Hanging by a little thread
Spotted in the wild, in a teaser on the McClatchy Washburo frontpage, the dangling ... well, what should we call this? A dangling gerund phrase?*
McCain has big lead in South,
poll of 11 states shows
McCain enjoys a 16-point lead — 51 percent to 35 percent — among Southern voters, the poll, by Winthrop University and ETV, shows. And the further into the South you go, the larger McCain's lead. Eight-six percent of voters said race was not a factor in their preference, but one-quarter said having a Muslim parent was.
It's not a dangling participle, as in: "Having a Muslim parent, the campaign grew complicated." (If you'd like to score "Having a Muslim parent, Bubba said he could never vote for Obama" as a dangler too, go ahead; some danglerphiles disagree.) But it has the same sort of problem: Whose is the Muslim parent? The voter's (or voters'), or the mysterious candidate who goes unnamed in this blurb?
I really don't want to get into another round of survey nitpicking (hie yourselves to the originating paper and have at it, if you're inclined), but this is worth singling out:
And the further** into the South you go, the larger McCain's lead.
That's simply not true -- at least, whether it's true in some way or not, it isn't a finding of this poll. McCain's advantage in the whole sample is significant at 95% confidence. His advantage appears to be*** significantly larger in the Deep South states. It sounds cool and authoritative and intuitively sensible to state that as a linear relationship, but there's no evidence to support it. So don't.
* Nominations and analyses welcome.
** Ahem.
*** I am getting really, really tired of writers (and editors) who don't bother to report subgroup sizes.
McCain has big lead in South,
poll of 11 states shows
McCain enjoys a 16-point lead — 51 percent to 35 percent — among Southern voters, the poll, by Winthrop University and ETV, shows. And the further into the South you go, the larger McCain's lead. Eight-six percent of voters said race was not a factor in their preference, but one-quarter said having a Muslim parent was.
It's not a dangling participle, as in: "Having a Muslim parent, the campaign grew complicated." (If you'd like to score "Having a Muslim parent, Bubba said he could never vote for Obama" as a dangler too, go ahead; some danglerphiles disagree.) But it has the same sort of problem: Whose is the Muslim parent? The voter's (or voters'), or the mysterious candidate who goes unnamed in this blurb?
I really don't want to get into another round of survey nitpicking (hie yourselves to the originating paper and have at it, if you're inclined), but this is worth singling out:
And the further** into the South you go, the larger McCain's lead.
That's simply not true -- at least, whether it's true in some way or not, it isn't a finding of this poll. McCain's advantage in the whole sample is significant at 95% confidence. His advantage appears to be*** significantly larger in the Deep South states. It sounds cool and authoritative and intuitively sensible to state that as a linear relationship, but there's no evidence to support it. So don't.
* Nominations and analyses welcome.
** Ahem.
*** I am getting really, really tired of writers (and editors) who don't bother to report subgroup sizes.
Labels: grammar
posted by fev at 11:40 AM
6 Comments:
Since I think we can safely assume the question was about what affected their choice, I think it's a huge strain to get "MY having a Muslim parent" instead of "You-know-who's having a Muslim parent" out of that. After all, are you asking if it's the voter's race? Same thing.
Sometimes context really does remove ambiguity.
The voter's having a Muslim parent looks like an implausible reading even without context -- it's pretty obviously a dangler. On the other hand, without context the obvious reading is that it's McCain's parentage that is the issue.
Well, unless you think it's the voter's race in the first clause, I think this is a parallel that's too easy to understand to call any kind of problem (and it's not a dangler, as fev points out).
On that topic, though: nothing in English syntax prevents "dangling modifiers"; our grammar and our syntax no longer match. So we have to rely on semantics to sort them. As they say on LL, danglers aren't ungrammatical, they're just rude (because they make your reader work too hard), and when they're clear no one notices them.
But honestly, I read that paragraph several times before proceeding to the next one and still couldn't find the problem.
Who's will win for this election.
Ah, but that's the fun part! There is no next graf for this one; it's a standalone reefer for the story inside. So from your Euro-discourse-analytic-sorta perspective, it's an interesting example of what can and can't be left out of news, based on what the audience is assumed to share: You can say "Muslim parent" without saying "Obama" in the same place.
Agreed, most of the dangler family is in the ambiguous or WTF category, and that's a long scale. The Logsters do give an example of a really bad -- though "grammatically" flawless -- preposed modifier error: "Without U.S. support, Saddam's forces quickly crushed the rebellion," or something like that.
Sorry to keep returning to this, but I've been assembling notes for an all-in thumbsucker about news language, and some of 'em relate to this (and no, this is not what that nice editor meant by "revise and resubmit").
I don't think "race" is as clear-cut as "Muslim parent"; a lot of readings might very well involve the voter's race (or perceptions of it). Some possible issues:
1) Race is a property we all have, but it isn't always treated that way in discourse. The gender marking in "male nurse" (or, in Fox World, "female homicide bomber") suggests that the natural condition of nurses is female: The two sorts of people in that profession are "nurses" and "male nurses." A persistent criticism of U.S. news language has been that it tends to divide the world similarly into, say, "candidates" and "black candidates." So when people are talking about race, it's hard to know whether they mean "black and white" or "having 'race' and not having 'race'" (you can wear out the quote key in a hurry in criti-culti studies).
2) "Muslim" parent is pretty one-dimensional in what it means. "Race" isn't. Granted, it probably doesn't mean respondent's race alone, but we don't know whether it means "candidate's race (black)" or "candidate's race (white)." More to the point, we don't know if it means "candidate's race is same as respondent's" or "candidate's race is different from respondent's." Leading to ...
3) Right-wing discourse about race in the presidential campaign suggests some other stuff entirely; a lot of it is on the order of "how dare they complain about racism when blacks are 9-to-1 for Obama" (I think this is a subset of the bigger "racism isn't a problem anymore" category; others' mileage may vary). That suggests a bit of a third-person effect in there somewhere: An attitude sort of like "Race isn't an issue for people like me, so why don't those [derogatory culture-of-complaint noun here] just get over it?"
"How scared are you that Obama's dad was a Muslim?" yields pretty straight-ahead answers. Almost anything involving race is going to be trickier, especially when respondents have so many meanings available to them.
End of screed. No, I haven't become a peevologist overnight, but I don't think the bit of news language under discussion is as clear as it looks.
Post a Comment
<< Home