So I went to a workshop yesterday on dyslexia and second language learning and it was, overall, really interesting. However, the bit that really stuck in my mind was the observation/recommendation for learners with Specific Learning Differences (SpLD) to use the kind of learning where learners work out the rules for themselves was potentially less effective and that for clarity and accuracy one should be using direct explicit instruction. This was an interesting point, and certainly an idea that had never occurred to me, although it does make some sense. It did, however, stir the embers of a little “what if” internal dialogue I have had about the comparative values of discovery learning and of direct instruction.
First, to set out my stall slightly here, unless otherwise stated I am talking about post 16 ESOL learners. I’m making no claims to anyone else. This is important because this forms, for me, the root of the issue.
My interpretation, and my brief Google researches would suggest that I’m about right, direct instruction means telling people stuff. Hattie ranks it as having an effect size of 0.59, which means it is a fairly good thing to do. 0.4 is the average, suggesting that anything below this is of a neglible effect, and we should be focussing on the higher end stuff. Hattie puts inductive teaching at .33 and problem based learning right down at the bottom – an effect size of .13. This is damning, really, suggesting that telling people stuff, asking them good questions about it, and giving them feedback on their performance is what works. Getting students to sit down and work it out for themselves isn’t.
Ok, so I’m basing this on certain assumptions, but essentially as things stand, the evidence would appear to be stacked against what, for me, is one of the cornerstones of how I have always viewed English language teaching and learning. To my mind learners need to work things out for themselves because telling students stuff assumes a shared language. This, ultimately, is what sets ESOL/EFL teaching apart from pretty much everything else in education. Telling people stuff and checking it through the various types of questioning is not always even possible, and even where explanation is possible there are still questions to be asked about how to make the shift from explicit, conscious meta-knowledge (“the past simple regular verb structure is formed by adding -ed to the base form of the verb”) to unconscious use (“I walked to college today”). The obvious one is degrees of controlled and freer practice (said the good little CELTA trainer). So sometimes I do wonder whether, with the right level of students and effectively designed practice tasks, direct instruction could work for some ESOL students.
But, and of course there’s a but, there remains a strong case for inductive learning activities in an ESOL class. When talking of Hattie, for example, one has to remember that a lot of his work was based on school age children in native language environments. So it is perfectly possible that direct, explicit instruction will fail because the learners simply don’t have the language with which to deal with the explanation, and you don’t speak the language(s) of the learners. You can explain as much as you want but, to be honest, you might as well whistle. Then there’s the case made by Richard Schmidt for what he called “noticing“. Unconscious exposure to language is one thing but decent learning only happens when you start to become aware the systems and rules governing grammar, including those slightly epiphanic moments of “oh, that’s what the rule is!”
Inductive learning, or discovery learning creates the conditions for this to happen. I don’t mean the slightly vague discovery learning espoused in task based learning, but careful, guided and structured discovery. So students are exposed to examples of the grammar, for example, then through carefully designed questions work out what the rule actually is. To use past simple regular verbs, for example, students read a text and underline all the verbs. Students then answer questions: are these things happening now? Did they happen before now? etc. The skill here is in the task design and in the management of that task: making sure that all students are answering the questions, making sure that they are clear questions, making sure that when the rule is finally elicited the students have good opportunities for effective practice of the language.
It’s quite a challenge too, because one can talk about form focussed instruction, which suggests that learners do need to be taught some stuff, whether it is integrated into contextualised, communicative focused lessons, but also an explicit form focussed lesson can be decontextualised – just teaching language for the sake of teaching language – although still to my mind we have to bear in mind the focus still needs to be on getting learners to produce the language. Certainly in my own practice I have become less of a communicative-context obssessive, liking grammar lessons where contexts and situations arise as the lesson develops.
So I’m going to do an experiment. I’m going to use my level 1 class as my guinea pigs for this, and I’m going to see which of these two things work better. First of all, I’m going to devise a simple test of two specific grammar points. The grammar points will be things they have covered before but at something of a remove: passive voice and reported speech. This is because I want them to come at the language with something resembling the same prior exposure, but also because these involve manipulating verbs outside of a time-tense context. I’ll administer the test – probably a simple transformation or gap fill, and record the results. Then in week 1 I’ll teach one structure using a direct instruction followed by controlled and then freer practice, and in the following week I’ll use an inductive approach, followed by practice activities as close as possible in format to the first week. At the end of each lesson I’ll also get feedback from the students on how much they felt they had learned.
In the following lesson, I’ll give the students the same test as they did before the experiment.
I know it’s not conclusive in terms of making a grand statement in favour of either, and a little flawed (someone pointed out to me the possibility that one of the forms may be harder than the other to learn, which is a good point) but as a piece of action research I think it’ll be interesting to explore.
Watch this space.
(By the way I found a whole bunch of interesting links on these themes, some of which I read properly, some of which I just skimmed. Definitely I want to read a bit more on this, but I present the links here for your enjoyment!)