Let’s start, if we may, with the rather tender concept of “meta”. Frank Coffield talks about students “going meta” in “All you ever wanted to know about learning and teaching but were too cool to ask…” In his case it was the idea that we become better learners if we know what it is we are learning about and what we will gain from it: sitting ourselves outside of our own learning and looking in at the learning and what there is to be learned. This is the basic philosophy behind target setting, and behind the idea of the learning outcome, and would, at first glance, seem pretty sensible.
Then there is the issue of meta-language: “the language used … to talk about and describe language”, as it is excellently summarised by the British Council. Terms like verb, noun, present perfect, indefinite article that sort of thing. These are the bones, the framework of language description upon which we hang classroom practice, but, and we shall explore this later in a little more detail, but for now I’d like you to reconsider the distinction between being able to use the language and knowing the rules of the language. (And if you don’t think there is a difference, then maybe you and I need to sit down for a long chat).
One area where these two concepts meet up is in the development of learning outcomes. These are the focus of your lesson, the things you expect learners to be able to do: the visible products of learning. They are usually written in some form of meta language (learners will be able to use past simple irregular verbs) and they channel the spirit of going meta but in a teacher managed way: the teacher has the opportunity to manage the whole process of learners reflcting on thier learning and identifying what they need to do to improve. When learners can tell you what they have learned, demonstrate it and reflect on it, this also provides a quick win for assessment for learning, indeed, in some minds, this very teacher-led assessment for learning is the only assessment for learning because it provides visible, measurable evidence of learning: “Have the learners achieved the learning outcomes?” becomes the question, not “Have the learners learned something as a result of being in the lesson?”. All you do is share the outcomes with the learners at the beginning, and get them to self assess against these outcomes, and then reassess at various crucial points of the lesson.
It’s not a bad idea, at least not when you are teaching skills based lessons where the outcome is an improvement in, say, reading or listening. It can be problematic, of course: a SMART learning outcome like “be able to read a text and identify 4 facts from it” is only going to be practice in that skill: it doesn’t allow for the fact that the skill may not be repeatable to order with any given text. It’s partly transferable but not entirely. You also can’t just write “read a text on fly fishing and be able to identify four facts about fly fishing” because although it is, technically an outcome (it is the specific skill that the learners will have demonstrated an ability to do), it is also pretty useless – its far better to be (and this is where SMART bites you on the backside) a touch more general, and say what transferable skills the learners will have developed.
That said, getting learners to think about how to apply the skills and sub skills of language can have value: a little (graded) instruction on how to read can be very useful. It’s not a bad idea, for example, to couch reading in terms of gist and detail, skimming and scanning, as long as the learners have a good idea of what these things mean. I would personally never bother, myself, except for scanning and skimming: certainly I never ask students to “read for gist” because even when you understand the concept it’s still pretty vague – a much better task would be “read it and tell me a good title” or “read it and decide if the writer likes fly fishing”, that sort of thing. (Although the learning outcome would still be “read and identify the gist of the text”).
Grammar and vocabulary based lessons should, in theory, be easier: you start with the language point (e.g. Present Perfect”) then simply choose the form of evidence of learning (ask and answer questions in a “find someone who…” task) and work back from there: “use present perfect to ask and answer five questions” is a workable outcome, and easily “differentiable”.
There are problems, however. For one, as I’ve seen on more than one occasion, explicitly sharing the grammar to be taught can create antipathy, even resentment, amongst the students (“I know present perfect. I study it before.”) or it can totally destroy any sense of discovery, of working out what the grammar is. Sometimes, too much “going meta”, especially at the beginning, can actually spoil the learning that happens. This is especially true of problem areas like present simple third person singular, or indefinite and definite articles, which have fairly neat, memorable structural rules, but which are very hard for some learners to ever use naturally and easily (I have known some very high level Polish users of English who slipped on articles from time to time). In these examples, as in many examples, the lesson structure benefits from the learners working out the rules inductively first before applying them and then finally (as an optional extra) putting a meta-language label on it all.
Sometimes too much explicit knowledge of a given language point can be a bad thing. It can disrupt the classroom processes, interfere with students ability to engage with the learning, because they may dismiss it (“Teacher, I know present perfect”) or they may be confused by it. Yes, it creates a situation where the learners can hang their learning on a framework, and expand their learning with wider reading and engagement, but this shouldn’t be the starting point of the learning.
The final issue with presenting the language to be learned first like this is that the learners develop an impression of language learning that takes place in a false universe. There is being able to use the language, and then there is being able to tell people the rules: these are different things. I can tell you how to saw a straight line, prune an apple tree to encourage fruiting, or fix the brakes on a bike. But I can’t do any of those things, at least not terribly well. My knowledge of those things is far greater than my knowledge how, and no end of reading, study and so on is going to help me improve, because I need to practice using those skills.
It’s not deep learning to be able to tell the teacher the rules, and I would be highly critical of an observer of the lesson who mistakenly identified this as learning. There isn’t a clear theory/practice distinction at play here. Meta-language, or rather meta-knowledge, perhaps, is learning of a sort, but in an ESOL context, and maybe any context, deep learning occurs at those lightbulb moments, where a task has been structured carefully and managed by the teacher that the learners work out the rule and then apply it correctly – “Oh, I see!”. They may take a bit of time and recap to get to the point where they can apply it unconsciously, and it is unlikely to happen in the first lesson in which they are exposed to the language, but the first step has been made.
The challenge here, however, is that to some minds, sharing learning outcomes is Best Practice – a thing which must be done at all times to all learners. But as I’ve explained above, sometimes this isn’t appropriate – because the meta-knowledge required can get in the way of the lesson structure, and even when you make the learning outcomes learner friendly, and remove the meta-language, learners can still spend the whole time trying to second guess the learning: “teacher, I know this!” is a cry that fills me with dread, and to which I always reply “then show me by getting this task right” at which point they fail to do so and you start again. Sometimes a lesson isn’t about the product, but the process – and by their very nature, learning outcomes detract from this. There should still be opportunity for reflection and discussion of what learning happens in a lesson, mind you, it’s just that it shouldn’t be seen as requirement at the beginning of the lesson: why not develop the learning outcomes as the lesson progresses, rather than rely on their conscious application at the beginning?