But didn’t you do grammar-translation?

I get this a lot, of variations on it. Yes, it’s undeniably true that I studied Greek and Latin via grammar translation methods. I did 5 years of Latin at university, all G-T. And I was taught Greek via G-T. And at the end of 5 years of Latin I wondered why I couldn’t read anything, and at the end of 4 years of Greek I was fine with a New Testament and lost without it.

And now I’m quite a few years down the track, and people both ask me, “well, aren’t you at the point where you are because you did grammar-translation and then you went on from there?” I think this is really a case of post hoc ergo propter hoc, and it’s a logical fallacy. But to recognise that you need to come to terms with at least one thing.

In all sorts of contexts I keep telling people that it’s entirely possible to learn ancient languages without learning ‘grammar’. And they don’t believe me. Which is understandable, because (a) most people have very little idea about how languages are acquired, but (b) most people think they understand how languages are learnt. (b) is really quite problematic precisely because people thing, ‘well, you learn some grammar, and some rules, and some vocabulary, and you get better and faster at internalising those things’. And, more often than not, this corresponds to their experience of language learning situations.

This is also compounded because people generally think learning a first language is radically different to a second language. Which, it isn’t. It’s a little bit different but the process is mostly the same.

So, let me bring you to a conversation I had recently with some Latin learners. This is an abstraction of a real conversation.

Me: You don’t need to know what an ablative is to learn Latin.
Them: what??
Me: well, you don’t. most native latin speakers probably didn’t have a specific notion of ‘ablative’ that they learnt.
Them: Ørberg doesn’t even introduce the ablative for a hundred pages or so.
Me: Except, chapter 1 of Familia Romana is full of ablatives!
Them: what? oh yeah, I didn’t even realise.
Me: Exactly. You understood in Italiā just fine, because it was comprehensible in context. You didn’t need to be told what an ablative was beforehand.


How is this relevant? Well, if you circle around to the initial point here, it’s that it’s not only entirely possible to acquire a language through comprehensible input, it’s that acquisition and explicit grammar instruction are such different processes that result in entirely different outcomes, that even if you do explicit grammar instruction, it doesn’t necessarily help language acquisition.

And, we have a whole field of research that supports this. That’s the whole field of Second Language Acquisition. And, at the very least the vast bulk of that research suggests that explicit grammar instruction aids very little to zero the process of acquisition. I’m very happy to concede that it may help somewhat, although some linguists in this field would say “no, not at all, at best it doesn’t hinder”, but well, let’s be generous at this stage.

That’s why I advocate for Communicative Language Teaching and for acquisition – because I read introductory material in SLA, and then I went on and read research papers, and I keep reading as much research as I can find time for. I’m interested in teaching in a way that produces acquisition rather than explicit knowledge, because I’ve experienced both, and I believe genuine acquisition is a more worthwhile goal, that sees students reading texts without translating, with direct access to the language, and understanding with fluency.

However, I do teach grammar. Sometimes, for some contexts:

  • When I have students who need grammar for their courses and employ me to help them learn grammar to pass a grammar-based exam.
  • When I have learners who have already done grammar, and find it useful to use grammar as a meta-language to illuminate things in a text. That’s precisely what grammar is useful for – talking about Ideally this can be done in the target language – then you’re both talking about language while still getting input in the language.
  • People often like to have grammar, in either language, so that they have an explicit knowledge of what they are figuring out implicitly. It is affectively helpful for them to feel like they know what’s going on. I’m fine with that, because I think ‘feeling like you are understanding’ is important for learners.
  • When I have learners who want to be equipped to deal with commentary-type material that uses grammatical meta-language. In this case, I am training learners to acquire a competency in a different area – how does one learn the explicit knowledge of language required to engage in conversations about explicit knowledge of language. To the extent that that’s a goal, that can be taught. It’s not acquisition though and it doesn’t lead to acquisition.

There are good reasons to teach grammar, so that’s why I sometimes do. At the end of the day though, neither research, nor my experience on both sides of grammar-translation and communicative-language-teaching (with historical and contemporary languages, as a teacher and a learner), suggests or supports that G-T leads to acquisition.

7 responses

  1. Pingback: But didn’t you do grammar-translation? — The Patrologist | Talmidimblogging

  2. Greetings,

    I am a classical languages educator and a doctoral candidate in historical theology. I, like you, first learned Greek, Latin, and Hebrew in the grammar translation method. I particularly excelled because I was good at memorization and test taking. It only occurred to me that I might have something wrong in teaching those languages when I began to teach French to American students and had far more success with my students ability to read and speak French than Latin. That is, same kids, same teacher, different method (it was a small school). I was given a grant to spend the summer at the Accademia Vivarium Novum and I have never wanted to teach grammar-translation again. This is a very long winded way of saying I agree with you nearly 100%, but I have one nagging question: what do we say about our teachers, especially at the graduate level, who have a tremendous grasp of the language, but don’t speak it or think in it? My advisor has done many translations and seems to have a thorough grasp of Latin and Greek, but never writes or speaks in the language. Your last line says, “neither research, nor my experience on both sides of grammar-translation and communicative-language-teaching (with historical and contemporary languages, as a teacher and a learner), suggests or supports that G-T leads to acquisition.” What does grammar-translation lead to? Or what would you call an academic who can translate and understand the language seemingly well, but they have not acquired it in a C-I fashion, which leads to speaking, writing, and fast reading in the language?

    Keep up the great work. I have followed you for a while without engaging much, but I love the kind of work you’re doing.


    • Hi Chad,

      I think that is a great question, which is why I anticipated it by writing next week’s post about it. But here’s my short answer: translating in order to understand a text is one way of rendering incomprehensible language input, comprehensible. Do that enough (oh, say, years of translating masses of text) and you will provide yourself with a lot of comprehensible input, which is going to fuel implicit learning. But it’ll feel like one applied explicit learning to get there (which they did, but perhaps not in the way they think they did).

  3. I second Chad’s question and am looking forward to next week’s post!

  4. Pingback: But didn’t you do grammar-translation? — The Patrologist – Mon Blog

  5. Seumas I know there isn’t enough data to conclusively show one way or another, but in your experience teaching Greek students, does taking a more communicative approach shift the entire bell curve right, with individual students more or less keeping their relative positions within the curve? (Does that make sense?) In other words, do you find that almost any student would do better with a communicative learning approach than the same student would have with a GT approach? Or do you find more variance – that overall, the communicative approach shifts the median right, but some students do worse with it than with a GT approach?

    • There definitely isn’t data for Greek, and I don’t have enough data for Greek personally either. There are also just several methodological issues: if you test for grammar, explicit grammar students will do better because that’s what they have been taught and tested on. If you test for implicit knowledge, e.g. by reading comprehension, or assessing unforced output, then acquisition students generally do better. If you teach for acquisition for a reasonable period, and then teach some grammar on top of it, those students can usually equal or outperform grammar-translation students. At least, based on what I’ve read for other languages. I don’t think there’s good data on this for even large modern languages (Spanish, French).

      But one of the things we need to keep in mind, is that these are not two alternate paths to the same destination, they are two types of cognitive work that result in two separate, and minimally-overlapping, types of knowledge.

      Anecdotally, grammar-translation has high drop-out rates and low ‘conversion’ to upper level courses. It appeals to a certain type of person and type of learner, whereas CI-based learning has higher stay-in rates, generally better learner enjoyment, and is thus more inclusive. Because if you can learn an L1, you have the mental faculties to learn an L2, it’s not *as* dependent on, e.g. IQ.

%d bloggers like this: