trucks.
(much discussion on Language Log)
(4) negative concord (or lack of it) in declarative tags, as in
You shouldn’t play with alligators, I think / I don’t think.
(5) wonder whether NP can / can’t, as in
I wonder whether we can / can’t find some time to shoot pool this evening.
Items (1)-(3) probably have a common historical source, in “negation by association”, in which a concomitant to negation is reanalyzed as sufficient to trigger the semantics of negation (cf. French pas and a huge number of other examples).
Item (4) probably represents a choice between two treatments of tag elements, in particular those serving as epistemic modifiers (in effect, higher-clause epistemics demoted to adverbial status). One choice is to treat the adverbial as simple “verb-splitting” —
I think you shouldn’t play with alligators.
~ You shouldn’t play with alligators, I think.
and one is to treat the tag as subject to formal agreement (giving I don’t think).
That leaves us with (5). In such examples, there are three formal + semantic contributions:
(a) an embedding verb like wonder, see, ask, etc, with future-oriented, hence non-veridical, semantics;
(b) a connective whether or if ( these have rather different syntax and semantics) conveying a choice (whether/if … or not); and
(c) a modal of possibility or obligation: can, could, should (or the quasi-modal ought to).
These three elements — quite likely I’ve missed many wrinkles and more general phenomena here — combine to give a situation where the positive and negative modals can convey something very similar, but subtly different: in
I wonder whether we can / can’t find some time to shoot pool this evening.
I’ll see if we can / can’t fix this.
the positive variant (can) is noncommittal as to whether the event in question (finding some time to shoot pool, fixing things) is going to happen, but the negative variant (can’t) is, or can be, (positively) biased towards the event’s happening, conveying a hope that the event will happen. Similarly with:
I wondered whether we could / couldn’t find some time to shoot pool.
I’ll see if we could / couldn’t fix this.
The contrast here is familiar, from yes-no questions:
Can / Can’t we find some time to shoot pool this evening?
Could / Couldn’t you fix this?
The positive versions are neutral, the negative versions positively biased. (Note: these are not new observations of mine.)
So what we’re seeing in the embedded if/whether examples is the importation of the semantic distinction between positive and negative clauses in main-clause questions.
No doubt there are whole continents of data and interpretation I’m ignorant of. But this is a start.
[Further comment: The positive vs. negative modal contrast in case (5) illustrates Bolinger’s Dictum (or the program of “unfree variation”) — see here — expecting a semantic/pragmatic difference to accompany a syntactic or lexical difference. Subtle differences might also turn up in cases (1)-(4), for at least some people who have both variants.]
June 26, 2011 at 5:13 pm |
I think this one can be resolved (at some level of resolution) by pointing out that if you discover the value P of some variable, you discover the value of 1-P at the same time. So finding out whether ♢p also tells you whether ¬♢p; thus either works as an idiom.
June 28, 2011 at 6:07 am |
Well, that’s a beginning. But there are a lot of details to work out.
First, I see now that it’s actually not possibility, but ability, that’s at issue. The effect doesn’t show up in pure possibility constructions like “I’ll see if it’s possible for us to help you” (cf. “I’ll see if it’s possible for us not to help you”).
Second, there’s still the difference in biasing to explain (which is why I brought in questions).
June 29, 2011 at 7:19 pm
Shouldn’t that be “cf. ‘I’ll see if it isn’t possible for us to help you'”? That’s what I thought John Lawler was saying.
June 28, 2011 at 7:31 am |
Do you see intensified “not” as negative concord, a separate category, or not relevant to this discussion?
For example, I would interpret these two sentences to mean the same:
1. Pupils must not run in school, especially on the top floor.
2. Pupils must not run in school, especially not on the top floor.
June 28, 2011 at 2:38 pm |
Nice example, of still another type. Sentence-final contrastive adverbials can reproduce various spans of the main clause that they’re contrasted to; in effect, these are ellipses of various sizes. In this case, on the top floor is paired with in school, and the ellipsis can include not or not. (Ok, it’s a lot more complex than that, but that’s the crude outline.)
September 20, 2011 at 12:44 am |
My mother used to say, “That’ll teach you to talk with your mouth full.”, when I had ended up choking on my food whilst chattering.