Anki is an Open-source flashcard application, with desktop and mobile versions, and a sync state function between devices. The idea is simply to outsource physical flashcards into an application which then automates “spaced repetition”. The purported result is better memorization in less active time studying.

The flashcard process goes like this: Attempt to get it right, then award yourself a fail, where you put it back in today’s pile, or give it a pass, where you put it forward a certain number of days. If this is the first time you’re seeing this card, it might only go forward 1 day. But, given that you succeed with recall next time you see the card, you’ll send it forward the previous interval times 2.5. This continues the next time you encounter the card, inflating the interval each time (today I just sent one 1.5 years into the future).  You can adjust that down by marking it “Hard” or forward by marking it “Easy”. Fail the card and it all starts again. I do buy into this process well enough. I can’t tell you how much time and effort I’ve wasted in college going over stuff I already know and got fatigued/bored by the time I get to the stuff I actually need reminding.

I’ve used Anki for a bunch of years now, but let’s get to that in a minute. Even though I’ve used it for a while, I recently expanded my usage due to reading an article about it, which I can’t even find now. Many articles say stuff like “Superpower” in the title. There’s a ton of hype online when people think about the possibilities of nearly effortlessly learning random things like all the bones in the human body or every last street name in San Francisco. Sure. But, the article did make me think that I can memorize more things I care about, like Linux/Vim/tmux commands. Perhaps going for an augmented memory at age 45 doesn’t show great timing. Or, perhaps, it shows perfect timing.

I first started using Anki closer to a decade ago while trying to brush up on and maintain some of my Chinese character knowledge. You can find shared decks made by others, so I went ahead and looked up one for the 3000 most common characters. Easy-peasy, right? It actually turned out to be a disaster.

The problems were 2-fold. First, and you’ll see this posted everywhere as a critique and caveat of Anki, is that Anki is not a learning platform. It’s just flashcards, a way to solidify learned content into memory. Contrast this with some of the language learning apps, which present content contextually if not just in some sort of explanatory way. Once understood, then flashcards are appropriate.

Next, whoever made the Chinese character decks I used (and now use) clearly has no expertise in learning or teaching Chinese. Whoever made the decks clearly took lists of Chinese characters and made 2 cards for each: one for character recognition and one for character recall. Character recall is the problem here, because simply putting the single-syllable romanized “word” isn’t enough to differentiate it from perhaps many other characters with the same sound.

I could go on about this, but it recalls something Tim once noted to me about Wikipedia’s problem. It’s a great and valuable tool for being free, but it’s often written by non-experts and therefore contains a good deal of inaccurate or imprecise info. Fortunately for me, I am someone who happens to have some expertise in Chinese language learning. I was able to just add context to the single-syllable “words” in the character recall decks. The upside to Anki not being a specially-curated enterprise app is you can always edit decks as you see fit.

What all this has to do with my previous post should be pretty clear. I promised my minimal effort plan to learn and retain the a bunch of historical sword morphology, but I had to explain Anki a bit first. Next post, I’ll go over creating an Anki deck for the Oakeshott Typology.