ELT in Brazil

One of the most common ways people go about describing language learning is through metaphors, as if the pure concept lies a bit too much beyond reach. For example, learning a language is like learning to dance.

The dance teacher greets everyone and provides a brief introduction to the dance style we’re going to learn today. He asks us if we’ve ever tried it before and helps us connect and activate our prior knowledge of dancing.

Then, he presents the first couple of steps. Move here, move there, hands here, hands there, now spin. For him to be able to do that, he had to atomize dancing into pieces that compose dancing. He then had to decide which pieces were worth teaching and in what order, going from simpler to more complex, in his opinion.

After presenting the atomized steps, he gave us plenty of controlled practice. We practiced those steps for a couple of minutes, then he asked us to change pairs and repeat, repeat, repeat again. He would interrupt us from time to time to give us some corrections here and there because he expected us to be accurate from the start, before progressing to more complex steps.

Finally, towards the end of the lesson, the dance teacher plays the music, which sets a more complete context for what we were doing, and asks us to dance freely to it. Not really freely, because he is clearly checking whether we can perform the steps he taught us beforehand. The lesson ends in a feedback moment, and we all go home happy, knowing we learned something today.

Simple, isn’t it? Predictable also. Oh, and very teacher-controlled as well. Safe. Neat. This closely follows Skill-Learning Theory (SLT). This is the idea that learning a language is like learning any other skill in life, be it dancing, driving, swimming, or making pancakes. The procedure follows a predictable, rather stiff route: explicit presentation (be it inductive or deductive), a sequence of practice (often moving from very controlled to less controlled), and, finally, automatization.

This is the basis of mainstream ELT. You’re going to find this in your coursebook, in your TEFL certificate and diploma, you’ll find it in training sessions, in conferences, and you’ll find it being endorsed by exam bodies in the way exams are designed and their preparation materials. The problem is that SLT ignores recent findings in Second Language Acquisition and the fact that, deep down, we know that learning a language is an incomparable endeavor of the human mind. There is nothing as complex and sophisticated as learning a language.

There are several aspects of SLT that work wonders for teaching one how to Dougie, but are inefficacious when it comes to learning a language. For starters, SLT necessitates the chopping of language into discrete bits that can be ordered and delivered to students at a steady pace. We know language to be a complex, holistic system that is better understood through analytical lenses rather than by atomizing it and requiring students to synthesize the pieces back together.

We also know that Krashen was kind of right when he talked about a natural order of acquisition. Learners who acquire the language in naturalistic settings, for instance, those who migrate to a country where the language is spoken, acquire the language in pretty much the same order as those learners who learn the language in an instructed classroom setting – and mind you, this order is not the order in the grammar syllabus. This is an indication that, regardless of the order of presentation of the chopped-up language bits in a grammar syllabus-based course, the order in which learners acquire the language is not impacted in any way. Research in SLA has taught us that teachers cannot force a syllabus on their students, but we can help them progress faster through their own learner-syllabus and achieve higher levels of proficiency when compared to learners in a naturalistic environment.

Through our understanding of developmental sequences, which are documented sequences of stages a learner goes through when learning, for example, negations in English, we know that learners cannot be expected to be accurate first and fluent later. Learners must progress from one developmental stage to another, and there is nothing teachers can do to get learners to skip stages. Therefore, it is simply not possible for learners to be accurate with a grammar McNugget they have just been told about. This is simply not how learning a language works. Learners’ progress through developmental stages exhibits backsliding and U-shaped behaviors that, again, no teacher instruction can avoid – these are absolutely necessary developmental mistakes.

When it comes to explicit instruction, key to SLT, Krashen seems to have been correct once again. In his learning/acquisition hypothesis, Krashen differentiates the kind of knowledge that is brought about by explicit instruction (learning) and knowledge that is incidental, built during communicative events (acquisition). According to Krashen, a student’s body of explicit knowledge serves as a monitor, a gatekeeper who audits language that is about to be produced by the learner’s implicit body of knowledge. SLA seems to indicate that the knowledge we use to communicate, especially when we do not have the time to stop and think about grammar rules and whatnot, is the fruit of our implicit knowledge of the language. That is, knowledge that we have and that we do not know that we have. Contrary to this is explicit knowledge, which seems to be rather counterproductive when it gets in the way of natural, fluid communication. Additionally, going against what is believed in SLT, it seems that explicit teaching, which produces explicit knowledge of the language, cannot be turned into implicit knowledge through exhaustive sets of controlled, less-controlled, and ‘freer’ practice.

Of course, there are many arguments in favor of SLT. For instance, it makes it possible to design and market materials that provide students with input and controlled practice of chopped-up language. It would be challenging to sell language teaching/learning if you could not package it into A1, A2, B1, B2, C1, and C2 boxes; thus, atomizing language seems to serve a market demand. SLT-based teaching, such as the Present-Practice-Production framework, gives teachers a lot of control over what goes on in the classroom. The teacher decides what to teach (or has the grammar syllabus decide it for them), the teacher decides how this language will be presented, how it will be practiced, and then how it will be assessed. This level of control is appealing to the increasingly deskilled teacher and makes it possible that someone without any teaching experience or training can ‘teach’ the lesson in the book – whether this will be an effective lesson or not is a whole n’other story. This level of control also makes it possible that teachers who teach very large groups can maintain control of the class while giving stakeholders a sense of progress. After all, you’ve just finished book 3, so you must have progressed a lot from book 2! Making a pedagogical decision to try to band-aid a managerial issue does not seem to be the right course of action to me. Why are teachers teaching very large groups in the first place?

This seems to indicate the need to overcome some old (1960s-era) traditions in ELT to move forward and beyond SLT. In my opinion, teachers should realize that teaching and learning a language is unlike any other kind of learning and, through this realization, move away from teaching that is effective for baking, dancing, or computer programming. Our learners require teaching that is holistic, analytical, interactive, communicative, and implicit – and for this, we have to let go of teaching that is atomistic, synthetic, pseudo-interactive, practice-based, and explicit.

Now, “What then, Bruno?”. If you’re looking for language teaching that is in sync with recent findings in SLA, teaching that is holistic, analytical, interactive, communicative, and implicit, do give Task-Based Learning (TBL) and DOGME a try. You can find more about TBL with Rod Ellis, Mike Long, Geoff Jordan, and Peter Skehan. For DOGME ELT, go for Scott Thornbury and Luke Meddings.