What is "grammar-translation"

General discussion about learning languages
Online
Cainntear
Black Belt - 1st Dan
Posts: 1593
Joined: Thu Jul 30, 2015 11:04 am
Location: Scotland
Languages: English(N)
Advanced: French,Spanish, Scottish Gaelic
Intermediate: Italian, Catalan, Corsican
Basic: Welsh
Dabbling: Polish, Russian etc
x 3726
Contact:

What is "grammar-translation"

Postby Cainntear » Fri Feb 14, 2020 12:38 pm

Pilot_2270 started talking about "grammar-translation" in another thread, and this term is a bit of a bug-bear of mine, so rather than taking the thread off-topic with Another Cainntear Rant™, I thought it was worth a new thread.

pilot_2270 wrote:I find it interesting that this guy, Xiaomanyc, uses Anki flashcards to learn sentences from L1 to L2, e.g., from English to Chinese.

Meanwhile, Matt vs Japan uses Anki flashcards to learn sentences from L2 to L1, e.g., from Japanese to English.

And Luca Lampariello uses bidirectional translation where he goes from L2 to L1 and translating back to L2, e.g., looking at a Polish text and translating it to English and translating it back to Polish.

In the end, I think translation is a very useful tool (as an aside: is this any different from traditional grammar-translation exercises in textbooks?). But for some reason, translation is vilified and immersion is prioritized almost to the exclusion of other activities.

That depends by whether you mean actual traditional grammar-translation or the expansive definition immersive teachers tend to use.

The most specific description I've seen of "grammar-translation" is for it to mean the old-school approach to reading Latin and Ancient Greek, where you study grammar in order to be able to read extracts from classical texts, and translation exercises are used to force the student to process the meaning and allow the teacher to diagnose problems for correction. My father did Latin at high school, and that was the technique still used then. A large part of his course involved reading Caesar's Campaigns in Gaul in the original.

I believe that this is the oldest and original meaning of the term, and though I can't be 100% sure, I'll run on the assumption that it is.

That means grammar-translation is based on the following principles.
  • L2->L1 translation only.
  • Use of authentic materials i.e. materials written in the target language for a general audience, not specifically for language learners.
  • Goal is developing comprehension skills.

Of course, not all activities in a GT classroom would necessarily follow this pattern, and you might well see short sentence translations as grammar practice before moving on to the big translation, but that's the "grammar" part , not the "translation" part.

Now of course these same techniques were applied to living languages too at one point, so French, German etc courses tended to focus on reading and understanding, with translation to L1, but that was well over a century ago. Some (bad) teachers may have continued doing things that way, but things did change.

Even the seemingly simple change of switching from L2->L1 translation to L1->L2 translation fundamentally alters the whole thing, because now we're working on production skills, and by definition the learner's translation is not authentic materials -- so that's all 3 principles of GT gone, so despite superficial similarity, it is no longer GT.

Of course, modern language classrooms (ie. classrooms of modern language, not *modern classrooms of language) didn't go fully L2->L1, and the early approach adapted from GT incorporated both L2->L1 and L1->L2, but this was still different enough from Latin and Greek GT that calling them the same thing is misleading.
Learning to understand language is a whole lot quicker than learning to produce it, so the rate with which new features are introduced is much, much slower. That means that the L2 passages presented to learners aren't going to be authentic texts any more, but rather texts written specifically for learners, consciously demonstrating the language points that the learner is expected to be able to reproduce.

There is a world of difference between "The cat sat on the mat. It was a beautiful cat. It had black hair and blue eyes." and Cicero's accusation of Catalina in the Roman Senate.

To describe these two things as the same thing based on the superficial similarity in activities is to miss the point of both.

But it doesn't stop there.

Mainstream modern language teaching evolved and developed significantly over the course of the 20th century, with massive swings in technique. There were radical trends like "natural methods" and "the audiolingual approach" that got a strong foothold in education at certain points in history, making the use of the term "mainstream teaching" meaningless, but this term is typically used to describe teaching methods that have a line of inheritance from GT... and these are often lumped into GT.

The classes I had at high school in the 90s were mostly bookwork, with limited speaking practice. Most of the written exercises were translation, but the translation exercises were mostly individual sentences, and passage translation (the translation of one or more paragraphs of text) was included sometimes, but not in all lessons. People try to call it GT, but the "translation" in "grammar translation" was the end goal -- learning to translate. If you don't end every unit trying to translate a passage, are you learning to translate? It's very different.

Last time I saw a high-school language textbook, there wasn't a single passage-translation exercise in it.

So we're left with a very broad spectrum of techniques that people clump under a single label, and you can define 3 very distinct points on the spectrum that demonstrate the folly of trying to describe them as being the same thing:
  • Reading complex authentic texts in L2 and rewriting them in L1, preceded by grammar activities to make it easier to read the texts.
  • Writing translations of simple L1 texts in L2, preceded by grammar activities to develop the skills needed to produce the translation.
  • Doing short grammar activities involving translation.
...and then you've got to take into account that even courses mostly matching the description of the 2nd or 3rd point will incorporate communicative techniques like questions and answers ("Do you like dogs?" "Yes, but I prefer cats." etc), so in effect describing any modern classroom as GT is focusing on the superficial features of some of the activities carried out in the classroom while ignoring some of the others.

TLDR:
Grammar Translation is a term that has been broadened and broadened to the point of meaningless. Nowadays, it doesn't describe a method of teaching, but is just used as an insult by people who believe in immersive education to bash any classroom that uses any amount of translation, however little.
9 x

guyome
Yellow Belt
Posts: 62
Joined: Wed Jan 01, 2020 1:41 pm
Languages: French (N)
x 119

Re: What is "grammar-translation"

Postby guyome » Fri Feb 14, 2020 12:51 pm

Grammar Translation is a term that has been broadened and broadened to the point of meaningless. Nowadays, it doesn't describe a method of teaching, but is just used as an insult by people who believe in immersive education to bash any classroom that uses any amount of translation, however little.
Agreed. This has been exactly my experience while discussing Latin teaching: GT is used as a kind of catch-all phrase to vilify anything that the person using those word thinks was/is wrong with methods (s)he disagrees with.
1 x

Online
Cainntear
Black Belt - 1st Dan
Posts: 1593
Joined: Thu Jul 30, 2015 11:04 am
Location: Scotland
Languages: English(N)
Advanced: French,Spanish, Scottish Gaelic
Intermediate: Italian, Catalan, Corsican
Basic: Welsh
Dabbling: Polish, Russian etc
x 3726
Contact:

Re: What is "grammar-translation"

Postby Cainntear » Fri Feb 14, 2020 12:58 pm

Oh, and on a related, but slightly tangential note:
pilot_2270 wrote:I find it interesting that this guy, Xiaomanyc, uses Anki flashcards to learn sentences from L1 to L2, e.g., from English to Chinese.

Meanwhile, Matt vs Japan uses Anki flashcards to learn sentences from L2 to L1, e.g., from Japanese to English.

Memorising examples is fundamentally no different from learning rules -- either way, you are memorising something that you can consciously use to allow you to construct a sentence later.

The only difference between memorising examples and memorising rules is that a lot of language rules are badly written or badly presented and confuse rather than clarify. A good example when presented with a rule can help clarify the explanation, and can help you recall the meaning of the rule without having to recall the wording of the rule.

If you add a sentence to an Anki deck knowing why it's constructed the way it is, even if you never read a formally defined "rule", revising the card means recalling that understanding and therefore the informal form of the rule.

Which doesn't make it a bad thing. If the formal rule confuses and causes a mental block, you're better off without it.


The problem is that when people propose something like the "massed sentence" approach, they often talk as though it's a process of osmosis -- that simply knowing what the sentences mean will lead to acquisition of grammar. That only works if you get a high variability so that you see the language in enough combinations for patterns to emerge... and at that point, you're talking about a lot of massed sentences with very little repetition of any given one, so surely you'd be just as well engaging in extensive input instead, possibly with parallel translations...?

Then of course there's people who change their advice on pragmatic grounds.

Several online proponents of massed sentences made it clear that the important step was picking your own sentences to make sure they're meaningful, but then find that other people report poor results or just give up because it's so time consuming. Next step? Create pre-written sentence banks that other people can buy. Now the card genuinely is just exposure -- there is no associated rule or context to recall along with it... and it doesn't work either, because the minimal set created for someone with those associations is simply too small a set for someone without them to learn/acquire/notice the patterns.
4 x


Return to “General Language Discussion”

Who is online

Users browsing this forum: jackb and 2 guests