Distant Drums

« previous post | next post »

Jim Reeves' classic country song "Distant Drums" (written by Cindy Walker) expresses a marriage proposal by a young man who wants his true love to marry him right now, before the drums and bugles that he already imagines he hears in the distance arrive and he is conscripted and forced to go off to war. The main part of the young man's argument is expressed in the chorus thus:

So Mary marry me, let's not wait
Let's share all the time we can before it's too late
Love me now, for now is all the time there may be
If you love me Mary, Mary marry me.

I have a question about the underlined part, addressed primarily to professional semanticists. Some of our posts get a bit nerdy on Language Log, and this is one. Skip it if you hate to see a capital A written upside down.

OK, semanticists: think about a direct translation of the underlined complement clause into a quantified modal logic. Isn't the modal operator ◊ corresponding to may in the wrong position relative to the quantifier ∀ corresponding to all? Doesn't the young man actually mean that now may be all the time that there is?

He overstates, of course, as young men in love will. But the overstatement is just a routine neglect of restricting the range of a universal quantifier: just as Everyone's looking at you doesn't refer to everyone in the world, all in this context doesn't really mean all; it's restricted to just the relevant time points in the context. What he's talking about is all the time available for him and his true love to spend together. Let that adjustment be understood.

But my point is this: he surely doesn't want to claim that the singleton containing the present moment of time is identical with the set of all (relevant) time points in whatever worlds may exist; he's saying that right now it is possible (epistemically — compatible with all that we know) that the present (some reasonably small interval beginning with the present moment of time) is identical with the set of all relevant future time points in this world.

(He's certainly right to make that point; no one could see it more clearly than I do. At 3:15 on the afternoon of May 14 I was sitting beside Barbara's hospital bed just spending time with her and enjoying her company, wrongly believing that despite her terminal illness it was not in an acute phase and there would be plenty more time to be together in the coming months. And suddenly it turned out that now was all the time there was ever going to be for us, and she died right in front of me. If you love someone, love them now, for now may be all the time there is.)

I think the clearest way I can put it is that the song's young protagonist wants to tell his true love that it is possibly true that a narrow window containing the present is all of the time that has actual availability; not that the present is actually all of the time that has possible availability. And I just don't see how those truth conditions can be extracted from the syntactic form of the sentence used in the song.

Yet (and this is crucial) no one finds the line as written hard to understand, and I assume people get the right truth conditions, and correctly infer what it is that the song's protagonist wants.

So do natural languages leave the scope of modal operators and quantifiers open to such an extent that both meanings are available, a hypothesis you could call radical underspecification of scope in the semantic representation?

Or did the songwriter simply screw up and write the wrong sentence, and the entire country music-listening population were all too dumb to notice — an error theory of what's going on?

Or was the wrong sequence deliberately used to get the meter right and nonetheless we somehow effortlessly understand things as if the quantifier and modal had been the other way round, overriding the constraints of the language in our unstoppable drive to find coherent meanings — a performance theory of dynamic creative interpretation?

I don't do semantics, as you know: you don't see me at formal semantics conferences. But I do think I see a little problem here that has a certain interest. Email me (you all know where to find me). A suitably clear statement of what you think is going on here can be published here as guest posts if you wish (I will check with you to get your permission). Or you could dig in and do a full-scale journal article that explains what's going on in such cases (just credit me in a footnote as having noticed the problem, OK?).

Or if there's an account already written that covers it, point me to it (you know I'm not versed in the literature of your demanding subfield), and I'll add a note about it here.

[Comments are closed because just a moment ago was all the commenting time there may be.]

Update: a few people have been writing to me about the sense of may, suggesting that the answer could be that here it means might. This is not the solution. I don't think the rather subtle difference between these words (actually the difference between the present tense of the lexeme may and a special remote sense of its preterite form) has anything to do with this. I think there is a problem in semantic scope and logic here, and I don't think the answer will turn on anything about mere word meaning. Of course, I could be wrong. But no semanticists have written to me yet.

Comments are closed.