There are a limited number of hours in the school day, days in the school week, weeks in the school year. These are children’s precious years of education, so we have an absolute duty to ensure schooling is the best there is. Schooling requires curriculum content that is motivational for children, appropriate for what they need to learn and based on the best evidence of what works in the classroom.

In addition to more than 150 individually specified elements of spelling, and five pages of detailed specification of grammar (that are statutory), primary teachers in England are required to teach and assess their pupils’ knowledge of the following terms:

• letter, capital letter 
• word, singular, plural 
• sentence 
• punctuation, full stop, question mark, exclamation mark 
• noun, noun phrase 
• statement, question, exclamation, command 
• compound, suffix 
• adjective, adverb, verb 
• tense (past, present) 
• apostrophe, comma 
• preposition, conjunction 
• word family, prefix 
• clause, subordinate clause 
• direct speech 
• consonant, consonant letter vowel, vowel letter 
• inverted commas (or ‘speech marks’) 
• determiner 
• pronoun, possessive pronoun 
• adverbial 
• modal verb, relative pronoun 
• relative clause 
• parenthesis, bracket, dash 
• cohesion, ambiguity 
• subject, object 
• active, passive 
• synonym, antonym 
• ellipsis, hyphen, colon, semi-colon, bullet points 

And high-stakes statutory tests in England have up to now assessed this grammatical knowledge, typically using one-mark answers. What’s more, teacher assessment has also heavily emphasised grammar. We note the government’s welcome moves to change statutory assessment of writing, but, in general, the statutory specifications for teaching grammar, particularly the learning of much of this list of terms, fails all three of our tests of: motivational for children, appropriate and based on the best evidence. But does our argument about grammar matter? Is this just a “sterile debate” (as Debra Myhill put it in a recent article commenting on our work).  

A few weeks ago there was this headline on the Tes website “Teaching grammar does not improve children’s writing ability, research finds”. The piece was based on a summary of our recent research paper. A significant finding from the paper, and our research that underpinned it, was that England’s national curriculum requirements for grammar teaching, particularly in primary schools, are not supported by robust research evidence in relation to improving children’s writing. By robust evidence, we mean experimental trials that include as a minimum one intervention group (such as grammar teaching) and one control group (such as “business as usual” teaching). And preferably trials where the children/teachers in these studies have been allocated to control or intervention using random allocation. By robust evidence, we also mean basing decisions about effective teaching on multiple studies, not just single studies.

In a response to the Tes piece, Debra Myhill alleged that, “… the article makes other claims that are simply inaccurate and suggest a rather hefty misunderstanding of modern ways of thinking about the relationship between grammar knowledge and reading and writing”.

Not one of the three professors who peer-reviewed our paper prior to submission, nor the three international referees plus at least one editor, who reviewed our paper questioned our understanding about language.

‘Misrepresentation of our findings’

The next criticism made by Myhill was that we said there was “no evidence” to support the teaching of grammar to support writing, or as she put it, “Strange then to draw the conclusions about no evidence?” This is a complete misrepresentation of our findings.

In the paper we say this: “With regard to our substantive case of grammar, the current evidence from randomised controlled trials does not support the widespread use of grammar teaching for improving writing among native English-speaking children.” At no point in our article do we say there is “no evidence”; in fact, we present evidence from seven systematic reviews and meta-analyses (each summarising multiple experimental trials) and nine experimental studies, including Myhill’s and her team’s admirable, influential and rigorous study, of grammar in secondary schools.

The “no evidence” point is then pursued by Myhill through the argument that we have misrepresented the two studies that she and her team were involved in. Although we disagree that we misrepresented them, the more important point is that our article was based on the analysis of multiple studies, not just the two that Myhill chooses to defend. In fact, the overall conclusions of our work are that governments need to pay more careful attention to evidence, even if this evidence is contrary to ministers’ personal beliefs and ideology. And crucially, we clearly highlight some striking and clear differences between what the evidence shows about primary education and what it shows about secondary education.

The sharpest allegation in Myhill’s piece is this: “The real issue is the ongoing controversy about school grammar that seems to push people into ideological corners and biased reporting.” Who exactly is doing this “biased reporting”? If this implies that we are biased in our paper then we categorically refute the allegation and ask for evidence to be provided that we are indeed biased. Once again the seven reviewers of the paper did not allege bias, even if they did make a range of recommendations that is a normal part of peer-review.

A common criticism of critical analysis of policy, research and practice is that it is all too negative, and that no positive, constructive suggestions are made. Well, in this case, we also list a series of very clear evidence-based recommendations for better teaching of writing. And for those who are interested in the wealth of modern, and ancient, ways of thinking, you may be interested to read the outcomes of Dominic’s four years of research on writing published in his new book How Writing Works: From the birth of the alphabet to the rise of social media (Cambridge University Press).

Finally, an area where we have more agreement with Myhill is the idea of listening “to the voices of teachers”. We were pleased that more than 160 people commented on Adi Bloom’s excellent Tes article that summarised our research. We assume that many of these voices were teachers. And, of course, they bring a different kind of robustness in their reaction to our findings; some views along the lines of “Doh! I could have told you that!”

Professor Carole Torgerson is at Durham University’s School of Education and Professor Dominic Wyse is from University College London.