Right and wrong answers in grammar tests

The schools minister Nick Gibb, who is responsible for the Grammar, Punctuation and Spelling (GPS) tests that UK school children sit at ages 7 and 11, was asked a grammar question in a BBC Radio 4 news programme. The question had previously appeared on one of the GPS test papers. Here’s a snippet of the interview:

Interviewer Let me give you this sentence: I went to the cinema after I’d eaten my dinner. Is the word after there being used as a subordinating conjunction or as a preposition?

Minister Well, it’s a proposition. After – it’s…

Interviewer I don’t think it is.

Minister After is a preposition, it can be used in some contexts as a, as a, word that coordinates a subclause, but this isn’t about me.

Interviewer No, I think in this sentence it’s being used a subordinating conjunction!

The press had a wonderful time pouring scorn on the minister for not being able to answer a question that very young children need to be able to answer by the time they leave primary school.

There has subsequently been some debate over the question of whether the minister may actually have answered the question correctly. Oliver Kamm wrote a piece in the Times with this heading:

Screen Shot 2016-05-16 at 20.30.17

Kamm writes that Gibb was correct in saying that after is a preposition, and there “was no need for the minister to be defensive of his grasp of grammar.”

So was Gibb’s answer correct or not?

Well, that depends on the grammar you consult. The two most influential standard grammars of modern English, namely A comprehensive grammar of the English language (CGEL; by Quirk, Greenbaum, Leech, and Svartvik) and the Cambridge grammar of the English language (CamGEL; by Huddleston and Pullum et al.) have different analyses of the sentence that was presented to the minister.

In CGEL a distinction is made between the preposition after and the subordinating conjunction after, such that we have the two analyses below:

  • I went to the cinema [PP [P after] [NP dinner]]
  • I went to the cinema [clause [subordinating conjunction after] I’d eaten my dinner]

[In this notation the brackets indicate the grammatical units, called constituents. P = preposition; PP = prepositional phrase; NP = noun phrase.]

So in the CGEL framework after, as a preposition, takes a noun phrase as its complement, but as a subordinating conjunction it introduces a subordinate clause.

By contrast, in CamGEL (which was influenced by the Danish grammarian Otto Jespersen, as Kamm points out) we have this analysis:

  • I went to the cinema [PP [P after] dinner]
  • I went to the cinema [PP [P after] [clause I’d eaten my dinner]]

In the CamGEL framework after is always a preposition which takes either a noun phrase or a clause as its complement. Both after dinner and after I’d eaten my dinner are prepositional phrases. In my Oxford modern English grammar I adopt the CamGEL analysis.

It would be fair to say that the CGEL analysis is the dominant one and in this grammar, along with after, we also find words like although, asbecause, before, if, since, (even) thoughwhile, when, while, etc. within the word class of subordinating conjunctions. Because it is more widely adopted, and because the analysis is conceptually easier to grasp, the CGEL approach was adopted in the National Curriculum (NC) after a process of consultation which my former UCL colleague Dick Hudson conducted to find out what fellow linguists thought about the best way to present particular grammatical analyses for use in schools.

Now, because the National Curriculum adopts the CGEL analysis, we can say that Gibb was in fact wrong in saying that after is a preposition in the sentence I went to the cinema after I’d eaten my dinner.

Some commentators have been saying that we should not be teaching children about subordinating conjunctions because grammarians themselves can’t agree about how they should be defined. This is misguided.

The study of grammar is no different from other fields of enquiry in being replete with disgreements about how we should look at particular phenomena in the world around us. Think of the Expanding Earth Hypothesis in science, or disagreements among art historians about the significance of Renaissance art, or the different ways in which we can interpret the character of Iago in Shakespeare’s Othello. Such disagreements make these fields fascinating, and discussions of them propel our knowledge of the world forward.

But what does all this entail for the teaching of grammatical concepts such as ‘subordinating conjunction’ in primary schools?

One approach is to say that such notions are too hard for 11-year olds. This is what most commentators in the press and social media seem to have been claiming. Some of these comments were along these lines: “I have an English degree from Oxford and do not know what a subordinating conjunction is.” The conclusion that is then drawn is that nobody else needs to know either!

The alternative is that we do not shy away from teaching this notion, since it really is not all that difficult to grasp, at least no more difficult than learning about, say, the division rules at KS2. However, if we do decide to teach grammatical notions it’s best to use a particular model of grammar to teach the basic concepts, and to refine the concepts that are learnt in primary school at a later stage, e.g. at secondary or university level. Pedagogically it makes sense to use models, even if they are not agreed on by all specialists in a particular field. In science something similar happens, for example when the acceleration of free-falling objects is taught: we then pretend that there is no air resistance. In the classroom making use of idealisations and models is a very useful and effective strategy.

PS If you want to find out more about Otto Jespersen, have a look here.

PPS The basic concepts of grammar in the NC are all explained on the free Englicious site.

11 thoughts on “Right and wrong answers in grammar tests

  1. Thanks for a(another) really clear and helpful discussion. It’s a difficult topic to discuss right now because there are so many different sets of starting assumptions on this. I think the fun of studying grammar is all or mainly about trying to find evidence to argue for and against particular analyses etc.


  2. Hi Bas,

    I think you’ve got a bit of false equivalence in your arguments here. Just because there is disagreement amongst experts on a subject is indeed no good reason not to teach it. However, when taught, you’d want teachers to focus on the fact that there is disagreement in scientific subjects (and further hypotheses that need testing and evidence that needs collecting), or indeed room for individual interpretation in the arts. Generally, to encourage children to think critically about the subject. This isn’t how grammar is being taught in primary school, or how it is tested either. Read the grammar test again, you know it asks for the ‘correct’ answer. It is saying there is only one right way to answer the questions. It doesn’t engage with learning about different systems of grammar using critical thinking skills.

    And most of the examples you gave here, the significance of renaissance art, differing interpretations of Iago etc, are taught at secondary level in the UK (If they are taught at all).

    I also personally don’t disagree with the Spag test on the basis of it being too hard for pupils at KS2. It is something that can be taught and many pupils with good memories will be able to pass it. It’s just I question whether it is useful for them to spend so much time learning this particular interpretation of grammar, and in this way for the test. The childrens’ time could be so much better spent on other things and grammar could instead be taught in context, when needed, as part of a wider literacy programme.

    Why do you think this is a good way to teach and to test grammar? What evidence do you have to back it up?

    What’s your personal experience of teaching KS1 and KS2 pupils grammar?

    Do you have a background in educational research on the topic?

    What do your colleagues at the IoE, recently rated the no1 education department in the world, (which has recently merged with UCL) think about your work?

    Liked by 1 person

    • Hi Lucy,

      You write:

      “However, when taught, you’d want teachers to focus on the fact that there is disagreement in scientific subjects (and further hypotheses that need testing and evidence that needs collecting), or indeed room for individual interpretation in the arts. Generally, to encourage children to think critically about the subject. This isn’t how grammar is being taught in primary school, or how it is tested either. Read the grammar test again, you know it asks for the ‘correct’ answer. It is saying there is only one right way to answer the questions. It doesn’t engage with learning about different systems of grammar using critical thinking skills.”

      Somebody else asked me a similar question on a different forum and this was my reply (slightly adapted):

      At KS1 and KS2 grammatical terms first need to be learned. At that stage you present kids with a simple model of grammar, from which you can diverge at a later stage. In the same way that you need to tell kids about the difference between similes and metaphors before they can recognise them, you need to explain what nouns, verbs, adjectives, etc. are, before you can talk more meaningfully about them.

      However, even at KS1 and KS2 you can have good discussions with kids about e.g. which word class a particular word belongs to. In Englicious we have created activities where children are encouraged to think about word classes. Have a a look at the activity ‘Noun identification’ (which can be projected onto an interactive whiteboard by clicking on the little screen at the top right). By moving the words across the screen kids are encouraged to think about words like ‘show’ and ‘run’, which can of course be verbs or nouns, depending on the context. They really enjoy calling out sentences in which ‘show’ and ‘run’ are verbs and nouns, and learn about word classes in a playful way.


      As for the test having only right or wrong answers: yes, this is true and there are issues here that need to be addressed. I don’t think the current format is ideal.

      You then ask “whether it is useful for them to spend so much time learning this particular interpretation of grammar, and in this way for the test. The childrens’ time could be so much better spent on other things and grammar could instead be taught in context, when needed, as part of a wider literacy programme.”

      You can ask the same question about other subjects taught in school. Can you do without grammar? Sure, in the same way that you can go through life without studying history or literature. To my mind learning about the structure of the language you use every day enriches you. The different ways of teaching grammar are not mutually exclusive. They each have their own rewards.


  3. But Bas, I’m not arguing we should go without grammar being taught, only that it should be taught and tested in a different way. Also, I think there are pretty good arguments for literature and history being taught, enjoyment, beauty, utility in later life, understanding our past (and it’s influence on the present), development of critical thinking, evaluation of sources, to name but a few, though not all. I think you agree with me here.

    I don’t think all those arguments apply to grammar when its being taught and tested in this way though. Who, out of choice, would sit down and do a grammar exercise, of the type in the test, in the same way someone would sit down to enjoy reading a book? I believe it would be far fewer people, even if there are exceptions – perhaps academics in linguistics and your social circle are over represented here?

    What is your argument for teaching grammar in this way? Why is it useful? What’s your evidence that this method if teaching works? What’s your evidence that this way of teaching & testing grammar (and it is to the exclusion of other ways of teaching it, there’s only so much time in the day) enriches pupils?

    As far as im aware in no other subject do we teach only one interpretation of something controversial. In RE we don’t only teach one view of creation as the correct answer. You may simplify as a teacher but you don’t exclude in the way that’s happening with grammar. The government doesnt label something as correct where it’s only one of many interpretations in other subjects. I may be wrong, but can you think of any? In science I’m always careful to talk about scientific method and about how theories evolve over time depending on the evidence available. And science and RE are definitely not tested in the same way that grammar is at KS1 & 2.

    I’ve looked at your resources before, and whilst I think some of them are ok, and some quite good really, none of them are as good as teaching grammar in the context of children reading and writing themselves for a purpose. This is backed up by research! I read it on my PGCE, but don’t have access to it at the moment, so can’t link to it.

    I notice you didn’t reply to my other questions before, perhaps because the answers are:

    a) I like grammar and therefore I think its enriching and should be taught. But I can’t explain it beyond this.

    b) I have no reliable evidence.

    c) I don’t have much experience of teaching children of this age group before I started advising the government but I like to tell other people how to do it anyway.

    d) I don’t have a background in educational research, my expertise is solely in linguistics research and in teaching at an undergraduate and postgraduate level.

    e) They have a wide range of opinions but in the main they don’t like me very much because I don’t recognise or respect their expertise.

    Liked by 1 person

    • Hi Lucy,

      As for your questions, despite their rather off-putting tone, here are some answers:

      a) See my earlier post about this. And also David Didau’s blog: http://bit.ly/1WtOmMM

      b) Yes, more research is needed into grammar teaching and how it’s best done. Here are some links to recent publications:

      Kimberly Safford et al. (2015) ‘Teaching grammar and testing grammar in the English primary school: The impact on teachers and teaching of the grammar element of the statutory test in Spelling, Punctuation and Grammar (SPaG) in England’. Changing English 23, 3–21. (http://oro.open.ac.uk/44366/)

      Huw Bell (2016) ‘Teacher knowledge and beliefs about grammar: a case study of an English primary school.’ English in Education, 1–16. (http://bit.ly/1TSEie2)

      See also Dick Hudson’s page ‘Research supporting the teaching of L1 English grammar’ (http://teach-grammar.com/research)

      c) I have a teaching qualification, but I don’t have any recent experience in schools. You are implying that this disqualifies me from taking part in the debate. I disagree. And no, I don’t like to tell other people how to teach, but my team and I do think that Englicious offers great resources for teachers. They can pick what they like from the site, and it’s free. I’m pleased to see that you liked some of the stuff you looked at.

      d) True, but, as above, this doesn’t mean I can’t take part in a debate about grammar teaching.

      e) I work closely with colleagues in the Institute of Education, especially in the Department of Learning and Leadership and in the International Literacy Centre. Do they like me and what I’m doing? I think so, but why don’t you ask them yourself? I have no idea why you would want to suggest that I don’t recognise or respect their expertise.

      By the way, it would be a courtesy if you revealed your full identity to me and to the readers of this blog.


  4. Hi Bas,

    Sorry about the tone. I was annoyed as I felt I had asked reasonable questions and that you had ignored them, perhaps I went a bit far.

    My name is Lucy. I’ve got a degree in politics, have studied education at a postgraduate level and have recent experience teaching in the EYFS and KS1. I’m a nobody, with no particular axe to grind, just strong opinions about education policy. I’m taking some time out of the classroom at the moment for health reasons and just doing some volunteering at a school, teaching reading and generally helping out, as and when my health allows.

    I certainly don’t think you need to have qualifications to be part of the debate – I’m pretty sure I’d be excluded from many debates if you did! I’m all for freedom of expression! I don’t object to anyone talking or writing about anything and enjoy discussing things with all sorts of people.

    But I do think that your lack of relevant expertise is relevant to your job as an independent advisor to the Department of Education and the quality of the job that you’re doing there. This is something that impacts upon children across the country and that’s why I’m bringing it up.

    I personally believe that the best way of setting long lasting, good quality education policy, including the national curriculum and national tests, is for the Secretary of State for Education to take advice from a representative body filled with relevant stakeholders, including but not limited to, researchers in education, teachers, parents, children, headteachers, and teaching unions.

    I think it does show a lack of respect for others when you stray outside your area of expertise. If I were offered a job teaching undergraduate level linguistics I’d turn it down as I know I wouldn’t be able to do a good enough job. If I was to accept the job I think that would show a lack of respect for people who have the relevant expertise and qualifications and would do the job far better than me.

    I didn’t mean to imply anything about your behaviour or personality. You seem like a very pleasant and polite man, far better mannered than me! I’m sure your colleagues at the IoE are equally polite.

    It’s just that no one I’ve ever met at the IoE defends government policy on the spag curriculum & tests like this. They all believe in using evidence based teaching methods wherever possible.

    I see you think that more research is needed. I agree. But don’t you think all this is being done the wrong way around? Shouldn’t you have the evidence in place first before you bring in a policy, rather than do research to look for evidence to back up your point of view after its already been introduced? It’s all backwards. This isn’t how good public policy that stands the test of time is made.

    I’ll read the papers you linked to, they look very interesting thanks.

    I have read your other blog post about this before, but unless I’m missing something it seemed like a tautology on the enriching theme. Am I reading it wrong? I’m afraid I just don’t understand how or why the spag curriculum and test, as it is at present, is enriching. Can you explain it further? I’d be interested to read another blog post on the topic if you do requests.

    I think grammar education can be useful and empowering, particularly for children learning english as an additional language and children who don’t speak with received pronunciation and “standard” english grammar at home. As when grammar is taught in an embedded and contextualised way, it gives them the tools to access another register and all the opportunities that can bring.

    I didn’t find David’s blog convincing I’m afraid. I did read that a while ago and can’t remember the details now, so maybe I need to revisit it in light of your post.

    Thanks for replying, sorry again about the tone. I hope I’ve explained myself better here.


  5. I’ve just read the findings of the first two articles you point to. Neither of them shows that the way of teaching and testing grammar you espouse is the best or even a good way of teaching grammar.

    I can only conclude that you have no evidence to point to that shows that teaching and testing grammar in this decontextualised way improves the way children use grammar.

    The third link goes to a page that lists a whole load of papers but not all have links to the actual papers themselves so I can’t check if they say what Bas says they do. I haven’t had time to read all of them yet – I’m trying to find ways to read them for free.

    One of articles on that page is by Debra Myhill which is the one I was referring to earlier as reading on my PGCE about teaching embedded grammar. This is how I think we should be teaching grammar! If you google it you can find good summaries of her research. It does not support the Spag curriculum or test Bas is arguing in favour of.

    Debra Myhill has gone on the record as saying she thinks the tests are flawed and fly in the face of 50 years of research on the subject. She is a well respected researcher and this is her area of expertise. This is someone the DfE should be listening to instead of you Bas.

    Read about it for yourself here: http://www.telegraph.co.uk/education/primaryeducation/9975026/Government-adviser-criticises-flawed-primary-school-literacy-test.html

    Maybe you should read the articles you linked to again Bas? They don’t seem to support your claims.

    To anyone else reading, take a look for yourself if you have the time.


  6. Hi,
    I totally disagree with you and Pullum. Do you not think it’s perverse to call an adverb or a subjugating conjunction a preposition when every dictionary on this planet including the OED says you’re wrong?

    My response is too long for a comment and I’ve written it here in my blog: http://vivitelaeti.blogspot.sg/2016/11/puzzling-prepositions-why-grammarians.html

    I hope you will pardon the occasional facetiousness and irreverence in my blog but that’s the intended mood of my website.




  7. I think it’s surprising that anyone believes Nick Gibb could’ve been invoking sophisticated linguistics to defend his understanding of ‘after’!

    Although it’s *possible* to view it as a preposition in the original example, Occam’s Razor indicates otherwise. If it *is* a preposition with a sentence occupying the normal noun phrase position, then why does the original tense get changed (‘I’ve eaten’ > ‘I’d eaten’)? Moreover, you don’t have to look very far to find languages in which it’s impossible to interpret the equivalent of this ‘after’ as a preposition – German is an obvious example. In German the preposition generally has a related but different form from the semantically related subordinating conjunction (vor/bevor (‘before’); seit/seitdem (‘since’); nach/nachdem (‘after’)). More importantly, the preposition proper affects grammatical case, whereas the conjunction doesn’t; and the conjunction forces any verb to the phrase end, whilst the preposition has no effect on word order. Similar rules apply in Dutch.

    And why would Russian bother with the equivalent of “I went to the cinema after that-thing when I ate my dinner” – where the prepositional ‘after’ IS used but requires the insertion of an appropriately inflected pronoun before a conjunction and the rest of the sentence?

    I could give you many more examples, for example from Celtic languages, but the story’s the same. ‘after’ is both a preposition and a conjunction, not just the former.


Leave a Reply

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

WordPress.com Logo

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

Google photo

You are commenting using your Google 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 )

Connecting to %s