Two weeks ago, I attended a lecture by Prof. Michael McCarthy at Osaka Shoin Women’s University, and sponsored by Cambridge University Press. The title was “The Learner’s Turn”, meaning both “It’s the learner’s turn to say something” and “How learners take turns in conversation”. Michael McCarthy is a linguistics expert specialising in corpora, and has authored several books on corpora and language teaching, including From Corpus to Classroom and the recent Touchstone EFL series.

McCarthy gave several examples from spoken English corpora of how fluent speakers take turns in conversation. He said, “Every turn in the conversation, except for the opening one, is a response to the previous speaker”, and proceeded to demonstrate how this works from actual examples. “Well, actually…” was one example, showing a speaker signaling that his answer was not a straightforward answer to the question that was put to him. Remove the “well actually” and you get an abrupt change of topic that sounds like a non-sequitur, and jolts the smooth flow of the conversation. The examples were all very interesting.

“Flow” and “fluency” were key themes in McCarthy’s talk, and ironically it was at one such moment when McCarthy was talking about “confluence” and “fluency” that I wondered whether I wanted to “go with the flow”.

McCarthy stated that conversation must flow; it’s not just an exchange of information. An example he gave was a short exchange between two strangers, one of whom asks the other for the time. The respondent, obviously searching for some kind of timepiece, but unwilling to simply remain silent (thus possibly giving the interlocutor a wrong message, such as “I’m deaf; did not hear your request”), says “Erm… uhhh….” before finding the timepiece and announcing the time. No problem with this; I teach the importance of flow in conversation to my students, and it’s one of the reasons I like and use AlmaLang’s Conversations in Class, especially its Cultural Golden Rules of English conversation (pdf).

McCarthy then gave another example of confluence or cooperation between speakers in conversation: leaving a question half-finished with a hanging “or?”, such as, “So, you wanna have lunch, or…?” with several examples from the corpora of spoken English. McCarthy pointed out that it would be odd, strange, abrupt for the interlocutor to follow this with a cold, “Or what?”, and this is the point, dear reader, where a doubt crept into your author’s mind.

I don’t disagree that it would be odd, or strange or cold to ask “Or what?”, but does this mean that speakers should avoid using this, simply because the corpora shows us that 99% of conversational exchanges show the speakers politely taking their turns and showing cooperation and confluence? There may be cases where you want to be cold and abrupt. Should we pattern our ways of speaking on that of the majority? Does the fact that “this is how most people speak” automatically mean that this is how I should speak? Isn’t this conversation by numbers?

McCarthy’s other main theme, apart from “fluency”, was that teachers can make use of this vast amount of corpora-data to identify what fluent speakers do with the (English) language, and then teach this to learners of English. The justification he gave was that by this means, teachers can help learners at least give an appearance of greater fluency. In other words, you take data on what most, fluent speakers of the language do, and then teach this to language learners so that they will sound like fluent speakers as soon as possible.

This has been tried before, in the area of language-learning strategies. The idea was first proposed by Joan Rubin and Anita Wenden in the classic A. Wenden & J. Rubin (Eds.), Learner strategies in language learning (145-156). London: Prentice Hall. The idea was, first, study efficient EFL learners’ language-learning strategies; then, teach those strategies to other, less efficient, learners, and presto!

This idea morphed into teaching autonomous language-learning strategies to less autonomous students, as it seemed that efficient language-learners were autonomous. I have had less than stellar results with this theory.

McCarthy did not address the question of why non-fluent speakers do not use the strategies of fluent speakers. Is it because they are less quick-witted than fluent speakers, or is there something else at work? The parallel that occurred to me is that of infants learning to walk or talk. Why not “teach” crawling infants the techniques that “successful walkers” use? Then they could perhaps make faster progress towards walking.

Or, why not “teach” speaking strategies to babbling infants? The babblers could perhaps then speed through the annoying babbling stage and move more quickly into speech (or perhaps into “speech-sounding” verbal behaviour), that would most assuredly impress relatives and neighbours!

But perhaps there is a reason why infants go through a crawling phase before walking? Perhaps there is a reason why infants babble before speaking? (See Wikipedia’s entry on Jean Piaget’s developmental process model for children). Similarly, perhaps there is a reason why fluent speakers use such words as “actually” and why non-fluent speakers do not; i.e. because the use of “actually” in conversation requires a conceptual understanding of the target language that is beyond beginners or even intermediate speakers. Perhaps the progression from beginner to fluent speaker requires going through stages, like the stages of physical and mental development in infants; stages that can neither be rushed nor bypassed.  (Cf Second Language Acquisition: Developmental Patterns).

Although the theory sells a lot of books, I remain sceptical. And rather than training speakers to “sound” fluent when in fact they are not, I  prefer a more cognitive approach: helping people to express what they want to say, rather than “selling” them the most popular expressions and turns of phrase.

Tags: EFL, language + rhetoric