Scott Thornbury, Freelance, Barcelona, Spain

The case that is often made for the importance of teacher language awareness typically follows this logic: just as other professionals (lawyers, doctors and so on) claim authority by virtue of the knowledge they have of their subject, so too should teachers. Since the subject of language teaching is language, it follows that this is what they should know about.

Ignoring for the moment the question as to whether the subject really is language (as opposed to, say, communication), we are still left with a number of problems, specifically: Where does this knowledge come from? How much knowledge is sufficient? And, what does the teacher do with this knowledge?

Let’s dispense with the last question first, since it is not the major thrust of this article. If ‘language knowledge’ is understood as being explicit knowledge of the systems that comprise the target language, as described, for instance, in a grammar or a dictionary, what do teachers do with this knowledge? Presumably they do not simply transmit it, unmediated, undigested, in the same form in which they originally encountered it. Even if you are in favour of the explicit teaching of grammar, you are unlikely to see much value in simply declaiming the rules of grammar to your learners, in the manner of Thomas Gradgrind, in Charles Dickens’ Hard Times: 'Now, what I want is, Facts. Teach these boys and girls nothing but Facts. Facts alone are wanted in life.’

Language learning, after all, is not the cumulative acquisition of ‘facts’ about the language, but the capacity to adapt the resources of the linguistic environment to one’s own communicative needs, and language teaching aims to provide the optimal conditions for that process to develop. Whether or not this necessarily involves the picking up of some isolated facts along the way is still a hotly contested point. (If, like Gradgrind, you are teaching ‘boys and girls’, it’s likely that the facts will fall on deaf ears.)

But, even if you don’t teach the grammar explicitly, you will still need to provide feedback on learners’ output, to anticipate learners’ problems, to assess the difficulty of reading and listening texts, and to make sense of syllabuses and assessment procedures, all of which require a degree of language awareness.

So, on to our other two questions, which I see as being interrelated: Where does a teacher’s language knowledge come from? And how much knowledge does a teacher need?

For teachers whose first language is not English, their language knowledge is likely to have been the residue of their many years studying the language, often quite formally, including what ever preparation they may have had to become teachers. For those teachers whose first language is English, on the other hand, knowledge about language is more often acquired on the job, principally from the books they teach from, perhaps reinforced by reference to pedagogical grammars, although nowadays it’s likely that they go online to find answers to their questions.

Either way, the quality of the information available varies considerably. Take, for example, these ‘rules’ from online grammar sites:

• We can't use this tense [present continuous] (or any other continuous tense) with stative verbs.
• Do not use the with uncountable nouns.
• Shall is used with first-person singular and plural, and will with second and third person.

A few moments reflection or corpus searching should be sufficient to find plenty of counterexamples!

To be fair, pedagogical rules involve a degree of simplification, as Michael Swan (1994, p. 46) reminds us: ‘One will often need to compromise with truth for the sake of clarity, simplicity, conceptual parsimony or relevance.’ Nevertheless, the above ‘rules’ would seem to be a far remove from what might constitute ‘sufficient’ language awareness for teachers.

Coursebooks are not exempt from such criticisms. Nor even many pedagogical grammars. Dave Willis (1990) challenged the view, for example, that ‘reported speech’ merited the attention it gets in coursebooks and descriptive grammars. Based on corpus data, he concluded, ‘It is difficult to sustain the argument that reported statement is a useful grammatical category at all… Reported thought is much more common than reported speech. But reported thought does not figure in pedagogic grammars with anything like the same inevitability as does reported speech’ (1990, pp. 21-22).

In similar fashion, Graham Burton (2021) used corpus data to disprove the frequently expressed prohibition against combining the future form going to with the verbs go or come – as in I’m going to go out or They’re not going to come to dinner. Burton found that these forms were significantly frequent and concluded that ‘this commonly given rule should no longer feature in explanations of the use of going to used by learners of EFL.’ As to its persistence in pedagogical grammars, he theorized that it may be considered true ‘simply by virtue of the fact that it appears so frequently in print’ (p. 23). He suggests that ‘materials designers and teachers may benefit from at times taking a more critical approach to the grammatical explanations available to them, and be ready to test them against empirical evidence, such as that offered by language corpora’ (2021 p. 24).

A second, and, arguably, even less reliable source of teacher language awareness is the teacher’s own intuitions. These are frequently prompted by learners’ spoken or written output, where snap decisions need to be made about accuracy or idiomaticity. They are also the stock in trade of teachers’ staff room conversations or – more typically nowadays – online social networks.

A recent example: a question was posted on Twitter querying the grammaticality of the following fragment of a movie subtitle: ‘But because if it had have happened…’. The post received over 60 responses, not one of which referenced attested data in the form of corpus findings, but instead drew solely on the respondents’ intuitions, such as:

• I’d prefer “had to have” I think but it just strikes me as a pretty normal omission
• I was about to answer with a no, but then tried it with stress on had (rather than on have), and it surprised me with how natural it felt
• It almost makes sense but I can't seem to figure out how this is different than just "had happened" instead of "had have happened"
• Totally grammatical for me as long as ‘had’ carries stress
• Clearly a mistake. Should've been 'if it had happened...'

Intuitions like these are notoriously unreliable. As John Taylor (2012, p. 11) notes, ‘We should take “what speakers say they say” and “what speakers think they say” with a large grain of salt. We need to turn to actual data, records of what people “actually say”.’ At the very least, a corpus search would have rapidly established whether the example cited above was a one-off, and hence perhaps a typo or a learner error. In fact a quick search of the 15 billion-word News on the Web (NOW) corpus produced 29 examples of the string had have happened such as:

• If this had have happened 10, 12, 18 months ago it would have been a red card.
• Imagine if this had have happened in London, or anywhere else in the UK.
• If that had have happened in any other country she might not even be alive.

A comparison search of if this had happened produced 806 occurrences in the same corpus, compared to just 9 for the complete string if this had have happened, suggesting that the latter form is rare but not so rare as to be considered an aberration – if we concur with Charles Fries’s edict that ‘there can be no “correctness”, apart from usage’ (1940, p. 15). Closer examination of the data would show whether there is a bias towards spoken rather than written English, towards US rather than UK varieties, and towards particular verbs. (The British National Corpus has 44 examples of if I’d have known…, for example). The point is that this kind of information is freely available at the click of a mouse, and yet is very seldom appealed to in these online and staffroom discussions about language.

A further problem with intuition, in the case of speakers whose first language is not English, is the possible interference of the first language. Recently a teacher emailed me to query the claim that I had made in a book on grammar that the learner utterance I had a big surprise was not idiomatic. Without knowing the teacher’s first language I emailed back:

Regarding I had a big surprise, there is only one occurrence of this in the 1 billion-word Corpus of Contemporary American English, whereas there are 275 instances of I was very surprised. This suggests that while I had a big surprise is well-formed, it is not idiomatic. The way that the concept is typically realised is I was very/greatly etc surprised. In the related Corpus de Español,(700m words) however, there are nearly 200 examples of the different forms of tener una gran sorpresa. In Spanish this is clearly idiomatic, in the sense that it is a conventional way of expressing the same idea. The learner who originally wrote I had a big surprise was a Spanish speaker, suggesting an effect of first language transfer.

The teacher wrote back, conceding that he too was a Spanish speaker.

Corpus data, of course, are not infallible, and need to be interpreted cautiously. And just because a sequence doesn’t appear in the data doesn’t mean that it may never appear one day. Notwithstanding, corpora are often the best, cheapest, quickest guide we have available in order to answer such questions as: ‘Do you say A or B?’ ‘Is C correct?’ ‘What’s the difference between D and E?’, as well as for verifying the received wisdoms of coursebooks and pedagogical grammars. A little knowledge (of how to use them) is not necessarily a dangerous thing.

References

Burton G. (2021). ‘Are you going to go?’ Putting a pedagogical grammar rule under the corpus spotlight. Glottodidactica XLVIII/1. Adam Mickiewicz University Press, 7-26.

Fries, C. C. (1940). American English grammar. Appleton-Century-Crofts.

Swan, M. (1994). Design criteria for pedagogic language rules. In M. Bygate, A. Tonkyn, & E. Williams (Eds.), Grammar and the language teacher. Prentice Hall, (pp. 45-55).

Taylor, J. R. (2012). The mental corpus: How language is represented in the mind. Oxford University Press.

Willis, D. (1990). The lexical syllabus. Collins COBUILD.