After a career spent trying to drum padding and hesitations out of my own speech and that of trainee broadcasters, it seems I might have been contributing to lower communication values after all. This story in Nature covers the field.
'Ums' and 'uhs' contain meaning, say US psychologists. 'Uh' and 'um' send information to listeners just like proper words, say Herbert Clark and his colleague Jean Fox Tree. They analysed transcripts of conversations between academics and from phone conversations and answering-machine messages.
English speakers lob in 'um' before a long pause and 'uh' in front of a brief hiatus, the analysis revealed. People even create compounds such as 'the-um' or 'and-uh', says Clark, showing that speakers know that there is going to be a problem after the word even before they begin it.As I said, as soon as you "notice" them, the task becomes impossible, like the centipede asked how he walks with all those legs. It argues for distributed processing of language rather than hierarchical, the "umm engine" (h)umms away doing its own thing, applying the accelerator, the brakes, signalling into the congition stream without us needing to worry about oversseing it or applying conscious resources to it, but just reaping the benefit of its service.
The researchers believe that speech contains two streams of information, which speakers blend and listeners unravel. One strand contains the meaning. With the other - asides such as 'um', 'uh', 'like' and 'y'know?' - speakers comment on how smoothly their train of thought is running. "Remarkably, we do these things more or less simultaneously during conversation," says Clark.
Speech researcher Robin Lickley agrees that 'uh' and 'um' should be treated as genuine words. "People tend to think of these things as sloppy, whereas they're perfectly normal," says Lickley, who works at Queen Margaret University College in Edinburgh.
He also likes the idea that speech contains parallel strands. But he doubts that 'uh' and 'um' really perform the function that Clark and Fox Tree claim for them. "I don't think they're inserted to help the listener - about half the time people don't notice them," he says. "They just keep the flow of speech going." (There's a difference between noticing something and it contributing to the meaning of the conversation, tsk, can do better)
Other studies have shown, however, that listeners process speech more quickly with the 'ums' and 'uhs' left in than when they are taken out. And beginning an answer with 'um' is interpreted as showing greater uncertainty than a silent pause of the same length.
So has studying 'uhs' and 'ums' made Clark more conscious of his own hesitations? "If you aren't careful it's a killer, but I try and keep it from becoming one," he says.
What interests me is that, even among professionals, this can still be difficult to accept. After all, we drive cars without being aware of what we are doing and we've had a lot more experience with language which has co-evolved to match, as perfectly as possible, our congitive capacities.
Now Gary blogs a little about prescriptive types and rightly gives a serve to those dedicated to rules of effective speaking but I would also distinguish between conversational and colloquial speech on one hand, and effective presentations on the other. He asks us to consider a couple of hours of bad standup comedy as demonstrating that there is more to language than words and I WHOLLY agree, but I'd also be interested to do an Umm, err count on a top standup like billy Connolly or a traditional one like Les Dawson.
Billy's language especially is well larded with wonderful and expressive, not to mention emphatic language, but it is also spare and elegant and beautiful to listen to, and there is almost no padding. When confronted by that kind of mastery it seems to me that there are at least two ways of talking about the same thing, one that most of us use, littered with props, crutches, ellipsis and other organic processes that get there in the end, and another that uses language like a scalpel and bucket, the one to open us up and the other to pour meaning into us.
NB, the scalpel for opening and the bucket for pouring, not the other way round.