Believed to be an F-18

« previous post | next post »

Anyone following the national news in the US yesterday — and perhaps many of you following international news elsewhere — has undoubtedly heard about the tragic crash of an F/A-18D fighter jet in San Diego yesterday morning. (The pilot managed to eject safely, but the plane crashed into a house, killing "[a] mother, her young child and the child's grandmother".) The fighter was on a training mission over the Pacific Ocean, and according to reports had already lost an engine over the ocean. Nevertheless, the pilot was apparently instructed to fly the jet to the Miramar Marine Corps Air Station, which requires flying over residential and business areas just south of UC San Diego (where my basement office in Language Log Plaza is located). This is where the plane crashed.

Wait, this is Language Log — what's this post about? Well, it turns out that this unspeakable tragedy has a couple of linguistic twists. First and foremost, two UC San Diego linguistics graduate students live a stone's throw away from where the plane hit; fortunately, both students (and their home) are safe and sound. Second, one of those students (hat-tip, Cindy) noted an early news report with the following structurally-ambiguous sentence:

The pilot of the plane, believed to be an F-18, reportedly ejected safely and parachuted to the ground.

(The source was our local ABC Channel 10 news website, but the text has since been changed; I found a clipping of it here.)

Of course, the intended, sensical reading is for the verb phrase believed to be an F-18 to modify the noun plane, like so:

But plane is embedded inside a prepositional phrase that itself modifies the noun pilot, and for reasons I won't go into here — go nuts, syntactically-oriented commenters — (many) English speakers are (often) initially tempted to have the verb phrase modify that 'higher' noun, even though the result is non-sensical:

So, as Cindy put it (in her message letting us all know that she and Rebecca were safe), this wording "make[s] it seem like the pilot is an F-18".


  1. Mark P said,

    December 9, 2008 @ 12:45 pm

    One word – how about "which"? – would have made it clear. I think reporters tend to use an artificially abbreviated style that can lead to things like this.

  2. Ryan Denzer-King said,

    December 9, 2008 @ 12:57 pm

    I remember Arnold doing a post on "low attachment", which if memory serves means that we want to associate constituents to the lowest possible node, so why in this case do we (incorrectly) what to attach "believe to be…" to the higher NP rather than the lower one?

  3. Jens Fiederer said,

    December 9, 2008 @ 1:02 pm

    Those of us who program computers fervently miss the use of the parenthesis, and wish there were a good way (we do "open" and "close" or "lpar" and "rpar", but it lacks fluidity) to render this in speech.

    "The pilot (of the plane believed to be an F-18) reportedly ejected safely and parachuted to the ground." is much less ambiguous in writing, but no better when read out loud.

  4. Eric Baković said,

    December 9, 2008 @ 1:03 pm

    Mark P — I think you'd need two words, "which is/was" … :-)

  5. Mark P said,

    December 9, 2008 @ 1:08 pm

    Eric – yes, of course. I was artificially over-abbreviating.

  6. J. said,

    December 9, 2008 @ 1:09 pm

    Interesting that an "unspeakable tragedy" has any "linguistic twists".

  7. John Cowan said,

    December 9, 2008 @ 2:25 pm

    In spoken language, or at least my idiolect of it, sentential stress deals with the problem: "The PILOT of the plane, believed to be a lieutenant" vs "The pilot of the PLANE, believed to be an F-18". But both of them sound like newscasterese, probably because they are too terse to be natural.

    I found the sentence in the linked article "It was unreal, except this was real" (said by one of the witnesses) to be interesting too.

  8. Nathan Myers said,

    December 9, 2008 @ 4:49 pm

    It was kind of stupid to mush these unrelated facts into one sentence. Cf. "The plane is believed to be an F-18. The pilot reportedly ejected safely and parachuted to the ground". The language is still questionable. Better, "Reports suggested the plane was an F-18, and that its pilot ejected and parachuted down."

    Who would be able to say how safe the ejection was? Ejection is almost always fraught with dangers known and unknown.

  9. Mark P said,

    December 9, 2008 @ 4:55 pm

    There is another language and journalism point here, the use of "believed to be." Who believed it? The civilian authorities at the site? Navy personnel? The reporter?

  10. June Casagrande said,

    December 9, 2008 @ 4:59 pm

    As a former reporter and editor and current sometimes copy editor, I can tell you that the job is a constant struggle to avoid little problems like this, made much harder by deadlines and breaking news and a writing style that pushes for an ever greater economy of words.

    I suspect that, under calmer circumstances, the copy editor would have inserted a "which is." I certainly would have — assuming I didn't have a managing editor and Web editor breathing down my neck. Regardless of whether a copy editor would have considered this a "mistake," the goal of the job is to assure precision and clarity (to minimize opportunities for confusion). As you pointed out, this sentence fell short of these goals.

    (Not to excuse it, of course. The story just conjures up such an eerily familiar "oops" feeling for me.)

    On a separate subject, when I saw you highlighted the words "a mother, her young child and the child's grandmother," I thought you might be about to address something I notice.

    Seems to me that women subjects of news stories are much more likely to identified based on their family relationships. That is, in news stories, a woman is often called a "mother" whereas in similar stories a man would be just a "man" — not a "father." Of course, that's completely subjective and disregards issues of newsworthiness that may make motherhood more relevant to certain stories. But still …

  11. Craig Russell said,

    December 9, 2008 @ 5:58 pm


    Of course, there is much that can be said about the differences in the ways men and women are represented in the news. But what interests me about the phrase "a woman, her young child, and the child's grandmother" is that there are other ways this group could be identified:

    -A young child, his mother, and his grandmother
    -A woman, her young grandchild, and the grandchild's mother

    The choice made by this story of how to identify these anonymous people seems to make a vague implication that, with a group like this, the "main" person should be considered the non-elderly adult, and the others should be labeled by their relationships to her.

    I wonder, if the victims had been a couple, if the plane would have perhaps killed "a man and his wife" or "a woman and her husband"?

  12. Tim said,

    December 9, 2008 @ 6:12 pm

    It seems to me that they would likely have said "a husband and wife", but I don't know.

    I also find it interesting, after looking farther into the article, that the grandmother was the mother of the woman who was killed. The fact that they said "a mother, her young child and the child's grandmother" led to me to assume that the grandmother was the father's mother. Why, I wonder, did they not say "a woman, her young child, and her mother"?

  13. Mark P said,

    December 9, 2008 @ 6:21 pm

    The identification of the victims might have been an almost random choice among the ways they could be identified, or it might say something about the unconscious way we view women. The choice might also have been intentionally or unconsciously made to emphasize the emotional content of "mother." Or the reporter might have run into a cul-de-sac where the only exit seemed to be one awkward phrase or another.

  14. Ellen said,

    December 9, 2008 @ 7:08 pm

    Nathan Myers wrote: It was kind of stupid to mush these unrelated facts into one sentence.

    Perhaps. But it also seems to be a pretty common feature in journalistic writing — adding parenthetical comments to convey information, even when it's really unrelated to the rest of the sentence.

  15. Merri said,

    December 10, 2008 @ 3:59 am

    AG : the mathematicians' way to express the dilemma in deciphering the abovementioned sentence is that, while there are commutative (e.g. Latin) and non-commutative (e.g. English) languages, no known language is associative, i.e. many sentences are understood in two or more ways, according to where you stop for breath (or parentheses). There is the well-known "time flies like an arrow, fruit flies like bananas", but also umpteen involuntary sentences from all written sources.

    The last examples from press I've seen are :

    – J ury Convicts New York Man of Killing Wife for the Second Time.
    – Panda moved after China quake gives birth to twins.
    – Teacher OK after Crashing into Bear on a Bicycle

    and 'World Wide Words' finds new ewamples each week.

    The problem is compounded, in English as in Chinese, by the fact that words' grammatical categories are loose : there are few languages where 'eye' may be a verb.

  16. Karen said,

    December 10, 2008 @ 5:41 am

    "I also find it interesting, after looking farther into the article, that the grandmother was the mother of the woman who was killed. The fact that they said "a mother, her young child and the child's grandmother" led to me to assume that the grandmother was the father's mother. Why, I wonder, did they not say "a woman, her young child, and her mother"?"

    Wow. That's staggeringly bizarre.

  17. Alex B said,

    December 10, 2008 @ 7:17 am

    To Karen,

    Because that would make the woman the focus of the report, and only mention the child in relation to her. Instead, the story is "Jet kills child". Hence, the child is taken as point of reference and the others are mentioned in relationship to the child. Perhaps grammatically not as clear, but arguably more emotionally effective.

  18. outeast said,

    December 10, 2008 @ 7:21 am

    I think you'd need two words, "which is/was" …

    Is, shurely? And I'd have wanted to see 'which is believed to have been an F16'… is it still an F16?

    Hey ho, so much scope for pedantry. Or pedanticism.

  19. Jorge said,

    December 10, 2008 @ 8:05 am

    Why, I wonder, did they not say "a woman, her young child, and her mother"?

    Maybe that could be a lesbian couple and their child?

  20. Chris said,

    December 10, 2008 @ 9:13 am

    Who would be able to say how safe the ejection was? Ejection is almost always fraught with dangers known and unknown.

    "Safe" is often used of a past event to mean "no injury actually resulted", regardless of what the risks were. Thus you hear of, e.g., drunk drivers getting home safely (instead of getting home fortunately unharmed, which would be more accurate).

    I suspect the safe ejection is the same phenomenon. It must have been safe, because the pilot wasn't hurt by it (this time).

  21. Merri said,

    December 10, 2008 @ 11:29 am

    @ Outeast : no, it isn't a F-16 anymore. It is a heap of bent steel. So 'to have been' is right. That's where the past infinitive's resultative meaning comes in handy.

  22. David Clausen said,

    December 10, 2008 @ 2:57 pm

    Seems to be a case of a reduced relative clause which are notorious for attachment ambiguity. The second interpretation corresponds to the extraposition of the reduced relative from its original attachment site on 'the pilot'. Whereas the first interpretation corresponds to a local, non-extraposed, attachment to 'the plane'. Interpretation tends to favor local attachment of the reduced relative, although additional information can contribute to disambiguation such as explicit relativizers and contextual information so local attachment is by no means necessary, giving rise to the possible, but in my opinion implausible 2nd interpretation.

  23. dr pepper said,

    December 10, 2008 @ 3:44 pm

    Today, CNN referred to the bereaved survivor as "The husband and father".

  24. Barbara Phillips Long said,

    December 10, 2008 @ 10:15 pm

    Merri —

    Pittsburgh Business Times headline:

    Wall Street closes up on Monday


    For those debating the way the women were described, I think sometimes writers fall back on stereotypes and sometimes writers rely on information from police, firefighters or other emergency personnel. If rescuers didn't know exactly what the relationships were, they may not have been able to provide information about whether the grandmother was the mother of the young child's mother.

  25. rpsms said,

    December 13, 2008 @ 10:52 am

    My first guess was that the report was written before they knew for sure the relationship between the grandmother tand the child's mother.

    It seems nitpicky, but relatives and friends read these reports and might come to the wrong conclusion about which of their loved ones just died.

    The article may have inset a picture of the husband, so the wife may have lead first to establish a connection between the image and the story.

  26. Baylink said,

    December 14, 2008 @ 5:07 pm

    The pilot of the plane — believed to be an F-18 — was….

    Parentheticals always modify the closest noun, no?

    (Yes, I am not a linguist, but perhaps that precisely why I don't have any trouble with this sentence, though I would punctuate it differently.)

  27. Arnold Zwicky said,

    December 14, 2008 @ 6:04 pm

    Baylink: "Parentheticals always modify the closest noun, no?"

    Not exactly. The generalization is often stated in terms of *words* ("nearest *noun*"), but the phenomena have to do with *phrases* (constituents). In NPs that have post-head modifiers ending in a NP, there are then (at least) two possible candidates for "nearest NP": the whole thing, or just the NP-final NP — "high" modification vs. "low" modification. (The point is that both of these NPs immediately precede the modifier; both are "nearest".) There's a huge literature on high vs. low attachment, a topic that's come up here on Language Log a number of times over the years. But there are certainly cases of high attachment and also cases of low attachment, and the difference between them seems to have much more to do with non-structural factors (like discourse organization) than with structural factors.

    In the case at hand, the most likely interpretation depends in part on what went before. But the fact that "the pilot of the plane" is the subject of the sentence, and therefore likely to be topical, does make it a plausible candidate for the modified NP (even though that doesn't take sense).

RSS feed for comments on this post