Supplementary apposition train wreck

« previous post | next post »

Bob Bechtel was reading James Warren's article "The Potentially Revolutionary Political Role of Fried Chicken" (in The Atlantic) when he stumbled on this sentence:

Reaching for a New Deal is supported by the Russell Sage foundation, a bastion of research in the social sciences and due out in August.

The Russell Sage foundation is due out…? What has gone wrong here? Was he looking at a sentence with a missed comma (perhaps from a typo)? Or an overapplication of the rule saying there should be no comma before a final and?

Both, I'd say. His intuitions about the etiology of the error seem spot-on correct to me, and they're compatible. It certainly is a train-wreck of a sentence. The attempt was obviously to write a sentence of the form Reaching for a New Deal is [F1 and F2], where the bracketed part is a coordination of two predicative complements: F1 = "supported by the Russell Sage foundation" (a passive VP) and F2 = "due out in August" (an adjective phrase).

It would normally be wrong to have a comma before the and in the phrase supported by the Russell Sage foundation and due out in August. Even advocates of the "Oxford comma" (people who write Tom, Dick, and Harry rather than Tom, Dick and Harry) always write Tom and Harry, where there are only two coordinates, with no comma.

The trouble started when the writer (or an editor) inserted an appositive noun phrase a bastion of research in the social sciences after the name of the foundation. A noun phrase of this sort is what The Cambridge Grammar of the English Language (chapter 15, section 5, p. 1357) calls a supplementary appositive: it is a loosely attached noun phrase constituent conveying a piece of supplementary information about the referent of the noun phrase that precedes it. And any supplementary constituent must be flanked by commas. Thus in a sentence like the following, both commas are obligatory:

The Russell Sage foundation, a bastion of research in the social sciences, will publish the book in August.

In the intended sentence, then, there seems to be a position where a comma before the word due might be said to be both forbidden and required: on the one hand, you don't put a comma before the second coordinate in a binary coordinate structure; on the other, you do have to have a comma at each end of a supplement.

The resolution is simple enough: the rule isn't that a comma is forbidden before the second coordinate in a binary coordinate structure; it's that you don't use one in that position unless there is some independent reason for it. A supplement attached to the first coordinate provides that independent reason. There can be other reasons too. If you want to mark the second coordinate as an afterthought, that also motivates a comma; you can see a case of that sort in the first sentence of Alice's Adventure's in Wonderland:

Alice was beginning to get very tired of sitting by her sister on the bank, and of having nothing to do…

Somehow James Warren (or his editor) lost track of the fact that the supplementary appositive after the Russell Sage foundation needed to be closed with a comma. It was left out, and once that mistake had been made the resultant structure made a certain parsing mistake very natural: it is syntactically tempting (though it turns out semantically hopeless) to read the sentence the way Bob Bechtel did — as if everything after the comma were a single supplementary appositive:

Reaching for a New Deal is supported by the Russell Sage foundation, [a bastion of research in the social sciences and due out in August].

It is perfectly natural and grammatical to have a noun phrase and an adjective phrase coordinated like the part in brackets above; examples are easy to find in corpora or on the web: This user is [a father, and proud of it!] here, for example; Barb is [a gay man and happy to be so] here; and so on.

It's only when you try to make semantic sense of the notion of something being both a bastion of research in the social sciences and yet not out yet (due out in August) that you draw a blank — you pursue a parse that ends up with an incoherent meaning, and you have to retrench. You read the sentence over again and conclude that the best hypothesis is that the writer was in error: there is no parse that accords with the standard rules of punctuation for written English.

Even a small error of this sort is a disaster for fluent reading and uptake. It's only a comma, but it's extraordinarily important, and leaving it out was a bad, bad mistake. Don't ever think that if a descriptive linguist were grading your writing you would always get an A+.

[Comments are a bastion of the blogosphere but closed on this particular occasion; send email to if you wish.]

Comments are closed.