Thanks! I couldn't stop until I'd completely read both articles. They were very insightful, and I share bananiel's curiosity about what implications they might have for learning languages with ideographic writing systems like Japanese.Bumping this cause it needs to be seen more.
The main point of the first article is that all languages divide the world differently, and therefore words in one language never have exact equivalents in another. All the advice about translation, the use of dictionaries, learning from context, etc. flows from this basic premise. The second article reinforces the first, only disagreeing about how far these ideas should be applied; ie. to what extent one should avoid translation and learning vocabulary from dictionaries.
These articles strongly advocate the immersion approach to language learning, and give examples of the superior benefits of simply reading in the target language to build vocabulary and an intuitive feel for the grammar. I've read other articles about people doing this with great success in European languages, but I've yet to see someone say they've applied this approach to reading Japanese.
But now that I think about it, is reading Japanese really so different from reading Spanish or French? When you see a new word in kanji, unless there is furigana, you are not sure how to pronounce it, but you can say the same about unfamiliar words in French and Spanish, since their spelling systems are hardly phonetic. Alphabets *seem* less threatening because at least you can make a good guess about how the word is pronounced based on past experience. But is Japanese really any different? If you've seen those kanji before, you can take a guess about how they would be read, and be right a good percentage of the time. Either way, you still need to look up the correct pronunciation to be sure.
If anything, Japanese is probably easier to read for contextual meaning than French or Spanish because the kanji make it graphically obvious what the parts of each word literally mean. A foreign student of English who has studied for several years may not know that "hydraulic-power" has to do with "water", but even a first year Japanese student realizes what 水力 means, even if he might not recognize the word when spoken.
If the authors of these articles were advising Japanese language students, I think they would probably tell them it is perfectly alright to use a dictionary to look up the pronunciations of words that don't have furigana given. In fact, you certainly must do this if you expect to be able to recognize the word again when you hear it. And likewise, definately DO look up the correct stroke order for a kanji if you are unsure you are writing it correctly. But definately do NOT memorize long lists of kanji with their English equivalents.
I think this also applies to the popular James Heisig approach of learning an English reading for all the characters first, then learning the Japanese readings later through context. This extra step of translating kanji into English would almost certainly be viewed as a big waste of time by the authors of the two articles.