Would you like to learn a language without spending time on grammar training? Or would you prefer to focus on grammar rules, hoping that they will bring you closer to your communication goal?
In my opinion, these are two rather polarised positions. The first is strongly influenced by the ideas of Stephen Krashen and his acquisition hypothesis, in which he states that explicit knowledge does not convert to implicit knowledge, and that the two elements are distinct. Implicit Knowledge is the knowledge that you get when you are doing something, and you do not notice that learning is taking place, for example, learning how to walk. Explicit knowledge can be codified, stored and accessed and easily be transmitted to others. Students can learn grammar explicitly, but perhaps they cannot actually use it to produce language. This is known as the “no interface” position, in which learners are expected to acquire language through comprehensible input. In this doctrine of teaching, the learners are exposed to a large amount of input, and are then expected to pick up grammar structures naturally and intuitively. Furthermore, there is a natural order of learning which we engage in intuitively and explicit grammar training has no value in teaching methodology.
On the other end of the spectrum is the “strong interface” position which argues that students should study grammar, which will lead to them gaining the abilities they need to speak a language. According to this theory, explicit knowledge can convert to implicit knowledge through repeated practice drills and exercises. The course book is almost always organised around a grammar syllabus with a sequence of arbitrary grammar points. The teaching methodology is usually Present – Practice – Produce, with the fast majority of time spent on grammar explanation, drills and grammar exercises with very little time left over for natural language production. This approach is found in the vast majority of grammar workbooks.
A much more useful theory is the “weak-interface” position, which acknowledges that the study of grammar can increase the speed of learning. This approach draws from both extremes of the spectrum, so that both implicit and explicit knowledge are seen to work together, but with a recognition that implicit knowledge is vitally important, and that explicit knowledge functions as a facilitator of implicit knowledge. This approach suits task-based learning, in which context and function arise out of a task. The teacher does not provide explicit grammar instruction, and the learner is required to “notice” the target grammatical structures, which can then be practiced in more depth. Grammar exercises are understood in the context of emergent grammar need.
How does this work in a lesson? The teacher will first provide a context with examples of the target language, which the learner is then required to notice, discover and analyse. A similar procedure is a Test-Teach-Test approach, in which the learners themselves will be asked to produce language in the first test-phase, the teacher can focus on real language deficits in the teach-phase and suggest improvements, followed by a repetition of the “test-phase where improvements are implemented. This is a much more reactive approach to teaching, where the teacher focuses on slowly emerging functional needs, rather than on an arbitrary grammar syllabus. An even freer approach is Dogme teaching, where experienced teachers engage students in communicative tasks, without relying on the crutch of printed training materials. In this process, grammar and vocabulary are allowed emerge and are dealt with by the teacher in a very natural, communicative process. Here learning takes place without the learner even noticing.
In conclusion, thankfully we don’t have to adopt extreme positions, and it is possible to find a middle road, in which the learner can benefit from the interaction between explicit and implicit knowledge, learning from context, emergent grammar, and learning items in a more natural way based on student need, rather than learning from a random list of grammar functions.