Language Journal: April 14th 2019 - The "trick" trap

I think it is a necessary evil when learning language that people advertise things as "a cool trick to learn French" or "the easy way to learn Japanese" or "hacking learning Kanji" and so on.

The truth is, on some level, the tricks and hacks are real. And on some level it is a festering pile of rubbish.

I learned Kanji using a variation of the RTK method. And, many people in the social groups I'm a part of do as well. But they are SO hung up on the particularities of these methods that they lose sight of reality. I've talked about this a bit, but I started seeing it in another form today as well (N+1 decks).

The problem with Heisig's method is EVERYONE. GETS. HUNG. UP. ON. THE. KEYWORDS. "How is a village different from a town?", "what's a decameron?" and so on.

The point of Heisig's method IS NOT to burn these keywords into your memory. The Kanji DO NOT MEAN these keywords. These are simply placeholder words. There absolutely is value in the fact that these placeholder words are related to the meanings of the real words. But, what is MOST important is identification. That you can tell the kanji you call "village" apart from "town" and apart from "decameron". Nothing more.

And then the rabbit holes people go down get even crazier like questioning why there are even two different Kanji for "town" and "city". It doesn't matter! There ARE. And, you're learning them, not only because there are 2, but because there are two in the list of general use Kanji. So you'll see them a LOT. Hence, you NEED to know them, regardless of whether YOU think it makes sense or not. Japanese people probably wonder why the heck we have "hamlets" and "townships". In short, English is a FAR worse offender. And it STILL DOESN'T MATTER.

In short, Heisig's method is GREAT. But, it isn't everything. And you A) Need to know what it IS good at and what it IS NOT and B) You need to "graduate" from it eventually. The Heisig method prioritizes recognizing Kanji so that you can accelerate your vocab acquisition. It ignores Kanji readings to this end. But, obviously, if you want to attain any reasonable level of proficiency in Japan, you need to start learning those readings from the vocab. And then, the understanding of the Kanji's meaning will REPLACE the keywords and through that practice you'll also be learning the readings (in theory).

You read that right. All of those keywords you committed to memory? Once you feel confident with them, the next step is to use that knowledge to obliterate those keywords and replace them with understanding.

THIS is why it doesn't matter what the difference between a city and a town is. You will understand the difference between the two Kanji with exposure. Even if you never actually learn the difference in meaning between those 2 English words. And this is why keywords are useful, but also a distraction.

You COULD replace the keywords with their numeric ranking in terms of how commonly used the Kanji is. But, frankly, keywords are easier to build mnemonics around than arbitrary #s. And keywords with at least some potential relevance to the vocab you'll be learning eventually is somewhat helpful.

My suggestion when learning though is, do one of two things when you hit synonyms or words you don't understand; choose one and number them (city1 and city2) or lookup real world vocab and remember it as the "machi" in 下町 which is the lower/older part of a town. In other words, emphasize that it is just a placeholder so you don't get hung upon it, or "shortcut the shortcut" where it isn't working and jump straight to some REAL application of the Kanji, complete with readings.


N+1 decks are a similar beast. A lot of people love decks where each card builds on what was already taught and each new card contains only 1 new term or concept. This is a fantastic way to acquire a lot of early knowledge. It feels good. It is quick. But then the same people try decks which don't do this and say that they can't tolerate it.

Guess what? Reality? It doesn't care if you can tolerate it or not. Your first steps out into the "real world" of  native language materials will be RIFE with vocab and possibly even grammar you've never seen. Being too dependent on heavily structured decks robs you of the ability to plan for such difficulties.

If you think "oh well, I just need more cards", then consider this. I'm over halfway through the Core 10k decks. I've started reading manga, I watch anime without subtitles and I've worked with numerous learning materials. I probably have a vocabulary of between 5-15k Japanese words... I still REGULARLY hit sentences where over the half of the sentence is legitimately new words. Sentences where there are words I know, but the placement in the sentence or the way they are used are foreign and I don't catch them, or I just can't remember them in the moment.

N+1 decks simply don't give you the exposure to overcome ANY of those problems. And it emphasizes wrote memorization over actually learning anything.

Keep in mind as well, a great many words are written in various forms. Some have multiple Kanji for the same meaning and reading. Some switch to Hiragana or Katakana at times. And, if we're talking listening instead of reading, it is even harder with differences in tone, pitch and timing.


I agree, structured decks are great early on. Just as I agree that Heisig's method used properly is good early on as well. But N+1 decks have more than one problem. And, one of them relates EXACTLY to this city/town problem. Your brain doesn't care at all that you want to learn all of the Kanji and learn them well. It recognizes that you're trying to learn them, and tries to help you "cheat".

I'd noticed this myself with kanji with keywords like "grove" and "forest", or "city" and "town". Eventually, I my brain picked up on Kanji I learned around the same time, that I also tended to get wrong. And so, I caught myself thinking one day "wait! saw that other Kanji I normally see around the same time as 'town' so this is probably 'town'". And it was.

See the problem? I wasn't actually identifying the Kanji based on the Kanji at all. And, sooner or later, the cards I fell back on to help me through? They fell out of sync. And some times I'd get both of the ones I screw up in the same review session and only get the other after screwing up the first one.

My brain had picked up on patterns to get me the answer I wanted, even though it was the manner in which I got the answer was more important.

So, I started doing new decks in random order and judging my progress based on the whole card. Randomizing alone doesn't alleviate the problem, it just changes which cards are in the pattern. But, it makes wrote memorization much harder because I often didn't know most of the card and would need to remember every word, how it was conjugated, etc...

I also pretty much had to do it because the problem was even more pronounced in vocab decks. The added context of the cards made it easier for my brain to remember which cards came up around the same time. And, the decks often had cards for both vocab and a sentence, and the vocab and sentence were always separated by a single day when done in order. So, if I wasn't sure but I had seen it's partner card recently, I could make a reasonable guess.

By randomizing AND forcing myself to tackle more than just the single target word, I was also training myself to spot the other words on the cards when they eventually came up out of context as well. Which is an essential skill. In the beginning it was hell. My accuracy shot down to like 40-60% for a long time. But, now, I get fewer wrong than I have new cards. And I'm better able to detect the words I learn down the road.

To cope I learned legitimate tricks to deciphering the language. I switched from memorizing vocab words as I did in the beginning to ACTUALLY doing what I should have; associating the words with their kanji and their readings. 

And, since I learned these additional "deciphering skills", I'll now, some times, even guess a totally new word and it's Kanji from audio alone based on the sounds it contains on occasion.

If I were just relying on N+1 type decks still, I suspect I'd be focusing on memorizing a single word at a time in the exact context it exists in. My comprehension of Kanji readings would be lower. My accuracy would suck. And for the same time and effort I'd be further behind in my studies.


I hope the message comes through. I'm not actually against these techniques. But, you really need to understand their purpose, along with their strengths and weaknesses so you can leverage them while they're still valuable and discard them before they start holding you back.

Heisig's method is something you should be abandoning entirely once you start getting deeper into vocab. N+1 decks should only be used while you still feel like you're at a foundational level. They have a place. And they are great tools. But, Heisig's method is easy to misunderstand, and N+1 decks are so efficient because they exploit your brains innate ability to cheat. Those cheats won't yield long term success after a point though.

Comments

Popular Posts