63 Cards
This is a log of my feeble attempt at Thai.
Ideally I would try this on a language that I have no previous exposure, but I'll cut myself some slack and choose Thai this time. The last time I studied Thai was almost ten years ago, and there's a little personal motivation towards this language as well.
The rules are going to be mostly the same with the Spoiler Alert: Turkish, except that I probably won't post much about the grammar or even word usages. Not only that I have some idea how the language works already, Thai has very little morphology to speak of, and so it's more about getting the feel for different function words, serialization and speech particles.
There will be another little rule: the daily review count stays below 30.
Things may happen and sometimes it may go over 30, but when that happens I'll stop adding until the count goes down. The daily limit may get adjusted over time; I read Thai very slowly now, but as I grow more and more familiar with the script and the language, then the limit may get raised, but that's for the future. The general idea is that I don't want to spend too much time reviewing Thai on Anki.
As a result, the rate of deck growth will be necessarily much slower than with Turkish. I want Thai to remain as a side project where I can retreat to as a break in between my main project, that is, Turkish for now. I toyed around the idea and created templates for Middle Korean and Gothic decks, but they got scrapped again: I feel like a moron for not having thought about this, but dead languages offer little lighthearted escapist entertainment to be "breaks". And keeping the arbitrary ceiling of 30 will result in an extra side effect.
That is...
Well, I don't think going purely audiovisual is an efficient method. That was how I tried Thai at first, and I gave up and picked up textbooks. I want to try this again, going primarily audiovisual while building an Anki deck slowly -- Anki cards will be the anchor points where I solidify what I look up, but it will be supplementary while the project itself remains primarily audiovisual. If I listen to Thai without creating too many cards, this will happen by default, whereas if I let myself build a larger deck faster, then the project will end up being primarily Anki-driven, that is, text-driven. That is the idea. An idea is an idea, it may fail spectacularly but I need to try and see it.
The Anki cards and the listening materials may not even come from the same sources, but that's fine. 80/20 -- it doesn't matter where I mine sentences, high-frequency words will repeat themselves. But this is also due to my ineptitude and laziness - I can't type Thai as I listen, and finding Thai script for Thai audio is a bother (unless it's a song).
Why not just log the hours then? I don't think I can keep track of the hours I listen to Thai. That's hard and it's going to be inaccurate. And if I don't listen to one song because I don't want to bother logging those 3 minutes, well, that's a loss of the audio time. I'll keep track of the deck growth as the measure of my progress instead, since that is a much easier stat to keep track of.
Anki as the measure
Now might be the time to talk about this. Maybe I should have explained this on the Turkish thread, but when I started that thread, this was still an untested idea. I'm more certain about it now.
I don't worry about going past the beginner stage. In learning any language, the problem hits when we reach the cursed "intermediate" stage. We feel we're not making progress. We lose motivation, and it's a matter of willpower, determination and/or obligation. It may be many things, but what it stopped being a long time ago is -- fun.
I think this is a perceptual problem. Say, putting in the same amount of time and effort as a beginner and as an intermediate learner should result in the same amount of improvement, but we don't perceive it that way. Why?
Going from 10 to 20 is not a +10 increase. We perceive it as a 100% increase.
Going from 100 to 110 is not a +10 increase. We perceive it as a 10% increase.
Going from 500 to 510 is not a +10 increase. We perceive it as a 2% increase.
By the time we are on the intermediate plateau, the marginal improvement falls to near zero, and we feel like we're not making any progress despite trying hard. We just soldier through, and we eventually reach where we become more capable of reaping the benefits of using the language for the things we want. The process feels more enjoyable by this "advanced" point, though more often than not, we don't make through the intermediate no man's land in this trench warfare.
One of the ideas I had in my mind was to slap one number that is representative for my growth in the language to minimize the effects of the subjective logarithmic growth. If I can keep a visual reminder of the linear growth, I can maintain the sense of progress. I still feel doubts, there were (and are) moments where I felt I wasn't going anywhere with Turkish, but the increasing number does not lie. After having added and learned 100 more cards, I know that I know that much more Turkish, I know the effort was not in vain. By keeping a log, I can look back when the deck was smaller and be assured that I am moving forward, and Anki being a spaced repetition program adds to that confidence in a way that a printed book cannot.
The idea is the same. The Anki deck size is an arbitrary number. But it's a number that increases linearly in proportion with the effort, and it's a number I can see. It's the EXP that I can grind in the game of languages, the only difference here is that I'll grind just a bit more slowly with Thai.
But before getting to that fancy topic, there's something else I need to do.
Tutorial
However, in order to have 30 cards per day to review, I need to have added so many cards already, and given this is Thai, that already is a challenge. Yes, I'm talking about the goddamn writing system which I believe is the most needlessly complex alphabetic writing system on the planet. ("No but it's an abugida" -- Oh, shut up.)
Brushing up the Thai script took a while, and I had to go very slowly with this. Today is Day 25, and I have 63 cards, making it less than 3 cards per day added on average. I was honestly doubting if I could do this in the first place, really, even with some faint memories from the last decade, which is why it took so long to start the thread. After all, there is a limit to how much information I can hold with this cute, sexy but freaking bizarre script without overloading myself. Especially when a card looks like this:
Amen. That's not even all of it, it scrolls down even further. Though this is a particularly bad example, you get the idea. This card is still learnable because I know what it should say from Japanese. The initial sentence mining strategy will remain more or less the same with what I did to Turkish: i.e. being an assholistic Japanophile ignoring the language's proper culture and heritage.
Still I had to write the card multiple times, and pressing "Again" on Anki was not a rare occurrence at all. My Thai deck has the Interval modifier of 80% (the review intervals are 80% of the default), and I still had to press "Again" and "Hard" a lot of times. I did not add new cards every day, because creating one card could easily take half an hour at first, and dumping too many new words that I can't remember is a foolproof way to fail.
This Tutorial will last until I have a steady stream of ~30 cards per day, and I feel confident I can read Thai. (And I'll also need to learn to type Thai, that's another concern.) I will remove that transcription field on the Anki deck at the end of the Tutorial. I expect this stage to last another month or so. Let's see how this fares.