tadoku and the forgetting plateau

SRS systems operate on a more or less exponential curve; every time you pass a card, the next review is scheduled at a time in the future that is the previous interval multiplied by some calculated constant. And if you pass the card every time with the same difficulty ranking, the review schedule will continue along the same formula until the interval reaches into multiple years, multiple decades if you bother with SRS long enough.

The trouble is, I don’t think that actually works. Yes for fairly short time periods it does work; up to a few months. But there comes a time when a card that you haven’t seen (and let’s limit this to simple vocabulary or kanji cards for discussion’s sake) outside the SRS system will be forgotten no matter what the elves inside the black box say (that’s how computers work right?).

I began to notice this with my vocabulary deck (since abandoned), which I kept up for probably close to two years. I also definitely notice it with my kanji recognition deck using the RTK keywords – there are some pretty rare characters in there. Take a look at the lifetime stats for this deck:

写真 2013-09-03 9 52 18 PM

You can see that while for both young and mature cards the “good” button is the large majority, the overall percentage correct is significantly lower for mature cards; as well, the “hard” button was proportionately used a lot more, nearly double in fact. I doubt I’m the only one whose charts will look like this. And the reason for that is exclusively, I believe, that the standard SRS curve breaks down over multiple years; that no matter what the numbers say, if you haven’t seen a kanji or a word for a year or more you’re very likely not to remember it.

And as to what’s to be done about this problem, if in fact it is a problem? You might quite persuasively argue that if you haven’t seen a word for two years you likely don’t need to know it. But that depends on whether you’re sufficiently engaged with the language or not. For example I don’t really do anything at all with German any more so there are many very basic words that I don’t know, but if I were actively reading, or conversing, or whatever, it would very quickly become apparent that I did in fact need them. So if mastery of the language is your objective then yes, this is a problem – and the solution I completely unsurprisingly postulate is simply this, read more! The algorithm might be a bit off but if you see the word even once or twice in between that two, three, five year review interval, you’re certain to remember it; and the very best way to do that is to maximize your time actively engaged with the language; and the very best way to do that is to read more.

Additionally, I would caution (voice of experience – hey, it’s all I’ve got to stand on here) against adding very rare words or characters to your decks in the first place. Sure it might be a cool word, but if it shows up in one spot in one book and you aren’t likely to see it elsewhere for years, just enjoy it in the moment and then let it go. Once you’re more advanced and more likely to see it a bit more frequently, then feel free to add it, if you still think you need to. And of course knowing which words fall into this category only comes from doing this wrong multiple times, so maybe just ignore this paragraph.

Advertisements

unlimited tool works

Go to a photography forum and most of the discussion is usually about cameras.  Hang out with motorcyclists and most of the discussion is about bikes.  And I’d venture to say that in a samurai dojo, there was a great deal of talk about swords.

Likewise, in groups of people learning languages, the dominant topic is tools: which books, or no books at all; whether to SRS or not; what style of SRS cards; immersion versus intensive study; dictionaries, programs, apps, and on and on.

It could be considered that all this is much of a waste of time. The true expert thinks about his practise and his technique, and not much about whether the camera or motorcycle or sword he is currently using couldn’t perhaps be replaced by a better model, or upgraded somehow.  However, I can’t say I fully agree with this.  How many of us are at that level?  How long did it take Kiyonari to master the track, or Musashi to master swordsmanship, and how much experimentation with the equipment took place before it ceased to be a consideration?

There’s one sword that fits you and only you perfectly.  Don’t get attached to the old ones, and don’t be afraid to try new ones, until you find it.

Unlimited Blade Works

don’t fear the void

Having recently plucked up the courage (or more accurately, resolved to put forth whatever effort was required) to read whatever Japanese text I liked whether or not it had furigana, I have found firstly that it isn’t all that hard and secondly that it has some real benefits.  I probably know 300 or 400 kanji readings, at a guess, just for perspective – as often as not incomplete, as well; just kun-reading or maybe just one on-reading.

This started when I began my second reading SRS deck, which is four-panel manga.  Azumanga Daioh was the choice, for a few reasons; I loved the anime, I already had the books, it was by the same author as Yotsuba& which I also enjoyed immensely, and it didn’t seem very difficult.  (It actually has some unexpectedly uncommon kanji, I’ve found, but still it isn’t hard.  Grammar is as simple as you’d expect in a four-panel.)  Azumanga Daioh does not have furigana, and that was another reason I picked it for this deck.  If my intention was to reread the text as often as needed in order to remember all the words whether they have kanji or not, furigana would simply be a distraction, and a possibly deceptive one at that.

Then I also began the second volume, just reading through it, not SRSing it, in a feeble attempt to get my pagecount for this month’s ReadMOD out of the “embarrassing” zone.  (It’s as I expected a little difficult when most of my reading time is spent on the SRS decks, which I’m not counting.)

I found it quite easier than I expected.  I know enough kanji readings to input an unknown word as “this kanji is the first kanji of this word I know, and then the next one is the first of this other word” much of the time.  When that fails, either using the SKIP input in Kotoba! on my iPhone, or the IME pad and Tangorin on the computer, gets the job done infallibly and doesn’t take very long.

The small amount of additional effort, though, seems to trigger something in the brain that aids memory.  If you have furigana, you just glance up and there’s your reading, no extra time or effort involved, and the kanji barely registers.  But if you have to think about the kanji, consider where else it’s used, or count the strokes for a SKIP input, or even draw it on the IME pad, there is a great deal more involvement.  Maybe it could be compared to sketching a landscape as opposed to taking a snapshot of it.  Both will help you remember, but putting in the time and paying attention to every part of the scene will help you remember much longer and in more detail.

instinctive srs

The other day it occurred to me that I’ve been using SRS my whole life to learn language.   Chances are, you have too.

Think for a moment about what the average kid newly fascinated by books does, without ever having heard of Anki.   Show her a new book and it is immediately devoured, with frequent questions being asked of the parents.   But it doesn’t end at the one reading; the book is read over and over and over, several times a day perhaps, then less often as other books show up, then whenever she remembers it, until the time comes that she has outgrown it entirely.  This process continues (for the more bookish sort of kid like myself) well into the teen years – I probably read Treasure Island a dozen times, for example.  And the last few times I read it were further and further apart.   I would suggest that this is a typical, normal pattern; and is the mind’s way of instinctively reinforcing what it has learned.   While there might not be a conscious moment of thinking “ah, this memory is getting fuzzy, better reread”, I believe that this is actually what is going on subconsciously; although, if you asked the child in question why she hadn’t read such a book lately, she’d just say she was “tired of it”.   At that point, the reinforcement that comes from randomly seeing the words in the immersive native-language environment is good enough to keep the memory alive.

Using an SRS program can sometimes give the illusion that a memory has a defined point, precisely calculated by the computer, where the fact is deleted from the drive.   Of course, that isn’t actually the case; rather, they get a little fuzzier every day until at last they can’t be recalled at all.  Since the precise date of being reminded isn’t nearly as critical as the program suggests, it might well be that this sort of instinctive SRS is just as efficient as the more regulated and high-tech variety.

Adults, you might think, don’t tend to come back to books or other media again and again, and therefore it might not work too well to depend on this sort of instinct.   That may well be true mostly, but there’s one area where nearly everyone follows this pattern throughout their lives, and that is music.  So here’s your homework (and mine): take the newest Japanese songs that are in heavy rotation on your playlist, get the lyrics (maybe from jpopasia or goo), and make sure you understand them fully.  This isn’t extensive reading where you can skip over things; you’d want to get every word down.  Then, just listen to music as you normally do, no need to enter the whole song into Anki.   I suspect that a few months down the road, when you think of listening to that album again after it’s been collecting dust for a while, you’ll still remember everything you learned.

it’s simple math time again

The scholar who studies for four hours at 80% efficiency is far ahead of the one who tinkers with his method for three hours and then studies for one hour at 95% efficiency.

Of course this is overstated; but as much as Khatz and many others (rightly) encourage experimentation, there comes a point where your method is Good Enough.  And then it’s time to stop worrying about how you’re studying, and just get on with it.

fluency is not the goal

In the society at large (which, be it said, I am not usually a part of), when it becomes known that one is learning a foreign language, generally the first question one receives is “so are you fluent yet?”

And, unfortunately, this question is usually taken seriously and answered in all earnestness, whether “yes” or “not yet”.

I say unfortunately, because this matter of fluency is both something of a myth, and – I was about to say, seldom anyone’s actual goal, but the definitions of “fluency” vary so drastically that it can’t even be determined honestly whether it is anyone’s goal or not.  You might think it’s your goal, and then you might think you’ve reached it, and then the next person might come along and think you have very far still to go.  Or perhaps that you are far beyond what they would consider fluent.

Some would say that fluency means that you are, linguistically, in every way indistinguishable from a native.  If your accent is notably unusual*, or if you commit the occasional amusing error**, you are not fluent, they say.  On the other end of the spectrum some say that if you can comfortably get by in ordinary daily life without getting too lost or putting too much additional wear on the pages of your dictionary***, there you go, fluent.

And then let’s say some day you arrive at the point of fluency that you set out towards – will there still be more to learn?  Certainly yes, so how will you know when you have “arrived”?

I will have to disappoint anyone hoping that I will propose my own definition.  I just don’t think it matters, at all.  I would suggest that learners rather focus on what they would like to do in the language; not some arbitrary definition that no two people have ever agreed on.

For my part, I would like to be able to read Japanese as easily as I read English.  That’s pretty much it.  Will I be fluent? I’m not even going to answer that.

*natives have accents too

**natives do this all the time

***natives look up words too

toolset refinement

I was once amusingly characterized as mistaking the acquisition of learning tools for actual learning.  In fact, it almost seemed like a fair point at the time.  However, I’ve noticed that my toolbox is becoming ever more lightweight as time goes on.

I started with a few textbooks and reference books, most of which I no longer use.  Also, at one point near the beginning, I had sixteen apps on my iPhone relating to Japanese study.  Now I use three.  As for websites, the sites I use have also gotten fewer in number, as have the blogs I read.  It’s like an apprentice woodworker, who might start with a dozen planes in his toolbox but eventually finds he never uses ten of them.

All that notwithstanding, I don’t think there were any mistakes in that approach.  Some of these tools, like the excellent Human Japanese app (an introductory text with quizzes and games), were only useful at certain times.  When they were useful, I got a great deal of use out of them, but they have been outgrown.  Others had duplicate functionality, or near to it – and how would I know which I liked had I not tried both?  I don’t think language learners should ever be concerned about getting too many resources, too many tools.  You don’t know what will click with you until you try it, so if it looks even remotely useful, by all means acquire it and give it a go.  If it wasn’t right for you, there’s an easy way to tell – it’ll collect dust.  And if it was, it just might become indispensable, at least for a while.

I think towards the end of one’s language learning curve all you’d need would be something to read, something to review, and a dictionary.  And once the language is mastered, you’d have your library and nothing else.