Repetition disfluency

« previous post | next post »

Modern mass media expose us to a lot of political speech, and therefore to a lot of journalistic commentary on politicians' individual speaking styles.

Regular readers know that I don't generally have a lot of sympathy for attempts to tag Politician X with his or her allegedly characteristic X-isms, whether it's the collections of Bushisms and Palinisms, or Barack Obama's supposed over-use of first-person singular pronouns. This is partly because the empirical support for these attempts is generally below even the usual punditorial standards, and partly because style ought to be less important than substance.

However, I've already commented on Gov. Rick Perry's fondness for a certain hedge/filler ("If you will", 7/29/2011), and this post will describe (what seems to be) his characteristic mode of disfluency when extemporizing. Hypocrisy? We blog, you decide.

Most of us speak fluently when we're comfortable with our topic and our audience. We have our well-rehearsed stories and our ready arguments — but when we have to frame a series of thoughts that are new or difficult, or when we're especially concerned to avoid misunderstanding, the compositional process slows down below our normal speaking rate. This creates "dead air" that we need to fill in some way.

The commonest solution is "filled pauses", ums and uhs and the like.  Another solution is to throw in a few words or phrases that are largely independent of the syntactic and semantic structure of the argument or explanation: hedges like "like" or "if you will"; audience appeals like "OK", "right", "you know", or "if you see what I mean"; and so on. And another option is just to let the compositional dead air surface as silent pauses.

All of these options have other functions, or at least other effects. For a start, any interpolation tends to highlight whatever follows it. And different interpolations have different pragmatic and sociolinguistic associations — see the list of links at the bottom of this post for discussion of some examples. Most people use such interpolations even in fluent and well-rehearsed performances, and some people use them a lot. But whatever else is going on, these various interpolations all do fill up dead air, and therefore are likely to be more frequent when (for whatever reason) we're composing our contributions more slowly.

A way to fill up speaking time that I don't think we've discussed before is to repeat short words or phrases, or the beginnings of longer words or phrases. Sometimes these "false starts" precede a self-correction, where a compositional option is started and then abandoned in favor of another path entirely. But there's another pattern, in which the repetitions precede a fluent continuation.

These repetition disfluencies seem analogous to stuttering, but I don't perceive them as stuttering — and they don't seem to be described as stuttering by others. From time to time I've noticed particular speakers in real life who are especially prone to this kind of disfluency, but I've never had a significant sample of recordings to study. So I was interested to find this pattern in several of Gov. Perry's  interviews, for example here.

Here's an example. The interviewer's question:

Governor, why does Texas cons- continue with abstinence education programs? But we are the third highest teen pregnancy- we have the third highest teen preganancy rate among all states in the country. The questioner's point is ((it)) doesn't seem to be working.

The start of Gov. Perry's answer:

It- it- it- it- it works
uh maybe it's uh it's the uh
maybe it's the way it's being taught or the way that it's- that it's being
a- a- applied out there, but the fact of the matter is
uh it is the best
form of
((t- t-)) to- to teach our children

A bit later:

I'm- I'm a- I'm just going tell you
from {laughs}
I'm going to tell you from my own personal
abstinence works
and- and- and the point is
if- if- if- if- if- if we're not
teaching it
and if we're not
impressing it upon them
then no

One reason this doesn't sound like stuttering is that it affects mainly phrase-initial (and vowel-initial) function words like "it" and "and" and "if", as opposed to syllable-initial consonants in content words. It's likely that the temporal profile of these compositional repetitions is also different — but unfortunately there are no published corpora of stuttering, as far as know, so there's no way to test this hypothesis.

[I should emphasize that Gov. Perry is usually a very fluent speaker, even when extemporizing, as you can find out for yourself by looking at some of his clips on youtube or elsewhere. The point here is not that he's unusually disfluent, but rather that his disfluency manifests itself in a way that is linguistically (though probably not politically) interesting.]

Some other LL posts on disfluencies of various sorts:

"And uh — then what?" 1/5/2004
"Reanalysis — and not", 2/8/2004
"um, em, uh, ah, aah, er, eh", 5/2/2005
"The the the and the thee the", 7/26/2005
"Young men talk like old women", 11/6/2005
"I mean, you know", 8/19/2007
"Speaking (in)coherently", 11/20/2008
"Filled pauses and faked audio", 12/28/2008
"More (dis)fluency and (in)coherence", 12/31/2008
"Who knows?", 1/7/2009
"Uh accommodation", 4/2/2010


  1. David Donnell said,

    August 15, 2011 @ 10:42 am

    I've observed that virtually no political junkies, even if they're otherwise intellectual people, are interested in these topics. In the polarized world of politics, "journalistic commentary on politicians' individual speaking styles" is red meat that feeds hungry partisans on both sides, and such partisans couldn't care less about facts that contradict received beliefs about the speech (and therefore intellect) of politician X on the opposing team.

  2. Dan Lufkin said,

    August 15, 2011 @ 10:46 am

    Yeah, right, but abstinence, if you will, didn't, did not work — that is to say, it wasn't effective, entirely, for Gov. Perry"s parents, his mother and father, that is..

  3. HP said,

    August 15, 2011 @ 1:06 pm

    The point here is not that he's unusually disfluent, but rather that his disfluency manifests itself in a way that is linguistically (though probably not politically) interesting.

    I wouldn't be so sure about that. Given the subject matter, I'd say it's not only politically interesting, but psychologically interesting as well. But those are topics for other blogs.

  4. slobone said,

    August 15, 2011 @ 1:36 pm

    I would have called this stammering, rather than stuttering, but I have no idea what the technical distinction is between them, if any.

    [(myl) To a first approximation, people in the UK use "stammering" where people in the U.S. use "stuttering". But what's going on in these clips is neither stammering nor stuttering, as far as I can tell.]

  5. Janice Byer said,

    August 15, 2011 @ 2:46 pm

    Might we call it "struggling" to say what apparently doesn't feel quite right? As Mark notes, we're fluent when we're sure about what we're saying.

  6. Janice Byer said,

    August 15, 2011 @ 3:00 pm

    Struggling to find the right-wing words is what I really want to call it.

  7. Dakota said,

    August 15, 2011 @ 4:18 pm

    The thought that someone with two children could utter those words without experiencing cognitive dissonance is…inconceivable.

  8. mgh said,

    August 15, 2011 @ 9:02 pm

    I shared slobone's impression that this would be called stammering, and it brings to mind both Jimmy Stewart and Bob Newhart. A quick google finds Bob Newhart's own distinction between stammering and stuttering:

    "Stammering is different than stuttering. Stutterers have trouble with the letters, while stammerers trip over entire parts of a sentence. We stammerers generally think of ourselves as very bright. My own private theory is that stammerers have so many ideas swirling around their brains at once that they can't get them all out, though I haven't found any scientific evidence to back that up."

  9. marie-lucie said,

    August 15, 2011 @ 10:57 pm

    I think that stuttering is a pathological condition which affects the sufferers under most circumstances, while stammering occasionally occurs in otherwise normal, even very fluent speakers (as in Perry's case) when faced with an unexpected and emotionally charged event or topic which causes a physiological reaction which interferes with their usual control of their speech apparatus. The difference may be one of degree rather than of nature. (There was an excellent, very well-written article by a stutterer in Slate a few weeks or months ago, describing what happens to the afflicted person).

  10. Jake Nelson said,

    August 16, 2011 @ 12:19 am

    I agree with those who label this stammering- my own idiolect distinguishes stuttering and stammering the same way.
    I've generally noticed it among those used to conversation styles where they have to interrupt to get a word in- they start a statement and keep starting it repeatedly until it seems like there's an opening (as in, people have noticed they're trying to say something and paused to let them do so), and then they proceed with the rest of the statement. After long enough, it becomes habit even when there's not crosstalk. (I know a couple people who do this, and they have very talkative family members.)

  11. Jerry Friedman said,

    August 16, 2011 @ 12:42 am

    @slobone: I was brought up with a similar distinction: stuttering is repetition of initial consonants or consonant sounds; stammering is silences or prolonged vowel sounds in the middle of words, or things like that. When this came up here before, it seemed the professionals in the field make no such distinctions. (I'm reminded of the "technical" distinction I so carefully made between midgets and dwarfs, only to find that in medicine the conditions are considered different types of dwarfism. But specialist usage doesn't make common usage wrong, maybe not even when common usage attempts a technical distinction.)

    On another tangent, I heard a bit of discussion today on whether forms of fuck (a word nobody said in full) are used to fill pauses, like um and you know. The presenter said he didn't think so, but I feel sure I've heard it. Not in political speeches.

    Back on topic, Kipling's characters in Stalky & Co. sometimes repeat initial words in moments of strong emotion (such as hilarity) or when taken off guard. Search for "The-the", "let-let", "stinkers", and "but-but". I suppose repeating an interjection, such as "Why—why—why", is a little different?

  12. LDavidH said,

    August 16, 2011 @ 3:15 am

    @Dan Lufkin & Dakota: The issue isn't whether abstinence works or not (of course it does: no sex, no baby – there's only ever been one exception to that in the history of humankind), but whether teaching about it results in less teenage pregnancies, which is a completely different matter.

  13. maidhc said,

    August 16, 2011 @ 4:03 am

    As an ordinary person, if you asked me a question that required me to think about something I hadn't thought about before, I probably would um and ah and generate lots of fragments. On the other hand, if I'm expecting a certain question, I generally run through an answer beforehand in my head so I could produce it fairly fluidly.

    I expect that typically politicians have their staff produce lists of likely questions, so they could work out answers ahead of time. One can see this in action sometimes when politicians launch into their canned answer which is similar to but not quite the same as the question. A good interviewer would press the point, but this rarely happens in the US.

    Many politicians also have a set of non-committed responses that they trot out when they get a question that they don't really want to answer. "I'm glad you asked that, Bob, because that's a question that concerns many families in this great country of ours, etc., etc."

    Neither of my examples is any evidence of sincerity by the politician, but it might be considered part of the job description that a pollie should be able to come out with some reasonable kind of statement when called upon in public. Of course the extreme end of this is the speaker who rolls out total falsehoods with such an air of sincerity that it leaves one doubting the fabric of reality.

    Perry's performance finds me somewhere in between "Did he seriously think no one would ever ask him about that?" and "Here is a guy who is so sincere that he answers every single question from first principles".

    I think that oratory of the Ted Sorensen variety is seen nowadays as a mark of being a "Washington insider", but a certain amount of tongue-tiedness imparts a sort of Jimmy Stewart aura. But an excess of folksiness makes you sound like a blithering idiot.

    I may be too cynical, but I see an army of consultants measuring their candidate's position on the fluency continuum and continually applying corrections based on the latest polls.

  14. GeorgeW said,

    August 16, 2011 @ 5:11 am

    It seems that we accept hmms, uhs, stammers and the like as allowing the person to formulate a coherent response, but we find a period of silence embarrassing and assume the person is stuck, frozen, unable to respond.

    [(myl) Speak for yourself. Being "blocked" can be a symptom of a stuttering-like condition, but the different between that and pausing for thought is usually obvious, certainly to the person affected and also to their audience.]

  15. GeorgeW said,

    August 16, 2011 @ 6:35 am

    myl: I was thinking of filling a pause while thinking with sounds (hmms, ums, etc.) as opposed to silence. I think — and speaking only for myself — that they can be interpreted differently; the former contemplative, the latter stuck.

    [(myl) Presumably people fill dead air with semi-contentless vocalizations because they feel that silence might be misinterpreted — to the extent that such behavior is chosen rather than automatic, anyhow. But several other (mis-) interpretations are available, depending on the situation — not paying attention, choosing not to answer, giving up one's conversational turn, and so on.

    Literally being "stuck", in the sense of having something to say but being unable to get it out, is generally pretty obvious at least in situations where vision is also available.

    I've read and heard many times that there are large inter-cultural differences in the size, distribution, and interpretation of conversational silences. I've never seen any very persuasive evidence of this, but this is probably because the people who study such things generally rely on impressions and anecdotes, which are probably correct in this case.]

  16. david said,

    August 16, 2011 @ 6:51 am

    The 'repeating entire phrases' reminds me of this Fry and Laurie clip:

  17. GeorgeW said,

    August 16, 2011 @ 8:25 am

    myl: "Literally being "stuck", in the sense of having something to say but being unable to get it out, is generally pretty obvious at least in situations where vision is also available."

    I was actually thinking of being mentally stuck, i.e. unable to spontaneously formulate a response.

    I think the contemplative pause can also be signalled by physical gestures like the thinker finger-to-lips move, lips tightly pursed, etc.

  18. Eric 0nes said,

    August 16, 2011 @ 9:38 am

    @Jerry Friedman
    Yes. I have used that particular disfluency before, having picked it up from others, I suppose. It's usually a grammatically-placed "fuckin'," and realized (sorry for my lack of IPA or anything unamateurish) as "fuckiiiiinn'…" taking up roughly as much time as an "uh…" with the second syllable a pronounced with a higher tone than the previous.

  19. Peter Trudgill said,

    August 16, 2011 @ 2:18 pm

    I was just wondering if Americans are aware that the filler "if you will" is peculiar – or so it it seems to me – to North American English. The rest of us don't use this.

  20. Jerry Friedman said,

    August 17, 2011 @ 4:50 pm

    @Eric Ones: Thanks for the confirmation.

    I'm an amateur too, but my impression is that IPA isn't very good for the kind of thing you were showing. Maybe I'm wrong.

  21. Kaitlin Duck Sherwood said,

    August 23, 2011 @ 9:37 am

    Jean Fox Tree published an interesting paper on speech disfluencies about ten years ago which I interpreted as saying two things. 1. Ums and uhs were signals to the listener that the speaker was still working on that utterance and had not changed his/her mind or gone mentally missing. 2. The length of the pause was signaled by the utterance: um for short pauses, uh for longer pauses.

RSS feed for comments on this post