Can they even prove that?

« previous post | next post »

In my last post puzzling over even ("What does 'even' even mean?", 2/8/2011), I suggested that there's a new even idiom, exemplified in phrases like "How does that even work?" or "What does that even mean?", in which even has become simply an intensifier rather than a scalar focus particle. If this is true, it would be a rebirth of even's pre-16th-century use as (in the OED's gloss) "an intensive or emphatic particle" that can be "prefixed to a subject, object, or predicate, or to the expression of a qualifying circumstance, to emphasize its identity". However, the 94 helpful comments on that post left me wondering whether this is really happening.

So here's another even example that brought me up short — Michael Hinkelman, "Feds unveil 50-count indictment against 'Uncle Joe,' 12 others", Philadelpha Daily News 5/24/2011:

But Joseph C. Santaguida, Ligambi's attorney, said the reputed mob boss, who pleaded not guilty to all charges and has a bail hearing Thursday, claimed that the feds' case was "weak."

Asked if Ligambi, dressed in a white polo shirt and jeans, was the mob boss portrayed by prosecutors, the defense attorney said: "I don't know if [prosecutors] can even prove that. I don't think it's that strong a case."

I'm not going to go over the standard treatment of even as a scalar focus particle, and its interaction with questions and negation — you can read the earlier post or (for a more complete and technical treatment) see e.g. Anastasia Giannakidou, "The landscape of EVEN", NLLT 25:39-81, 2007. But I'll give what I hope will be a helpful example for comparison, one where the scale involved is simply numerical. Consider someone discussing a basketball player's listed height:

I don't know if he's even 6' tall, much less 6'4".
??I don't know if he's even 6' tall, much less 5'8".

As far as I can tell, Mr. Santaguida's comment similarly goes in the wrong scalar direction. What's provably true, it seems to me, is higher on the relevant implied scale than what's true:

I don't know if that's even true, much less provable.
??I don't know if that's even provable, much less true.

Of course, it's possible that a different focus-conditioned scope for even was missed in the quote's transcription:

I don't know if [prosecutors] can even prove THAT [much less the other, more specific charges].


  1. Ø said,

    May 24, 2011 @ 8:32 am

    Does "I don't even know if they can prove that" play a role here?

  2. William Ockham said,

    May 24, 2011 @ 8:48 am

    I interpreted the comment to mean "can even prove THAT". Too bad we don't have audio.

  3. Jayarava said,

    May 24, 2011 @ 9:34 am

    This new usage resembles the Sanskrit and Pāḷi evaṃ/eva E.g.
    Pāḷi: kiṃ tass'eva atthaṃ atthi? 'What does *that* even mean?'
    Or kiṃ tassa atthameva atthi? 'what does that even *mean*?'
    – probably no relation though.

  4. Cecily said,

    May 24, 2011 @ 9:35 am

    A four-year-old I know has recently started using the construction "what even IS that?"

  5. Jon Weinberg said,

    May 24, 2011 @ 9:46 am

    I suspect that MYL's closing suggestion, and Wm. Ockham's, are correct: "I don't know if they can even prove that he was a mob boss, much less that he committed the acts alleged in the indictment."

  6. Ben Hemmens said,

    May 24, 2011 @ 9:54 am

    This is making me very curious about whether I finally need to do Semantics 101 or possibly just lack the gene for tasting the different flavors of "even" ;-)

    But I tend toward your last possibility. Assuming the lawyer was being careful with his words. Maybe the "even" was just there for phonetic/rhythmic reasons. Another person might have inserted an ahh.

  7. GeorgeW said,

    May 24, 2011 @ 10:06 am

    I interpreted it as in myl's final suggestion: "I don't know if [prosecutors] can even prove THAT [much less the other, more specific charges]."

    However, it might have been stronger for the lawyer to say, 'they can't even prove the other, more specific charges much less that.'

  8. John Roth said,

    May 24, 2011 @ 10:07 am

    We've got a lawyer for a mob boss. I suspect he thinks that being able to prove something in court for a jury doesn't mean it's necessarily true.

  9. Jason said,

    May 24, 2011 @ 10:25 am

    The other day I found myself saying something similar to the following to someone:

    Ok, this idea of a "structuring absence?" What even is that, let alone what relevance does it have to stereotype theory?

    And then I paused to ponder the strangeness of this utterance. And went off to google, and discovered that I was by far not a strange freak of nature, but had muchness company indeed:

    Emotional Cheating? Yeah… What Even IS That?…/emotional-cheating-yeah-what-even-is-that/

    What even is that? I don't know what is funnier. His expression or that he's holding a drink.…/what-even-is-that-i-dont-know-what-is-funnier

    What even? Is that a skeleton?

    I can't even find it in myself to get Facebook Connect (what even is that?)

    Offered for data points. What even, right? It's invaded my head, at least.

    [(myl) I don't think there's any question that "what even *is* that?" is an example of a new set of even-expressions. But the semantics of this example seems congruent with the traditional scalar-focus-particle interpretation: you're asking something like "Is 'a structuring absence' even a coherent and comprehensible concept, much less one that's relevant to the topic under discussion?"]

  10. J. W. Brewer said,

    May 24, 2011 @ 10:32 am

    Jon Weinberg's analysis seems most plausible, but myl's sense of scalarity is inapplicable to this discourse context if he is assuming that, because truth seems logically prior to provability,* a defense counsel is ever under any circumstances going to voluntarily disclose or confirm a true-but-unprovable fact which would be disadvantageous to his client's legal position (there might ultimately be tactical reasons to make such an admission at trial, but a pretrial exchange with a reporter is an unlikely context to do so). Changing the subject from truth to provability is the lawyer's job, and his quirks of usage are going to reflect that, although obviously it's optimal for the language not to be sufficiently jarring to the listener/reader that the subject-changing is unduly conspicuous.

    *And the fact that a jury unanimously concludes that a particular proposition has been proven beyond a reasonable doubt does not provide a 100% guarantee that the proposition is, in fact, true. Indeed, the so-called Alford plea is designed for use in situations where a defendant wants to assert "although X is not in fact true, I accept that a reasonable jury might well conclude from the prosecution's evidence that X has been proven beyond a reasonable doubt."

  11. Shmoo-El said,

    May 24, 2011 @ 10:57 am

    I think the idea may be that lawyers *can* prove something that is not true. That is, it's a comment on the strength of the argument being put forward ("the case").

  12. John Cowan said,

    May 24, 2011 @ 11:01 am

    We all hope that nothing is provable that isn't true, but ~G (the negation of the Goedel string) might turn out to be provable in elementary number theory, which would mean that number-theory is ω-inconsistent, and non-standard analysis is the only kind there is. Every sane number theorist assumes it's not — but there is no actual proof.

  13. Shmoo-El said,

    May 24, 2011 @ 11:02 am

    We've got a lawyer for a mob boss. I suspect he thinks that being able to prove something in court for a jury doesn't mean it's necessarily true.

    Sorry John Roth. I see you beat me to more or less the same point. One quibble however. Aren't all defense lawyers required to put the best case forward for their client, regardless of what they think the "truth" may be? A defense lawyer, unlike a scientist, is not required to voluntarily put forth evidence that could damage their client's case.

  14. ironhorse said,

    May 24, 2011 @ 11:46 am

    Something like "I don't know if he can even prove that…" much less persuade a jury to convict.


  15. Rube said,

    May 24, 2011 @ 11:56 am

    I think there's something subtler going on. If he simply says "I don't know if they can prove that. I don't think it's that strong a case", he sounds weak, as if he doesn't know if he can stand up to the prosecutors or not.

    By adding the "even", he somehow makes the prosecution sound weak. They can't "even" do something. He's shifted the ground through the addition of one word.

    But what do we call that word?

  16. Ran Ari-Gur said,

    May 24, 2011 @ 12:05 pm

    @John Cowan's "Every sane number theorist assumes [~G]'s not [provable in elementary number theory]": is that because sanity causes the assumption, or because the lack of the assumption causes insanity? (Or is your statement just vacuously true, there being no sane number theorists to begin with?)

  17. Nick Lamb said,

    May 24, 2011 @ 12:07 pm

    Shmoo-El, as far as I know the lawyer's ethical duty to their client is restricted by their knowledge of the facts. Usually all they know is what the client tells them, which is fine. But if they know the crime was committed by the client, either from being there or because the client confessed, that's a problem.

    Remember although the accused can't commit perjury (unlike other witnesses the accused doesn't address the court under oath) their lawyers can.

  18. Jens Fiederer said,

    May 24, 2011 @ 12:29 pm

    I think the "different focus-conditioned scope" explanation is the only one that even makes sense.

  19. iching said,

    May 24, 2011 @ 12:31 pm

    @John Cowan: It's gonna take me a while to get my head around that. If there are true statements that can't (even in principle) be proved, what does "true" even mean? And if there are false statements that can be proved, what does "prove" even mean?

  20. J. W. Brewer said,

    May 24, 2011 @ 12:49 pm

    Perhaps it would help to gloss "prove" as meaning something like "establish the likely truth of to the greatest degree of reliability that is typically feasible in the particular context." That's going to be one thing in mathematics; something else in other fields of human endeavor.

  21. Mr Fnortner said,

    May 24, 2011 @ 1:14 pm

    I first encountered "A lie is as good as the truth if you can get someone to believe it," in a Peanuts comic as I recall probably 40 years ago (I think Linus was commenting on someone else's prevarication). I think this is the sort of truth at play here: something you can get someone–a jury perhaps–to believe.

  22. Eric P Smith said,

    May 24, 2011 @ 2:01 pm

    @ itching: Yes, there are true arithmetic statements that cannot (even in principle) be proved. That is paradoxical, and yet it belongs at the heart of Mathematical Logic. There are many books that try to explain the matter to the non-mathematical reader, but my favourite is one of the earliest: Gödel, Escher, Bach, An Eternal Golden Braid by Douglas R. Hofstadter. I can recommend it.

  23. Brett said,

    May 24, 2011 @ 2:11 pm

    @iching: Questions about true but unprovable statements are a very tricky area of mathematics, where there is no clear agreement. There isn't generally a problem with false but provable statements, however. If a false statement is provable, then every statement is both provably false and provably true; in other words, the system is inconsistent.

    Regarding true but "unprovable" statements, the question really comes down to the definition of "provable." For Godel's incompleteness theorem, this means "provable using first-order logic." I believe that the correct interpretation of Godel's theorem is that it demonstrates that first order logic is incomplete, in that it cannot prove some statements that are clearly true. (Many others disagree with this characterization, however.)

    Consider a statement that claims to hold for all the integers (or on some other countable set). If it were false, there would be some integer for which it does not hold. This counterexample could be found, and it would constitute a disproof. Conversely, if no disproof exists, the statement is true. This proves the statement, but this method of proof cannot be executed in first order logic. Hence a proposition like Godel's can be but undecidable in first order logic but provably true in a more general system of logic.

  24. D.O. said,

    May 24, 2011 @ 3:31 pm

    I suggest, without a shred of evidence, that "if [prosecutors] can even prove that" is a shorthand for "if [prosecutors] have enough evidence that we need to bother putting forward a defense".

  25. Chris Holdaway said,

    May 24, 2011 @ 3:53 pm

    On the subject of mathematical proofs:

    The mouseover being particularly relevant:

    HARD: prove that 1+1=2
    REALLY HARD: prove that stuff can equal other stuff

  26. Mr Punch said,

    May 24, 2011 @ 4:52 pm

    I agree with the final interpretation, and think "even" is misplaced – it should be "they can't prove even that." The key to the overall "even" issue is, I believe, the phrase "not even wrong". "What does that even mean?" means "That's not even wrong, it's meaningless." It's not exactly an intensifier – it has the effect of shifting the focus of the discussion.

  27. Marge said,

    May 24, 2011 @ 5:56 pm

    To prove something beyond reasonable doubt it need not be true. So "true" > "provable in a court of law", even if "mathematically provable" > "true".

    Wait… more than true? What does that even mean?

  28. Dave said,

    May 24, 2011 @ 6:39 pm

    @brett, @iching:

    There's plenty of "clear agreement" about true but unprovable statements, but it is a topic that seems to invite cranks, sometimes creating the illusion of serious disagreement.

    Brett's comments are simply mistaken in a few regards. First: if a false statement is provable, that doesn't have the result that the system is inconsistent; it might only be omega-inconsistent. See John Cowan's earlier comment.

    The biggie: first-order logic is not incomplete. It's complete. Godel proved this in 1929, before he went on to do the incompleteness stuff. It's arithmetic that's incomplete.

    Third: checking every integer is not a proof procedure. Suppose a statement in fact holds for every integer, and you set off to check it. At every step, you'd be left wondering whether it really holds for every integer or whether you just hadn't found a counterexample yet, even though one might be right around the corner. Checking more integers is not going to get you out of that situation. What you need to prove that a statement holds for every integer is, well, a proof.

    For all concerned: Eric P Smith's book recommendation is worth following up on; I'd also recommend Francesco Berto's There's Something about Godel. Both are approachable and cover the basics of this stuff. For a super-quick-and-dirty overview, there's this link:

  29. DaveK said,

    May 24, 2011 @ 8:21 pm

    Does anyone else think that this use of "even" might owe something to a confusion with "ever"? "I don't know if they can ever prove that".
    "What ever does that mean?"
    I doubt it's the only reason for the usage but it may be an ingredient in the psychological stew.

  30. micah said,

    May 24, 2011 @ 8:31 pm

    This article reminded me of this post, though the key word in it is "already" rather than "even":

    "The already small crop of candidates running for the 2012 Republican presidential nomination might be bolstered by one more if former New York mayor Rudy Giuliani decides to jump in the ring."

  31. Brett said,

    May 24, 2011 @ 10:23 pm

    @Dave: Since mathematical logic is not the main topic here, I am not going to get into a long discussion. As I conceded, there is significant disagreement about these issues. All (or almost all) of it is related to disagreements about what techniques of "proof" are allowed. (Equivalently, it depends on what one means by a statement being "true.") I am a strong believer that first order logic is not the only proof technique that is allowable. This is not a common viewpoint among logicians, but it is common enough among mathematicians in general; many respected individuals have lined up on each side, and the arguments back and forth have been rehashed many times.

    Three specific points: 1) The distinction between omega-inconsistency and inconsistency does not survive if you allow certain proof techniques beyond first order logic. The proof is readily available via Google. 2) Godel's completeness and compactness theorems are compelling sounding, but given the known failures of first order logic to prove otherwise provable statements (as in points 1 & 3), the two theorems could simply be recast to show that there is a fundamental incompleteness in model theory. 3) Any statement about the integers that is proven not to be disprovable is true by the proof I outlined (which is not a proof using first order logic).

  32. Jerry Friedman said,

    May 24, 2011 @ 10:39 pm

    I suspect that D.O. is right that the intended meaning was something like "They can't even make a case for that." And I suspect even more strongly that Rube is right; the most important thing about even in this sentence is that it's dismissive.

  33. maidhc said,

    May 25, 2011 @ 2:51 am

    Getting back to the question of whether there are words whose meaning we no longer know…

    In one of the weekend crosswords there was the clue "Zesty in Paris". That led me to the realization that I don't even know what "zesty" means in English, much less what it would be in French. I know what "zest" is, but "zesty" has appeared in so much meaningless advertising describing tasteless snack foods, bland sauces and nondescript fast food items that I no longer perceive it as having any real meaning at all.

    (The answer turned out to be "piquant". In the same crossword the answer to the clue "sickening" was "nauseous". To my mind it should be "nauseating". I'm almost motivated to write to the editor.)

    Another word that seems to have lost its meaning is "hearty". As applied to canned soup, it nowadays appears to mean "containing extra cornstarch".

    As for "even", it reminds me a bit of the scene in Gravity's Rainbow, trying to explain the grammatical structure of "ass-backwards" to a German learning English. "It's an intensifier!"

  34. emar said,

    May 25, 2011 @ 3:47 am

    I don't think the scale involves truth being ranked higher than provability. If truth is ranked wrt proof, it's obviously lower. The relevant higher ranked alternative is getting someone convicted. Then it's a perfectly normal scale:

    "I don't think they can even prove that, much less convince the jury/judge, much less get him convicted and actually behind bars."

    (the reading pointed out by ironhorse above)

  35. Andrew said,

    May 25, 2011 @ 5:13 am

    Another apparently non-scalar use of "even": In [i]Attack the Block[/i] there are numerous scenes in which a main character (a member of a south London youth gang) must explain something unbelievable (an alien invasion) to another character. Each time he is met with scepticism and follows up with a plaintive "I'm not even lying".

  36. RP said,

    May 25, 2011 @ 5:55 am

    Interesting stuff.

    I think there are instances where something like "I don't know if he's even 6′ tall, much less 5′8″" could make sense.

    What if you were to demand that I find you someone as small as possible, and at any rate no taller than 5'? I might bring you a candidate, and you could reply "I don't know if he's even 5'3", let alone 5'".

  37. Dave said,

    May 25, 2011 @ 6:28 am

    @Brett: Not sure if crank. Where can I look to see the details worked out?

  38. The Ridger said,

    May 25, 2011 @ 1:06 pm

    I think there's a certain desire to put the adverb between the modal and the verb, rather than between the verb and the object, in play here, too.

  39. Katje said,

    May 26, 2011 @ 6:44 pm

    Here's another instance:

  40. Albert Vogler said,

    May 26, 2011 @ 7:59 pm

    @maidhc: I thought I'd previously seen "nauseous" with the meaning "sickening," so I checked. According to OED, first definition is "inclined to nausea; fastidious" (first attested 1604). Second is "causing nausea or squeamishness" (1612). After that come: nasty, unpleasant, loathsome, disgusting. For whatever it's worth.

  41. This Week’s Language Blog Roundup | Wordnik ~ all the words said,

    May 27, 2011 @ 9:32 am

    […] argued against the em dash, while the bloggers at Language Log wondered what “even” even means; explored the rejection of the power semantic; pondered the U.S. North Midland dialect (“You want […]

  42. Kee said,

    May 28, 2011 @ 7:36 pm

    I think I noticed a comparable example of this recently. I'm pretty sure this was relatively common with my fellow teens in '90s Southern California: "That isn't even funny!" (As in, "Stop that! That isn't even funny!" or "I have so much homework it isn't even funny.") There may be a bit more of a meaning change than just intensifying, particularly when the phrase is used to mean something like, "It's ridiculous in a really bad way", but it doesn't seem scalar.

    It pops up in an online discussion here:

    06 Mar 2011, 12:34 AM
    I am currently using Avira Antivirus but I notice that it hasn't updated in a few days and I have to manually click update for it to update

    Plus, it has a very high number of False positives that isn't even funny."

  43. Hermann Burchard said,

    May 30, 2011 @ 3:20 am

    I don't know if prosecutors can even prove that [Ligambi is a mob boss, much less all the charges in the 50-count indictment ]. I don't think it's that strong a case.

    This is just longer a version of what MYL reads and what the lawyer said [filling in the gaps] using the cryptic, highly contracted, deleted style typical of the careful attorney (no special acoustic evidence needed), who, incidentally, is not conceding anything about his client.

RSS feed for comments on this post