What’s that supposed to mean?

If you watch any episode of the British soap EastEnders, you can guarantee that someone will come out with the line “What’s that supposed to mean?”. The EastEnders scriptwriting team employ this expression so frequently that I suspect they have a button on their laptops that generates it at a single keystroke. There is no point in trying to analyse “What’s that supposed to mean?” into its component parts: it is what we call “non-compositional” because you can’t guess the meaning of the whole phrase (or its communicative function) from the individual words. This is an example of  “chunking” – the tendency of words to appear in prefabricated strings. And chunking has been the subject of a flurry of recent blog posts – see here and here.

None of this is new. Over 80 years ago, the linguist and language-teaching pioneer Harold Palmer recognised the importance of ready-made sequences in the way we store, process, and understand language. The arrival of large corpora in the 1980s gave this theme fresh impetus by providing empirical evidence for the notion that word combinations are anything but random. This undermined the idea that language consists simply of “grammar” (a set of rules) and “vocabulary” (individual words that combine with one another according to these rules), and gave way to a new model based on what John Sinclair called the “idiom principle”.  Sinclair identified recurrence as an essential feature of language: much of what we say consists of frequently recurring combinations of words or – as he put it – “semi-preconstructed phrases that constitute single choices”. These phrases (or chunks) are seen as the key to achieving naturalness and fluency.

But although Sinclair’s ideas have been extremely influential, some experts remain sceptical. Michael Swan worries that an emphasis on chunking risks sidelining what he sees as more central features of a language-learning programme: grammar and vocabulary.  It is true that learning chunks may not seem so different from a superficial “phrasebook” approach to language. There is also, in some cases, a fine line between a frequent and typical word combination and a worn-out cliché which students would be better off avoiding. Finally, those who promote the idea of ELF (English as a Lingua Franca) argue that in ELF interactions “idiomatic” language is best avoided: it may get in the way of effective communication, and it increases the learning load for non-mother-tongue speakers, without any obvious benefit. In this spirit, Brett Reynolds, another “chunking sceptic”, recently wrote that memorising a collocation such as strong wind (just one of many thousands of collocations) may not be worth the effort when there is so much else to learn in terms of grammar and vocabulary. He asks: “Wouldn’t ‘big wind’ or ‘heavy wind’ get them by just fine?”

Well, up to a point. My first response would be that most of us learning a second language want to sound natural. We may not always achieve it, but it’s a good aspiration, and this means more than simply “getting by” with expressions like “heavy wind”. But there are more fundamental objections, and I will deal with these in another blog on this important subject.