February 08, 2007

Droning on

People who rant about the way their language is used come at things with theories about how the language works that are firmly held, but are not made explicit and not examined.  The problem is, their premises are so often wrong.

With this in mind, let's take a brief look at Drone Man on "pilotless drones".

Drone Man eventually froths himself up to the raving "Is it sinking into your thick skull, you high school drop-out?", over the San Francisco Chronicle's having printed a reference to "pilotless drones".  The idea is that a drone (in the context of aircraft) is by definition pilotless, so that "pilotless drone" is tautologous; "You tell me, is there any other kind of a drone, other than a pilotless drone?", Drone Man demands as he works himself into a frenzy of language rage.

Hidden in this is a theory about how the semantics of modification works:

Intersective modification: the denotation of an Adj N combination is the intersection of the denotations of the Adj and the N.  That is, Adj N has the same denotation as N plus a restrictive relative clause containing Adj: N that/who is/are Adj.

On this theory, "pilotless drones" means 'drones that are pilotless', and that's just stupid, because "pilotless" doesn't restrict the denotation of "drones".

Now, I grant right away that there's plenty of intersective modification around.  My "a brief look" above is understood intersectively, and so for that matter is "intersective modification" itself.  The thing is that there's ALSO a fair amount of non-intersective modification around.

If I tell you, "My supportive friends helped me through tough times", you'll probably understand me to be asserting that my friends are (all) supportive and that they helped me through tough times, not that only my friends who are supportive did so (implicating that I had non-supportive friends, who were of no help), which would be the intersective reading.  In fact, "supportive friends" is ambiguous between an intersective reading ('my friends who are supportive') and this appositive one ('my friends, who are supportive'):

Appositive modification: the denotation of an Adj N combination is the same as that of N plus a non-restrictive (a.k.a. appositive) relative clause containing Adj: N, which/who is/are Adj.

Plenty of Adj N combinations are, out of context, ambiguous between intersective and appositive modification; but context, background information, and reasoning about other people's intentions are usually enough for us to decide which reading is the appropriate one.

Back to Drone Man.  What's his problem (beyond being appallingly short-tempered)?  He's assuming that all Adj N modification is intersective.  But this is just false.  Drone Man's tirade is entertaining, but it's based on a misunderstanding of English grammar.  (Cue Emily Litella: "Never mind!")

Now for the subtlety.  You might think that even the appositive reading of "pilotless drones" would be stupid, since drones are all pilotless.  But look at the explicitly appositive version: "drones, which are pilotless".  This isn't stupid at all; it REMINDS us, in a helpful way, that drones are pilotless.  In general, even when the denotation of Adj is included within the denotation of N, appositive Adj N can do useful discourse work.  As a bonus, since intersective Adj N is stupid in this situation, the potential ambiguity is eliminated in practice, in favor of the appositive reading.

Now whether this is what the Chronicle writer intended is another matter.  "Pilotless drones" could have been an error.  But I understood it, charitably, as having a reading that would make sense -- the appositive reading.

Now an example with appositive Adj N in this inclusion situation, in a context where the writer's intentions are pretty clear.  This is from a comment by "waxwing" on Dave Barry's blog, following up on another poster's report that, omigod, there are earthworms up to 11 feet long:

OK, I must say it...does it bother anyone else that legless earthworms are measured in feet?

Very effective, I think.

zwicky at-sign csli period stanford period edu

Posted by Arnold Zwicky at February 8, 2007 11:25 AM