Thoughts on ELT, English and whatever else comes into my head
@Muranava suggested I write about the process of grading texts from a writing point of view and, since I started writing stories and articles for the British Council Learning English Teens site, I’ve become much more aware of the mechanics of how to do it because I’ve written each piece at three different levels. B2, B1 and A2 which run at three different lengths: 900 words, 700 words and 500 words respectively. I don’t know if the way I do it is how anyone else would do it, but there’s very definitely a systematic approach.
I’ve written about some aspects of writing Graded Readers before in a post on The difficulty of writing good “bad” writing because you have to break some rules that are usually considered Doing It The Right Way. But levelling for EFL is another thing entirely and not just about sticking in a load of otherwise unnecessary adverbs. Which, by the way, I still hate doing and don’t, therefore, do enough of.
So here’s how I go about it.
First, I write the high level one, pretty much just writing it naturally with no thought to level. There’s a certain amount of grading I already do automatically as I often find anything I write for EFL never comes out much beyond B2, sometimes even lower, as I tend towards shorter, simpler clauses. The benefit to this is it’s much easier to grade down than up. I can write anything at any level, and my first published Reader, The Tomorrow Mirror, was Beginner with a wordlist of just 300 words to choose from, but to change that to A2 I’d have to sit down and start it afresh.
Then I turn to what has become a double-edged sword, the Cambridge Assessment English Vocab Profiler (EVP) ***, to weed out the vocab that’s higher than B2. Pretty simple and it’s easy to switch words out for synonyms at this point. The EVP is easy to use and gives clear results, once you’ve allowed for it not recognising short forms like “won’t” and being out of date on usage of words like “Twitter.” I don’t know what ratio of vocab makes a text B2 though as the majority of words we use to create sentences, at least in non-academic writing, is A1 and A2. This paragraph so far, for example, is 2/3 made up of A1 and A2 words with another 17% unlisted as I just mentioned. But it’s completely natural writing with no levelling on my part even though only 9% of it is B2 – C2.
NB The EVP’s first analysis defaults to displaying the word at its lowest level meaning. You have to check any words that have multiple meaning to make sure the way you’re using it is not a higher level. That’s also partly why texts can initially appear to skew low.
*** Turns out what I was referring to as the EVP is a separate tool based on the EVP, which now charges for the kind of analysis I am describing above.
What probably matters more in EFL terms, though I’d be with you in arguing against this as such a massive basis for determining level that it’s how we teach, is the grammatical structures and the complexity of sentences. But, there I have only an imperfect way of measuring as I have to run it through a Readability tester like the Flesch-Kincaid scale (no, I’d never heard of it before and still don’t really know who that person with the double barrelled name/those two people is/are). The one I use, Readable.io, rates that same paragraph as 10.3 which is pretty high because my sentences are long and is equivalent to C1, according to the site. I usually find when I know I’m writing for EFL that I struggle to get above B1 in terms of Readability. But I tend to know which grammar structures I can use because I have absorbed the CEFR from course books over the years and my mind now works like an English File contents page. Second conditional = B1, third conditional = B2, present perfect = A2, chuck in the continuous at B1 etc. When in doubt, I cross reference with the lists of grammar according to the CEFR on ExamEnglish.com and delete, modify
or add structures accordingly.
Here is one of many inconsistencies in this whole categorisation thing, though. For example, the grammar structures list has the modal verb “might” as B1. That seems stupid to me as surely students need to, and can, express the idea of uncertainty before then (though maybe they stick with “maybe”). The Profiler certainly thinks so as it has “might” as a piece of vocab at A2, much more reasonable and the one I choose to go with. Another annoyance is how it deals with suffixes. “Drive” and “driver” are both A1, but “sell” is A2 and “seller” is B1, “hunt” is B1 while “hunter” is not listed. It also hates adverbs and nouns over adjectives when I am sure a student would be able to understand them in a written text. But unless I have a wordlist as authority, how am I to argue my case for putting them in a Reader other than my intuition?
This is all because the Profiler is based on what students produce in Cambridge exams, ie their active knowledge, not their passive knowledge. So, surely “might” is passive knowledge even earlier and should be taught at A1? But it’s not going to be on a syllabus or in Graded Readers until higher levels so how are those exam students producing it? And, anyway, why are we basing input on output … isn’t that a circle that will make someone’s head combust one day? I assume this flaw is why it doesn’t seem to like suffixes. Maybe students don’t use them, but this doesn’t mean they couldn’t make the leap of inference to understanding them. Maybe they, like Stephen King, shun adverbs and who could argue with them for that? These days when I write texts for any project I rarely get given a wordlist and I just have to use the Profiler to stay on track with level. But I can only suppose that means the CEFR level of writing produced by writers is lower than it used to be because everyone is relying on this one source since it’s easy to use and free. Not that I have any idea how different publishers were creating their target vocabulary lists before the EVP.
Anyway, back to the method. I take a quick look at my list of B1 and A2 words and see if there are any synonyms that might be added at B2 but, mostly, stage two is shortening and levelling down. I look at the entire text for a general idea of where the hardest parts are and see if they can be cut entirely, which might only mean parts of sentences but can be a whole paragraph. These cuts tend to be descriptive material or additional to the main points. That way I can work on simplifying the rest, rather than wasting time simplifying something that will be cut for length anyway, kind of like when the hairdresser lopped off the end of my ponytail when I hadn’t had a haircut for a year. No point styling the ends if you’re going to give them the chop entirely.
Then I go through, pruning clauses, breaking up sentences and replacing obvious words with their B1 equivalents. There isn’t a massive amount to do here from B2 to B1 because of the way the text is made up in the first place with the ratios of necessary low level words. Grammar wise, levelling is easy too. I think any idea can be worked somehow into the lowest level. Here’s Regret, for example.
She wished she hadn’t told him. B2
She shouldn’t have told him, she thought. B1 (Or “I shouldn’t have told him,”she thought.)
She thought about their conversation. “It wasn’t a good idea to tell him,” she said to herself. A2
And here’s how that looks if we take a section from the B2 ghost story Mystery Train …
A man stuck his head out of the window of the engine and shouted, furious. ‘Don’t you know it’s dangerous on the tracks?’
‘But … but … there aren’t any trains!’ Claire said, feeling stupid as the words came out of her mouth. There was clearly a train now, whatever had been true before. The train shouldn’t be, couldn’t be, there. It’s not real, she told herself.
‘This isn’t real,’ she repeated out loud. But the heat coming from the engine certainly felt real enough.
… and compare it with its B1 partner.
A man put his head out of the window of the engine and shouted. ‘Don’t you know it’s dangerous on the tracks?’
‘But … there aren’t any trains!’ Claire said. But there was clearly a train now, whatever was true before. It’s not real, she told herself.
‘This isn’t real,’ she repeated out loud. But the heat coming from the engine felt real.
The last one is the hardest, both vocab-wise and in terms of preservation of the narrative. Going from 900 words to 500 but telling the same story is a real exercise in distilling the essence of what you’re trying to say, though having already gone via the halfway house of 700 helps. And there is a lot more vocab to simplify. Grammar isn’t that hard to change. I use little tricks like turning passives into actives (generally good writing practice anyway) and often find I can avoid past perfect just by sticking it into past simple and putting time markers in; reported speech (a teaching point I loathe) becomes direct speech easily because then you can “forwardshift” to a simpler tense eg present perfect to past simple. Here’s how that section ends up at A2.
A man put his head out of the window and shouted. ‘What are you doing? It’s dangerous on the tracks!’
‘But … there aren’t any trains!’ Claire said. Maybe there were no trains before, but there was a train now. It’s not real, she told herself.
‘This isn’t real,’ she said. But the heat coming from the train felt real.
That’s about it! I like the discipline of writing at three different levels and I also like that a story is not limited to only one set of learners. If Graded Readers were more popular with publishers, I’d love to see mine with high and low versions.
So, far, I’ve got three pieces out on Learning English Teens with three more written and another three at planning stage which I’ll add links to as they go live. They all have pre-reading and after reading exercises for use in class or as homework.
There’s nothing to fear but … – a horror thriller (link to post about this story which contains the links to all 3 levels)
Mystery Train – a ghost story (links above)