New twist on a classic misnegation

« previous post | next post »

Jared Dubin, "The NBA’s Other Offensive Revolution: Never Turning The Ball Over", FiveThirtyEight 3/14/2018 [emphasis added]:

We’re in a golden age for NBA offense. Teams are scoring 110.1 points per 100 possessions during the 2018-19 season, according to — a full 1.3 points per 100 possessions more than the previous high of 108.8, which was set two years ago.

This is largely — and rightly — credited to the boom in 3-point attempts. […]

But while the genesis of the other offensive changes can be neatly traced, the decline in turnovers is a bit more puzzling. […]

Regardless of why, the impact of turnovers cannot be undersold. […] You can’t score if you don’t have the ball.

Our interminable catalog of misnegations has many entries dealing with analogous cases, e.g. "We cannot/must not understate/overstate", "Overstating understatement", "'Cannot underestimate' = 'must not underestimate'", "…not understating the threat", "(Not) Underestimating the Irish Famine", "Overestimating, underestimating, whatever", "'Impossible to understate' again", "'Hard to understate'", etc. etc.

For a simple explanation from first principles, see "The Estimation Game", 4/3/2014.

I think this is the first example we've catalogued that involves over-/under-selling rather than stating, estimating, emphasizing, and so on. However, it's far from the only example Out There.  Here are the other five top examples of "cannot be undersold" from the current Google New index — and note that every one of them is upside-down:

[link] After the Flowers news was revealed, Renners called the deal a home run in a piece, and said the impact that it makes for the defense cannot be undersold.
[link] Solange’s ability to elevate southern black aesthetics as high art, illuminate the oft-forgotten work of DJ Screw, and build space that allows for black femininity to become an entryway to spirituality cannot be undersold.
[link] And while that's certainly a worthwhile debate to have, the importance of that moment, a kiss shared between two proud warriors in love with each other, cannot be undersold.
[link] He knows this offense inside-out and that cannot be undersold.
[link] It cannot be undersold just how good the Pittsburgh Steelers offensive line has been this season.

So we need to repeat a question asked many times before: if (almost) everyone (almost) always uses a turn of phrase "wrong", doesn't that make their usage actually right? After all, we linguists believe with Horace that language changes "si volet usus / quem penes arbitrium est et ius et norma loquendi" ("if it be the will of custom, in the power of whose judgment is the law and the standard of language"), right?

I don't think so. Norma loquendi certainly governs word meaning and morphosyntax, which are basically matters of cultural convention. Things are arguably different in compositional semantics, which is the logical combination of simpler meanings to form more complex ones.

Of course, I've argued elsewhere that "could care less" is an idiom, and therefore OK.  So why not treat "cannot be under(estimated/stated/sold/etc.)" the same way? For me, it just doesn't feel right to do that — but maybe I'm being a hypocrite.

The obligatory screenshot:

[h/t Rick Rubenstein]


  1. MattF said,

    March 17, 2019 @ 12:08 pm

    Hmm. OK, rather than 'cannot be overstated', I guess. But maybe there's some interference with 'cannot be denied', which is not a misnegation.

  2. James said,

    March 17, 2019 @ 12:14 pm

    Yeah there's always a sneaking suspicion that the modal could be deontic, in which case the negation isn't mis- at all. (I know this has been discussed more than once in the earlier threads.) I can't really hear it that way in this case.
    Isn't there a performance/competence distinction here? I would give good odds (appropriate or 538) that Jared Dubin would recognize his own expression as over-negation if he thought it over. (Unlike, I am assuming, most uses of "I could care less…")

  3. Gregory Kusnick said,

    March 17, 2019 @ 12:40 pm

    Bleedover from "We will not be undersold!" perhaps?

  4. Paul Kay said,

    March 17, 2019 @ 2:03 pm

    I find James's comment thought-provoking. I wonder if some clever person could design an experiment to test how the Jared Dubin's of the world would respond to queries along the lines James proposes. i can imagine replies like, "Yeah, I'm always a little uncertain whether it's overX or underX."

    I agree with Mark that when it comes to logical composition a usage can just be wrong.

    BTW, I've often wondered about sentences like "Romain Gary was born Roman Leibovich Katsev" (Wikipedia). Chances are the author was only named after he was born. There may be odd cases these days when a fetus's sex is known and a name is given before a baby is born, but surely that is rare even now. This turn of phrase is not restricted to English.

  5. KevinM said,

    March 17, 2019 @ 3:46 pm

    It is incorrect, but didn't seem so to the writer. The reason may have something to do with the "forbidden" vs. "impossible" sense of can't. "Can't be undersold" doesn't raise a red flag because the speaker agrees that underselling is something that shouldn't be done here. That there is no degree of "selling" that would be excessive is a different thought, but not an inconsistent one. So I wouldn't quite equate this to classic "misnegations," such as the use of a double negative when only one was intended.

    [(myl) The trouble with that explanation, as discussed many times in earlier posts, is that every example of "cannot be over/under-Xed" or "cannot over/under-X" has corresponding phrases "impossible to over/under-X", "hard to over/under-X", "difficult to over/under-X", etc., with exactly the same pattern of apparent errors.]

  6. Viseguy said,

    March 17, 2019 @ 4:29 pm

    I'm not even sure what "selling", whether under- or over-, has to do with any of the cited examples. Who's selling what, to whom? It just seems like a malapropism to me. (Gregory Kusnick is onto something, I think; the TV hucksters' "we will not be undersold!" may be fueling this usage.) On the other hand, maybe "undersold" means something like "undervalued". In which case we're back in underestimate/understate territory, aren't we?

  7. Michael said,

    March 18, 2019 @ 3:20 am

    Paul: that "born" expression is framed after "nee" which is correct.

  8. Stan Carey said,

    March 18, 2019 @ 4:25 am

    One I read recently:

    All of which is to say that we can never underestimate the psychological impact of language’s massive migration from the ear to the eye, from speech to typography.

    It very much fits the pattern, but the culprit was a bit surprising: Neil Postman, in The Disappearance of Childhood.

  9. Rodger C said,

    March 18, 2019 @ 7:23 am

    that "born" expression is framed after "nee" which is correct.

    I don't follow at all why one word for "born" is more logically correct than another.

  10. Cervantes said,

    March 18, 2019 @ 8:30 am

    Well, the writer's intent in "cannot" may be that it would be a mistake to do so, that it should not be done. That seems perfectly logical.

  11. Robert Coren said,

    March 18, 2019 @ 10:23 am

    Paul. Michael, Rodger: The issue isn't the (non-existent) distinction between born and nee, but the inclusion of the given name. My ragbag of a brain contains a line from George Bernard Shaw's The Devils' Disciple, in which a lawyer reads from a will that the deceased has left a bequest to "My wife, Annie Dudgeon, born Annie Dillard", and pauses to note that the phraseology is faulty, saying to the eldest son, "Your mother was not born Annie; she was christened so."

  12. Ursa Major said,

    March 18, 2019 @ 12:37 pm

    One could probably make an argument that the act of bestowing a name on a new-born child is retroactive to (or even before) the birth. Consider a hypothetical exchange: "What is her name?" "We haven't decided yet" – implying that the baby has a name already.

  13. Keith said,

    March 18, 2019 @ 4:01 pm

    I disagree with the idea "could care less" is OK, simply because there are lots of people who say that. You may as well argue that a whale is a fish, because there are people who claim that to be true.

    (Although there is, apparently, no such thing as a fish.)

    Maybe a better analogy would be to claim that because there are lots of children who claim that there are monsters under the bed, that this is therefore true.

    Well, I say no.

    There are no monsters under the bed, no matter how many children claim that they are there.

    And no, "could care less" is not OK if you are claiming to have no opinion about something. If you state that you could care less, you have some opinion and could, possibly, have less of an opinion.

    Come on, it's not rocket surgery. It's just apathy. Maybe.

  14. Keith said,

    March 18, 2019 @ 4:10 pm

    @Michael and @Robert Coren

    This idea of "born" being equal to "nee" is nonsense.

    The term "née" is the feminine past participle of the verb "naître" and is used in French to indicate the family name at birth of a female child, as distinct from the woman's name after marriage. It was carried over into English usage where is it the strict equivalent of "maiden name".

    I suppose its adoption was influenced by Anglicised Norman-French legal terms very common in American legal usage, such as lessee, tansferee, licensee…

  15. Philip Taylor said,

    March 19, 2019 @ 3:57 am

    'And no, "could care less" is not OK if you are claiming to have no opinion about something. If you state that you could care less, you have some opinion and could, possibly, have less of an opinion'.

    Great to see a contributor here being willing to condemn something as "wrong". More proscription and prescription, please — description is often little more than a fashionable way of throwing in the towel.

  16. Robert Coren said,

    March 19, 2019 @ 9:43 am

    @Keith: It's my impression that in the context we were discussing, "born" is being used in the same way.

    I remember long ago reading a Ring Lardner story in which he indicated that a male character had changed his surname (to something less Jewish-sounding, if I remember correctly) by including a parenthetical "(né [original-name])", and being impressed that he omitted the feminizing final e.

  17. ktschwarz said,

    March 19, 2019 @ 11:41 am

    The interesting scientific question is how, even though misnegations are compositionally wrong, they go unnoticed and readers gather the intended meaning anyway. Compositional semantics is not the only strategy for understanding, not even the most important. It can be overridden by the emotional tone of the context and by real-world background knowledge. These misnegations suggest that faced with more than one negative, or a negative combined with a threshold ("enough", "too", etc.), our monkey brain gives up on composition and falls back on context and background.

    I'd like to call misnegations bad style, since they force readers to filter out words that contribute nothing to the sentence.

  18. Andrew Usher said,

    March 22, 2019 @ 7:44 am

    I have to agree that the biggest problem with the original example is using the metaphor of 'selling' in a context where it seems to make no sense. I can't even tell what it means to be 'undersold', though I imagine from context it is a synonym of 'underestimated', in which case it's wrong.

    As for born/nee, that quibble is silly: in English we've long since reserved 'nee' for women's maiden names, so using 'born' for any other occasion of name change make sense and soon I think 'nee' will disappear except perhaps in the law. The expression isn't literal anyway: one can be 'born' male or female, but not 'born' a name, even if it has already been chosen – but everyone knows what it is shorthand for, and that's enough to make it acceptable (and certainly grammatical).

    k_over_hbarc at

  19. TIC said,

    March 24, 2019 @ 5:34 pm

    FWIW, I tend to bristle at and disapprove of those mis-phrasings whose words combine to literally denote the *opposite"* of the intended connotation… Clearly, I'm a prescriptionist/proscriptionist (or is it proscriptivist/prescriptivist?) in this regard…

    In the same vein, and probably not surprisingly, misprhasings such as "three times closer" and "ten times smaller" always catch in my ear and drive me to distraction…

    I also agree that the fascinating question is why/how such misphrasings, including but not limited to quasi- and actual misnegations, are so easily and commonly spoken and written *and* apparently so easily and commonly understood — seemingly without hesitation or conscious thought — with their intended meaning…

    Using 'born X' as shorthand for 'named X at (or shortly after) birth' is not in the least problematic for me… And if I ever heard, in response to "What is her name?", the response "We haven't decided yet", it would never in a million years occur to me that the speaker was 'implying that the baby has a name already'… I'd assume, as I'm sure most hearers would, that the parents simply haven't yet agreed up on a chosen name…

  20. ktschwarz said,

    March 25, 2019 @ 2:06 pm

    Recommended reading on "X times smaller": Arnold Zwicky in Language Log (2008) and further discussion by Victor Mair and others (2012). In short: it's an established idiom, scientists use it in scientific writing, and it's been used in English for hundreds of years going back to Newton, Jonathan Swift, and David Hume.

  21. TIC said,

    March 25, 2019 @ 6:13 pm

    Thanks, ktschwarz… I've read both of those posts before, but I dutifully read them again… This time I set a personal record — I made it through about half of the (67!) comments following VHM's post… My previous record was about a quarter… And I'm now convinced that it's perfectly logical, understandable and acceptable to say that this time I didn't read about two times fewer of the comments than last time…

RSS feed for comments on this post