This blog post was brought to you by the generosity of AJATT's patrons!

If you would like to support the continuing production of AJATT content, please consider making a monthly donation through Patreon.

Right there ↑ . Go on. Click on it. Patrons get goodies like early access to content (days, weeks, months and even YEARS before everyone else), mutlimedia stuff and other goodies!

Khatz, If You’re Fluent, Why Do You Still SRS?

“Dearest Khatzumoto,

If you’re fluent, why do you still use SRS? I mean, it’s not like you use SRS for English words 1“.

I get asked this question or some variation of it quite a lot, so I thought I’d answer it here. In twelve words:

“Before enlightenment: chop wood, carry water. After enlightenment: chop wood, carry water.”

When I first came to Japan, I started to feel a George W. Bushian sense of “mission accomplished”. So I stopped SRSing altogether.

What happened?

Well, my core, daily-use, 日常 vocabulary remained more or less constant.

But what I like to call my 教養 vocabulary, my “cultural literacy” vocabulary, like knowing what 魑魅魍魎 means…that fell apart. Also, knowledge of somewhat rarer kanji and their readings — 夕凪 — that fell away. Even certain expert terminology —  words like “輻輳” (a term you’ll come across in networking and queueing theory) — started to escape me.

And that’s just loss. My new acquisition and retention rate was more or less zero. So not only was I losing both active and passive vocabulary, I also wasn’t gaining ground 2; I wasn’t progressing; I wasn’t moving forward: I was shrinking. Whatever new words I learned or was taught simply did not stick.

In short, the knife still cut in the middle, but it was blunt and rusty along the edges. I wanted a beautiful, gleaming, dangerously sharp blade.

I stopped sucking at Japanese but I didn’t lose the ability to relapse into suckage. I didn’t lose the ability to forget. And I didn’t lose the will to and enjoyment of learning new words. It’s kind of a living testament to Carol Dweck’s work — skill is not fixed: it’s dynamic, alive, mutable.

So I started SRSing again and haven’t really looked back.

It’s  been suggested that there is such a thing as a law of diminishing returns with respect to learning vocabulary. Certainly, this is true with regard to frequency of usage. However:

  1. The socioeconomic rewards that a larger, deeper, more precise vocabulary brings far outweigh any frequency differential. At least in my experience. In fact…
  2. The socioeconomic rewards that any given word brings 3 are distributed in inverse proportion to the frequency of that word.
    • I’m not advocating the use of unnecessarily grandiloquent phrasing here; you’ve seen me talk; you know I speak and write quite plainly, regardless of the language in question. But when you need to know what 齲蝕, 領土的一體性 and 家事消費 are…you freaking need to know.

Pretty much all societies reward the articulate profusely and punish them only lightly (if at all). Those who communicate ideas and concepts in great volume and with great precision — those who have expressive power — enjoy many benefits. This is true not only of words but also of physical gestures — so, like, sport and dance.

In a very real sense, words are almost — perhaps not quite, but very nearly almost — a quantifiable economic asset, somewhat like stock or real estate 4: you don’t buy books for the paper, you buy them for the words. So words are literally assets. SRS, then, is a tool that helps you build and maintain your portfolio of assets.

Your nice, round, J-Lo assets 5.

When Pablo Casals reached 95, a young reporter asked him “Mr. Casals, you are 95 and the greatest cellist that ever lived. Why do you still practice six hours a day?” Mr. Casals answered, “Because I think I’m making progress.”
They Did Not Give Up


  1. Actually, lately, I do, but…that’s another story 😛
  2. Yeah, this ended up unintentionally sounding like a Yogi Berric tautology and for that I am sorry
  3. not that it makes practical sense to discuss this kind of thing at a single-word level, but for the sake of the fake mathematics
  4. Confucian (in mindset) proverb: “There are golden houses in books and there are beautiful girls in books“, quoted in Understanding Chinese Culture and Learning by Ting Wang of the University of Canberra
  5. Don’t act like you didn’t see this coming

  18 comments for “Khatz, If You’re Fluent, Why Do You Still SRS?

  1. ブライアン
    October 15, 2011 at 14:25

    For what it’s worth, fluent or not you’ve still probably not been exposed to Japanese as much, overall, as English.  (7-8 years versus, what… 20+? (Yes, years is a bad metric but in this case it’s hardcore immersion on both sides, so…(recursive parenthesis woo!)))  I would think there would come a time when the SRS is a lot less useful… but at that point your card intervals will exceed your expected lifespan anyway.  Which is kind of the point.

    • October 16, 2011 at 04:16

      Originally, I sort of disagreed with this, but now, I think you may have hit the nail on the head here. Khatz said that he went back to SRSing, which would imply his card intervals didn’t exceed his lifespan. If he had, I bet he wouldn’t have forgotten the things he already learned.

      • October 16, 2011 at 08:17

        I have to disagree. Sometimes, I forget words in my native language. They’re on the tip of my tongue. They’re very basic words, but I just can’t recall them. It may be what Khatz is referring to.

        • October 16, 2011 at 12:24

          This is where I had to think twice about how I felt about ブライアン’s comment. The phenomenon of having something on the tip of your tongue is a rather mysterious one, however, it’s not related to retention of vocabulary. (It’s more about your brain trying to make a connection, but can’t quite recall part of the connection fast enough.)
          What I took from Brian’s (ブライアン) comment is that if your SRS says that the card will appear again in 1.3 years, fluent or not, you need to review it in 1.3 years. Once it says, “next review in 94 years”, there’s no need to reset it, you can probably assume that you’ve burned that kernel of language deep enough into your mind.
          This is not to say that this will be the case for 100% of the words you learn, however, I think we could agree that it would be for 90 – 95%. This is also not to say that if you abandon a language, you’ll never forget it as long as everything you learned was SRSed to the point where the next time it’ll appear is after your death.
          Learning is life long, so you can call it a victory when an srs interval is *longer* than your *life*.

          • ブライアン
            October 16, 2011 at 22:11

            Not exactly what I was getting at.
            Forgetting/remembering any given word is pretty much irrelevant — it’s not a good metric of progress as a whole.  (Keep in mind that SRS aims for 90% retention, not 100%.)  I think there’s a baseline level of exposure to a language you have to achieve before it will maintain itself.  SRS is a catalyst that lets us get to that level a whole lot quicker than usual, but in return the SRS becomes necessary to retain that information.  Until you reach the baseline, at which point the environment takes over for your “reps” and the majority of your cards will be far in the future.  (FWIW, I think it would take at least a decade to actually get a card interval to 50+ years… it was exaggeration, any useful word will be encountered in the environment before that point anyway.)
            …That was long-winded and not well written.  Think of it this way.  If you’re more than a few dozen sentences into learning Japanese, you’re never going to need the SRS to remember the meaning of 何.  It’s common enough that the environment will remind you before the SRS gets a chance.  My point is that at some point most of your vocabulary will be like that.  But that’s after a long time.  (And doesn’t hold outside of an immersion environment.)  I’m suggesting Khatz wasn’t at that point when he went to Japan, and therefore needed the SRS to retain his fluency.
            (And none of this applies to new, freshly-learned words.  Those take repetition to learn no matter what, so you may as well use the SRS for it.)

            • ブライアン
              October 16, 2011 at 22:19

              I swear this had paragraph breaks when I posted it… sorry.

  2. Henry
    October 15, 2011 at 16:44

    Dear Khatz, I am using your articles and concepts as best aid in my learning English journey. This is the first time I have written to your blog during two years follow up your posts. 
      Now,here is my point. A language is like a car: the grammar is its engine,pronunciation is the wheels and vocabulary is its body.  If one mastered grammar and pronunciation,one can be fully functional in the language for years before reaching adult-level vocabulary acquisition.
     Although,the downside of poor vocab is poor reading/listening comprehension. Once grammar and pronunciation were mastered,there would be no need for SRS as learning vocab would progress constantly as long as you live in full immersion environment.  
    Please,feel free to comment on my ideas.

  3. hermanblue
    October 15, 2011 at 17:16

    I understand what Khat’s trying to say. However, I think I’m more for the approach of native speakers when I get fluent, that’s to keep reading as much as possible, to maintain a good vocabulary in the language. Not to be against SRS, but it always seems to be a bit not natural.

  4. October 16, 2011 at 07:22

    What I’ve learned from learning Japanese for these past 2 year+ is: you won’t really ever stop learning, even if you reach fluency or native level fluency. There is still much to be learned in any language, even your native language. I’ve come a long way from learning Japanese 2 years back, when I started. I know now, I must maintain doing my srs reps and immersion otherwise, I will slowly start to loose both passive vocabulary and active vocabulary. It’s true what they say, if you don’t use it, you’ll loose it. Same apples to languages. 
    I’ve actually come to love using anki and srsing. It keeps my mind fresh, learning new words, learning new contexts and experimenting with learning Japanese. I’m currently playing around with production type cards and MCD type cards and 2type monolingual definition cards(trying to find more ways of learning more and have fun reading my srs+it can be used at a source of double immersion). What I’m trying to say is, it doesn’t matter if you are fluent or not, you still have to keep learning+maintaining. I’m 2/3 of the way to my goal of fluency and I have no end in site for my learning.
    I’ve learned a lot from this blog and learning Japanese but one thing stands out the most: you have to have fun with learning.

    • Caren
      October 17, 2011 at 09:36

      I hate to be the spelling nazi, but I have to do this. It’s a disease, I know.

      Loose = not tight.
      Lose = to not have anymore. <- this is what you meant. Stop using “loose” when you mean this.

      Site = a place
      Sight = vision <- this is what you meant. Please use this instead.

      Now, I shall go punish myself for breaking the “no online nazying” streak I was on for over 4 months. It really does become a disease.

      • October 17, 2011 at 10:32

        No problem, I don’t mind being corrected. The more I get corrected, the less chance I’ll repeat the same mistake, over and over again. Being a grammar nazi isn’t that bad but some people can become annoyed by it. I typed this one pretty fast and now, when I look over it. I could have phrased what I said better.

  5. L
    October 17, 2011 at 16:49

    I am taking Khatz’s “getting used to Japanese” and Krashen’s idea of “comprehensible input” much more seriously.  For this reason, a program like anki is a misnomer.  I no longer use anki as a program to “remember” Japanese sentences or a thing to translate language facts from English to Japanese (which is the worst way to learn), Japanese to English, or even Japanese to Japanese now.  The back side of the card doesn’t matter anymore–it could be mangled English, mangled Japanese, perfect English; the back side just serves as a leverage to help you get familiar with the sentence on the front side.  The point of doing reps now is to test whether you are “familiar” with the grammar structure, the vocab, the nuance of the sentence, etc..  The words or sentences or grammar structure that you pull from your environment and put into anki are language facts that you don’t recognize in the beginning, but through concentrated, frequent exposure, you start to recognize it (frequent exposure to a language fact serves to solidify these facts into your memory bank).  Reading books and listening further exposes you to these formerly unrecognizable language facts and further solidifies these language facts in your head.  So if you treat anki as a tool to increase your opportunities to get familiar with the language, then it is not so unnatural, but an easy, concentrated, game-fied way to “study” Japanese.  If you combine this with your environment and using your environment as well as a tool to get familiar with your target language, you cannot help but get familiar with your target language.  And the more familiar you are wth your target language, the more natural it wil roll off your tongue when it comes to crunch time (communicating with native speakers). 

  6. L
    October 17, 2011 at 21:21

    Sorry wanted to add to the comment I wrote above.  This is in regards to inputting facts on the front and the backside.  The backside must capture the nuance of the sentence to facilitate the familiarization of the sentence.  It varies also whether you want to directly translate the unrecognizable words to English (or your L1).  In any case, the nuance-fact doesn’t have to be a perfect translation, nor does it have to be in English–this is what I mean by “it doesn’t matter what the back side is”.

  7. November 18, 2011 at 10:41

    Hello there
    I’ve been reading AJATT on and off for years, and the main reason I’ve finally got round to commenting is just to say that I’m a fan and have taken a lot of inspiration from it.
    Even though I live in Japan, have married into a Japanese family and have been studying Japanese for eight years, my Japanese still sucks, and I have to say that the process of trying to improve it can at times be utterly soul destroying – I think you have to be incredibly thick-skinned and an incurable optimist to get up every morning and battle your way through what seems like an unending series of misunderstandings and miscommunications, and I can quite understand why so many people simply give up.
    But anyway, I have a question that you’ve probably been asked 50,000 times already, but which has been bugging me for a while:
    If you were the first person to AJATT, where did your SRS sentences come from? Did you find them somewhere, or write them yourself, or as the original AJATT pioneer, did you not have SRS sentences at all, and just realised at a later date that they would have helped you, and would therefore help AJATT-ers of the future?
    Yours chicken-and-egg-ingly

    • ahndoruuu
      November 19, 2011 at 10:50

      He picked them out of his environment.  In fact, he still recommends doing this rather than using a pre-mined source such as MFSP, though of course it is seen as more “difficult” and you really have to have your environment going full-steam-ahead.

      • November 20, 2011 at 09:10

        なるほどね – thanks for clearing up that mystery for me! Makes perfect sense now that I come to think about it, and is something I’ve been trying to do recently myself – ie. rather than making a note of vocab or grammar points, listening for whole sentences or collocations instead – 『まじかよ!」「なんだそれ?』and so on.

Leave a Reply

Your email address will not be published. Required fields are marked *