Pullum’s Non-Parallelism
Posted by Neal on May 29, 2006
In a recent Language Log posting, Geoff Pullum has this to say about a movie reviewer who read the book The Da Vinci Code and made some observations about Dan Brown’s strange and klunky syntax, observations already made by Geoff Pullum two years earlier:
[Anthony Lane] sounds a tiny bit like an intelligent literary stylistician who has just been awakened from a two-year coma and thus attracts a certain amount of eye-rolling at conferences as he brings up points that he thinks are new but they’re not.
I was interested in the points that he thinks are new but they’re not bit. Pullum uses a relative clause to zero in on what kind of points he’s talking about. Specifically, he’s talking about points that have two properties: 1) He (Lane) thinks they are new; and 2) they’re not new. The first property appears in the relative clause as he thinks [ ] are new, with silence instead of a subject for are new. The missing subject corresponds to points. The second property appears in the relative clause as they’re not. No missing subject here; points corresponds to the actually-spoken subject they.
So why didn’t Pullum leave out the they here, just as he did in he thinks [ ] are new? That is, why didn’t he write this?:
…points that he thinks are new but aren’t.
I could ask him, but hey, Geoff Pullum is a busy man. I’m sure he won’t mind if I just speculate a little. He’s a pretty tolerant guy, after all. Maybe he wanted the sentence to end with a stressed not for emphasis (which you don’t get if it’s swallowed up in the contraction aren’t) and it sounded better as but they’re not than but are not. Or, maybe he wanted to avoid having points fill different kinds of gaps in the two phrases. Sure, points corresponds to a subject in both he thinks [ ] are new and [ ] are not, but in the former, it’s the subject of the embedded verb after thinks that’s being left out, while in the latter it’s the subject of the main verb are. Perhaps in Pullum’s grammar, this kind of double duty sounds as strange as a noun corresponding to both a subject and an object in a relative clause, like some of the ones here.
Whatever the reason, it evidently sounded better to Geoff Pullum to keep the they in the second clause. The result is that we have another coordination in which something is marked only on the first coordinate which nonetheless has scope over both coordinates. We’ve had examples involving negation, modality, and question formation, as illustrated below:
- Negation: “It was fun to run into someone who [wasn’t stodgy] and [thought at some point you should call it quits],” remembered Ellen. (link)
- Modality: “[They must have loosened the hooks] and [Mr. Cleaver didn’t notice it],” Jerry said. (link)
- Question formation: [Did you make your own contributions to a complying superannuation fund] and [your assessable income is less than $31,000]? (link)
Pullum’s is the first example crucially involving relative clause formation.
Literal-Minded 
Bridget said
I definitely agree that “they’re not” is much better than “are not” there. For what it’s worth, if I were just saying the sentence without a pause (as it is punctuated), I would really want to use “aren’t.” If, on the other hand, I were to separate the last part from the rest of a sentence with a beat to emphasize it (“that he thinks are new — but __ not”), “they’re not” would sound the best.
Geoff Pullum said
Ooh! After all these years as the subject, the syntactic investigator, now all of a sudden I’m the object, the syntactically investigated! I’m source material! I’m data!!
Ingeborg S. Nordén said
Here’s another non-parallel coordination that drives me up the wall. I’ve heard it in so many drug commercials that I’d call this a “pharmacist’s coordination”.
“[Drug name] is not intended for women who are nursing, pregnant, or may become pregnant.” Come ON… women who “are may become pregnant”? Sounds a lot clumsier than “there was a farmer had a dog”!
Neal said
I agree with Runelady about the pharmacist’s coordination. For other “coordinations at unlike levels”, check out this post.
ACW said
I think (pero Geoff sabe mas) that the “they” in Pullum’s second clause is something called a resumptive pronoun. As I must have mentioned here before, at the age of four my son performed the following heroic extraction: “I don’t know where my shoes are, but guess what I do know where is?” After laughing at this for a while, we tried to rephrase it so that it didn’t make our WH-island neurons hurt. And we couldn’t. Here’s a thought that can’t be expressed in English without lots and lots of circumlocution. “I don’t know where my shoes are, but I do know where something else is. Can you guess what that something else is?” But our first attempt involved a resumptive pronoun: “I don’t know where my shoes are, but guess what I do know where it is?” This is also bad, but maybe slightly less bad. I can’t tell, because my WH-island neurons were injured when my son was four.
Neal said
ACW, the idea of they as a resumptive pronoun is a good one. I still favor my analysis, since my sense is that resumptive pronouns tend to be used as a rescue strategy, in places where it would seem you don’t need a pronoun, but leaving it out sounds awkward (because of island constraints, as in your son’s example). In Pullum’s example, though, leaving the pronoun out sounds OK to me. But then again, if something in Pullum’s grammar rejects that option, then the situation would be ripe for a rescue by resumptive pronoun. This is where an eager syntactician can step forward and present different predictions that the different analyses would make, and then test them out.
Post-Election Post « Literal-Minded said
[…] Even Geoff Pullum does it: [H]e brings up points that he thinks are new but they’re not. […]