Thornbury: Fence-Sitting & Slime-Sprawling

Bad Mediators 

In Part 1 I suggested that those who write books and give teacher training courses in ELT have a duty to act as mediators between  researchers and teachers, and that most of them make a mess of the job. This opinion was supported by the mini study Thornbury carried out and then reported on at the 2017 IATEFL conference. The study looked at four top-selling “How to Teach English” books which are recommended reading for hundreds of thousands of people studying to get a qualification in ELT, and it found that all four books are based more on the authors’ biases, intuitions, feelings, and what somebody else told them, than on any serious attempt to critically assess what research findings tell us about how people learn languages. In a post on these mediators  I suggested that Thornbury took a disappointingly uncritical look at the data that his study had produced.

Staying on the Fence 

Unlike the four writers he reviewed, Thornbury himself has discussed research findings that challenge ELT orthodoxy more than once, so if he thinks it’s important for him to keep in touch with research and to use research findings to inform his views on methodology, why doesn’t he expect the same of others? And since he’s been so outspoken in his criticism of coursebooks, why didn’t he mention this when discussing his findings? The answer seems to be that Thornbury has developed the unique knack of not just sitting on the fence, but actually living perfectly perched on it. He’s become so adroit at deftly ducking controversy, so practiced at never getting drawn on the political issues raised by the matters he discusses, that he makes the UK liberal democrats look radical. He knows perfectly well that the bosses of the British Council, the publishing houses, the exam bodies, the training outfits and so on will simply not allow any serious attacks on current ELT practice to be made – witness his own publishers’ making it clear to him that they’re “not interested” in his McNuggets views or in what he really thinks of the CELTA course. He knows that the ELT educational system is set up in such a way that teachers are unlikely to hear about “inconvenient” research findings which challenge coursebook-driven ELT, or which show that the Pearson Test of English is built on sand, or which describe the Common European Framework of Reference as “a prime example of in the way political and social agendas can impact on language testing, and how language testing can be made to serve those agendas” (Fulcher, 2005). I suppose Thornbury thinks, like many reformers, that he can be more of a force for change by staying inside the tent than pissing on it from outside. I think that this argument is demonstrably wrong, but never mind; even if that is Thornbury’s view, it doesn’t explain why he doesn’t adopt a more critical stance. In the end, maybe it’s just that he’s a really nice guy and he doesn’t like upsetting people. Well, I can certainly relate to that. 🙂

Misrepresenting Chomsky 

Still, there’s another bone I have to pick with the loveable Thornbury, and that is his continued misrepresentation of Chomsky’s work. If you look at the “daft things the experts said” at the end of my last post, it was Thornbury who said “The NS-NNS distinction is absolutely central to the Chomskyan project”. It isn’t, of course, and, pace Thornbury, the onus isn’t on Chomsky to perform the logically impossible task of proving that some aspects of the knowledge of language that children demonstrate couldn’t have been acquired from input, and it isn’t the case that there’s no empirical evidence to support Chomsky’s theory of UG. In my post Treatise on Thornbury’s view of SLA I pointed to some mistakes in Thornbury’s account of what Chomsky says about language and language learning, and also the faults in his arguments about UG in general and the poverty of the stimulus argument in particular. It’s important to stress that none of the emergentists who Thornbury now seems to think offer the best explanation of SLA, least of all Larsen-Freeman, has offered an explanation for what young children know about language. As Eubank and Gregg (2002) argue, to suggest that language learning is explained by a general theory of associative learning is to leave unexplained

1. the fact that children know which form-function pairings are possible in human language grammars and which are not, regardless of exposure.
2. The countless cases of instantaneous learning .
3. The knowledge children have in the absence of exposure (i.e., a frequency of zero) including knowledge of what is not possible.

Furthermore, to quote Eubank and Gregg (2002, p. 237)

Ellis aptly points to infants’ ability to do statistical analyses of syllable frequency (Saffran et al., 1996); but of course those infants haven’t learned that ability.  What needs to be shown is how infants uniformly manage this task:  why they focus on syllable frequency (instead of some other information available in exposure), and how they know what a syllable is in the first place, given crosslinguistic variation.  Much the same is true for other areas of linguistic import, e.g. the demonstration by Marcus et al. (1999) that infants can infer rules.  And of course work by Crain, Gordon, and others (Crain, 1991; Gordon, 1985) shows early grammatical knowledge, in cases where input frequency could not possibly be appealed to. Landau & Gleitman (1985) even document lexical acquisition in spite of frequent input, where a blind child acquired (her own interpretation of) verbs like “look” despite frequent training  under a different interpretation. 

In a comment on the post about Thornbury’s view of SLA, Gregg wrote this:

Hi Geoff,

I think I’d revise one bit of your discussion: Where you say

Thornbury’s unqualified assertion that language learning can be explained as the detection and memorisation of frequently occurring sequences in the sensory data we are exposed to is probably wrong and certainly not the whole story.

I’d change ‘probably’ to ‘definitely’. It’s striking, and depressing, to see how purveyors of ’emergentism’ continue to ignore the mountain of research showing the complexity of language, and the other mountain of research showing the kinds of linguistic (and other) knowledge young children show, knowledge that no one has been able to account for on an empiricist learning theory, and how they continue to blithely assert that it’s all done by generalization across input samples, without showing how. I’m again reminded of the story … of how Rockefeller became rich. One day as a young lad he found himself with a penny in his pocket. He walked down to the farmer’s market and bought an apple, walked to Wall Street and sold it for 2 cents. Then back to the market to buy 2 apples, back to Wall Street, … At the end of a week he’d bought an old wheelbarrow, and after a month he’d earned enough to put down the first month’s rent on a small fruit shop. But then his uncle died and he inherited everything.    

  

Sprawling in the Primeval Slime 

While Thornbury’s remarks about emergentism are slightly less preposterous this year than they were in 2016 (he’s moved on from the Larsen-Freeman and Cameron’s (2008) nonsense about complex systems to slightly better-argued stuff by the likes of Nick Ellis and Tomasello), he continues to incite a younger generation, who after a quick perusal of Samson, Everett, Wolfe and other reliable sources, share their ignorance with others in the comments sections of the A to Z of ELT blog. Thornbury being Thornbury, he doesn’t tell the young uns that they’re talking baloney, he actually encourages them. In one of his posts this November, Thornbury cheerfully quips that, given the choice between Chomsky’s self-proclaimed triumph of “human reason” on the one hand, and “beastly grovelling in the primeval slime” on the other, he’ll choose the slime every time. The trouble is, he invites the younger generation to join him in the beastly bog; he encourages them to think that their ignorance of Chomsky’s work should be worn like a badge of cool, and he confidently assures them that SLA is best explained as the complex result of a simple process of “reinforcing contingencies set up by the verbal community”. You couldn’t make it up, so it must be true. Well, for the time being we’ll have to leave them to it, happily frolicking in the slime, unconsciously strengthening the associations between who knows what cues, and trust that before they get too much older, the brighter ones will get tired of it, climb out, and leave their genial hero alone with his dirty bucket and spade, there to finally appreciate the power and utility of non-communicative uses of language, or ‘thinking’ as Chomsky refers to them.

Reference 

Eubank, L. and Gregg, K.R. (2002) News Flash: Hume Still Dead.St udies in Second Language Acquisition , 24, 2, pp. 237-247.