Schoolchildren should learn more about grammar. Of course they already know English (if it's their first language) but they also need to be able to talk about it - to talk about prepositions and pronouns, modal verbs and relative clauses, tense and subject-verb agreement.
Suppose everyone knew about subject-verb agreement; the change of verb- form in "I like" and "he likes". One thing that would be obvious is that some verbs do "agree" in this way with their subject, but others don't - nobody makes ordinary past-tense verbs agree, so it is "I liked" and "he liked". But more interestingly, most non-standard dialects don't even do it to present-tense verbs, giving patterns such as "I likes" and "he likes". The point is that we can say all that without using the words "bad" or "poor" at all, because we are just stating the facts. It is no worse for a non-standard speaker to say "I likes" than it is for a standard speaker to say "I liked".
If only. Recently the Daily Mail carried an article about an Ofsted report on a primary school near Portsmouth, where some classroom assistants were heard to use local "Pompey" forms such as "I likes". We accept there is an issue about the use of non-standard English in schools, but that is not how the Daily Mail presented it. For the paper's reporter it was an issue of "poor" grammar, but "I likes" is precisely what Pompey demands. And for the headteacher it was "bad" grammar, but he also said that he was not "denigrating the Pompey accent or dialect - we are all proud of where we come from". If you are proud of it, why call it "bad"?
The point is that if Pompey children learned about grammar through the study of Pompey, they would be able to think about subject-verb agreement in standard English. Instead of learning about it as a moral issue of accuracy or quality, they would see it simply as a difference - and as adults they wouldn't get into the same muddle as Ofsted and the Daily Mail.
A case of context
We are not saying that anything goes. Communication is increasingly occurring in many forms and through many channels. While clarity is vital, context is also important. Studying the forms and structures of regional varieties, street slang, text messaging and spoken language alongside the forms and structures of standard English - as a basis for mastering standard vocabulary, spellings, punctuation tricks, style-shifts, rhetorical devices and grammatical patterns - should be seen not as dumbing down, but as a positive move into the real language of real speakers. Linguists know that young people are adept at code-switching - shifting between different forms and registers of a language - but to do this they need at least two codes, and one of them should be standard English.
The muddled thinking illustrated in the press response to the Pompey story shows how far education is from achieving our goal. And at the heart of the muddle is a great big gap where grammatical analysis should be. Analysis objectifies; it asks "what is?" rather than "what should be?" and its aim is understanding rather than judgment. Judgment can come later, but let's get the facts right first. If we are exploring the language of text messaging, it is not really enough to say that it is just different from the language of speech and writing - we need to be able to identify exactly how and why it is different and which grammatical patterns are apparent.
It is the same with slang. While the fun of slang is its creativity and flexibility, there are some interesting grammatical points. One is about the way in which particular slang terms occupy the same grammatical categories - terms of approval such as "choong", "sick" and "peak" functioning as adjectives, for example. Another is about how new words tend to be assembled from elements of existing words, offering us the chance to look in more detail at morphology, the study of the building blocks of words.
Likewise, the study of regional and ethnic dialects allows us to look at grammatical areas such as subject-verb agreement, pluralisation and double negatives. Once children can analyse their own language, they have a foundation for learning about standard English in all its glory, including all the bits that open up completely new linguistic territory for them - such as prepositions front-shifted with relative pronouns (for example, "the book in which I read it"). Some children can work out the details of these `literary' patterns for themselves on the basis of their reading, but the details are not obvious and may need to be taught explicitly.
That this can be achieved by the explicit teaching of grammar has recently emerged from an important research project at Exeter University. Professor Debra Myhill and colleagues designed a curriculum in which explicit teaching about grammar is central, but is always very clearly focused on some aspect of writing. For instance, in teaching about tense and person, the focus is on the choices available in telling an adventure story. Having learned about the grammar, the children immediately apply it to their own writing. They discuss the pros and cons of "present-tense" or "first-person" narrative, and then play with the alternatives in their own production.
In the project, the strict emphasis on relevance produced a significant improvement in children's writing. This outcome is a significant reversal of the decades-old negative conclusions about grammar teaching. It (at last) answers the complaint that the national curriculum for English had no research justifying the central role it gave to grammar teaching.
Isn't this in the curriculum?
If the national curriculum already gives such a central role to explicit grammar, what are we complaining about? Children are already being taught about grammar, so we're already heading in the right direction, aren't we? Once again, if only. The reality in schools is different. While word classes are taught at key stage 1, the experience of most A-level English teachers is that very few students come to them with much knowledge of any grammar beyond nouns being naming words and verbs being doing words - unhelpful labelling at the best of times.
School leavers seem to know even less grammatical terminology now than they did before the national curriculum was introduced. Our evidence for this claim is a very simple test of grammatical knowledge that was taken by incoming first-year undergraduates in 1986, and then again in 2009. The test is based on a 23-word sentence in which students are asked to identify one example of a noun, one of a verb, and so on through 15 grammatical categories, none of which are more erudite than "finite verb" and "definite article".
In 1986, the typical undergraduate in a language-orientated degree such as linguistics or French could identify 12 categories correctly, compared with only eight for the non-language undergraduates. But in 2009, language specialists in the same institutions (Aston University and University College London) could only manage ten categories, and no institution out of the 13 that took part matched the 1986 figures. The mean scores ranged from ten down to five and below. Not surprisingly, non-language undergraduates found the test even harder.
Why should official policy have had so little effect on teaching outcomes? Could it be because official policy is unrealistic in expecting children to learn such a difficult thing as grammatical analysis? The answer must be no, because other countries achieve much better results. Spanish- speaking language students in the University of Zaragoza took exactly the same test, in English, and recognised a stunning 13 out of 15 categories, with non-language students hard on their heels at 12 categories.
Another possibility is that teachers aren't actually teaching much grammar. Unfortunately, there's no way to know for sure what is actually happening across all schools, but we have the impression from friends and colleagues that many teachers at key stages 3 and 4 are giving lessons about word classes and clause structures, and some are even looking at morphology and non-standard varieties. But this often depends on the linguistic confidence of the individual teachers, rather than on a systematic approach to language that is embedded in the curriculum and grows cumulatively from year to year. Likewise, when students come to be assessed at GCSE we know very well that even at the top ends of the mark schemes for English, linguistic analysis plays only a minor part in the band descriptors.
But we are still hopeful. The new GCSE spoken language study could offer the potential for some systematic study of spoken language, its grammar and structure and the attitudes people have to different forms of it. It is too early to assess its impact on the GCSE curriculum but there are some positive signs so far. Moreover, the huge success of the English language A-level has shown us that there is an appetite among sixth-form students and teachers for a course that gets to grips with grammar and interrogates language.
Even more promisingly, the new UK Linguistics Olympiad has revealed an unexpected enthusiasm among pupils for grappling with the intricacies of grammatical systems. This enthusiasm appears in some schools as early as Year 7, and among boys as well as girls. Children who can work out for themselves the rules for a language they've never heard of are surely capable of understanding the rules for their own language, and are likely to enjoy the experience just as much.
Serious, coherent and cumulative grammar teaching should start in a key stages 3 and 4 curriculum that puts the structure and use of language at its heart. Students in Years 9 and 10 who have already done this kind of work have proved enthusiastic and skilful. They likes grammar and we likes it, too.
Dan Clayton is a research fellow at University College London and has taught English for 15 years. Dick Hudson is emeritus professor of linguistics at UCL.