You like to think that you have avoided the trap - that, despite all the pressure, you have gone your own way.

You know that the grammar, punctuation and spelling (GPS) tests introduced by former education secretary Michael Gove in 2013 strip grammar back to a set of “right” and “wrong” answers and an excessive list of clause-level terminology. You know that there is an undue emphasis on standard English. But you hope you have managed to contain all those issues to the test, and that your classroom is free of them.

And yet: what has the pedagogical impact of the tests really been? My research suggests that rather than improving literacy for all, the tests are having the opposite effect: creating artificial bits of writing, warping teachers’ and pupils’ understanding of what language really is, and intimidating teachers into pedagogies that they do not value or believe in. England may have gone up the ranks for reading in the Organisation for Economic Cooperation and Development’s Pisa (Programme for International Student Assessment) results, released this week, but that does not even begin to tell the whole story.

Teachers, linguists and commentators expected problems right from the start. The decision to introduce the tests and to focus so heavily on drill and identification was triggered by a discourse of “falling standards” and children “not being able to speak properly”. It was Lord Bew’s independent review of key stage 2 in 2011 that explicitly recommended the introduction of a test at the end of primary school on the grounds that, in grammar, “there are clear ‘right’ and ‘wrong’ answers”.

The fightback has been fierce. Criticisms have centred on the way that the tests reduced grammar to “feature spotting” - a pedagogy reminiscent of 18th- and 19th-century classrooms, whereby pupils identify grammatical constructions in artificially constructed sentences and correct “errors” in non-standard grammar. In short: no creativity, no opportunities to explore linguistic meaning, and no recognition of the value of linguistic variation. A question from the tests illustrates this well:

Circle the adverb in this sentence: “Of all the toys in his large collection, Karl’s little brother liked the cuddly rabbit best.”

Out of all of the test questions over the years, this “identify” type has been by far the most prominent. In the 462 test questions children have sat over seven years of papers, 269 have been in the “identify” style of the above example.

Meanwhile, 134 have been “complete/correct/rewrite” questions, in which pupils are asked to “correct” examples of “incorrect” language (which is typically the correction of a non-standard form). For example, in this question, the non-standard “was/were” variation is framed as being “wrong”:

Circle the correct verb form in the sentence below: “The last place I saw Jack and Gwen was/were in the playground.”

There are multiple issues with both questions, even if their impact were confined to just the tests. For example, questions such as the latter force children to label what might be legitimate features of their own dialects as “wrong”, even though the majority of UK school children speak a non-standard variety of English (Cheshire, 1982). This “binary” view (ie, standard versus non-standard; right versus wrong) reproduces hierarchies in language, which only serves to further marginalise speakers of non-standard English, in what is called linguicism: discrimination based on the way that people use language.

But how far do the issues presented in the tests permeate through the way grammar is taught? Do the tests have the potential to coerce teachers into reproducing those issues in their classroom practice?

In the past couple of years, I have surveyed and interviewed hundreds of teachers, scrutinised the test questions and guidance for test creators, revisited political discourse and visited classrooms in order to understand the “power” of the tests and their ability to shape, influence and manipulate pedagogies.

This is what I found.

‘So fake’

My research suggests that teachers’ grammar pedagogies are often controlled and regulated by the nature of the test questions and the way that language is conceptualised within the tests. It also found that the tests have narrowed the curriculum, meaning that teachers spend significant amounts of time teaching decontextualised grammar in preparation for them.

Two examples taken from the data illustrate this nicely (all names are pseudonyms):

• “My teaching of grammar is now all purely focused on the test and making sure that the students know the terminology and how to identify it. It’s not teaching as I want to or know it - it’s teaching to the test, pure and simple. They make me do things I don’t want to do.” (James)
• “The format of the test is about not understanding; it is a feature-spotting activity. I think the grammar teaching I end up doing is really quite artificial because it’s all about underlining features and using overly technical terms just for the sake of doing so. It turns into a box-ticking exercise. I wonder: is it doing anything useful? I don’t think so. It’s not teaching anything about language apart from the names of little tiny bits.” (Julie)

In a similar way, my research suggests the emphasis on standard English and the framing of language as “correct/incorrect” seems to be crafting cultures where language “policing” in the classroom takes place. Language policing, or the surveillance and regulation of young people’s language, was a common theme in the data. I saw and heard about examples of “word jails”, “ban the slang” posters on classroom walls, “grammar police” badges given to pupils, and pamphlets stuck on to desks that listed non-standard forms that were “outlawed” from classrooms.

Quite often, teachers justified these policies in reference to the emphasis on standard English in the curriculum, the tests and the Teachers’ Standards. For example:

• “At my school there is an expectation that the correct use of English is modelled when speaking to children. This is new for us since the tests were introduced. This can be hard for people who speak with particular dialects/regional accents.” (Ruth)
• “The way that the tests describe language is so fake. I know that language doesn’t work like that but I’m kind of steered into saying that it does in my classroom because of the language of the tests.” (Carl)
• “I think the curriculum changes and the tests and all of the emphasis now on standard English has really made me change the way that I react to dialects and things in my classroom. You know, now I feel like I have to kind of tell students off about the way they use dialect or whatever. And I feel really not comfortable in doing that. I don’t want to be that person who does that because of course I value their dialect but it’s difficult to handle that when you’ve got those tests looming over you.” (Alice)

Let’s be clear: teachers are not choosing to teach in these ways because they think these pedagogies have any meaningful value. Teachers don’t go into the profession to teach bits of grammar by rote. They don’t go into the profession to teach to the test. And they certainly don’t go into the profession to tell children how they should or shouldn’t be speaking, to pick young people up on their use of “incorrect” language, and to devalue linguistic variation.

Instead, teachers operate within a system whereby their pedagogical autonomy and choices can be influenced by systems and structures from “above”, be they management, parents, Ofsted, tests and so on. Teachers are under pressure to deliver. Their reputation, their self-identity, their pay and their careers are on the line.

And so it is little surprise that my research has shown that the way language is being framed in the curriculum and the tests is coercing, socialising and intimidating teachers into teaching decontextualised grammar, and often engaging in language policing. And this happens despite teachers knowing that this kind of grammar teaching or way of thinking about language isn’t particularly valuable in terms of developing critical reading and creative writing abilities.

Of course, there are lots of teachers doing brilliant work, resisting the decontextualised, clause-level, right-wrong view of language that the tests promote. I heard about and saw grammar teaching that was integrated seamlessly into lessons, with teachers and pupils using grammatical knowledge and terminology as a way of exploring texts in investigative, meaningful ways. Not all schools are engaging in “ban the slang” policies, and regional linguistic variation is generally celebrated and respected by teachers, even if not by the tests and by government.

But even in those instances, students still have to take the tests, and still have to engage in the kind of reductive grammatical work that the tests demand of them, even if teachers want to resist these kinds of pedagogical impositions. For instance:

• “I do want to teach grammar in context and get away from the rote learning of the test. But my school insists on discrete grammar lessons where we have to teach them what is ‘right’ and ‘wrong’ and so I end up doing that.” (Sarah)
• “What the tests have led to, which I don’t necessarily agree with, is a lot of standalone grammar teaching where children complete worksheet exercises or tick-box things or fill in the gaps …What I then sometimes don’t see is that knowledge crossing over into the writing, which is actually where the grammar is the most important.” (Billy)

In short, the tests are changing the way our young pupils are being taught grammar - and the implications of those methods are deeply worrying.

The government argues that, as a result of the tests, writing has improved (Gibb, 2018). If that were true, then perhaps some would argue that the methods would be justified (though I would not see it that way).

However, my research and that of others suggests strongly that decontextualised grammar teaching has no positive impact on writing. Moreover, teachers in my data talked extensively about the damaging impact on writing that the tests and the way that grammar is framed in the curriculum have had. For example, I saw and heard about lots of writing activities that were framed like this:

Write a descriptive sentence that includes one verb, two adjectives and at least three nouns.

I don’t think any writer would recognise this as a particularly helpful model for generating good writing. The message about language here is clear: no need to worry if it flows or if it successfully conveys the passion, concern or anger you might wish to portray. Just make sure it is packed to the rafters with a predetermined number of different grammatical elements.

It’s akin to a music teacher urging their pupils not to consider the feel or timbre of a new composition, but to simply make sure they have used 20 crotchets, and a minimum of 30 semiquavers. It is the writing equivalent of painting by numbers. Take this sentence, for example: “Very slowly, the terribly ugly monster, who was hungry for human flesh, came crawling out of the dark, cavernous pit.”

Here, we have a sentence that ticks all the grammatical boxes: it has a fronted adverbial (very slowly); an expanded noun phrase (the terribly ugly monster); a relative clause (who was hungry for human flesh); and another adverbial (out of the dark, cavernous pit). But is this good writing?

That’s subjective, of course, but I would argue in this case that it is not. It’s clunky, unnecessarily detailed, and peppered with “features” for the sake of it, rather than being driven by a sense of readership or authorial craft. But, the way that grammar is conceived and tested within current policy would steer a marker towards valuing this because of its inclusion of grammatical features rather than its meaning or creativity (for a developed criticism, see Barrs, 2019).

Data from secondary schools supports the fact that writing is not improving and also suggests that the way we are now teaching grammar has a broader negative impact.

Secondary school teachers I spoke to often talked about how their Year 7 students excelled at identifying grammatical features of language, but typically had very little experience of applying and using this knowledge in any meaningful way. Students are desperate to hunt for and identify, for example, preposition phrases in texts, but have no idea of how or why preposition phrases might be used by writers, or their potential significance as a resource for making meaning.

The language ideologies and pedagogies found within the primary school grammar test then, can “bleach” into secondary school, with teachers having to then spend time emphasising that interpreting texts is not about feature spotting and listing grammatical terms.

Ways and meanings

So, what should we be doing? Research (Myhill et al, 2012) has consistently demonstrated the value of contextualised grammar teaching, characterised by exploring grammar in relation to authentic texts and genres, looking at how different linguistic choices can create different meanings and allowing students’ responses and interpretations to lead the pedagogy rather than grammatical terminology. Good grammar teaching relies on an “integrated” model of language and literature, where grammar is conceived of as a system of social meaning, rather than a list of constraining rules.

Grammar under this pedagogical model becomes much less about “rules” and “correctness”, and much more about choices, meaning, genre and patterns. Grammar here is not just about the clause - it is about discourse.

I wouldn’t deny that all students need access to standard English, purely because of the ideological power that it holds in wider society. But banning non-standard forms from the classroom and punishing students for using their own language is not the way to teach students about standard English. Instead, classrooms should be places where language is seen as a fluid, dynamic resource, where students draw on their own linguistic repertoires as one way of expressing their identity (Drummond and Schleef, 2016; Snell, 2018). All users of a language are able to style-shift and code-switch, ie, adapt their language according to the context in which it is used - conscious choices that index and carry social meaning (Moore, 2012). But the binary, reductive view of grammar on the tests fails to acknowledge this linguistic reality, and so automatically disadvantages young people from the outset.

More “meaningful” language work in schools, then - and there are obviously schools that are already doing such things - might involve studying regional variation, looking at how language establishes and maintains social power, exploring how language intersects with other social factors such as race, class and gender, and so on.

But for all this to work, this practice would need to be reflected in the assessments, too. The test would have to go. And I believe very few teachers would mourn its demise.

Ian Cushing is a lecturer in education at Brunel University, London

This article originally appeared in the 6 December 2019 issue under the headline “Grammar rules, OK?”