The question was recently raised on a teachers’ list

Is word-for-word translation (“glossing”) or translation into natural English “better” in the TPRS classroom?

(This translation being referred to is for comprehension checks; no one is talking about having kids translate long passages of text in writing.)

This kind of “translation” is intended as a quick brain dump. The purpose is to answer the question: did the student accurately comprehend the meaning of what was just said or read? So in one sense, it probably doesn’t matter whether the English the student outputs to answer the question is good or bad, as long as the teacher can be sure the meaning has been comprehended.

But the question involved a practice in class where the teacher requests that the kids gloss (state the meaning of each individual word or unit in English, regardless of correct English word order or the way English would naturally be expressed) rather than translate. The rationale is that students can better understand the word order of the unfamiliar target language this way. But we need to ask: do we want kids to understand word order (which is learning), or acquire it?

What will this sort of translation accomplish? It will let the teacher know whether the student knows the meaning of each individual word, but at the cost of switching the student’s mind over to doing just that: individually translating each individual word. This focuses the student on the word order or the individual words, not on the meaning of the language as a whole. It necessarily takes the student out of the zone of unconscious comprehension and thrusts him into an analytical place. Because it is not natural for a native English speaker to say “The balloons to her they-please” instead of “She likes balloons.”

If the student is going to output that sentence in Spanish, and the word order is not already “in the ear” (acquired), so that it falls out of the mouth naturally, then the student has to use the Monitor, or fall back on the native language. The Monitor isn’t only for writing, but it’s really only useful for writing, or maybe decoding reading, which can be done at one’s own pace. For speaking and listening, unless someone is a really unusually fast processor, knowing rules or mnemonics does not help, because how is there going to be time for a student to stop and say, “Okay, the thing to-the-person it-pleases” and then substitute those words? There isn’t. Not if there’s going to be a normal conversation going on.

But, people say, it is useful for writing. So, shouldn’t we push this sort of translation in class, to help with word order in writing? Maybe. But given the limited class time in most K-12 programs, I want the majority of writing I read to reflect what has been acquired. It’s easy for me to be lulled into a false sense of security if everyone can output something in writing, even if it’s through the use of non-acquired language scaffolded by rules or mnemonics. And I want to always know what the students have not yet acquired.

Comprehensible input is what drives acquisition. More comprehensible input is what’s needed before a student has acquired a particular structure, whether that’s word order, verb endings, or whatever. While CI is being provided, we do comprehension checks to see whether the comprehension is taking place correctly (if students match language with the wrong meaning, it’s not helping much.) But a comprehension check should be checking the comprehension of the language, not its parts.

When someone has truly comprehended something, it’s effortless to output it in the native language, because we can express whatever we like in our native language without effort (there may be some limitations, depending on information density and detail, but for the kind of language encountered in the K-12 TPRS classroom, I think it’s a safe assumption). It’s counterintuitive, but the kind of short translation (comprehension checks) we ask for in class is really NOT translation as most people think of it (going from one language to another). It is the sort of pure oral interpretation that comes from what interpreters call “detextualization”.

Detextualization is just a fancy term for isolating pure meaning — “removing” the text, which consists of the specific words from some particular language. So that process has an extra “step”: one language –> pure meaning –> another language. (As a point of trivia, this is why interpreters do not use shorthand to take notes; instead, we note down pure meaning, not in any particular language. That way, I can look at my notes, see meaning alone not tied to any language, and then speak whatever language I happen to have a fluent control of to express that meaning without effort.)

That’s exactly the sort of thing we’re looking for with a comprehension check. I don’t want a decoding of the sentence I’m checking. I want to know what meaning was immediately “gotten” by the student. That tells me a couple of things:

  1. The student has enough working memory in the target language to retain the meaning of the pieces unconsciously until a holistic meaning has been constructed in his brain. This is a big impediment to people who don’t have a firm grasp of vocabulary; they will often get “hung up” on an unknown or unfamiliar word that can’t be decoded immediately, and they then lose track of the language that is coming in after that word. So good working memory performance shows me something about how deeply vocabulary and grammar have been acquired.
  2. The student is not linking the two languages. There’s no need for that (nor does any standard recommend or require it), and it’s really detrimental to becoming proficient in a new language. Under old teaching methods, students had to do that, because they were constantly asked to output things in speech and writing that they had not yet acquired, and they could do nothing else but fall back on the native language.
  3. The student is focusing on meaning, not form. This indicates that the student is speaking/hearing/reading/writing the language, not about the language. Contrastive grammatical analysis supports activities where there’s time to engage the Monitor, but it doesn’t drive acquisition to the degree to which comprehensible input (focused on meaning) does.

As a pop-up, glossing by the teacher strikes me as just fine. The purpose of pop-ups is to focus attention on form, in very short bursts. But it seems to me the core purpose of comprehension checks is pure meaning, and glossing makes many things unnecessarily complicated as a result.