My current understanding of how grammar is mastered – the ‘theory’

This is the first post in series of two posts in which I review my current understanding of how grammar is mastered. In this post I’ll overview some research on grammar acquisition and in the second one I’ll give a concrete example of how some of these ideas prompted me to tweak the way I teach a particular grammar topic (the patterns used with ‘I wish’). To summarize these changes, I have

• stepped away from the ‘pure’ PPP lesson shape with its initial ‘presentation’ stage
• modified some of the practice activities in order to change the way the structure is retrieved from memory

What do we mean when we say ‘mastering grammar’?

• Probably everyone who’s involved in language teaching knows the term ‘interlanguage‘. The idea is that a learner has an ‘interim’ version of the target language that’s being developed to approximate the target. I got deeper insight into the concept of inter-language having read about two types of memory/processing operating in our brain:
  
  Grammatical processing belongs to the so-called implicit knowledge and operates automatically, mostly without recourse to attention, unless the person stops before formulating the utterance and rehearses it in their short-memory. This means that the students do not have access to ‘grammar rules’ while they’re speaking spontaneously. Instead, they use so-called ‘routines’ for grammatical processing, and these routines might not correspond to the rules they know.
• It has been said that we learn the language in the same way as any other skill: awareness raising -> massed practice -> automatization and that thus Skill learning theory justifies the PPP lesson shape (Dornyei, 2009). However, I personally have found the PPP lesson shape somewhat problematic. The way I see it, the devil lies in the detail – in other words, what happens during the first two stages of PPP might fall through if they do not in fact correspond to the stages outlined by the skill learning theory.

• The first P (‘presentation’). I’ve been experimenting with my own language learning (learning German) and intuitively I’d say that if the lesson starts with straightforward ‘presentation’ and goes on to controlled practice in which s/s practice applying the rule (whether given explicitly or induced by the s/s from input), this encourages them to start constructing a new routine, completely unrelated to the one that is currently stored in their interlanguage. ‘Eliciting’ the rule does not save the situation here, because when we elicit the rule, the students retrieve explicit knowledge of the rule, and the ‘meta-thinking’ involved does not kick in the routines the students use subconsciously. Also, even in the realm of explicit knowledge some confusion is still possible. One of my students once told me that questions in present simple are formed with ‘do/does’, only to go on and write a few questions for a partner with no auxiliary whatsoever. When I asked the guy why he was not applying the rule, he told me that he was using a different structure. Over time I got convinced that rules should be presented after production has taken place so that the students are
  1. encouraged to connect the new explicit knowledge to the knowledge they’ve already got
  2. have a chance to ‘review’ and modify the routines they are currently using (of course, ‘reviewing’ an automatic routine sounds oxymoronic 🙂 but what I mean is that the s/s will learn to monitor the right routine and ‘weaken’ connections in the brain through self-correction. See also the article by Nick Ellis in Sanz, C. and Leuw, R. ed (2011).
  I like the term ‘noticing the gap’ that Scott Thornbury uses in his books and in fact nowadays, when I’m planning my lessons, I think of the first stage not so much as a ‘presentation’ stage as a ‘noticing the gap and learning to deal with it/getting the crutches’ stage.
• I tend to think now that it’s ok to give the students ‘rules’ (downright deductively) so long as they enable the students to understand the essence of a grammar point (e.g. through highlighting the way the choice of a structure is determined by discourse, as in the case of the passive voice, or by highlighting the contrast between  structures, e.g. ‘past simple = tied to the past, present = relevant; present perfect = relevant + there’s an element of ‘before’). The main point of the rule is to allow the students to review the language they produced at a previous stage, genuinely see a problem with it and improve it (simply noticing that it contradicts the rule given by the teacher won’t do the trick, of course, but seeing that the wrong choice of structure impacts on the message, if T highlights the contrast of structures, will). However, so far I haven’t come across any theoretical grounding of what type of rules would be most helpful and why.
• In order for the s/s to notice the gap, some record of their use of TL is needed. I generally prefer to set a discussion question that would require the use of TL. However, such tasks are not so easy to formulate. Alternatively, there are writing activities, such as grammaring, two-way translation and dictogloss. In my opinion, grammaring can be used for any pairs or sets of structures that get confused (e.g. will/going to, the passive vs active voice, past vs present perfect), although, as the choice of form often depends on the discourse/co-text, it’s probably better to give s/s highly contextualized sentences to reconstruct. For instance, students could process a text for meaning and then get a copy of the text with several sentences tippexed out and replaced with the lexical items from these sentences.
• As for the second P (practice), there are two sides of the coin to think about and these seem to be difficult to reconcile. On the one hand, there seems to be evidence provided by attention theory that practice has to be focused and intensive (as, trying to practice two different routines at the same time, a student wastes a considerable proportion of attention resources). On the other hand, the teacher has to be extremely careful in ensuring that the controlled practice activity really does model the way the skill is operated in the brain. I’m in no way a psycholinguist, but as far as I understand it in real production there has to be some kind of retrieval of the automatic routine and choice going on. If a students is involved, say, in a ‘have you ever’ drill (however communicative), the structure is likely to be stored in their short-term memory and thus the skill is not being practiced. I find it difficult to reconcile this with the fact that practice has to be focused in order to make full use of the s/s’ attention resources (hence ‘distracting’ the student to kick the structure out of their short-term memory is a bad idea). Hence, I think that all controlled practice has to involve a contrast of structures, so that the students are required to make constant choices.
• In their article called Implicit and Explicit Corrective Feedback and the Acquisition of L2 Grammar in Rod Ellis et all (2009) Rod Ellis, Shawn Loewen and Rosemary Erlam present evidence that when practicing a grammar point, efficient feedback targeting the grammar point
  1. is provided on the spot
  2. identifies the mistake for the student (e.g. the teacher repeats the mistake with questioning intonation)
  3. encourages the student to make the conscious effort to retrieve the correct form (e.g. [T: ‘go? you need the past‘] when the student is telling a story in the past and tends not to use the past tense).

That said, it transpires from SLA literature that productive practice is actually not the only, or even the main process through which the correct automatic routines are formed. It seems that a large part is played by noticing the structure in comprehensible input. This means that unless the students are encouraged to read and listen widely, and also re-read and re-listen, directing their attention to the form,  they might get stuck with ‘incorrect’ routines despite practicing grammar in class.  

References

Sanz, C. and Leuw, R. (editors) (2011). Implicit and Explicit Language Learning: Conditions, Processes, and Knowledge in SLA and Bilingualism. Georgetown University Press

Rod Ellis et al. (2009). Implicit and Explicit Knowledge in Second Language Learning, Testing and Teaching. Multilingual Matters

Dornyei, Z. (2009) The Psychology of Second Language Acquisition. OUP

You can partially view all these books through the ‘click to look inside’ feature.

Also, I’m currently watching a fascinating series of lectures on Cognitive Science – here are the lectures on the explicit and the implicit: