24 Mar 2007

Don't Buy This Book

Submitted by Karl Hagen
If you are a high school student studying for one of the big standardized tests (the SAT or the ACT), or if you are the parent of such a student, I have one important piece of advice for you: don't waste your money on commercial test prep books, especially when it comes to the material for writing.

I have just sat down with recent editions of a bunch of SAT books to begin a study of their content adequacy with regards to writing. My hypothesis going in is that they foist a lot of nonsense onto students, and within 10 minutes of starting to read, I've already found one that irritates me so much I must share.

The following comes from Princeton Review's, 11 Practice Tests for the SAT and PSAT. 2007 Edition, p. 25. The context is that the authors are explaining why certain writing questions are difficult:

Read this sentence: Peter Jackson's movies have made him one of the most successful directors working today. Sounds perfectly fine, right? Wrong. In SAT-land, this sentence is incorrect. Specifically, the pronoun him has no antecedent, meaning that there is no noun

Yes, yes, I know what you are thinking. You and I both understand from context that him refers to Peter Jackson, but in formal terms, Peter Jackson's is an adjective that describes the noun movies, and movies can't be the thing to which him refers. This is an obscure rule but one that is fair game on hard SAT questions.

There are multiple layers of crap in this explanation, so let me get out my shovel and start digging.

First, possessive noun phrases are not adjectives. It simply isn't true that anything that modifies a noun is an adjective, and those who can write such a thing have clearly never been exposed to serious scholarship about the way that English really works. In real formal terms, Peter Jackson's is a genitive noun phrase functioning as a determinative[1], and there is absolutely no reason it can't serve as an antecedent. In fact, it does so regularly in educated usage and always has.

Notice that this rule is justified not on the basis of some potential lack of clarity, but in an argument based on necessity from a mistaken premise about parts of speech. Never mind that all the evidence shows the assumption and the conclusion both to be mistaken: it must be so. Some book somewhere said it. My English teacher said it. So it must be so. They, of course, know better than the vast majority of speakers of English, including writers like Jane Austen.

Second, the rule is ultimately sold to students based on the fact that it is followed on the SAT. In other words, no matter how much you think the rule reeks, you've got to learn it anyway, because it's on the test. If that were true, I would actually find it a good argument. But there's no evidence that the writers of the SAT do follow this rule.

The example chosen here makes it clear that the Princeton Review authors have in mind a particular question from the 2002 PSAT that became a matter of controversy.

Toni Morrison's genius enables her to create novels that arise from and express the injustice African Americans have endured. No error

(If you're unfamiliar with the format of this sort of question, you have to pick the underlined segment that contains an error, or "No error" if there is none.)

As written, the question was intended to have no error. A teacher, however, argued that the right answer was the first underlined section (choice A), based on the suspect rule above. After review, the College Board canceled the question (i.e., it did not count towards students' scores). This event was widely (mis)reported as an instance of the College Board changing the answer.

Looked at that way, you might jump to the conclusion that the College Board was agreeing with the teacher, but that's not what happened.

All three experts who reviewed the question after the challenge said

1. For them, the sentence had no error
2. If you follow the rules of some (not all) usage guides, A appears to be an error.

That's hardly an endorsement of the rule. The College Board did not change the answer to A, they dropped the question because it is ambiguous. You get different results depending on what authorities you follow. (The teacher who reported the problem was apparently campaigning to have E scored as incorrect, but he was rightly ignored.)

How can anyone possibly conclude from this event that the rule is endorsed by the writers of the SAT? They didn't intend this sentence to reflect the no-possessive-antecedent rule, and when challenged on it, they didn't actually change their minds about the grammaticality of the sentence.

This is precisely the sort of sentence that the SAT is obligated to avoid because some widely used usage guides disagree on the point at issue. (The fact that one side of the disagreement may be utterly full of it is neither here nor there, since the College Board has not outlined a specific set of principals as to what it considers standard usage to be.)

As part of the project I mentioned, I am also reviewing the content of every released SAT and PSAT writing section I can get my hands on in order to have some basis for comparing these books. So far (and I'm about 2/3 done), I haven't found a single instance of this rule, either deliberately or inadvertently triggered.

So the good folks at Princeton Review are bludgeoning poor students into learning a useless and counterintuitive rule that contributes nothing to the clarity of one's writing and has no chance of ever appearing on a real test. What an utter waste of time! The fact that they are doing so while adopting an I-know-this-sucks-but-you've-got-to-learn-it tone, as if they were the students' friends offends me all the more.

Arnold Zwicky has gone into the origins of this rule in great detail. You can find his three-part response in the archives of the American Dialect Society mailing list here, here, and here.

Note 1: For readers of CGEL, note that I continue to cling to the older usage of determiner for the lexical category and determinative for the function. I have never seen a motivated explanation as to why Huddleston and Pullum chose to swap the significations.