Spoken style correction: the iPeeve™

« previous post | next post »

I just had a terrible idea that could probably make someone a modest fortune. I was inspired by Erin Gloria Ryan, "My Love Affair With 'Like'", Jezebel 6/26/2011:

I use the word "like" with embarrassing frequency. I've started paying attention to how other people talk as well, and it's amazing how many women who I know are very smart are similarly infected with like-itis.

Where does this come from? Why do we do this? […]

Since we know that saying "like" too much leads others to negatively judge our intelligence, maybe inserting "like" into a sentence is something that we do to purposefully make ourselves sound less intelligent and forceful and therefore less formidable than we actually are. We're sabotaging ourselves! […]

Maybe women of my generation have been taught, through positive social reinforcement, that we're supposed to pepper our speech with meaningless modifiers that make us sounds a little less sure of ourselves, a little less credible. No one likes a show off or a know-it-all. Better temper your smart-talk with assurance to whoever you're speaking that you're not, like, a threat or anything. Any girl who's been teased for middle school nerdery has likely developed a long standing aversion for the feeling of being excluded for being too smart or opinionated. This is the way that socially acceptable people talk. This is the way that pretty people talk. Women are taught that it's more important to be pretty and socially accepted than it is to be smart. Ergo, like.

Never mind for now whether vernacular like is genuinely female-associated.  Alexandra D'Arcy  ("Like and Language Ideology: Disentangling Fact from Fiction", American Speech 82(4) 2007) found that the discourse-particle version (as in "They're like representatives of their whole like clan, but they don't take it like really seriously, especially like during planting season") was used more by males than by females:

And never mind the broader question of whether there's a female speech style characterized by features that weaken assertions — this is a widespread and plausible idea, but most attempts to find empirical confirmation have come up empty (see e.g. "Gender and Tags", 5/9/2004).

Gender aside, let's be clear:  incessant repetition of a low-information-content filler is annoying to listeners and should be avoided, whether the filler is "like" or "literally" or "you know" or just plain "uh". And operant conditioning, like a teacher who rings a bell whenever a student uses vernacular like, is an effective training method. That's why my terrible idea is so seductive.

What am I talking about? Well, now that speech recognition has gotten to be pretty good and very cheap, it's only a matter of time before someone combines a speech recognizer with a style checker, and creates an app for your smartphone that will make it vibrate (or beep, or flash) whenever you indulge in any of the verbal tics that you've asked it to watch out for.

Of course, other functions could be added — detection of stigmatized pronunciations,  taboo words, political incorrectness, … But detection of over-used discourse particles and other fillers should be especially easy. With Andreas Stolcke and Jiahong Yuan, I've been working on detection of filled pauses, and it's not hard to achieve a high success rate, easily good enough to be used for operant conditioning.

I think that this is such a bad idea that I would be conflicted about revealing it to the world, if I weren't certain that its discovery and implementation are inevitable. For all I know, someone has already submitted a patent application.

[I should make it clear that I'm not against intervention to eliminate verbal tics. But I can see many problems with constant digital monitoring of verbal behavior.]


For those who are specifically interested in vernacular like, I can recommend Muffy Siegel's paper "Like: The Discourse Particle and Semantics", J. of Semantics 19(1), Feb. 2002, as well the following relevant LL posts:

"It's like so unfair", 11/22/2003
"Like is, like, not really like if you will", 11/22/2003
"Exclusive: God uses "like" as hedge", 1/3/2004
"Divine ambiguity", 1/4/2004
"Grammar critics are, like, annoyed really weird", 9/13/2004
"This construction seems that I would never use it", 11/11/2004
"Looks like a reference problem", 11/11/2004
"Seems like, go, all", 11/15/2004
"I'm like, all into this stuff ", 11/15/2004
"I'm starting to get like "this is really interesting", 11/16/2004
"That's such a like, coincidence", 5/31/2005
"Like totally presidential", 8/17/2007
"It's, like, war", 12/10/2007
"'Like' youth and sex", 6/28/2011

Note especially the discussion ("Grammar critics are, like, annoyed really weird", 9/13/2004) of a  method in which a speech therapist uses unpleasant smells to condition her clients not to use vernacular like. Luckily, digital-to-odor conversion is not likely to be a standard feature of smartphones for some time.



52 Comments

  1. UK Lawyer said,

    June 27, 2011 @ 7:12 am

    IMHO, as they say, this is, like, a great idea, know what I mean?

  2. Tim Miller said,

    June 27, 2011 @ 7:23 am

    >>With Andreas Stolcke and Jiahong Yuan, I've been working on detection of filled pauses, and it's not hard to achieve a high success rate…

    I would be very interested in learning more about this work.

  3. Chris said,

    June 27, 2011 @ 7:27 am

    "…incessant repetition of a low-information-content filler is annoying to listeners and should be avoided"

    Then why do we do it so much?

  4. Matthew Wright said,

    June 27, 2011 @ 7:37 am

    As a UK dweller without teenagers my main exposure to 'like' is via Friends. But there it seems more often to occur when reporting speech, so that "he said X" becomes "he was like X" or "he was all like X". The second usage seems to me to convey more disapproval than the first. In neither case is like filler, though.

    [(myl) That's the function known as "quotative like". Read D'Arcy's paper.]

  5. Tim Martin said,

    June 27, 2011 @ 7:39 am

    "And operant conditioning, like a teacher who rings a bell whenever a student uses vernacular like, is an effective training method."

    The problem I see with this is that operant conditioning with words tends to make people avoid a word based on phonological assocations – i.e. the sound of it. In other words, people who try to avoid the discourse marker "like" end up avoiding certain other uses as well, such as comparative like. I have actually witnessed this. A person trying to reduce the amount of meaningless "like" in their sentences will say something such as, "It's like when two people…", and then they will stop and say "oops I did it again," as if the like they just used was incorrect in some way.

    For a further example, look at what has happened with "X and I." Instead of learning to say "X and I" as the subject of a sentence and "X and me" as the object, operant conditioning has produced lots of speakers who say things like "between you and I." Again, the operant conditioning seems to act on the level of sound as opposed to grammar.

    [(myl) Well, in my limited personal experience (one recent report from a male teen), the method (of having a teacher make you stop talking whenever you use the discourse-particle version of like) works, in the specific sense that it gets rid of the targeted behavior.]

  6. Martin J Ball said,

    June 27, 2011 @ 7:52 am

    See also John Maidment's blog entry on 'like': http://blogjam.name/?p=5336

    [(myl) In my opinion, the traditional taxonomy (as in D'Arcy's paper) is a better one.]

  7. Pflaumbaum said,

    June 27, 2011 @ 7:55 am

    I'm not sure people are appreciating the hideousness of this development. MYL's just invented the iPeeve™.

    And in a hundred years it'll probably be embedded in your head, so you never even come close to 'splitting the infinitive' or whatever. A microscopic Simon Heffer, right inside your brain. *Shudders*

    In fact, this might be the end of language evolution.

    [(myl) Very nice. I've added your coinage to the post's title.]

  8. BeSlayed said,

    June 27, 2011 @ 8:11 am

    As a quotative, presumably like is not a "low-information-content" filler, and therefore it's interesting that this seems to be the function which it fulfils most for both males and females.

    [As I believe I've noted on Language Log before, the extension of a comparative with a "like" sense to quotative-type functions is not unparelled; Sanskrit yathā underwent a similar extension. Briefly:

    Though yathā is, properly-speaking, a relative pronoun used in relative-correlative constructions of the form yathā X…tathā Y "As X…., so Y", it also be can used by itself as a comparative like English like:

    mansyante mām yathā nṛpam
    "They will consider me like a king." [Mahabharata 4.2.5]

    And we find yathā appearing also with the function of a quotative, in examples like:

    viditam eva yathā "vayam malayaketau kimcitkālāntaram uṣitāḥ".
    "You certainly know, like, 'I stayed for some time with Malayaketu'."
    ("It is certainly known (to you) that I stayed for some time with Malayaketu.") [Mudrarakshasa VII]
    ]

  9. Pflaumbaum said,

    June 27, 2011 @ 8:29 am

    @MYL – thanks, I will accept as payment the waiving of my LL subscription fee for the year.

    @Tim Martin –

    I'm not sure that [X and I] in object position is just a hypercorrection motivated by the stigma against [X and me] in subject position – plausible though it seems. See Chapter 4 of Thomas Grano's thesis on co-ordinated pronouns.

    His conclusion is that it's a natural extension of subject-position nominatives in common collocations like [you and I] and [my wife and I], and goes back to the 16th century – though it's been reinforced by later prescriptive pressure.

    There may be fuller treatments that disagree with that, I'm not a linguist and that's the only paper I've read on the subject.

  10. Brian said,

    June 27, 2011 @ 8:40 am

    This is what I love so much about language log: no matter how good a theory sounds, the first thing you have to ask is "Is there any evidence to support it?"

  11. JFM said,

    June 27, 2011 @ 9:01 am

    Abverb?

    At first I thought I had missed something in my linguistics classes, and tried in vain to figure out what it might mean.

    [(myl) Nice catch. This is definitely a term in search of an application.]

  12. Andrew Dowd said,

    June 27, 2011 @ 9:24 am

    The problem with operant conditioning is that it eventually wears off as the subjects get tired of playing the game.

    [(myl) Dunno about in general, but it's clearly possible to train certain linguistic behaviors — or lack thereof — in a way that lasts, at least if the trainee is willing to buy into the enterprise.]

  13. Niall McAuley said,

    June 27, 2011 @ 9:45 am

    Joseph O'Connor does a radio diary on RTE (that's in Ireland), and "Like" is probably his most popular piece from it:

    http://www.youtube.com/watch?v=Yx-siLoRmgc

    Well, I liked it.

  14. C Thornett said,

    June 27, 2011 @ 9:58 am

    And presidential speech-writers could be trained to avoid uppity personal pronouns.* I have no hope of training media pundits to consider genuine evidence before stating their opinions as facts.

    *Ahem. This is a joke.

  15. Marta said,

    June 27, 2011 @ 10:05 am

    @ Tim Martin (and myl's reply):

    My training as a speech-language pathologist tells me that Tim's concern with overgeneralization is legitimate. Unless the program is sophisticated enough to differentiate acceptable and unacceptable productions of "like," and then provide feedback only for the unacceptable productions, overgeneralization to the acceptable productions is predictable because the feedback isn't based on the correct contingency. I am sure that the teacher in Dr. Liberman's anecdote only corrected the young man's unacceptable productions.

    Of course, the non-contingent response version provides an excellent business opportunity. The people who bought Line Corrector provide a ready-made market for LikeCorrector2 – the corrector that corrects overcorrections. If you build version 2 the same way as version 1, you could sell updates in perpetuity.

  16. Eric P Smith said,

    June 27, 2011 @ 10:09 am

    Spoken style correction works. At Primary School (Elementary School) in Scotland in the 1950s one teacher would make us stand on our chair for a minute if we said anything that wasn’t a sentence. Years later, a friend said to me, “Eric, you’re the only person I know who always speaks in sentences.”

  17. Jens Fiederer said,

    June 27, 2011 @ 10:34 am

    It doesn't take all that much sometimes to eliminate a verbal tic.

    My father used to constantly insert "an und fuer sich" (roughly "in and of itself", although the actual prepositions are "on" and "for") until my little brother asked "What is this peach you are constantly talking about?" ("einen Pfirsich", "a peach", sounds pretty similar.

    He stopped because every time he started to use the phrase a picture of a peach popped into his head.

  18. ronbo said,

    June 27, 2011 @ 10:48 am

    It's a great idea! We are giving the analog version a try.

    Our almost-11 yo daughter is a like-aholic and my wife and I have taken to counting "likes" on our fingers when we are with her. We don't say anything (unless asked; she didn't realize the quotative use was incorrect) but a couple of fingers in the air will stop her short and get her immediately to start over. So far, she has been a good sport about it, but we will see how long that lasts.

    This approach is really recent (<2 weeks) and began spontaneously, so who knows how it will turn out. Still, we'd rather get her to self-correct than try more obvious forms of negative reinforcement. For one thing, her allowance wouldn't last very long at current rates if we, like, docked her a quarter for every improper instance.

  19. Timothy Martin said,

    June 27, 2011 @ 10:50 am

    @Pflaumbaum: Interesting. I've taken a cursory look, and may go back for more later (although I can tell that this paper is going to be hard for me to understand as a non-linguist).

    This is just anecdote, but it comports with Marta's experience as well:

    It seems to me that hypercorrection is at work here because I occasionally hear instances of object-position "X and I" that intuitively sound wrong to me. Why should that happen, if such grammar has been in use for centuries?

    It seems, rather, that people have learned that "X and me" sounds bad and "X and I" sounds good.

    Support for this hypothesis also comes (again, in anecdotal form) from my observation that formality affects object-position "X and I" use. People seem to me (myself included) more likely to use object-position "X and I" when speaking in formal contexts than when not.

    Just some thoughts.

  20. TonyK said,

    June 27, 2011 @ 11:09 am

    @ronbo: "We are giving the analog version a try." So something that's even more digital than digital has to be called analog! We need a new word, 'fingery'.

  21. Spell Me Jeff said,

    June 27, 2011 @ 11:48 am

    Am I correct that "low-information-content filler" often serves a "holding-the-floor" function in conversations where interlocutors might otherwise exploit a "pause-to-collect-one's-thoughts" in order to interrupt the current speaker?

    If so, the real problem is having minimal control over the tic, such that one exhibits it in contexts where serves no function at all (lectures, interviews, etc.). I suspect these are the contexts in which listeners are most apt to notice the tic and be annoyed by it.

    I wonder if there are any studies on the "adjustment" of such tics in the broader matrix of style switching.

  22. Thor Lawrence said,

    June 27, 2011 @ 11:49 am

    Would the iPeeve pick up the supposedly passive? ;>}
    Thor

  23. Trimegistus said,

    June 27, 2011 @ 12:13 pm

    The theory that it's a deliberate "dumbing down" of one's dialogue doesn't match what I've observed in the wild. My kid attends a geek camp, was home-schooled through middle school — and still, like, says, like, like, like all the time.

    I think it's semantically equivalent to the "uh, uh, uh," beloved of politicians — a way to indicate you're still talking while your brain sorts out the next remark.

  24. Ray Dillinger said,

    June 27, 2011 @ 12:40 pm

    As a child I was specifically taught never to trust anyone who speaks faster than I believe he can think. "Like", "Uh", etc, were given me as prima facie evidence of such a problem. And so, I resolved to stop using them myself.

  25. Pflaumbaum said,

    June 27, 2011 @ 12:56 pm

    @ Timothy Martin –

    I think the problem is that this area is now so fraught and over-policed that what 'sounds wrong' is not a good guide to what is unnatural. It may be that there are several distinct dialects in this respect, each with its own rules. One may be the underlying, uncorrected one (I think mine is!), or it may be that they've been in free competition among different regional social groups for centuries.

    For instance [X and I] 'sounds wrong' to me not only in object position but also in subject position. Like (I believe) most kids down my way, I grew up with me, him, us, them as disjunctive pronouns, the default case whenever they're not the whole subject of a finite verb, i.e. in the various contexts given by Grano in Chapter 2, or pp.459-463 of the Cambridge Grammar.

    Various adults then attempted to make me conform to a system in which me is always accusative, and in very formal contexts I do. But he and I have been friends since college still sounds really affected to my ear.

    Recently I attempted to do an experiment on my friends to see what their underlying case system was. I waited till everyone got tipsy, and then started seeding the conversation with co-ordinated agents in an attempt to get them to use pronouns, which I then wrote down on my phone. Unfortunately the experiment failed because a) they started making all sorts of speech errors that almost certainly weren't 'underlying', 'uncorrected' forms, and b) I also got drunk and stopped writing things down.

  26. Matt said,

    June 27, 2011 @ 1:00 pm

    I think it's a great idea, and think it would be a great prank to change the tics that someone's phone is programmed to warn them off. Want a phone that buzzes every time you say 'be'?

  27. Dakota said,

    June 27, 2011 @ 1:09 pm

    In Arabic "yanni" يعني is the filler word. I have only heard it used by men.

    I too thought of Spell Me Jeff's theory of preventing interruptions. A vocalized pause is certainly a way to signal that one is not finished speaking, where an eloquent pause might be taken advantage of by someone with an opposing agenda. There also may be some kind of charisma-generating effect – this is one mark of "transformational" leadership – but how do you quantify charisma?

    @ C Thornett: Those who count presidential pronouns would hit pay dirt if only they would switch to counting the "um"s. Obama is a brilliant speechifier, but his vocalized pauses are quite distracting once you become aware of them. Not sure how you would count them though, transcripts usually filter them out.

  28. J. K. Gayle said,

    June 27, 2011 @ 1:40 pm

    What am I talking about? Well, now that speech recognition has gotten to be pretty good and very cheap, it's only a matter of time before someone combines a speech recognizer with a style checker, and creates an app for your smartphone that will make it vibrate (or beep, or flash) whenever you indulge in any of the verbal tics that you've asked it to watch out for.

    What about that tic that starts a sentence, "Well," as in "Well, now that speech recognition has…."? Maybe the whole point of some of these "fillers" is to indulge and not to annoy the listener(s). Maybe it's all about "sounding" as if one is "talking" about something and not just "writing" about it. Maybe Erin Gloria Ryan and the friends with who she speaks like and really love sounding like Valley Girls. Couldn't it be that there is a purposeful sociolinguistic function of such words? Well, if you asked me, I'd say Yes. (And, well, then, of course, I certainly might write Yes too, for sure.)

    [(myl) Sentence-initial well is what's generally called a "discourse marker" — it wouldn't become a "verbal tic" unless it was over-used. And if you read Ms. Ryan's article, you'll see that her (self-)conscious reaction is being told about her over-use of like is quite negative — she closes by saying that "I've realized, in reflecting on the sting of understanding that the way I speak negatively influences people's perception of my worthiness to be listened to (regardless of the quality of my actual ideas), that it's time to expel "like" from my vocabulary."]

  29. Dan Hemmens said,

    June 27, 2011 @ 1:54 pm

    Would the iPeeve pick up the supposedly passive? ;>}

    Surely the real question is whether the supposedly passive would be picked up by the iPeeve?

  30. Tim Martin said,

    June 27, 2011 @ 2:23 pm

    I don't know how I ended up signed in first as Tim and then as Timothy. Fear not, for we are one person. Feel free to call me Tim.

    @Pflaumbaum: Lol. Good points, also. I do recall that when I was in elementary school, "me and him" type constructions were much more frequently heard than they are now. Now I'm wondering how much prescriptivism and overt teaching have altered my ideas about what sounds good….

  31. Bobbie said,

    June 27, 2011 @ 2:32 pm

    Flashing lights or beeps might work. So could electroshock therapy! Whip lashes! Slaps across the face! Not as subtle, but probably a lot more effective in producing instant results. (I AM joking!)

  32. Peter Howard said,

    June 27, 2011 @ 2:40 pm

    How does one distinguish between discourse marker and approximative adverb in a sentence like: "The snake was like six feet long."? It could mean: "The snake was (and what I'm about to say next will really impress you) six feet long." Or it could mean "The snake was about six feet long."

    Or it could have elements of both, I suppose.

    [(myl) Good insight. See "Divine ambiguity", 1/1/2004.]

  33. GeorgeW said,

    June 27, 2011 @ 2:49 pm

    @Dakota: "In Arabic "yanni" يعني is the filler word. I have only heard it used by men."

    I have heard it used by Saudi women as well. Also, this might be a Gulf thing. I haven't noticed it so much in Egypt.

  34. Nyq Only said,

    June 27, 2011 @ 3:03 pm

    Your iPeeve is out of step with computer culture. Consider the spell-checker: you could build a spell-checker that simply provided negative feedback when you spell incorrectly (the dreaded red wiggly underline) but typically they also correct your mispelling – and in some cases without your intervention.
    Instead of the iPeeve you should have an iPretendPedant that simply deletes your likes, uhs, and ers from your speech and replaces them with meaningful pauses in your conversation.
    "I like so loved that movie" would be heard at the other end of the conversation as "I [brief pause for thought as if considering the issue] loved that movie".

    [(myl) Alas, free-field sound cancellation is a difficult problem, not (as far as I know) solved in the general case with any available technology, and unlikely to be available as a smartphone app for many decades if ever.

    (You may be imagining an app that modifies what you say into the phone before transmission, so that the party at the other end of the call hears silence instead of pause-fillers. This could be done, but it's likely to sound weird most of the time. A bigger problem is that it assumes people actually talk on their phones. None of the cool kids do that anymore, so you'd be selling into a dying market. Pretty soon your only customers would be the people shouting into their phones on commuter trains, and it's clear that they don't care what other people think of them anyhow.)]

  35. GeorgeW said,

    June 27, 2011 @ 3:44 pm

    This discussion begs the question as to why the 'overuse' of like would be peevogenic. I often find myself trying to think of a synonym to avoid repeating the precise word that I think best fits.

    I would suggest that there is a discourse constraint, 'Don't be Repetitious' which 'like' violates excessively.

    [(myl) But like is not unique in annoying some listeners — people have comparable reactions to overuse of literally, you know, etc. Ms. Ryan's claim in this case is that there are other stereotypes at work as well, so that when someone — especially a younger woman — overuses like, she's perceived to be unserious and even somewhat stupid. This is plausible, though of course there are similar stereotyped reactions to other sorts of verbal characteristics, including regional, ethnic, or class accents. And people subject to those stereotypes have to decide whether to knuckle under and change their speech.]

  36. Dakota said,

    June 27, 2011 @ 3:57 pm

    @GeorgeW, "yanni" is very big with Jordanians, also Palestinians, if that isn't repetitious.

    My grandmother used "you know", it was just something characteristic I came to associate with her, rather fondly.

  37. GeorgeW said,

    June 27, 2011 @ 4:11 pm

    [(myl) "But like is not unique in annoying some listeners — people have comparable reactions to overuse of literally, you know, etc."

    I'm sorry, didn't intent to suggest that 'like' was the only peevogenic word or expression. There seems to be no limit to speech forms that can annoy.

    In the case of 'like' in this particular discussion, I think it is the repetition that is so annoying.

  38. GeorgeW said,

    June 27, 2011 @ 4:18 pm

    @Dakota: Yes, I consulted my wife (Egyptian-America) who just arrived home. She finds 'yanni' overused by Saudis as well as others you noted. She mentioned a Lebanese friend (female) who 'overuses' yanni.

    Maybe the Levantis took it to the Gulf with them. She confirmed that it is not common in Egyptian speech.

  39. Henning Makholm said,

    June 27, 2011 @ 4:43 pm

    Those who champion the theory that "like" is merely a pause-filler or floor-holder should read D'Arcy's endnote 5. ("… most instances of like are prosodically integrated; co-occurrence with self-repair and hesitation phenomena constitute the exception, not the norm …")

  40. Chris Waters said,

    June 27, 2011 @ 5:26 pm

    MYL: interesting you should mention free field sound cancellation. The BBC had an article Friday, "Acoustic 'cloaking device' shields objects from sound".

    http://www.bbc.co.uk/news/science-environment-13905573

    Acoustics aren't my field, so I'm not qualified to separate the BBC's science from their pseudo-science here, but if there's anything to this, then it seems to me that the sole remaining obstacle would be persuading folks to wear "stacked sheets of plastic with regular arrays of holes through them." If you could overcome this obstacle, then perhaps an auto-correcting iPeeve could become, like, a reality! :)

    [(myl) The cited work has to do with making objects acoustically transparent, in the sense of making them (say) invisible to echoes. It doesn't address the problem of how to arrange for a sound source to be adaptively cancelled everywhere in spatial volumes — even modest-sized ones — at a distance from the cancelling apparatus. There's certainly work on that problem — e.g. here and here — but …]

  41. Matt McIrvin said,

    June 27, 2011 @ 9:39 pm

    Just don't use old Ludwig Van! He never hurt nobody!

  42. Marc said,

    June 27, 2011 @ 10:20 pm

    @Mark Liberman

    Luckily, digital-to-odor conversion is not likely to be a standard feature of smartphones for some time.

    You may have spoken too soon on this point. See this press release, dated June 14th, 2011, from the UCSD Jacobs School of Engineering.

    http://www.jacobsschool.ucsd.edu/news/news_releases/release.sfe?id=1082

    To wit:

    In a proof of concept paper published online today by the journal Angewandte Chemie, the researchers demonstrate that it is possible to generate odor, at will, in a compact device small enough to fit on the back of your TV with potentially thousands of odors.

    and

    “This is likely to be the next generation TV or cell phone that produces odors to match the images you see on the screen.” said Jin. The multi-odor concept was initiated by Samsung’s research and development group, headed by Jongmin Kim at SAIT. They came to UCSD with a request for a practical means of accomplishing such a vision.

  43. Nathan Myers said,

    June 28, 2011 @ 1:54 am

    Haven't you heard academics (and would-be academics) use "i.e." much the same way in conversation, and at the podium? It's hard to say which is worse.

    For the case of "like", it seems that in many uses, and in various grammatical positions, its purpose is to reduce the degree of implied precision, as if the speaker would not want to be held to account for the accuracy of the expression or quote. Does "he was all like" really mean "he said", or does it mean "he said what seemed more or less to the effect that"?

  44. Ken Brown said,

    June 28, 2011 @ 5:33 am

    My iphone is already nannying me. It (or more likely the 3G provider) won't let me see the D'Arcy paper because of "content control".

    Quotative and non- quotative "like" seem alive and well in England. The blog linked to from my name here reports a rather marked example I overheard on a train a few days ago.

  45. jan said,

    June 28, 2011 @ 7:58 am

    Yanni is the filler word in Arabic? I immediately thought of the New Age Greek singer, Yiannis Hrysomallis, aka Yanni. According to Wikipedia, others with the same first name are also Greek. I wonder what the connection is, if any. And where does "yada" or "yadda" come in?

    Regarding girls who don't want to seem too smart and don't want to be know-it-alls–
    When somebody asks them out on a date do they still act hesitant, uncertain and inarticulate? One response I remember is that the guy should ask her boyfriend since he would be able to answer the question better than she could. Yes, it's a hint, but still…

  46. GeorgeW said,

    June 28, 2011 @ 8:13 am

    Jan: I don't think there is any connection. Arabic 'yanni' is the 3rd-person, masculine, imperfect inflection of the verb 'ana which can literally mean 'to mean.'

  47. Chris said,

    June 28, 2011 @ 12:11 pm

    File this under mediocre science blogging: This LL post just got linked to by Andrew Sullivan earlier this morning, along with the D'Arcy chart. Unfortunately, the Sullivan post provides exactly ZERO analysis or discussion of the chart itself (even though it is posted as quite a large image). Another example of Sullivan's tendency to re-post science related quotes without much consideration of what they mean (he rails against political bloggers when they do the same kind of thing).

  48. Matt Stuttle said,

    June 29, 2011 @ 5:01 am

    See, it starts off with the iPeeve, then you get your phone giving you a Flesch score. Next thing you know your phone will chirrup in a meeting and a voice from your pocket says "Hi! It sounds like you're trying to negotiate a deal! Would you like help with that?"

  49. Link love: language (32) « Sentence first said,

    July 5, 2011 @ 7:38 am

    […] and the colours of the alphabet. Railspeak should be terminated. Spoken style correction: the iPeeve™. Top punctuation boffins sort out the multiple shriek stop!!!!! ScriptSource: documenting the […]

  50. George D said,

    July 21, 2011 @ 9:37 pm

    It's my (unconsidered) hypothesis that 'like' , 'you know' and similar fillers lower the ostensible truth value of a statement by qualifying it or implying familiarity with its content. This reduces the social gap between the speaker and the listener.

    In situations of high social status competition this would be particularly valuable. This is how I see them used in classic 90s and 80s film and television of the Valley Girl genre.

    The statement that these hold no purpose seems to come from those who are unfamiliar with their use. Which is kind of fair enough on one level – if they annoy you then you're going to block them out. But if you're going to make strong statements about them, then familiarity with context is essential.

  51. Mary Apodaca said,

    July 30, 2011 @ 4:40 am

    Two observations and then a question:
    I remember first realizing, c. 1959, that I was hearing cool people, not just kids, but musicians, for instance, using 'like', so I started trying to do it myself. it never, like, entered my everyday usage. Or is it "entered my everyday usage, like?"
    I haven't seen any messages bringing up the French use of 'genre.'

    Question: Does the fact that it was hip to use 'like' in the 50s and/or the French use of 'genre' that seems so like 'like' change the dynamic of the analysis of the usage?

  52. John said,

    September 12, 2012 @ 6:26 pm

    What is the source of the use of the phrase "was like" as a substitute for the verb "said?" Instead of saying "I said" I say "I was like…" This is strange.

RSS feed for comments on this post