Over on the TESL-L mailing list, Mert Bland, Dean of Graduate Studies at CBU (That's Central Buganda University, not Christian Brothers University), is making his same tired claims about vocabulary teaching. One of my responses, which garnered a lot of positive response, is reprised here:
The folks who buy into the anti-memorisationist agenda tend to be the same who hold anti-translationist ideologies. Yet the arguments they marshal in their attempts to rationalise these extreme positions do not bear scrutiny and, to the extent that they are true, can usually be applied equally to learning vocabulary without the L1.
Take, for example, the recent claims by Mert:
"no concept in the the target language is a direct translation of a concept in the native language."
First of all, this is simply false. Admittedly, some words will not overlap well. However, to the extent that any two native speakers of a given language share a common understanding, the meaning of the vast majority of nouns, verbs, and adjectives will be the same in translation. For example, 'book' will have a one-to-one translation in almost every major natural language.
Anti-translationists will say "gotcha! What about book used as a verb, as in 'book him for murder'?"
The *noun* 'book' will have a one-to-one translation.
"Ha!" say the anti-translationists. "My concept of book includes electronic books, while that of pre-industrial hunter-gatherers in the Amazon rainforest does not." To which we calmly reply, "nor does that of my grandfather, though he is a native speaker of English."
Not only is the argument false, but even if one were to imagine it to be true, it is only relevant if we take the position that it is best to fully and accurately specify the meaning of a word when teaching/learning it. Yet, clearly, here, translation does a better job--not the straw-man single-word translation example being bandied about here in the form of English 'leg' to Japanese 'ashi', but the use of the L1 to translate AND explain the word. English synonyms and paraphrases will typically be only as good as, and likely worse than the L1, and are almost sure to be less comprehensible to many learners.
Mert goes on to argue:
We try to bifurcate the languages, completly separate retes. This can't be done if individual vocabulary items are tied together in the two concept clusters.
This takes a very naive view of the brain and attempts to make arguments based on ideas that even neurolinguistic experts are unclear about. It also conveniently ignores the fact that after the teacher spends 10 minutes gesticulating, explaining in English, and searching for props, the students will finally go, "Aha! He means (L1 word)." This activation of the L1 translation is almost certainly automatic and unavoidable in most cases, yet it is just what this teacher was trying to avoid.
Mert also makes the point that "A connection once made is hard to break; these things tend to fossilize."
If this is true, then it is a good reason to make sure the connection is the correct one by using the L1 and not a mistaken guess about what a given pantomime means. If it is false, then there is no reason to think that the gradual changes and refinements of a meaning that accrete with exposure to the word in multiple contexts will be any more difficult if one starts with an L1 connection than if one could somehow begin with a pure, non-linguistic concept.
(Please note that I'm not arguing that teachers should be experts in the L1s of all of their learners, just that they should allow the L1 its place when possible.)