Headline FLoP

« previous post | next post »

A headline from today's cnn.com entertainment page:

(1) Brinkley spouse slept with, gave teen $300K

This is a lovely example of FLoP coordination, what would be a routine Right Node Raising (with the NP teen shared between the two conjuncts), except that something extra, $300K, follows teen in the second conjunct, so that the two conjuncts are not parallel.

There are several extra twists in this one.

(Hat tip to Tim Miller.)

The background of the story: Brinkley is the model Christie Brinkley, whose husband, architect Peter Cook, admits (in cnn.com's telegraphic wording):

(2) I had sex with teen, paid her $300K to keep quiet

(This has a perfectly ordinary coordination of two VPs.)

First twist: as Miller noticed, (1) is entirely well-formed as a RNR example, with $300K as the shared NP, but on this parsing the meaning is preposterous in the real world: Cook slept with $300K (that's the absurd part) and gave a teen $300K.

Second twist: Miller suggested that the glitch in (1) is the result of space constraints on headline writers. And in fact, there's a well-formed RNR version of (1) with the intended meaning —

(3) Brinkley spouse slept with, gave $300K to teen

but it's one word longer than (1).

But we're talking about on-line headlines here, where space isn't such an overwhelming consideration; it's easy to run on to another line (see below). So my guess is that the headline writer just got tangled up, in the way that people who write non-headline FLoPs do.

Third twist: someone at cnn.com apparently noticed the problem, so that by the time I got around to checking the story, the head had been fixed, with the RNR replaced by an ordinary coordination of VPs:

(4) Brinkley spouse had sex with teen, paid her $300K

Now, this is two words longer than (1), and it's split between two lines (with a break between with and teen).

Of course, what's shocking about the story isn't the coordination, but the $300K in hush money.




  1. Bunny Mellon said,

    July 3, 2008 @ 2:19 pm

    What shocked me was that it was Peter Cook, but this must be some other guy, not the famous British architect who drew moving buildings.

  2. Melynda Huskey said,

    July 3, 2008 @ 2:42 pm

    This construction is clearly related in some way to a book title which plagues me each time I visit the grocery store: How to Make Your Children Mind Without Losing Yours. I can't put my finger on just what it is that nags at me in that title, but there's something amiss.

  3. Sili said,

    July 3, 2008 @ 2:53 pm

    Architect? Satirist, surely.

  4. hjælmer said,

    July 3, 2008 @ 3:07 pm

    Only if the teen is into necrophilia.

  5. Bunny Mellon said,

    July 3, 2008 @ 3:15 pm

    Peter Cooke the satirist died several years ago.

  6. Kris Rhodes said,

    July 3, 2008 @ 3:20 pm

    Is it accepted by most linguists that usages like (1) are in some way non-normative? Because when I am confronted by examples like (1) I usually have to puzzle out a little bit what is even supposed to be wrong with them. They read fine to me at a first glance. Is there anyone who takes examples like this not to be examples of non-normative speech (or anyway writing) but rather as unexplored examples of a previously unexamined _normative_ kind of language production?

    They way it "feels" to me is, you've got "…sleeps with []" and "…gives [] $300 million", and you string them together like so: "…sleeps with [], gives [] $300 million." Then you replace one of the brackets with the actual term the brackets "stand in" for. But you don't need to replace both brackets with that term, for the same reason you don't need to in "…crashes into [], damages []." I think it is usual to "fill in" the second bracket in cases like this, and so that's the one that gets filled in in (1).

    I take it that there are syntactic rules that linguists think hold over English which make the meaning of (1) strictly out to be that the $300 million was both slept with and given to the teen. But why not take examples like this to indicate that it turns out these syntactic rules are not valid for English after all? Why not read (1) as syntactically ambiguous?

    Sorry for the confused language. I took several linguistics classes as an undergrad, but that was long long ago.



  7. Tim M said,

    July 3, 2008 @ 3:28 pm

    >>But we're talking about on-line headlines here, where space isn't such an overwhelming consideration; it's easy to run on to another line (see below).

    I don't know, I get the impression from the cnn.com homepage that they try to keep the headlines in the "More News" section (where I saw this item) to one line so that their headings all align – whereas the headline you noted occurs with the actual article, and is able to make use of breathing room.

  8. Mark P said,

    July 3, 2008 @ 3:49 pm

    This Peter Cook is an architect, but not the British architect.

  9. mollymooly said,

    July 3, 2008 @ 4:15 pm

    @Kris: I can only say that I had to read (2) to understand (1), probably because this use of commas is restricted to the American dialect of headlinese.

  10. JBL said,

    July 3, 2008 @ 4:31 pm

    Melynda Huskey –

    I think it's because "mind" doesn't have an object. So it's not using it as in "Mind your manners", it's using it in "Do you mind?"

    So the title could be restated as "How To Annoy Your Children Without Going Crazy". But that's probably not really what the book is about.

    At least, that's what does it for me.

  11. Russell said,

    July 3, 2008 @ 4:35 pm

    Kris >

    It's true what you say, syntacticians generally take English to have a rule ("right-node raising") whereby some element at the right edge of a series of juxtaposed structures can be put off until the very end. Thus AZ notes that

    (3) Brinkley spouse slept with [], gave $300K to [] teen

    is well-formed, because in both cases "teen" would appear at the right edge.

    The question then becomes whether "FLoP" coordination (or, Right-node wrapping, as Neal Whitman now calls it) is then (i) an anomaly, an error in production, a blend of multiple structures, (ii) an reason to reject the "right-edge" part of the rule as too severe, or (iii) a similar but different construction (with particular contextual, genre-based, and possibly syntactic and semantic restrictions on it).

    One might compare this in broad strokes to if whether or not, previously described on LL.

    (ps. I tried to have a list of (a) (b) and (c), but you can see what that does…)

  12. Ralph Hickok said,

    July 3, 2008 @ 4:49 pm

    Melynda Huskey and JBL:
    I think the problem is that "mind" is being used as a verb, but the second phrase calls for a noun.

  13. John Lawler said,

    July 3, 2008 @ 4:53 pm

    @Kris: Yes, syntacticians would say it was non-normative in some way; that's a nice way to put it. And your reaction is normal, pretty much the same as anybody's. The reason why we have to work so hard to understand it is in fact the non-normativity.

    As you point out, we can see what it means eventually, and the way we do it is the way you describe. But — and it's a big but — it does take extra work, because we're not used to this particular way of leaving out a word in English. If (1) were instead:

    (1) Brinkley spouse slept with, paid off teen.

    Then we wouldn't have to work so hard, since teen would be perfectly obvious as the [] object of both verb phrases. But the way it originally stood, the [] teen has a different role in each clause, and the $300Kdistracts from the calculations.

    It's not so much that the rule isn't "valid", exactly; it's just that violating it exacts a penalty in processing, and may not in fact be processed at all by some people.

    As for the syntactic ambiguity, no human would ever interpret it that way, because we know you can't sleep with $300K in the intended sense. But a machine wouldn't, unless you told it, and that's what "syntactically ambiguous" means — syntax is just the automatic, mindless rules that allow us to put sentences together and interpret them at the speed of thought, without getting into all the procesing details consciously.

    There are a number of types of ambiguity besides syntactic ambiguity; in fact Arnold is someting of an authority on the subject, having co-authored a classic paper on the subject some years ago, entitled "Ambiguity tests and how to fail them".

  14. Orbis P. said,

    July 3, 2008 @ 5:20 pm

    I find (4) ambiguous too! Now it seems like the spouse paid Brinkley $300K after sleeping with a teen.

    As for Melynda Huskey's "How to Make Your Children Mind Without Losing Yours" example, it's a kind of syllepsis, the word "mind" being made to stand for both a verb and a noun — a bit of clever wordplay.

  15. dveek said,

    July 3, 2008 @ 5:25 pm

    Mollymooly brought up the difference between BritE and AmE comma usage. I (an AmE suser) also had trouble with 1), but I think my trouble was not so much that it conforms to AmE comma usage (because I think it doesn't, at least not as I was taught), but because it's badly "paralleled", to verb a noun (sorry).

    I would have written for 1):

    Brinkley spouse slept with, gave teen, $300K

    which is almost like 3) but with an extra comma. It doesn't look very "headline"-y, though.

    Can anyone recommend a good summary of differences in BritE/AmE comma usage, or punctuation difference as related to clauses, subordination, etc.?

  16. Kris Rhodes said,

    July 3, 2008 @ 6:09 pm


    To clarify, what I said was that I have to work to see what's _wrong_ with it. In other words, I have never had a problem reading things like (1), and have had to think a little bit to figure out what others were saying were wrong with them. But I'm just one guy… :)


  17. Kris Rhodes said,

    July 3, 2008 @ 6:12 pm

    Also, again to clarify, by "syntactically ambiguous" I meant syntactically ambiguous, and did not mean semantically ambiguous. :)


  18. Adrian Bailey said,

    July 3, 2008 @ 7:29 pm

    The headline led me up the garden path: "Brinkley spouse (was) slept with". I even considered "(The) Brinkley (who the) spouse slept with" before twigging. I think this is because list-style headlines are uncommon here in the UK (and passive headlines are common).

  19. Arnold Zwicky said,

    July 3, 2008 @ 11:40 pm

    I see that I should have added a discussion of different styles of punctuation, since several of these comments are really about the conventions for punctuating RNR examples. Some writers set off the second conjunct in RNR with commas on both sides (my own preference, in fact), others thriftily omit the second comma (perhaps on analogy with omitting the "serial comma" in conjunction). You can argue (as with the serial comma and other conventions of punctuation) that one version is clearer for the reader or that the other is more parsimonious of printed characters, but there is no difference here in the syntax of the sentences.

    My discussion preserved the punctuation convention in the original. There is no point of syntactic consequence here.

  20. Bunny Mellon said,

    July 3, 2008 @ 11:49 pm

    Note that 'Brinkley spouse' is the same as Geoff Pullum's recent British example, 'Winehouse husband'. He described the grammatic difference between that and 'Winehouse's husband' in his Comments section, I think it must be the headlinewriters' conventional form for those men who are married to famous women, it's just a teeny kick in the balls.

  21. Arnold Zwicky said,

    July 4, 2008 @ 12:35 am

    To Bunny Mellon: you have two examples of "X husband/spouse" and jump to the conclusion that the construction is limited to picking out men married to famous women. You neglected to check whether it also works the other way. And it does: googling on {"McCain wife" -McCain's} (the exclusion is to subvert Google's treating "McCain" and "McCain's" as equivalent) pulls up a number of examples, and no doubt you can find many other instances of "X wife/spouse" referring to women. Here's just one example:

    Why doesn't McCain wife want to open up her finances/records?

  22. Arnold Zwicky said,

    July 4, 2008 @ 12:42 am

    Adrian Bailey said: "The headline led me up the garden path: "Brinkley spouse (was) slept with". I even considered "(The) Brinkley (who the) spouse slept with" before twigging. I think this is because list-style headlines are uncommon here in the UK…

    It's your impression that list-style headlines are uncommon in the UK. Is there any evidence that this is so? (Such impressions about frequency are notoriously unreliable. I'm not claiming that your impression will turn out to be inaccurate, only that we have no way of knowing, and that such impressions often are.)

  23. Bunny Mellon said,

    July 4, 2008 @ 1:09 am

    Gosh, Arnold. So strict and so grumpy! But I didn't jump to any conclusions, you did: I'm sure it works both ways round. And I didn't neglect to check, I just wasn't interested.

  24. Barbara Phillips Long said,

    July 4, 2008 @ 11:03 pm

    While the linguistic discussion about the headline is interesting, the headline itself is badly written. A decent copy editor should have caught the problem. I will spare you a rant about the need for copy editing in the news business.

    What intrigued me about the headline was the aseptic use of "spouse" and "teen." At first glance, I'd have gone with "Brinkley's 4th husband paid girl $300K to hide affair."

    Of course, spouse is shorter than the word husband, and teen is more accurate, given that the girl was 18 when she and Cook met. Girl implies a greater age gap. The problem is, my suggested headline has a count of 49 and the article's original headline had a count of 41.

    Revising further, "Brinkley spouse paid teen $300K to hide affair" comes in at 43 and "Brinkley spouse paid teen $300K hush money" is 41.5. "Brinkley spouse slept with teen, paid $300K" has a count of 40.5.

    I might have made some counting mistakes, but it is possible to write a more readable headline that's close to the original in content and length.

  25. Kathleen said,

    July 5, 2008 @ 1:21 am

    Arnold's not grumpy, he's just a scientist who thinks critically. In my book, the words "I think it must be…" are a sure sign that a writer is poised to jump to a conclusion.

  26. Stephen Jones said,

    July 5, 2008 @ 7:32 am

    I had sex with teen, paid her $300K to keep quiet

    Can I sue my the career advisers I had at school when I was a teen for bad advice?

  27. Aaron Davies said,

    July 6, 2008 @ 10:20 pm

    Maybe there's something about the Brinkley case that attracts poor users of parallelism: I clipped the following lede from a July 1 New York Post article ("The other other woman") on the case:

    Christie Brinkley's lawyers may have one more weapon in their attempt to paint her wayward hubby as a callous cad — a married, 31-year-old fitness model with whom he had a relationship and then dumped, leaving her marriage in tatters, sources told The Post.

  28. Arnold Zwicky said,

    July 7, 2008 @ 1:40 pm

    Neal Whitman has a nice follow-up to this posting on his blog, here.

RSS feed for comments on this post