Sanctioned behaviors/ideas/methods?

« previous post | next post »

In the comments on Tuesday's "Come and go" post, Andrew Gelman wrote

Here's an example: the statistician Steve Stigler quoted as saying, “I don’t think in science we generally sanction the unequivocal acceptance of significance tests.” Unfortunately, I have no idea what he means here, given the two completely opposite meanings of the word “sanction.”

and Philip Anderson responded

In British English at least, it’s possible to sanction people, or organisations/states, with the sense of imposing sanctions on them (although it sounds strained to me), but if behaviour, or an idea, is sanctioned, it can only mean permitted. So I see no ambiguity in your example.

So this morning's Breakfast Experiment™ is a preliminary peek into this issue.

tl;dr: I share Andrew's intuition rather than Philip's — but the data seems to offer Philip (at least statistical) support.

The Corpus of Contemporary American English has 3019 hits (in a billion words) for sanctioned. I took a random sample of 100 (a nice feature of the search interface at Mark Davies' BYU corpus site), and classified each one as bearing either the meaning "authorized" or the meaning "punished", applied either to "people/organizations/states" or "behavior/ideas". (I've taken Philip Anderson's terminology for the sanctioned entities, though as you'll see if you look at my sample, those glosses give somewhat incomplete coverage to what still seems like a pretty clear distinction…)

The results:

 people/organization   behavior/idea 
authorized 12 68
punished 19 1

The one exception to Anderson's generalization is from "Sexual Diversity in Urban Norwegians", Journal of Sex Research 2002:

The gender differences may be explained by the fact that female sexuality is generally more strictly controlled and sanctioned than male sexuality is.

A quick search for similar conjoined participles yields, among others, one from a 1993 letter to the editor in the San Francisco Chronicle:

The way President Bush and our government have handled the issues with Iraq has reminded me of the authoritarian traditional family, one that criticized, admonished, condemned, sanctioned and punished.

And of course the phrases "criminally sanction" and "criminally sanctioned" are common, and are often applied to behaviors as well as individuals or organizations, e.g. "Out of Joint: How the Growing Disconnect Between Federal and State Marijuana Laws Impacts Employers", Bloomberg Law 2/8/2013

It is possible that there will be a “constitutional showdown” between the federal government and the states before the laws even take effect. Shortly after the election, Colorado’s Attorney General, John Suthers (R), cautioned Coloradans that:

the ability of the federal government to criminally sanction possession, use and distribution of marijuana, even if grown, distributed and used in a single state, was recognized by the U.S. Supreme Court [in Raich]. Therefore, absent action by Congress, Coloradans should not expect to see successful legal challenges to the ability of the federal government to enforce its marijuana laws in Colorado.

Or this:

As the criminally sanctioned behaviors often deal with ethics, a breach thereof is susceptible to affecting the image and earnings of the company.

And this:

While the motives of researchers within this paradigm often aimed to demystify and destigmatise sexual minorities, it has been argued that the binary distinction between ‘deviant’ and ‘normal’ it tacitly promoted may have contributed further to conflations between homosexuality, sex work, and other culturally and criminally sanctioned behaviours and identities (Sumner, 1994). 


  1. Breffni said,

    February 11, 2022 @ 12:35 pm

    In the Stigler quote, it seems clear from the context that he means "approve of":

    Stephen Stigler, a statistician at the University of Chicago, agrees with the general premise that "you can have a real effect which is nonetheless trivial in the practical sense." He doesn't think this is widely misunderstood, though: "I don't think in science we generally sanction the unequivocal acceptance of significance tests."

  2. J.W. Brewer said,

    February 11, 2022 @ 4:42 pm

    It seems potentially relevant that the authors of the "Sexual Diversity in Urban Norwegians" article which is the "one exception" in the sample seem unlikely (based on their names and institutional affiliations) to have been native-speaker Anglophones. I have no idea to what extent the Journal of Sex Research as of 2002 had the resources or inclination to copy-edit articles submitted by pretty-good-ESL authors into more idiomatic English. And there's the further problem that academic-journal English has its own turgid-jargon style and/or register, so I'm not sure I would have pegged that quote (without checking on the authors' background) as ESL-ish rather than just turgid-jargonish.

  3. AntC said,

    February 11, 2022 @ 5:03 pm

    Thanks @Breffini, I was wondering if a wider context would clarify the Stigler quote.

    Although I'm a Brit, I'm in the dark with that sentence alone. The two negatives make it difficult to get a grip on its polarity; the ambiguity of "sanction" is in effect another flip of polarity.

    But your offered context is from some paraphrase, _not_ Stigler direct. How do we know the paraphraser has got the right sense? What's worse, there's another pair of negatives. (The WSJ is behind a paywall for me, so I can't look deeper.)

    So I'm still in the dark. (Another confusion is I'm not well up on the topic area: 'significance tests'/'confidence intervals' usually are like blinding with numbers — which might be what Stigler is saying.)

  4. Rick Rubenstein said,

    February 11, 2022 @ 5:57 pm

    As far as I'm concerned the point is moot, and this discussion should be tabled.

  5. David Morris said,

    February 11, 2022 @ 7:21 pm

    If Putin sanctions the invasion of Ukraine (behaviour), the world will sanction him (person).

  6. Michèle Sharik Pituley said,

    February 11, 2022 @ 8:49 pm

    I’m 52 & grew up in south western Ohio. Sanctioned has always had two meanings to me. It is usually used as one of the examples of words with two meanings.

    Now that I say that, I can’t think of any others; my mind is blank. Argh.

  7. Andrew Usher said,

    February 11, 2022 @ 10:53 pm

    I would certainly expect the meaning 'approve of' here. That is the only traditional meaning for the verb, and I would expect a scholar in a respectable field to follow it. Note that 'sanction' is quite rare in general use – it's possible for the common man to associate it with the 'punish' sense because that's the one head about in the media.

    'Sanction' meaning punish, at least without 'criminally' or such modifier, is still non-standard jargon for me, best confined to discussion of international relations where it no doubt originated (from the noun).

    k_over_hbarc at

  8. Trogluddite said,

    February 12, 2022 @ 8:57 am

    Andrew Usher said:

    it's possible for the common man to associate it with the 'punish' sense because that's the one head about in the media.
    Online sources (e.g. here) suggest that this is very much the case; the verb sense "to punish" possibly being coined as recent as the 1950s.

    In that case, the discrepancy in commenters' reading of the example may just reflect the conservatism of their idiolect or local dialect (FWIW, my BrE instinct matches Philip's). A similar example is "protest"; it is possible to protest in support of something (e.g. one's innocence), but the word is so associated with protests against things that "against" can be omitted in US-English, though not in British-English.

  9. Trogluddite said,

    February 12, 2022 @ 8:59 am

    *Oops, I must not have closed my blockquote tag in the previous post – my contribution begins at "Online sources …".

  10. Trogluddite said,

    February 12, 2022 @ 9:04 am

    *Oh dear, I messed up my hyperlink as well: it should have pointed here. Apologies to anyone whom I've confused!

  11. Terry K. said,

    February 12, 2022 @ 10:48 am

    Interesting that usage data shows sanctioned with behaviors is pretty much only used with an "authorized" meaning, but, it seems, anecdotally, that perception doesn't follow. That is, not all readers/listeners pick up on that.

  12. Ignoto Fiorentino said,

    February 12, 2022 @ 6:24 pm

    I speak American English, and I can't see any ambiguity at all either – I think Andrew is playing semantic games here. What would be the context in which someone would actually punish "the unequivocal acceptance of significance tests", as opposed to approving or disapproving it? It's not just that it's behavior; it's not the kind of behavior that plausibly deserves being sanctioned in the sense of punishment.

  13. Ignoto Fiorentino said,

    February 12, 2022 @ 6:26 pm

    Just to clarify – I don't think Andrew is being disingenuous. But I don't read him as making an actual claim about ambiguity either — I read him as engaging in wordplay.

  14. Terry K. said,

    February 12, 2022 @ 11:32 pm

    @Ignoto Fiorentino
    I think a key point you are missing is that not all readers/listeners understand the negative sense of "sanction" as narrowly meaning only to punish. Use and perception don't necessarily match for a couple reasons. One, it's possible for the range of meanings a person perceives as possible for a word to be wider than the range of meaning the person actually uses. Two, not everyone who hears or reads the word is someone who has that word in their active vocabulary. And those people will vary in how much they understand the nuances of the word.

    So the fact that the negative sense in actual usage isn't used when the idea of "punishment" doesn't fit doesn't mean that readers trying to figure it out understand that.

  15. AntC said,

    February 13, 2022 @ 9:06 pm

    To reinforce @TerryK, "sanction" might have only a mild sense of 'punish' that has no effect (and no expected effect) on the 'perpetrator'. Something like 'frown on' as a reaction to a claim/I won't read any more of their stuff.

    As I said above, the whole polarity thing in the Stigler example is confounded by there being so many negatives. (Also the abstract "acceptance", so it's vague as to agency.) I still can't tell whether:

    – Stigler frowns on accepting significance tests.
    – Stigler automatically accepts significance tests.

  16. R C Head said,

    February 13, 2022 @ 11:07 pm

    Is it relevant that the noun "sanctions" (at least in the US) is exclusively negative or punitive? If we say that an individual or organization faces sanctions, we know they have been punished, not approved. And I think that applies to behaviors as well. We would say "his crimes led to sanctions," but not "his accomplishments led to sanctions."

    If anything, the positive sense of 'sanction' feels to me like it's fading. The use of the term for, say, athletic scandals in college sports, is surely always negative, and that may be the context where most people encounter it. Equally, in international affairs, sanctions cannot be positive. Only the verbal forms can have a positive sense, but these can leak into the participles, with more ambivalence.

  17. J.W. Brewer said,

    February 14, 2022 @ 3:58 pm

    Contra R C Head, I suggest that if you skim the 132 instances of "sanction for" in BYU's Corpus of Historical American English from the 19th century through the 21st you will probably agree that the positive sense of "sanction" as a noun is not extinct, even though the negative sense has grown in prominence is recent decades. Compare one example from 2002 ("the lack of legal sanction for same-sex couplehood" – "sanction" is intended as positive or rather would be if it had existed at the time) with another from 2003 ("[t]he reputational sanction for complicity in fraud is severe," "sanction" intended as negative).

    I will admit I haven't checked COCA, which might give you a more meaningful sense of what the current ratio is. I will also admit that things may well be otherwise with "sanctions," perhaps because positive-sense "sanction" is not a countable noun? You might have royal sanction for your proposed course of action and you might simultaneously have papal sanction for it but it would still be odd-sounding (at least to my ear) to pluralize the positive-sense noun and say you had "sanctions for" your action because they had been approved by multiple different authorities.

RSS feed for comments on this post