On Dec 8, 6:57 am, Patrick Kenny <[EMAIL PROTECTED]> wrote:
> Greetings,
>
> I'm not sure what you're doing with Japanese, but I learned the writing
> system in about a year and a half using a combination of Mnemosyne and
> SuperMemo adding on average about 50 cards a day. The last eight months
> or so were exclusively in Mnemosyne, and I had no problems with the
> default algorithm.
Well, I'm not sure what aspect of the writing system you're talking
about, but.. I already know most of the Jouyou characters, and
probably quite a few over that. What I'm working on right now is
vocabulary. I can read Japanese pretty well at this point, but I still
have to refer to a dictionary more often than I'd like. This is
partially because my vocabulary isn't as large as I'd like it to be,
and partially because I've been reading some older works that use
characters to write words that are usually written in kana these
days.So I'm trying to learn as much vocabulary as possible, and learn
to read it in characters even if those characters are pretty obscure/
not used for that word these days. The problem here is that you're
never really done learning the writing system as long as you are still
learning vocabulary, which is what makes Japanese so difficult- even
in Chinese (at least in modern mandarin), once you've learned a
character you pretty much know how it is read in all situations (with
some exceptions, but generally.) In Japanese, of course, you have to
learn every vocabulary item de novo- in some cases knowing the on
readings of the characters is enough, but in others it isn't and you
just have to remember it by brute force- you can use mnemonics to
remember 2000 characters, but it's hard to see using a mnemonic system
to remember tens of thousands of vocab items (I do use some off the
cuff mnemonics to help with initial acquisition.)
I'm also conducting a bit of a personal experiment, trying to find the
limits of how quickly I can learn vocabulary. Since I'm trying to find
that out, it's not entirely surprising that I am not remembering
everything- this is why I said that I could be accused with some
justice of simply trying to learn too much at once. But I'm also
learning quite a bit a bit about where I, specifically, have problems
with this specific task. I wouldn't claim that what I'm learning about
that could be generalized to other people or other tasks.
Basically, I have yet to find a limit to how many words I can learn in
a day, and remember _on that day_. I think the most I've done in a day
must be close to 200, but I couldn't do that every day simply because
I don't have the time (and I'd likely have a nervous breakdown if I
did have the time and tried to do it ;) ). But when I am adding words
at that rate I have a real problem with retention. My recall is
generally pretty good after 1 or 2 days, but it drops significantly
after 5-7, and again after 11-15. I don't find this very surprising,
if for no other reason than that at that rate in that interval I will
have learned a lot of other words that are similar, and interfere.
But, it still makes sense to try to do as much as I can, even if my
recall will be lower than I'd like. There are a few reasons for this.
One is that some non-trivial percentage of the words I do just
"sticks"- in some (all too rare) cases so well that I'm pretty sure I
would remember the word in question in a year even if I didn't see it
once in that time. This drops with the number of words I do, but it
doesn't drop very quickly. So I am better off having 20% of 120 words
"just stick" than having 25% of 50 words do so. Another reason is that
I find it very useful to encounter words I'm in the process of
learning in running text. When I was first learning to read Japanese
most of the words I was learning could be counted on to pop up pretty
frequently, and that made learning them much easier- it provided a
very nice spacing effect. But I'm far enough down the word frequency
curve at this point that a lot of the words I'm learning don't show up
all that often. The more of them I have in a "half-learned" state the
more I will encounter. Sometimes running into a word in a text,
looking it up, and having a forehead-slapping moment is enough to make
that word "just stick".
Anyway, I'm certainly not saying that the default algorithm doesn't
work for me at all. I don't think it would be possible for me to learn
anywhere near as much as I'm learning without Mnemosyne. Even trying
to do something like a manual Leitner card system would just break
down under the overhead of managing the number of cards involved. I
will eventually get all of the words, even if some of them have to
make several trips through the not memorized pile. And I think the
default algorithm would likely be close to optimal for me if I were
learning, say, facts about geography, or even kanji in isolation.
But, I think it is more expensive to lapse than it is to review more
frequently, if you can limit the more frequent review to words you
find difficult. I actually think the default algorithm is likely very
good for me for words that have managed to successfully pass the 30
day mark- I will have a better feel for that in a few months. I need
more frequent review of some words up to that point to keep them from
lapsing though. I think I can achieve that effect to some degree by
"gaming" the algorithm a bit, and marking some things that should be
2s or 3s as 0s.
So I'm not suggesting that there's anything "wrong" with the default
scheduling algorithm, except insofar as I am not sure that one
algorithm can be optimal for all uses, particularly in the first few
repetitions. I would like more control over it, but I completely
understand why it's not desirable to add a bunch of switches and
buttons to Mnemosyne to control the scheduler. That would likely be
disastrous. I do think that the idea of having a plugin system that
allows a great deal of control over the scheduling algorithm would be
a great idea, and it sounds like that is planned for v2.0. I mentioned
this in the other thread, but I think it would be nice to be able to
have different schedulers apply to different categories.
I do think that the way that the hand size is limited is not ideal-
it's confusing, and it has some unfortunate effects in some cases. But
that's a separate issue, and I agree that it's not easy to say what
the ideal behavior would be. Anyway, I'm not trying to be overly
critical of Mnemosyne. It's a great program, and I'm finding it an
indispensable tool, so I greatly appreciate the work that's gone into
it. But I had also noticed the behavior Kenny was seeing, and that
aspect of the system struck me as a bit odd and inconsistent.
Thanks
Duncan
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"mnemosyne-proj-users" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at
http://groups.google.com/group/mnemosyne-proj-users?hl=en
-~----------~----~----~----~------~----~------~--~---
