I haven’t looked through much of the prescriptivist blather that no doubt filled the Internet last week as a result of “National Grammar Day”, but I have given some of it a once over. At first I was somewhat concerned, because most of what I read was metered and about obviously bad grammar (misspellings, misapostrophications, misconjugations, and the like). Could it be that National Grammar Day would be as unexciting to me as it would be to any normal person? Could it be that all the invective-filled overzealous prescriptivists I’ve grown accustomed to fighting had fled the web? Well, that would be dreadful, as it would put me out of a blogging job and force me to refocus on my research and other tasks befitting a graduate student.

A fate worse than death, that’d be.

Thankfully, no sooner should I start worrying than my fears were assuaged. I went to the Society for the Protection of Good Grammar’s (creators of National Grammar Day) website, and found that they were running a poll for the city with the worst grammar in the U.S. Then I felt that feeling, that wonderful feeling, where disgust at someone’s self-righteousness becomes epicaricacy when you realize they were wrong about the very point they were so self-righteous about.

Before discussing the grammar of the poll, let me just state that the poll’s methodology is fatally flawed. In this poll, you’re voting for the most egregious error from amongst a set of 19 grammatically flawed examples. The town from which the most egregious error comes is then considered to have the worst grammar. Now, I was a math major who recoiled in terror from statistics, and even I can tell you that this isn’t how it should work. It’s biased against large cities, doesn’t reflect average “bad grammaticality” of the city, etc., etc. I know that this is a potshot, because obviously prescriptivists have more important things to think about than experimental design, but the principles of good poll construction oughtn’t to take hiatus just because it’s not National Poll Design Day. I fully expect their next poll will be more robustly designed.

On the grammar end, of the 19 candidates quite a number are just misspellings or word confusion, such as “DEVELOPMENT PROPRETY FOR SALE” and “Please use the provided receptable [instead of receptacle]”. People differ, but personally I don’t consider misspellings or most word confusions to be indicative of poor grammar, but rather of failed lexical access. I am definitely reluctant to call that bad grammar.

However, the one particularly odd item in this poll was a headline from an Altoona, PA newspaper:

(1) Ex-Indonesian president dies

I read this at least five times before I figured out what the supposed violation was; you may be more finely tuned and have figured it out faster, but if not, it’s that the prefix ex- is attached to Indonesian rather than to president. The claim here is that “ex-Indonesian president dies” would not be equivalent to (2a) or (2b), only to (2c):

(2a) Former Indonesian president dies
(2b) Indonesian former president dies
(2c) Formerly Indonesian president dies

and therefore the proper headline should be:

(1′) Indonesian ex-president dies

Which is stupid. Obviously, ex- is intended to modify the phrase Indonesian president, and that’s what it does. ex- is capable of modifying a whole noun phrase, not just the word it’s next to. That’s not unique in English; possessive ‘s can also modify a whole noun phrase from the periphery of the phrase:

(3a) The king of England’s crown is jaunty
(3b) *The king’s of England crown is jaunty

Some might argue that (1) is suboptimal because it is ambiguous between (2a) and (2c) — assuming you agree with me that (1) can have the meaning of (2c). I’ll grant that, but (1′) is ambiguous too, between (2a) and (2b). And what’s it matter? In this case all three of (2a-c) are true, since it seems to me that Suharto ceases to have any nationality when he ceases to be alive. Plus, (3a) is ambiguous between whether the king of the crown of England is jaunty or the crown of the king of England is jaunty, and we’re fine with peripheral ‘s in that situation. So there’s no reason that ambiguity should make (1) wrong. And ambiguity is the only apparent potential problem with the newspaper’s headline as it’s written.

Furthermore, (1′) has additional problems. In the Suharto situation, ex- directly modifying president is okay because Indonesian can modify ex-president, but consider:

(4) ex-New York Gov. DeWitt Clinton pushed hard for the Erie Canal.
(4′) *New York ex-Gov. DeWitt Clinton pushed hard for the Erie Canal.

(4′) is just plain wrong to me, and if I give the benefit of the doubt and try to parse it, I’d say it carries a heavy implication that DeWitt Clinton was a New Yorker who was a governor, but not the governor of New York. So why say that (1′) and (4) are right, when we could more neatly say that (1) and (4) are right?

And to conclude this silly debate, the AP Stylebook says that (1) is the proper form, not (1′).  So regardless of whether or not (1) is ‘good’ grammar, it’s not Altoona’s fault that their newspaper followed the conventions of the Associated Press. (Altoona, by the way, is in Western Pennsylvania, so I’ve got its back.)

All that’s well and good, but what really interested me is that ex-Indonesian president dies is considered nearly grammatically correct at all. For me, dies is in an unacceptable tense; dies is in the simple present tense, which in English implies a habitual or repeated action (I eat every day at noon), and not even ex-Indonesian president Suharto repeatedly dies. In fact, I would be hesitant to call (1) grammatical because of this; I’d much prefer ex-Indonesian president died or ex-Indonesian president has died. But I suppose simple present is commonplace in headlines, so it’s considered okay.  Stupid headlines and their preferential tense treatment.  I’m off to sulk over that.

Summary: ex- can — and I think ought to — be used to modify an entire noun phrase, not just its head noun.  So ex-Indonesian president is fine to mean former Indonesian president.  If you’re really concerned about ex- usage, just use former and formerly instead; there’s no chance of mistaking former as modifying an adjective nor of formerly modifying a noun.