In the past, I've used SKIP codes to look up kanji, but I'm finding that as I work through RtK I can almost always recognize primitives even in kanji I don't know. With this realization, I'm finding multi-radical search in Wakan or PADict is usually a little faster for me than SKIP codes. And I could see it being a LOT faster if I only knew the stroke counts for the more complicated primitives/radicals.

It's like those aboriginal languages where they have numbers for 1, 2 and "many" - anything beyond 2-3 strokes I need to mentally draw the primitive to count the strokes to locate it in the lookup dialog, which slows me down dramatically. Repeatedly counting to 4 to find 日 is starting to drive me nuts, and while I can imagine that count eventually sticking just by repetition, I don't see that happening any time soon for 羊 or 隹.

So anyway, I'm considering adding a little side project to my kanji study of trying to memorize the stroke counts...but it's not obvious to me how best to do this. Numbers don't readily lend themselves to stories, but brute force rote memorization has never been my strong suit. Has anyone here made a systematic effort to memorize the stroke counts of radicals and/or primitives? If so, can you offer any advice on techniques that worked best for you?

Certainly don't claim to know stroke counts for all primitives off the top of my head but by virtue of going through RTK1, many of them will stick, e.g. as soon as I saw your examples of 羊 and 隹 it was almost as if I was looking at 6 and 8 respectively.

I don't think there's a need to memorize stroke counts, if you learn stroke order properly, you should be able to sketch them all and count the strokes using very little time or effort, even for the higher stroke counts.

if the two seconds it takes to draw 隹 is too slow, then you are in too big a hurry

And if certain stroke counts randomly stick in your memory, just think of it as a bonus

I think once you start looking up certain key radicals a lot in a dictionary, the stroke counts will stick in your mind with no extra effort.

johnzep Wrote:I don't think there's a need to memorize stroke counts

I agree, but I think it seems very cool know by heart the stroke orders. I remember 言, because I wrote ir a million times.

Well, so much for that idea. I still think knowing them would make dictionary lookups quicker, but if my senpai are doing well without actively trying to memorize, it cuts the urgency for me.

And I'm sure the most-used ones will eventually stick. In fact, the mere fact of publicly admitting I need to count to figure out where to find 日virtually guarantees I'll remember that one's four strokes.

When doing your reviews, just remember to count the strokes as you write each primative. If you can do it in Japanese that's even better!!

Yeah, I always count strokes when I review. Knowing the stroke counts 'by heart' isn't even an issue. A glance at mental picture alone produces the number almost instantly. It's a perk of RTK. And even complex radicals can be drawn in just a second.

[Katsuo]

For memory purposes it usually works better to swap abstract things for concrete ones. One method is to replace individual numbers by thing(s) you readily associate with that number.

Heisig sometimes uses this technique in RTK, e.g. he replaces "9" with "baseball".

You mentioned "turkey" 隹, which has eight strokes, so looking at that one:

The number eight, 8, is shaped like a pair of round-rimmed spectacles, so replace "eight" with "glasses". Now just imagine the turkey is wearing glasses.

Try applying this to some other eight-stroke radicals too:

金, 青 You have two pairs of glasses, one gold, one blue.
雨 When it rains you get lots of spots of rain on your glasses (and have to keep wiping them).
隶 Imagine a sieve made from pairs of glasses with the lenses removed.
長 Your glasses are too long and keep dropping off the end of your nose.

Etc.

Having said all that, I don't bother memorizing radical stroke counts too much myself. For multi-radical look-up, I usually type in radical names.

Katsuo, thanks for the radical list in the other thread, but even more thanks for offering a different approach to the problem than I'd considered.

It's almost like you applied a Fourier transform to my original question - going from "how do I remember the counts for each radical" to "here's an easy way to remember which radicals have a count of N". And I think that's an easier problem. Hmm.