What do we mean by "meaning". And does it matter?

 In language teaching, we seem to be struggling with two different meanings of the word "meaning".

On the one hand we have "I know that tortue means tortoise". Where pupils are tested on their knowledge of the meaning of words. It's an approach that believes in regularly testing pupils' ability to parse sentences containing known words and known grammar, to cement memorisation and conceptualisation. The language is selected (by frequency of vocabulary) and sequenced (to exemplify step-by-step grammar concepts), so that the pupil's knowledge is built and reinforced.

On the other hand, we have the idea that language should be for pupils to express themselves and understand each other, to create and take part in communicating "meaning". This reminds me of the Spanish and French word for "to mean" - querer decir / vouloir dire - to want to say something.

Does it matter in language learning that our pupils learn to use their language to say things they want to say? Does it matter that what they read or hear has something to say, rather than just to practise and test their knowledge of language features?

We see this in GCSE and A Level listening and reading papers. What masquerades as a comprehension question often turns out to be asking pupils to show they can parse certain language features, even when they are not relevant to the purported question. At GCSE, we have seen questions like these. Things like "He didn't get on with his teachers" (which accurately characterises the understanding of the relationship) being marked as wrong. Because it didn't accurately parse the word "badly". Or at A Level, this question about what someone did one day. Answering the question (she went to see the castellers, she took a photo, she posted it online) is not rewarded. Because it doesn't show knowledge of the grammatical features it was appetising to her, she decided to... And these are not random rogue questions. This is a feature of how the examiners see meaning as simple demonstration of knowledge of the "meaning" of words and grammar. Not the meaning of what someone said.

It is also in Ofsted's guidance on curriculum design for languages. They insist that language be introduced in a strict sequence, based on exemplifying concepts, not based on teaching pupils to say things. The example they give is to teach pupils to talk about red dogs and red tortoises, and to avoid teaching green dogs and green tortoises until a future step. Because the adjective rouge is invariable for gender. Whereas vert would require knowledge of adjectival agreement if applied to a tortoise. All of which ignores the fact that if you are going to teach pets, pupils will want to talk about a range of pets in authentic colours. At this stage, pupils are making links from the language to the real world, rather than links and patterns internal to the language. This kind of real meaning is important to learners. And you wouldn't want to lose that!

But is it important to learning? Maybe that kind of meaning, with lots of pupils all trying to say random different things they aren't ready for yet, leads to them being given a collection of one-off things to say, that don't stay in long term memory and don't add up to coherent conceptualisation of the grammar of the language.

Here's an example that happened with my Spanish class last year. So they were the last class to take the old style GCSE. Back in Year 10, we had done some work on Shopping. Improvising answers in speaking, then writing them up. Most pupils did something that rehearsed the repertoire of opinions, reasons, tenses with a bit of conflict, conversation and disappointment thrown in.

But one pupil came up with this:

You can see it does contain the elements of the repertoire we practise using across topics. It has opinions, with reasons, direct speech, an element of conflict of opinions, use of imperfect and preterite ending in disappointment. You can see the underlying structures of the model aquarium story, both in terms of the repertoire of language and in how they are deployed.

But this pupil's answer was different to many others in the class. Because they were telling a true story. And a painfully personal one, with genuine and lasting disappointment.

Does this make any difference in terms of language learning? I am tempted to throw my hands up in the air and raise my eyebrows. Because of course we want to be equipping pupils to say things they actually want to say. Not teaching it as some kind of sudoku where they concoct answers to successfully fit all the required specified pieces into a pattern to show they can solve the puzzle. But even just in terms of long term memory and internalisation of language, does this pupil's work show something important?

Well. Here's what happened in Year 11. As we prepared for the exam, I did not let them look at their work from Year 10. They should have not only internalised the language, but also have been able to still deploy it.

The pupils who had cobbled together an answer to show they could use the language features, were basically starting from scratch. They had no memory of the work that they had done in Year 10.

Whereas the pupil who had written the story based on a true story that mattered to them personally, wrote this.

They were able to quickly and fluently reproduce several of their answers from Year 10 because what they had written was memorable. But if you look closely, it's not at all a case of word-by-word memorisation. It's actually a different version of the same story. And I am pretty sure if I listen back to their recorded GCSE speaking exam, they came up with another spontaneous version on the day of the exam.

Where does this leave the current GCSE speaking exam? AQA's marking of the Conversation by counting conjugated verbs seems to have gone entirely towards the parsing of language features, rather than the creation of meaning. As we see in this post, candidates who want to say things in genuine response to the examiner's question, do worse than pupils who trot out a list of three essentially meaningless verbs.

This exchange from my mock speaking exam, would score in the bottom band:

Have you ever had a problem abroad?

Yes. Once in a shop in Spain my mum forgot how to say 'coat'  in Spanish.

So...?

So we went to a different shop to buy a coat.

The pupil made the unforgivable mistakes of giving a genuine answer to the question, interacting with the examiner, and putting too much information into one clause.

Had they just said, I went to Spain. I went to a restaurant. I don't like to eat in a restaurant, then they would have been in the top band.

All my instincts and experience, such as the examples of the pupils in this post, are telling me that this is bad for language learning.

AQA Conversation really not working

 At our latest ALL in the East meeting (scroll down for previous meetings on this ALL page), we discussed the implications of the AQA markscheme for the Conversation, in the light of conducting mock exams. The main area of concern was the way the need for information to be delivered in 3 clause chunks, each with a conjugated verb, did not make for a natural conversation with pupils interacting spontaneously with the teacher-examiner.

In the run-up to the mock, I had been confident that despite the confusion around AQA's marking of the conversation, it wouldn't in practice have too much effect. I imagined the stronger candidates would still do well. And the weaker candidates would still do less well.

I was wrong. Stronger candidates tended to put more information into a single clause. Stronger candidates tended to interact more with the examiner. Stronger candidates tended to give an answer that focused on saying something they wanted to say. All of these 3 things penalised them.

Here's an example.

Pupil: I always like to try to eat healthy food like fruit and vegetables.

Teacher: For example?

Pupil: For example yesterday I ate salad for lunch and a potato for tea.

Each of these answers is a single clause, putting this pupil in the bottom band. Their responses are both minimal. 

Had they said simply, I like to eat fruit and I love vegetables. I don't like salad, they would have scored in the top band, for an "extended" answer using three conjugated verbs.

One thing that is recommended in the AQA spec, is for the teacher-examiner to use short follow up prompts as I did in the example above, to elicit more information from the candidate. Things like, and...? so...? for example...? Why? And what if it rains...?

I have certainly used this in the previous GCSE, to interrupt and redirect a pupil who had a pre-learned answer to deliver, directing them away from stilted word by word regurgitation, and steering them into a more spontaneous interactive conversation. We also use it in teaching, for example with the conjunctions dice game described in this post, or in working with pupils explicitly on how an answer can develop logically and coherently.

After our discussions in the meeting, I contacted AQA to see if using these interjections to invite the pupil to develop their answer, would allow the two utterances in this example to count as developing the answer. So that my prompting, as recommended in the guidance on conduct of the exam, was allowing the pupil to show that they could continue their idea and spontaneously give further detail in interaction with the teacher-examiner. That's what I asked. And that is what I was hoping they might say.

Or, on the other hand, could it be AQA's decision that interacting like this actually penalises the pupil, because it means that what follows the "For example...?" doesn't count as developing the answer. It counts as a separate minimal answer.

That was indeed AQA's response. In this example from my mocks that I put to them, they determined that this is two minimal answers. This pupil would be in the bottom band.

This means several things.

Firstly it means that I am less likely to use these follow up prompts. Because a pupil who has already given information, could give some more. But may not have the required three items, if they already told me one or two things in their initial answer. Saying, "For example...?" may be trapping them into giving one further detail. So I will be pushed towards falling back on my list of starter questions, making the exam more of a predictable plod through a list of questions. Even though this is explicitly prohibited and undesirable.

Secondly, it means that I have to teach pupils to give answers in chunks of three conjugated verbs. This risks moving towards pre-prepared and over-rehearsed answers in order to achieve this. Again, undesirable.

Thirdly, it means that in conducting the exam, if a pupil only gives a single clause answer, or a two clause answer, I will have to sit and wait for them to add a third clause. Pupils will have to be trained to just say something. Without worrying if it is a logical development, or something they really want to say. As in the example above, just adding, I don't like salad would push you into the top band.

Fourth, and possibly worst of all, I have to train my best pupils to be more like the weakest. I have to train them NOT to talk naturally and put lots of information into a clause. I always like to try to eat healthy food like fruit or vegetables has to be replaced by minimal chunks of information each with a verb. I like cheese. I love cake. I don't like salad. 

Sorry. I just realised I accidentally and ironically used the word "minimal" to refer to what would be a top band "extended" answer. This markscheme is topsy-turvy.  In this universe, the clause that contained most information was the minimal one.

I must emphasise, that going into the mocks, I told myself that the marking wouldn't be too bad. But the experience did not, unfortunately, live up to that. There were numerous examples of strong candidates putting lots of genuine information into one clause. Here's an exchange with another pupil:

Have you ever had a problem abroad?

Err. Once, on holiday, in a shop in Spain, my mum forgot how to say "coat" in Spanish.

So...?

Oh. So we had to go to another shop to buy a coat.

This doesn't count as a pupil developing their answer in interaction with the teacher-examiner. This counts as two bottom band minimal answers.

I had to explain this to the pupil, using this example from their exam, to show them where they were losing marks. Their response, But that's not how a conversation works. 

Well. It is now.