Not OK.

« previous post | next post »

OK mapIn this week's online BBC News magazine, Alan Metcalf reprises his recent book OK: The Improbable Story of America's Greatest Word. I haven't read the book, but Prof. Metcalf is an established scholar as well as a successful popularist, and I have every reason to think that the book is well worth reading. Still, I have a little semantic problem with the article.

The article mostly discusses the history of OK, saying that its widespread circulation probably dates back to an unfunny joke in an 1839 article in the Boston Morning Post. Fair enough: he and the OED agree on this point. Then he goes on:

But what makes OK so useful that we incorporate it into so many conversations?

It's not that it was needed to "fill a gap" in any language. Before 1839, English speakers had "yes", "good", "fine", "excellent", "satisfactory", and "all right".

What OK provided that the others did not was neutrality, a way to affirm or to express agreement without having to offer an opinion.

Consider this dialogue: "Let's meet again this afternoon."

Reply: "OK."

I beg to differ. Consider instead this dialogue:

1) ME: Prof. Metcalf is wearing a nice red tie. YOU: OK.

Seem neutral?

I suggest that OK expresses acceptance and acknowledgement, e.g. of a proposal. And Prof. Metcalf is surely right that OK is neutral when it does this. Sure thing, lover boy!, for example, would be a non-neutral way of accepting an offer of help with a Linguistics 101 assignment. The thing is, acceptance and acknowledgement are not agreement, nor even exactly affirmation. So I cannot fully agree with or affirm Prof. Metcalf's analysis.

I'll now give some example dialogues to illustrate my claim that OK (unlike yes) does not express agreement, and then I'll come back to the neutrality issue. The examples all involve speech acts other than assertions, speech acts that we might imagine OK or yes being used to respond to. First, here's a pair of dialogues that begin with a (biased) question. Note that in dialogue (2) the response OK! is weird, which I've notated with a "#", whereas in (3) the response Yes! is natural. That's because the question demands agreement, and OK is not a straightforward way of agreeing.

2) # ME: Isn't he smart? YOU: OK!

3) ME: Isn't he smart? YOU:  Yes!

Examples (4) and (5) start with me making a promise. Promises are not ways of asserting things but of committing to doing things. In this case OK is a natural way of acknowledging the promise and of accepting that a commitment has been made. On the other hand Yes, as in (5) is odd, because it really doesn't make much sense for you to agree with my commitment: I've made the commitment whether you like it or not.

4) ME: I promise to give it back. YOU: OK.

5) #ME: I promise to give it back. YOU: Yes.

Finally, (6) and (7) are dialogues beginning with strong imperatives. You can accept my uncouth demand by meekly saying OK, or even OK, OK. But since there isn't anything you can reasonably agree with, Yes (or Yes, yes) is quite unnatural.

6) ME: Get the fuck out of my stuff!!! YOU: OK, OK.

7) # ME: Get the fuck out of my stuff!!! YOU: Yes, yes.

Surely my claim that OK doesn't express agreement is too strong? Here's a fine distinction. OK is certainly used to signal agreement, but does not express agreement. And even when OK signals agreement, it signals weak agreement. We're back to the neutrality issue. In (1), above, for example, the OK is particularly weak… as if followed by an implicit if you say so or whatever, and perhaps uttered with a resigned glance into the space about a yard above and to the right of the speaker's right shoulder.

Why weak? For the simple reason that OK doesn't (or at least didn't in the past) directly express agreement. Rather it expresses acceptance and acknowledgement. When an assertion is made, we are understood to tacitly agree to it if (i) it is clear that we have understood it, (ii) we were in a position to object, and (iii) we failed to object. Now, when I use an entire speech turn to say OK, it will be made clear that all of conditions (i-iii) hold, and so I am seen to tacitly agree to what has been said. But I have not explicitly agreed. The agreement is seen as weak for just this reason. I could have said Yes!, Absolutely!, Too true!, Indeed! or Right!, but I didn't. I merely acknowledged that I'd understood, or perhaps accepted your right to say it, with weak agreement following by pragmatic reasoning. (I'm reminded of the pragmatics of citation: when you cite someone, you acknowledge them, and you will be taken to have expressed some sort of approval if you do not disavow the cited work explicitly. But if you don't say that you agree, then you haven't.)

What we have, then, is an expression that neutrally expresses acceptance and acknowledgment, and which contra what Allan Metcalf says in the BBC piece, does not express agreement directly, though it can be used indirectly to signal weak agreement. That sound OK?


  1. passerby said,

    February 20, 2011 @ 6:56 am

    So another way of describing a weak agreement would be: I never said no… but I didn't say yes. (Goodness knows how many times I've heard that.)

  2. Rubrick said,

    February 20, 2011 @ 7:02 am


  3. Andrew said,

    February 20, 2011 @ 7:10 am

    My impression is that "All right" fits all the uses of OK listed here.

    "Let's meet again this afternoon." "All right."
    *ME: Isn't he smart? YOU: All right!
    ME: I promise to give it back. YOU: All right.
    ME: Get the fuck out of my stuff!!! YOU: All right, all right.

    Perhaps it's no surprise that OK comes from "oll korrect" which is basically the same as "all right."

  4. Ast A. Moore said,

    February 20, 2011 @ 7:15 am


  5. Richard Sabey said,

    February 20, 2011 @ 8:45 am

    So "yes" expresses agreement?

    "David Beaver makes some good points." "I agree."

    "David Beaver makes some good points." ?"Yes."

    To me, "yes" is not idiomatic in this context, even though an expression of agreement would be appropriate.

    Yes, it's true that a simple "yes" is not an idiomatic way of agreeing to something neutrally in speech. First it occurs much less frequently that "yeah" or "uh-huh", and secondly "yes" and "yeah" (unlike "uh-huh") are normally part of a more extended utterance.

    In a portion of the Switchboard corpus of telephone speech, "yes" occurred 99 times, whereas "yeah" occurred 869 times, and "uh-huh" occurred 482 times. And if "Yes" was in an utterance, it was part of a larger utterance about 80% of the time. Perhaps using "yes", in addition to agreeing, signals a very definite desire to take the floor, which can make it odd when you have not been offered the floor (in the form of a direct question), and you have nothing else to say.

    I don't know whether this is right, but I agree that there's more to "yes" than agreement, and moreover I'll certainly concede that it doesn't have exactly the same pragmatic function as "I agree" (which fwiw occurred only 9 times in the corpus, never as a single utterance).

    – dib

  6. marie-lucie said,

    February 20, 2011 @ 9:30 am

    Consider also:

    "David Beaver makes some good points." "Yes?"

    This "Yes?" does not necessarily signal or express agreement but rather a desire for clarification about the intent of the remark: "So what? What is your point? What are you trying to tell me?" And here, "OK?" would be appropriate from the first speaker (to establish basic agreement before offering more information), but not for the second speaker.

    Good point. The OED lists something close to this interrogative usage as a separate meaning of "yes":

    (Usually interrogative.) Expressing provisional assent, with desire for further information or statement; hence as an inquiry addressed to a person waiting in silence (= ‘what is it?’ ‘what do you want?’); also as a mere expression of interest (= ‘indeed?’ ‘is it so?’).

    Finding a common meaning of "yes" would be difficult, but it seems to me that the best line to take would be based on the old idea that "yes" is like a (re-)statement of the positive proposition just indicated, which could thus be in agreement with it or questioning it, depending on intonation and context. In the case of "a person waiting in silence", we might sometimes take the proposition indicated to be "I am waiting for you", in which case "Yes?" becomes equivalent to "You are waiting for me?"

    Homework exercise: unify all the uses of "yes", including of course the dialectal meaning "earthworm".

    – dib

  7. mgh said,

    February 20, 2011 @ 9:58 am

    which is older, "ok" or "uh-huh"?

  8. Colin Reid said,

    February 20, 2011 @ 10:10 am

    I'd read 1. as follows:

    ME: Prof. Metcalf is wearing a nice red tie. YOU: OK.
    I'm acknowledging that you have told me this, but giving no opinion on the information (perhaps I have never seen Prof. Metcalf's tie). I'm not saying I believe you, nor am I saying you're wrong. That is a kind of neutrality, but perhaps in a way that isn't socially expected (we are expected to believe people by default in conversation, or at least respond as if we believe them).

    ME: Prof. Metcalf is wearing a nice red tie. YOU: Yes.
    I already have the information and agree with the assessment.

  9. Eric P Smith said,

    February 20, 2011 @ 10:12 am

    The ambiguous use of OK was a contributing factor to the Tenerife airport disaster of 1977. The captain of a KLM flight, in heavy fog, mistakenly believing he had clearance for takeoff, radioed to the control tower “we're now at takeoff” or “we're now, uh, taking off”. Air Traffic Control replied “OK, stand by for takeoff, I will call you”, but because of radio interference the captain heard only the initial "OK" and this reinforced his misunderstanding. He continued with the takeoff and collided with a Pan Am plane on the ground. It was the worst accident ever in commercial aviation, with 583 fatalities.

    After the accident, aviation authorities round the world introduced requirements for communicating in standard phrases. OK is not one of them.

    (Source: Wikipedia, Tenerife airport disaster)

  10. Dan Lufkin said,

    February 20, 2011 @ 11:13 am

    I recently read the book and enjoyed it. It also has a very thorough study of the etymology of OK, which I appreciated even though it demolished my favorite Scots "Och, aye" theory.

  11. Aaron Toivo said,

    February 20, 2011 @ 11:13 am

    @ Richard Sabey:
    "David Beaver makes some good points." ?"Yes."

    I think you may be stumbling on lack of an appropriate context, because definitely works. Try substituting "yes, he does" for the response to see it. If that's still not coming clear for you, imagine the following conversation:

    Andy: "I'm not sure I understand all of Mr. Beaver's arguments, but I can see he does raise some good points."
    Barb: "Yes. In particular, the point about allowing tacit agreement without expressing it outright helped clarify this for me."
    Andy: "Yes. I thought that was helpful too."

  12. army1987 said,

    February 20, 2011 @ 12:18 pm

    I perceive "OK" to be at least slightly positive, using "fair enough" for more neutral cases. For this reason, having to click on "OK" to close the window on Facebook telling me that something went wrong seems slightly weird to me. Does anyone share this impression?

  13. marie-lucie said,

    February 20, 2011 @ 1:45 pm


    Coming to tthe US as a student many years ago, I was puzzled first to hear this, and second to realize that it could mean both "yes" and "no". At first I could not tell which one was which, until I realized that the meaning depended entirely on the intonation. So just talking about "uh-uh" is not enough, we would need to somehow mark tones!

    [(myl) Though I've made no special study of this, I've always assumed that the main difference is
    yes = [ʔʌ.hʌ]
    no = [ʔʌ.ʔʌ]
    Can you provide a link to an example where either of these associations is switched?

    (Each of these can also be felicitously performed with a variety of stress patterns and pitch contours, but as far as few seconds of intuition can determine, there aren't any rigid associations, though I suppose there are probably some statistical patterns.)]

  14. bfwebster said,

    February 20, 2011 @ 1:46 pm

    Unless I've overlooked it, no one has brought up the skeptical-while-being-nominally-polite use of OK, that usually requires it be dragged out in pronunciation (and often is used with an initial asperative):

    Alice: "Let's go skydiving for our anniversary!"
    Bob: "Ho-kaaaaaay." (Translation: I'm trying to avoid an outright rejection here, but are you out of your mind?)

    Dragged out enunciations of 'Sure' and 'Why not?' serve much the same purpose, but convey more of a literal acceptance than "Oooo-kaaaay".

    Note that the dragged out "OK" is used in a similar way to express opinions about third parties:

    Neutral: "Prof. Metcalf is wearing a red tie." "OK."

    Not-so-neutral: "Prof. Metcalf has stripped naked and covered himself with maple syrup." "Ho-kaaaaay…." ("The man has stripped a gear.")

    Unflappably cool: "Prof. Metcalf has stripped naked and covered himself with maple syrup." "OK." ("So?" or "And your point is?" or "Doesn't bother _me_ at all.")

  15. Josh McNeill said,

    February 20, 2011 @ 2:00 pm

    Doesn't 'all right' work in the same way in Metcalf's example?

    'Let's meet again this afternoon.'

    'All right.'

    Sounds just as neutral, in this context, as OK. I'm confused as to why he would suggest that OK filled that void of neutrality after acknowledging that people could have used 'all right'.

  16. Aaron Toivo said,

    February 20, 2011 @ 2:09 pm

    @ marie-lucie: the intonations used for the positive and negative meanings of "uhuh" are highly variable depending what sort of information you are conveying – skeptical surprise, bored assent, all sorts of emotional overtones like those. These two particles (positive and negative) might be best understood as dummy-words that serve mainly as a vehicle for carrying intonation.

    But there is at least one fairly reliable phonetic difference between them: the medial consonant is [h] in the positive "uhuh" and a glottal stop in the negative one. (The particles mm-hm and mm-m work almost identically, but the mouth even stays closed.)

  17. octopod said,

    February 20, 2011 @ 2:36 pm

    "OK, all right, fine!"

    Don't all three of these more or less serve the same role?

  18. ShadowFox said,

    February 20, 2011 @ 2:39 pm

    There is also ironic/sarcastic/puzzled "O.K." [often with "." representing likely prosodic pauses–or not]

    Reporter: Do you have any comments on negotiations?
    [Negotiations participant #1]: Get the fuck out of my face! I got nothing to say to you!
    Reporter: O.K.! Let's ask [Negotiations participant #2].

    Of course, "Right!" would serve the same function here, but "Yes!" or "Sure!" would not.

    Adding rising tone turns it into a semi-neutral query that is likely to turn into a disapproval.

    You: On April 1st the world is going to end! We must do something about that NOW!!
    Me: O.K.? What do you propose?

    "OK" here is not meant to signify agreement, but more of a placeholder to hear more (or not–sort of "whatever you say, buddy"). I am having a hard time expressing all the intended meanings here–perhaps I am not OK. Early senility?

  19. ShadowFox said,

    February 20, 2011 @ 2:41 pm

    @octopod–they do, when they are intended to serve the same role. But there are cases where you can insert OK and cannot insert the other two.

  20. ShadowFox said,

    February 20, 2011 @ 2:48 pm

    Here's a more neutral example of what I termed above as "placeholder".

    Student: Of course we can differentiate between us and chimps. There are many things we can do that chimps cannot!
    Professor. OK. Can you provide a specific example?

    Student: Higher primates are just like us! They can do everything we can do, given proper training.
    Professor: OK. Does anyone want to take issue with that?

    I suppose, the second response can be used in the first example as well–but would it have the same meaning?

  21. Ellen K. said,

    February 20, 2011 @ 2:59 pm

    I think people are being a little unfair to Marie-lucie. She's not imagining things, she's simply making the wrong generalization about the essential difference between the two. Seems to me she's correctly picked up on the intonation these two utterances tend to have, and is wrongly seeing that as the essential difference.

  22. baylink said,

    February 20, 2011 @ 3:03 pm

    "Uh huh" and "unh uh" are spelled the same now?

    [(myl) Well, as Yogi Berra used to say, you could look it up. You'd find that "uh huh" and "uh uh" are clearly different; "unh uh" is relatively rare. In the COCA corpus, the counts are 433, 58, and 0, respectively. In the LDC conversational transcripts, the numbers are 58,079, 9,419, and 0 respectively. In the NYT archive since 1981, it's 438, 228, and 0. In Google Books, it's 569,000, 98,300, and 982.]

  23. Mark F. said,

    February 20, 2011 @ 3:07 pm

    In Metcalf's list of potential synonyms, "yes" is actually the odd word out. All the other words (including "OK") have adjectival senses that do indicate approval, and I don't think those senses are completely divorced from the assent-granting senses. "Yes" does not. Also, all of the other words are interchangeable with "OK" in David Beaver's example sentences.

    It is interesting to ask whether he would have been right if he had left "Yes" out of his list, but I think that's impossible. In consecutive sentences, he denied that "OK" filled a gap, and then tried to characterize the gap that it filled. That can't be right.

  24. Paul Kay said,

    February 20, 2011 @ 3:57 pm

    @Eric P Smith. It's interesting that the ambiguity of OK that I think you may be referring to does not exist in all languages that have borrowed OK. In English (at least American English) OK has both the discourse particle use David Beaver has explained and a predicative use: "The movie was OK, but I doubt I'd see it again." I don't know anything about Dutch but I know that the predicative use of OK is entirely absent in French, although the discourse use is common. On this theory, had the pilot been French he might have been less likely to interpret the control tower's "OK" as a signal to proceed with takeoff.

    Not sure that it had any relevance to the crash, but yes Dutch does have predicative "ok": Ik vind het ok and Het is ok both mean "It's ok". Dutch tend to code-switch fairly freely, so I'm guessing that while this is definitely something native Dutch speakers would commonly expect to say or hear, they would nonetheless be very aware of the foreignness of the OK, like a hint of vermouth in a gin, but perhaps not as sacrilegious as vermouth in genever.

    – dib

  25. Chris Waters said,

    February 20, 2011 @ 4:33 pm

    Phone rings. I pick it up and say, "yes?" What am I agreeing to? Also, "Get the fuck out of my stuff!" "Yes, yes." seems perfectly reasonable to me, although it does suggest that the matter has been discussed previously. But I can't think of one example of something like "yes", "OK", "uh-huh", "all right", etc., that doesn't have both positive and neutral uses. It's true that I can think of cases where they don't seem to be perfectly interchangeable, but I don't believe it's something as simple as acceptance/acknowledgment vs. agreement. I think it's a lot more subtle than that.

    As an aside, the Japanese "hai" is commonly translated as "yes", but from what I've been able to gather, is used more like "uh-huh" or "OK". Someone whose knowledge of the language is deeper and not as purely academic as mine might be able to comment further.

  26. GeorgeW said,

    February 20, 2011 @ 5:00 pm

    Then, there is an 'authoritarian OK' that seems to command acceptance.

    Parent to child: You are going to get on that bus and go to school, OK, and you are gonna take that test.

    I have known people who employ this regularly in their speech even with peers (social or professional). And, it is annoying.

  27. GeorgeW said,

    February 20, 2011 @ 5:11 pm

    Is there a difference between 'yes' and 'yeah' other than register?

  28. Jason said,

    February 20, 2011 @ 5:49 pm

    @GeorgeW: The "authoritarian OK" is a question/demand that the other person say OK or at least indicate assent.

    Parent: You are going to get on that bus and go to school, OK?
    Child (grudgingly): OK.

    And of course, no discussion of "OK" is complete without a reference to Bull on Night Court:

  29. Janice Byer said,

    February 20, 2011 @ 7:20 pm

    "It's okay" in reply to being asked whether you like something is so far from neutral that not even tone of voice can mitigate the suggestion of damning with faint approval.

  30. GeorgeW said,

    February 20, 2011 @ 8:07 pm

    Jason: The "authoritarian OK" is a question/demand that the other person say OK or at least indicate assent."

    You are right, in some instances. But, I have heard this used with no expectation of a reply.

  31. Mark F. said,

    February 20, 2011 @ 8:08 pm

    What I said about substitutability was obviously wrong in the case of "excellent", "good", and "satisfactory". But I still think "yes" is special. It seems like there is a general pattern where a word that is used predicatively to indicate some level of approval of some thing comes to be used as a discourse marker to indicate some level of approval of the current situation. Often the approval weakens to mere assent, I suspect through politeness. I think "very well" is an older term that fits this same pattern.

  32. John Cowan said,

    February 20, 2011 @ 8:09 pm

    In International Morse Code, the sequence dah-dah-dah-dah-di-dah means "I/we agree"; it is not an accident that it is the letters "OK" run together without the normal interletter space.

  33. Michael Wise said,

    February 21, 2011 @ 12:28 pm

    I had an employee at an non-profit that I would frequently communicate with by IM. He infuriated me by responding to everything I wanted the non-profit to do, or wanted him to do, or proposed doing, with "k". I could never tell if he was agreeing to do it, agreeing that it was a good idea, agreeing to placate me as the boss, or even not agreeing. I'd find it even more exasperating when he'd disagree with something I'd done as treasurer, like pay for a business registration or a license that he didn't think we needed to have. He'd confront me with a question, and when I explained, his only response was, "k". The only reason I knew he was disagreeing was because he was asking the question in the first place, and my explanation ("Soliciting donations without a license is illegal, and will get us, me, in trouble") was falling on deaf ears.

  34. Read Weaver said,

    February 21, 2011 @ 5:56 pm

    Rather less significant than Tenerife, but a customer service agent for my ISP told me about some new thing the ISP was offering, which I wasn't interested in. I said "OK" when she finished describing it, and I later discovered I'd been signed up. (She hadn't finished by asking a question like "Are you interested?" or "Do you want to sign up?")

    My ISP's customer service people usually have South Asian accents; I've assumed that's where they're located. Anyone know if 'OK' functions differently in southern Asia?

    (It's entirely possible, of course, that the call center's instructions were "Sign them up unless they threaten bodily harm if you do.")

  35. Janice Byer said,

    February 21, 2011 @ 10:14 pm

    Rand weaver, my own experience with such customer reps strongly supports your parenthetical speculation. Seriously, so persistent can they be. it seems implicit something depends on it, which, to be generous, may have led her to wishfully extrapolate agreement from a mere voicing of comprehension.

  36. Janice Byer said,

    February 21, 2011 @ 10:32 pm

    Oops, my apologies, Read, for misspelling your name.

  37. jaap said,

    February 25, 2011 @ 4:55 am

    Ik vind het ok and Het is ok both mean "It's ok". Dutch tend to code-switch fairly freely, so I'm guessing that while this is definitely something native Dutch speakers would commonly expect to say or hear, they would nonetheless be very aware of the foreignness of the OK, like a hint of vermouth in a gin, but perhaps not as sacrilegious as vermouth in genever.
    – dib

    As a native Dutch speaker, I would say that OK has been fully assimilated into Dutch, and no longer has any foreign taint to it for anyone under 40. It now also has the Dutch spelling okee.

  38. Michael Cargal said,

    February 3, 2013 @ 6:26 am

    As a response to "George W. Bush was our greatest president," "Ooooooookay" shows at least doubt and maybe clear disagreement.

  39. Peter Tenerife said,

    November 27, 2013 @ 8:57 am

    I like that one Michael.:)
    UK English is good like that. It may sometimes take a while before you figure out that someone has insulted you because the language can be manipulated so well.

RSS feed for comments on this post