## Learning to Read and Write Kanji Tests

This is a post that covers 17 writing tests over all the kanji I've learned, 29-30 kanji per day.

The concept behind the big test is to test 30 kanji per day, excluding the one's I've already tested to get an idea of how many kanji I know. Since I learned 496 kanji, 30 kanji per day should result in 17 tests. Because I didn't have the optimization on the first three tests, there are a few repeated kanji in the first tests (五 具 典 所 語 高). It looks like 17 will still be the number of tests. Without the optimization I'd have to do a lot more tests. Since each test takes ~15-30 minutes (drawing kanji is difficult), this optimization is necessary (as more tests are done, the number of repetition per test grows).

I was really upset as I found that the rate of recall is close to 50%. While this is substantially lower rate than when I was learning rows and testing daily, I believe that it is accurate. You can see how many I don't try on -- because any attempt at drawing is a waste of time if I cannot draw the whole kanji.

The Kanji from each test:

The ones I got wrong:

Unlike the previous blog posts where I go through each one I got wrong, these tests I will say that the ones I missed speak for themselves for the most part. Most of these are plainly forgotten. Many I made attempts on. Is this a condemnation of my method of learning? I feel like it is not. For learning 50% I could have done it quicker for sure. For learning 100% I could have used a method that is more effective at cementing the drawing, but I am confident that future use and learning will help me towards my goal of being able to draw kanji. What I notice pretty clearly is that the extra day or two to get 100% on the row test resulted in me not knowing the kanji. The kanji I got wrong on these tests pretty clearly coincides with those I got wrong on the tests the day or two before I completed the row. What this means to me is that I need more days to really cement the row, but how do I keep it longer than just a few days? Spaced repetition clearly is the answer here. Alas, my spaced repetition system is not working, so I need to create a new one.

One thing I learned from Japanese for Busy People that has stuck with me and has stewed is the similarity between "it did not hurt" and "I did not want to go".

It did not hurt: 痛くなかったです。
I did not want to go: 行きたくなかったです。

So the main problem here is that the only difference is the き in 行きたくなかったです。 Because き is often muddy and because these sentences clearly work for the same type of situation, clarity is incredibly important in order to understand which of these two it is. "It did not hurt" could be said of being punched or kicked. "I did not want to go" could be said of going to an event. So the two aren't that easy to mix up, or are they? What if you were punched or kicked at an event? Indeed a funny quote from Clerks is where one of the main characters says "I'm not even supposed to be here today." That's plenty about that topic.

As I go on, I realize that 50% is alright for memory. Assuming that 50% is an F is really an invention for students learning at a reasonable pace things that their society deems dependencies for the rest of their life. Many courses focus on objective right and wrong answers which are an awful judge of many attributes of a person's journey. But subjective has the opposite problem -- can a professor or a TA accurately judge subjective quality of a paper? For many papers, the answer is clearly yes. That isn't a great measure of the quality of this type of grading because it's just saying that the paper they ask students to write is excellent, good, average, satisfactory, poor, or awful. If you're curious, I've gotten pretty good grades writing terrible papers in 6-12, community college, and university. If you're happy about my writing skill now (I kinda doubt it, I'm just typing whatever comes to mind) you can probably thank my decades of trying to organize my thoughts so that they fit the model of thinking that I work in.

That is -- Problem → Data → Method → Tangent (depth) → Realign → Conclusion (Success / Fail / Other)

Not all things follow this silly pattern of trying to understand a problem (education), collecting data (my memories, what people have recently told me when I asked questions or were caught in a monologue), my personal thoughts, the ability to determine if I'm solving this problem or not, and a satisfying yes/no answer to the problem. For any problem complex enough, it's almost certainly going to result in Problem → Data incomplete → Method (money bottleneck) → Tangent (depth) → Realign (I did not solve anything, it's too difficult) → Conclusion (Other). But there's no reason we should be tackling difficult problems like it's some sort of superhero comic. We aren't Superman, Batman, Wonder Woman, or Hit Girl -- at least most of us aren't. So perhaps we can make the world a better place by starting smaller.

So back to education, we may consider 50% to be pretty weak, but if we consider how little time I actually invested in learning these kanji, it might actually be acceptable. For a few more hours I could actually retain more kanji I learned, but I have years before writing kanji is actually a valuable skill -- if it ever will be. Reading kanji, recalling which kanji, and conversing are the critical skill that make language in the 21st century.

So we still don't know how many kanji I recognize. A single test would give a pretty good guess, but I'm spending a lot more time learning Portuguese. I bet you didn't guess that. I started using Duolingo (which is surprisingly good) to learn Brazilian Portuguese and I'm doing well. I'm learning about an hour each day (which is about how much I was doing Japanese a few months ago) and it's going well. I'm still doing Japanese everyday, but it's at the point where I am losing as much as I try to gain each day -- the podcasts are over my head and the kanji I'm not actually learning anymore. I'm able to use the podcasts to increase my speed in conversation, but it's difficult. But now that I'm done with finding out which kanji I know how to write and which I don't, I can start learning the important ones. While testing I also informally learned which ones I don't know how to translate from English to Japanese, which is good (anyone without furigana). I also informally learned which ones I can't tell the reading from the kanji -- those I had to look up in the dictionary. I can say that those were pretty few, so I have a pretty good handle on translating from kanji to readings.

Is that all? Not for Japanese, but for today it is. I'm excited to post this blog, so this is the end.

Javantea out.