Getting Away from SRS (Spaced-Repition Software) for Vocabularly Learning

Ok. I admit it. I have a problem.

I use spaced-repition software (flashcard software) way too much when studying Chinese.

And, I always have. Ever since I started learning Chinese.

I wasted way too much time on Memrise, Anki, Skitter, Pleco Flashcards, the flashcard/testing feature on FluentU and even the Remember the Hanzi flashcard site.

Way too much time. Much of it useless.

Countless hours. Hours I’ll never get back.

I first fell in love with the notion of spaced repetition software back when I first heard about Memrise before I moved to Japan a little over 3 years ago. I used the Memrise site — basically just a website to aid with mnemonics and spaced repetition — to learn Japanese kana (both hiragana and katakana) in a little under 3 hours.

It felt like a god-send. I never would’ve learned two new alphabets that fast with out it.

I did pretty good then though, at least with that “deck”, or set, of “flashcards”. After I initially “learned” the two kana alphabets, I never reviewed that deck again. I didn’t have too. Kana was everywhere in Japan — I could reinforce it by reading menus at coffee shops, reading Dragonball Z manga, or anytime I looked up, or studied a new Japanese word or vocabulary list.

That was a good use of spaced-reption. I used it to be more efficient, and then no longer used it when it was no longer efficient. That was good. That time. For that set of cards. That “deck”.

After that, I wasn’t so good. And that’s when the inefficiency started. The countless hours of lost time.

I poured through the JLPT (Japanese Language Proficiency Test) vocabulary list on Memrise until I “planted all the plants”. When I wanted to move more into grammar and sentences, I actually wound up abandoning Memrise for Anki — which was more customizable in what you could do and test yourself on.

When I decided to move to Taiwan and began studying Mandarin in Aug of 2013. I started with Anki, and Anki flashcard decks in Chinese.

Then, I switched to Pleco, then to FluentU, then I read Remember the Hanzi (RTH) and used a website that was an SRS system built for helping you remember your RTH stories/characters. Then, I switched to Skitter.

Then, most-recently, after about a month of being back in Taiwan I realized something. (I came back to Taiwan in November 2015, after being gone for over a year and a half)

I had a deck of cards in there, in Pleco, from the last time I was in Taiwan. A deck of cards I had built from words I came across back then. I looked through that deck and 1.) realized I didn’t know most of those words, and 2.) couldn’t even think of what context I’d come across them in, and couldn’t even think of what context they would even be useful.

So, I just deleted that deck. “If they are useful words, I’ll come across them again.” I thought.

A smart move. A step in the right direction.

The next step in the right direction was the next thought process. “Should I move the card decks I currently have in Skitter over to Pleco? (from Remember the Hanzi, recent ChinesePod lessons, etc.)?” “No, if I need those words, they will surely come up again in my studies.”

But, you know what they say: two steps forward, three steps back.

And, before you know it, I was reaching a point where I had 400+ vocabulary reviews a day again.

And I was saying things to myself like “Alright, I’m going to listen to this new ChinesePod lesson, right after I get these reviews out of the way.”

Only I would go through the reviews — and get a substantially large percentage of words wrong!

That’s not how SRS is supposed to work. You shouldn’t be getting 50% of your words wrong. Some words were just not sticking. Why? I started to dig deeper.

Luckily, I was putting words into different categories based on where I came across them. And, one thing was noticeably clear: words I learned from my girlfriend and her mom were sticking. I learned those words quickly, and tended to know them better when they came up in reviews. Words from other sources (reading material, ChinesePod lessons, even sessions with iTalki teachers) — not so much. Some of the words were sticking — but a lot of them weren’t.

Why is this? The biggest difference is that the words that were sticking were the words that were useful. Ie: they were the words I actually used after learning them. Because, they were words I actually learned. Words I learned in context, understand how to use in context, and then later actually used in similar context.

Other words weren’t. They were just words. Words from a list. A list my iTalki teacher gave me at the end of a session. Or a list of words from a chapter in a book. Even, if I was a list I made myself. It’s still just a list. And, a list is a devoid of a real context.

A vocabulary list is not a language.

So these lists were quickly becoming similar to the old list I found in Pleco, from two years ago. Lists of words, where I couldn’t even remember the context in which I added the words.

So, then, I abandoned SRS. For two weeks. I still studied Chinese, still spoke Chinese everyday, still did iTalki lessons, still (tried to) read the newspaper. I even still looked up words in Pleco and added them to lists.

I just didn’t study the lists. I didn’t do any flashcard reviews.

And what happened? I forgot everything and my Chinese atrophied to nothingness.

No, not really. Not much happened. My Chinese actually still probably improved. I still learned new words. Remembered some of them, even.

All that really happened was there was a couple of times, maybe three, maybe four. Four times, max, in two weeks, where I heard a word or came across a word in context where I felt like I know that word, but can’t remember what it is. Maybe if I had reviewed my flashcards I would’ve known those three or four words better.

Maybe not. Maybe I still would’ve struggled with them. It’s a toss up, really.

So, I’m getting off of SRS. Getting away from it, really.

Not entirely. It still has it’s use. One particular use case I’ve found, is to review words before and after a session. For example, I’ve noticed that if a review the vocab from a ChinesePod lesson, before listening to the lesson — I’m able to follow along with that whole particularly ChinesePod lesson much better than if I don’t.

So, yeah, SRS is useful in small limited doses.

But, SRS is not language learning.

(Also, see Hacking Chinese’s great post on this: If you think spaced repetition software is a panacea you are wrong)