How grammar-translation might lead to acquisition

Last week I wrote about how I did plenty of grammar-translation, and I don’t believe that it leads to acquisition. This often surprises people. Let me explain why it’s surprising:

Firstly, for those of us who did G-T and did a lot of it, and ended up with some language acquisition after all, it often looks like G-T lead to language acquisition. Just because after a whole bunch of G-T you ended up with some language doesn’t mean that G-T caused that language.

Secondly, for people in general who’ve had some experience of learning any language, grammar is so prevalent, that most people’s impression of ‘language learning’ involves grammar.

Thirdly, as adult learners in particular, we have a strong desire to understand what’s going on in a language. We almost can’t help ourselves. And I don’t actually think we should fight against this in learners. Knowing and understanding explicit structures is great for two reasons: (1) explicit knowledge of a language is not a bad thing! (2) we feel more in control when we have an explanation.

However, if the research consistently suggests to us that it is not grammar + memorisation + practice that leads to language acquisition, but rather exposure to comprehensible input in communicative contexts over time, how do we explain the scarred veteran survivors of G-T? You know the ones, the classics professors who can sight read Greek and Latin texts, or the biblical professor for whom the New Testament in Koine is as familiar as the English, if not more.

I wrote a while back about two different types of translation (, and I think this holds the key. What G-T is teaching you to do is: okay, I have a text in front of me, I don’t understand it, but I can apply (i) syntax + (ii) morphology + (iii) vocabulary = a translation. That is, people who practice G-T are applying a set of rules they’ve learnt, to translate a text in order to understand its meaning.

That, while we’re here, is very different from using grammar, or better yet linguistics, to analyse and discuss what a text is and does and how it means. This is why it’s really useful to get students to do some linguistic analysis on a native language – because they already know what the utterances mean, and so you can use the analysis to look at various parts of language.

So, you apply your G-T to understand a text, and you translate it, and then you understand its meaning. You’ve turned an incomprehensible input into a comprehensible one. Which means it’s now become primary linguistic data and your ability to process it as linguistic data for acquisition is engaged. But it took you X amount of time to do it this way.

Those who do tons and tons of G-T work – an undergrad program, grad school, work their way through a monstrous US college program’s PhD reading list, etc.. keep repeating this process over and over. And they (a) get very good at G-T, no doubt. Because that’s an explicit knowledge system and a skill, and they are practicing it constantly, but (b) they are also providing themselves with tons and tons of input.

Which explains why, in terms of SLA, they do end up with implicit acquisition.

That’s my general hypothesis. I don’t have any research on this, I’m just extrapolating from basic principles. And because it looks like G-T led to acquisition, and the other reasons I mentioned at the top, it’s very hard to dislodge this idea.

In a few posts time I’m going to talk about “what to do about grammar teaching”. Not that I have any definitive answer, but I have some suggestions. In the meantime, let me just ask:

What if input that leads to acquisition was not simply the by-product of the teaching methodology, as in G-T, but the core content of the learning methodology, so that acquisition was the goal, not an accident?

One response

  1. I can see from my experience how the G-T method can slowly lead to acquisition in some sense of the word, though not innevitably and in many ways despite itself. The process of translating and so going from greek -> english -> meaning will in some cases happen on phrases brief enough or repetitive enough that, as you suggest, it’ll start to make the connection between greek and meaning more directly. For example, in an easy text like Mark or John, in seeing the phrase ‘λεγει + αὐτῳ/αὐτοις/Pronoun’ often enough, it started to acquire real meaning for me.
    I’d speculate that it’ll depend somewhat on what the learner is trying to do. If they decode the sentence and think only about their english translation and quickly move on, I can see them never really acquiring anything, but if they engage in decoding and then go over the sentence again in greek, thinking about the meaning, that seems to be where someone would surely start to me making meaningful connections.
    The problem is that that kind of phenomenon is likely linked heavily to phrases/clauses that are very similar to ones native language, and so map easily. Situations where the word order is vastly different, or the nuance strange, I’m not confident that it’d ever really be something you’d be able to internalise just by reading too-difficult-texts.