Change the tests, but don’t ditch grammar

This time last year the Grammar, Punctuation and Spelling test (GPS, popularly called SPaG) was in the news almost every week. The DfE was mocked for not being able to prevent the leaking of a test, the Schools Minister Nick Gibb couldn’t answer a grammar question on live radio, and parents kept their 6-year olds at home to boycott the Key Stage 1 test. Numerous well-known writers, poets and education specialists railed against the tests and their inadequacy. Now, it’s spring and the debate has flared up again. Last month the House of Commons Education Committee published its report on primary assessment. After hearing a large number of experts who made clear that they did not like the tests, the Committee recommended that the Standards and Testing Agency (STA) review them, and that the KS2 test, like the KS1 test, be made non-statutory. (Did you spot the subjunctive verb? I’ll come back to it later.) In May the Guardian published a piece by Warwick Mansell on the SPaG tests in which the arguments in the debate were rehearsed. And the next day another piece appeared in the Guardian.

It is unfortunate that the debate on grammar teaching and testing is muddled because there are several strands to it which are confused or ignored. The first two strands concern whether grammar should be taught in schools and, if so, how. The third and fourth strands are the questions of whether and how it should be tested.

Let’s look at each of these strands.

Why do we teach grammar in schools? Critics say that the curriculum is already full and that we can do without grammar. Of course, it’s true that we can lead a happy and fulfilling life without knowing anything about subordinating conjunctions or determiners. But the same could be said for history. You can go get to the age of 95 without ever having heard of Henry VIII or the Great Fire of London or knowing what a parliament is. However, nobody would deny that learning about history enriches us and helps us to understand the world as it was, and how it is today. But have you noticed that nobody ever asks for evidence of the beneficial effects of teaching history? We just assume, quite reasonably, that knowing about history is important. For me learning about grammar is as important as learning about history. One reason is so obvious that it almost doesn’t need pointing out, and this is that grammar plays a central role in our lives, as part of the language(s) we speak. When we learn about grammar we learn something about ourselves because only human beings have language. We use language all the time, all day long, talking to friends and family, writing shopping lists, text messages and school essays. It makes sense for all of us, not just school children, to know something about how our language works, how the bits and pieces fit together. And of course, having a knowledge of grammar is also indispensable for learning foreign languages. If you teach Spanish and your students don’t know what an adjective is, it is harder to explain that they should order dos cervezas frías, not dos frías cervezas in a bar in Spain.

For a long time critics have argued that teaching grammar stifles children’s creativity. It’s better for them to engage with stories and poems and learn about language in that way. Grammar teaching is sneeringly called the ‘naming of the parts’, so that being able to label a word as an adjective is no different from knowing that the red car that just passed you in the street is a Ford Mondeo. Well-known writers such as Michael Rosen have made this point. E.L. Kennedy is on the record as saying that she has no idea what fronted adverbials are. It seems that for her the logical conclusion is that nobody else therefore needs to know what they are either. On Twitter people have said that they have an English degree from Oxford, and that they can’t answer the questions on the SPaG test. They seem to be saying that having an Oxford degree means that you know everything that needs to be known. As for the person who said that the primary tests would have stumped Jane Austen, well….

There is recent research that suggests that kids very much enjoy learning about grammar. When you point this out to critics and journalists it is steadfastly ignored.

Here are some quotes from Huw Bell (2016) citing teachers:

Screen Shot 2017-05-12 at 09.57.47

Screen Shot 2017-05-12 at 09.54.30

Screen Shot 2017-05-12 at 09.56.02

And in Kimberley Safford’s (2016) article much the same points are made:

Screen Shot 2017-05-12 at 10.01.54

Screen Shot 2017-05-12 at 10.01.27

It’s time some serious attention is given to this kind of research.

The last quotation above points to an issue that to my mind lies at the crux of the complaints we have been hearing about the tests: teachers feel insecure about teaching grammar. When the new National Curriculum was published people quickly noticed that apart from traditional terms such as ‘noun’, ‘adjective’ and ‘verb’, it also contained some new terms that are in line with the terminology used in modern linguistics, such as ‘conjunction’, ‘determiner’, and ‘adverbial’. From a linguist’s point of view this is very welcome, but it understandably caused alarm among teachers because they were suddenly faced with having to teach a subject about which they weren’t taught much themselves during teacher training, so they can be forgiven for feeling unhappy about it. It is a major failing of the government that it has not offered teachers any support on how to teach grammar.

Here we come to the second strand of the debate, namely how to teach grammar. Education specialists have pointed out that teaching grammatical terminology has no benefits at all. They say that it won’t help children to write better or read better. This may well be true, but much more research is needed to establish this. But even if turns out that grammar teaching isn’t directly beneficial, that doesn’t mean that we should ditch it. We teach history or literature, because these subject have an intrinsic value, and the same is true for grammar. When the critics talk about grammar teaching you find that there is often an unstated assumption, namely that when teachers teach grammar they must be teaching the ‘naming of the parts’ and no more than that. For older generations this was often what grammar teaching was like. But grammar pedagogy is much more than just teaching terminology, and has moved on from the way it was taught in the 1950s. Grammar terminology can now be taught using online resources which get children to learn about grammar in a playful and engaging way, using technology that is also used in games (see Englicious). Of course we do still need to teach some terminology, because it’s useful to have a metalanguage to talk about a subject, in the same way that it is useful to know what a metaphor is when you talk about literature, or what a monarchy is when you study history. There’s nothing to stop teachers from teaching grammar in a more contextualized way by using literary texts such as poems, plays and songs. As we have found at UCL in the CPD courses we offer, when teachers are offered the right kind of support to teach grammar they begin to enjoy it and see the benefits.

What about the tests? It is true that these are problematic in various ways. People’s views of the tests are coloured by the fact that they were brought in by Michael Gove, who was disliked by many teachers. It was Gove who insisted that the subjunctive be taught in schools, despite linguists advising him not to include it in the curriculum. Admittedly, the tests are somewhat unimaginative when children are asked to identify word classes or underline subordinate clauses, not least because it leads to teaching to the test, and the results of the tests affect how teachers and their schools are judged. So the tests need reform both in the way they are set, and in the way they are used.

Should we scrap the tests altogether, as some people have suggested? To my mind that would not be the right thing to do, because it would probably mean the end of grammar teaching in schools, and for the reasons given above that would be a shame.


1. As I have mentioned in different places on this blog, I should make clear that over the past few years I have advised the DfE on the SPaG tests by being a member of a Test Review Group. In this group a large number of experts, which includes teachers and literacy advisors, scrutinize the questions that will appear on the tests to ensure that they are clear and fair. The tests are looked at by other experts too, such as test specialists, statisticians, inclusion experts, and so on. There are several phases that draft tests go through over a two year period, before they are used in schools.

2. Huw Bell (2016) ‘Teacher knowledge and beliefs about grammar: a case study of an English primary school’, English in Education 50.2, 148-163.


3. Kimberley Safford (2016) ‘Teaching grammar and testing grammar in the English primary school: the impact on teachers and their teaching of the grammar element of the statutory test in Spelling, Punctuation and Grammar (SPaG)’. Changing English, 23.1, 3-21.

See also: How Bell and Steph Ainsworth (2017) ‘Grammar’s best kept secrets: what every primary teacher should know’, NATE Teaching English, Primary Matters.



4 thoughts on “Change the tests, but don’t ditch grammar

  1. Fascinating, Bas.

    All I would chip in is some research from a recent conference on literacy. The 2016 OECD literacy and numeracy tests show our 16-25 year olds are now comparatively the least literate in the developed world (23 of 23 nations) – PISA is similar in the downward trajectory it shows, although there’s so much argument about their methodology, it’s best to just look at the OECD research. In the same report, our pensioners are shown to be amongst the most literate compared internationally. Ostensibly, this suggests either we are less good now at teaching literacy than we have been in the past, or that the way we have been teaching literacy at some point became less effective, or what we meant by literacy changed.

    In addition, there is the fascinating research by the National Literacy Trust (2011), showing freakonomics-style connections between higher literacy and lifestyles: greater literacy rates = a greater likelihood of taking part in the democratic process, less likely to be on benefits, more likely to own a house/home, more likely to own a PC and use the internet – and that’s without the moral transmission of Literature and all it teaches us about each other.

    I think it’s easy to forget the absolute vitality of literacy, vocabulary and expression to a democratic society, and that it is the most influential factor in social mobility (not grammar schools!) whether here in a first-world country, or a third-world country where families often put their children learning to read and write before virtually anything else. Maybe these tests will at least address what is a very real problem identified both in large-scale research, or anecdotally, such as when my 17 or 18-year old bright students are unable to use capital letters properly, and do not know what a sentence is to avoid comma splices.

    Anecdotally again, I used Myhill’s Grammar for Writing research to create a mini SPaG unit for Year 11 Set 4s based on King’s Shawshank Redemption and cannot believe how much better they are now. Prior to this, when I asked them where commas should go they said ‘it’s where the reader takes a breath’. I asked when they were told that – and it was at Primary school, about 6 years previously. It wouldn’t happen now that we have overhauled SPaG teaching in KS3 in line with KS2 reforms, and upskilled the Dept, but it speaks to what we see in the wider research. The actual SPaG we agreed to sustain into KS3 is basic: the repeated teaching, every year, of things like what a sentence is, punctuation, homophones, etc – but it’s the KS2 tests that led us to this.

    Whether these tests will eventually fix everything, awkward as they are, I don’t know – but it did need fixing, and like water over a stone, I hope the demands on the kids will smooth out over time. It’s a shame we have become so utilitarian that we won’t teach what isn’t on a test, but how you solve that much bigger existential issue, I also don’t know!


  2. A ramble around your “tests” post

    Hi, Bas. Some of the anti-grammar arguments you refer to are plain silly. You went to Oxford yet you can’t answer certain questions? So what? You wouldn’t sneer if a child came home and talked of having learnt about quarks or nanomaterials or NMR scans: you probably didn’t learn about them at school or university either, nor contemporaries who studied physics or chemistry or neurobiology. Those subjects have moved on. So has linguistics. You may have learnt about grammar, probably at school – though at Oxford not so long ago it might well have been taught either not at all or much as the Victorians did.

    I usually admire the writings of Michael Rosen and A. L. Kennedy on other topics. I bet they can identify nouns and verbs without difficulty, though I wonder whether they have that cognitive dissonance so often brought on by traditional teaching: you think you know that a noun is the name of a person, place or thing, and that a verb must be a doing word, yet you *also* know perfectly well that the word _destruction_ is a noun, despite referring to an action and not being a name. Wouldn’t it be better to aim for a system which is coherent, self-consistent, and descriptively useful? And – crucially – defensible to an inquiring young mind? One that could be used equally well for a foreign language or your own? And one which – like any scientific system – is open to testing and refinement rather than unimpressive assertions like “That’s just how I learned about it x years ago, from teachers who were themselves taught x + y years ago, largely from books written x + y + z years ago”? – that is, authors and teachers who knew what they liked and imposed their prejudices on everyone else, or used it to judge if you were “one of us”. Their prejudices were and are of variable quality, by no means all useless, but many neither logical, useful or even historically well-informed.

    I feel strongly that it’s valuable to learn about language (not just, I hope, because of my professional interest). As you say, Bas, to discuss how language actually works you need a technical vocabulary. I’m doubtful whether young primary-age children need the more arcane terms – though ‘arcane’ is certainly not to be defined solely by “Well, in my day we never learnt that term and I’ve managed perfectly well without it”.

    But the whole testing thing raises real problems, many of which can’t be separated from policies to create many different kinds of school, allegedly to give parents more choice, incidentally destroying sensible local planning, and part of a cut-price attitude to most public services and public servants which imposes ever more demands while withholding adequate funds (and denigrating the practitioners). Oops, my prejudices are seeping out … But in present political conditions I’m not sure what to advocate. Which of these assertions is valid, which arguable?

    1. If something in the school curriculum is tested, it will play a part in school league tables, otherwise not.
    2. For practical reasons, you can only test things with simple right/wrong answers, especially when results have financial consequences.
    3. Much of language isn’t simply right/wrong, black/white: it’s too subtle, complex and interesting for that.
    4. If something doesn’t contribute to league tables, schools can’t afford to spend much class and preparation time on it.
    5. If teachers haven’t themselves been properly introduced to a subject, they can’t teach it with pleasure, enthusiasm and authority.
    6. Simply identifying word classes and grammatical functions is almost pointless unless harnessed to understanding of how varieties of English differ, how one’s own use of language may be made more effective in different contexts, how language can be manipulated for good and for ill – in literature and in life, and so on. But when so linked, it’s what any educated person should know.
    7. Students should learn how to deal with the complex facts of daily life, in biology, relationships, economics, etc. They must learn to discount nonsense.
    8. Students need a cultural life. Art, history, literature, music, etc. should be part of everyone’s educational diet.
    9. Language is both essential to culture *and* central to daily, practical life.
    10. Traditional grammar and prescriptive rules are full of old wives’ tales.
    11. Most students who really learn something about the structure, variety and history of language will enjoy it and derive benefit.
    12. Language study is important at both primary and secondary level, but maybe some of the grammatical detail could move from primary to secondary.

    What do you think?


    • Thanks, David.

      I think we are largely in agreement that the content of the curriculum and the tests need reform. As for your 12 points, I can’t answer them here, but there’s plenty of material there for future blog posts, so watch this space!


  3. Hi Bas.

    Thanks for another very interesting post. I agree that it would be an almighty shame if changing the tests somehow led to grammar being dropped from the curriculum too.
    I think that the negative comments you mention (usually along the lines of ‘well, I’m a writer and I don’t know what an adverbial is’, or ‘well, I’m 61 and I’ve never needed to know this stuff’) actually indicate two things.
    The first is that the reactive moves against grammar since the middle of the 20th century have been pretty successful – it’s now considered almost disagreeable to know about it, as if formal knowledge of grammar is entirely useless and can only lead to petty prescriptive squabbles or arguments about apostrophes. (And to be fair, there does seem to be a judgmental undercurrent to lots of discussion about language in the press and on social media.)
    The second main conclusion I’d draw is that linguists have not done well so far in showing what benefits grammar knowledge can have. Instead the false idea that knowing grammar terms will per se somehow improve writing (it won’t; it’s not even plausible) has been promoted from inside the National Curriculum, and because it is unsupported by evidence the counter-argument is strong. I think Dick Hudson has written extensively about the other possible benefits of wider language awareness.
    But providing evidence that grammar knowledge is useful is possibly besides the point. It’s clear that many children have not yet learnt to view grammar as hard or suspect, and do enjoy learning about it. Here is a brief extract from a discussion between two teachers that didn’t make it into my last paper; I had asked if the children found grammar interesting:

    T1 Not stood on its own, no. If they’re doing something with it in literacy I think they find it interesting.
    T2 Mine enjoy it like when they know they’re going to learn something new. […]
    T1 I think it’s they’re getting taught something new that they get to do.
    T1 Yeah, they love clauses [laughs] because it’s tricky. Well, I taught it to the highers and they taught it to the rest because they found it fascinating.
    Me The kids actually…?!
    T1 It spreads round the class. Like a rumour pretty much…



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s