So my conversation partner said something to me this week that made me think a little.

Her English was better than most Japanese I've met and although it's nowhere near perfect, she's very confident and can at least get her idea to anyone. During our meetings, she writes notes so that I can study them afterwards.

Today, she said "After you study these notes, I want you to throw them away. Just keep learning new things and throw them away as soon as you're done studying."

Is there anybody on this board that has taken a similar (or opposite) approach?
What do you think?

Thanks.

Old joke - The coolest thing about Alzeihmer's disease is you get to meet new people everyday.

I don't think she's suggesting to forget what you learn (hence having to learn it again). It's about not getting bogged down in what you learned at the cost of learning new things.

That's the great thing about an SRS. By it spacing out your reviews, it slowly "throws it away". And yes, one can put notes and reminders into an SRS. There's probably just no "answer" field.

Hmm, could one really call that an approach? Or perhapse it's just a rejection from established learning pratices in Japan......

Maybe its a sign she wants you to throw something else down on the table?

In all seriousness, I do something similar. But Nukemarine has got it I think. Her method is a somewhat old school approach though. That is to really memorise something and have it nailed in your head.

I think that's a valid approach. When I was learning English I never wrote down a single word or used a vocabulary list or tried to force myself to remember anything. It just happened naturally over time. If I forgot a word I'd look it up again.

I figured out the meaning of things, using a dictionary if necessary, and moved on to reading something else. I wasn't particularly immersed either. I had almost no access to any audio material and English was pretty much limited to what I could read on my computer screen until I felt ready to start reading novels.

Sometimes I wonder if I'm not just trying too hard with Japanese with all this time invested into SRSing sentences. Certainly repetition of the exact same sentences isn't optimal. It's sugar-coated dumb repetition. New material is much better.

However the SRS method has been very effective so far for me so I'm not going to give that up. But I think it's really important to never let the reviews get in the way of learning new things. Forgetting is normal and we should not be afraid of forgetting a lot as long as we replace that with even more new knowledge.

Codexus Wrote:I wasn't particularly immersed either. I had almost no access to any audio material and English was pretty much limited to what I could read on my computer screen until I felt ready to start reading novels.

I had to reread that because I always assumed you were a native English speaker.

chamcham Wrote:Today, she said "After you study these notes, I want you to throw them away. Just keep learning new things and throw them away as soon as you're done studying."

Reminds me of the old Lisztian, "Leave your dirty laundry at home," quote.

Codexus Wrote:Sometimes I wonder if I'm not just trying too hard with Japanese with all this time invested into SRSing sentences. Certainly repetition of the exact same sentences isn't optimal. It's sugar-coated dumb repetition. New material is much better.

I think this is where Khatzumoto's "regular SRS housecleaning" rants come from. If a sentence is too easy for you, or if you don't like it and you feel like it's wasting your time, throw it out! Kill it. If the words are important, you'll encounter them in another context some other time.

I think this is fantastic advice because it helps us to make our own personal "frequency lists". And learning a language is all about frequency. The more frequently used words you learn in the beginning, the more difficult and sparsely utilized words you'll be able to learn when the time comes.

I'm not sure I'd automatically add "too easy" to the kill list, though it'd certainly go into the do-not-add list. However, yes, on merest suspicion of time-wasting the sentence should be summarily executed without trial.

~J

woodwojr Wrote:I'm not sure I'd automatically add "too easy" to the kill list, though it'd certainly go into the do-not-add list. However, yes, on merest suspicion of time-wasting the sentence should be summarily executed without trial.

~J

Perhaps I should have phrased it, "If a sentence has *become* too easy for you..."

This is a fair description of my current approach, and why I dropped SRS entirely. If it's truly important you will see it again... if you don't, then you are wasting time in the SRS that could be spent in more productive ways..

mafried Wrote:This is a fair description of my current approach, and why I dropped SRS entirely. If it's truly important you will see it again... if you don't, then you are wasting time in the SRS that could be spent in more productive ways..

Whereas I picked up SRS because I'm at a level (passed JLPT 1kyuu) where a lot of the new words I see aren't going to get reinforced enough.

Look at the British National Corpus of 100 million words -- that's approximately equivalent to 1000 good-sized novels. Looking up some words at random:

slick -- 396
gravy -- 264
cheddar -- 145
economist -- 728
messenger -- 507

So, if you read 10 novels, on average you could expect to see 'economist' 7 times, 'messenger' 5 times, 'slick' 4 times. And the Japanese data may not be completely analogous to the English data, but like English, Japanese has a ridiculously voluminous vocabulary.

My point is, there are so many words out there -- and not even obscure words! -- that I'm not going to see often enough to reinforce them that well unless I put them in an SRS. It's only a tiny portion of my study time, because reading 10 novels is going to help my comprehension a lot more than just reviewing sentences over and over, but hopefully it will help my vocabulary retention a little.

The idea of an SRS isn't that it's essential, but that it's efficient. Sure, I would've hammered down the difference between a semigroup and a quasigroup just through encountering them eventually, but I credit the SRS with reducing the amount of time I spent sounding like an idiot.

~J

liosama Wrote:Maybe its a sign she wants you to throw something else down on the table?

At first she told me to hold the salami, but in the end, she couldn't resist.....

Haha....anyway, I think Nukemarine's first reply is more in line with what I was thinking.

I disagree with getting rid of too easy sentences. It takes almost no time to review those in the SRS. And the way the SRS works, probably after the point it becomes too easy the next review will be months or a year or more away.

I think its amazing that even if one were to have 10,000 accumulated sentences over the course of a year, you wouldn't really be studying that much per day, with each day less to study.

Think of it this way. As you become more fluent, and understand sentences better, you go through them faster. It's like reading a book. How long does it take you to read an understand a page of an average book in English? Look at how many sentences are in that sucker and I imagine once fluency is attained keeping up to speed with the SRS would take less than 15 minutes a day (if that)

i agree with jacf29, even words that might seem easy the 1st couple of times you see it often end up slipping my memory at some point along the line... just a quick glance at those old words every now and then keeps your mind sharp...

there are only soo many patterns to sentences out there. after a while it should eventually become more about the countless amount of vocabulary words you haven't yet cemented in your memory.

at that point sentences become as fast as vocabulary words for those you know well.

take this for example.

これはりんごです。
それはバナナです。
あれは魚です。

so easy it goes by as fast as a vocabulary word.

for now sentences like the following might be too hard for you.

彼は学校へ行くと言っていた。
彼女は銀行へ行くと言っていた。
お母さんはりんごを食べると言っていた。
(forgive me if my grammar is wrong)

But after you learn the grammatical patterns these sentences become just as easy as the first bunch of sentences. You just need to learn the new vocab words that come between the grammar patterns.

you are thinking of it like if i wanted to learn vocabs in my native language. since i have been using these sentence patterns for the last 20plus odd years of my life, there is no need to reinforce the grammar patterns.

Like hashiriya said, words get forgotten somewhere down the line unless using them often while reading etc. Same with sentence patterns. Unless you are at the stage where you are casually sitting down reading through stuff at close to native speed, I doubt you would get the necessary exposure to reinforce those grammar patterns to the point you dont need it in your SRS.

I've been doing this with my Drama deck. For those sentences it's more of "Can I understand this way too easily if I heard it without reading it?" which is key for me to just delete it. So a drama from four months ago that started out with 700 lines became 400 the first time through and is now down to about 200 sentences after 3 or 4 passes. This experience has just complimented what Khatz was saying about deleting sentences.

I just don't see myself deleting vocabulary cards. They're not there for the sentence, so the ease of the sentence is irrelevant. Same goes with grammar cards and kanji cards.

The purpose of the sentence is to show the proper context in which a word can be used... that's, I think, the reason vocab lists are frowned upon. That being said, I still advocate removing items that have become too easy. I mean, if, say, you're still reviewing a card that says "これは猫です。", then that is a waste of your time. I suppose it really depends on how good you are at figuring out what you *really* know, versus something that seems simple, but you may forget quickly.

IceCream Wrote:and at that point, isn't it better to use the srs only for words?
i mean, if i wanted to remember some difficult words in english, i don't think i'd enter them into anki in a sentence at all. it'd be superfluous, cos i'd already know how to use them in a sentence, i'd only need to know their meaning...
so, yeah, unless theres something about the sentence you dont really get, or can't use on your own, there's no point in having it in your srs, i think...

BINGO!

Let's be clear about our motivations. Are we learning Japanese to get by in daily life, be social, read manga and novels, understand dramas and anime? Or are we trying to get into Todai? (Hi, Yonosa!)

As someone said there's 100,000+ words in English (I think it's more in the millions if you count specialized, international vocabulary) and a similar number in Japanese. I don't know about you, but I don't know anywhere near 100,000 English words, nor do I have the desire to systematically study them. I did once, when I was studying for the GRE's but that's an exceptional circumstance. It's simply not an efficient use of my time to do so because I can always ask someone or look it up when encountered, at least when the meaning is not clear from the context (which rarely happens anyway outside of standardized or specialized tests).

SRS is a technology poorly adapted to language learning, at least for anyone who has set a reasonable goal like near-native fluency. If you're trying to win some memory contest or pass some obscure academic or legal exam, then yes, it may be advisable to study language with an SRS, but otherwise it's utility is too small to justify its use. You need about 5,000 words to achieve basic fluency (and yes SRS can be used for this task although again it has some serious deficiencies), but once you start doing things like reading novels and watching the news you will be encountering all of these words on a daily, or at least weekly basis. So why spend you precious study time reviewing them yet again? There was a post a while back titled 'do you have an exit strategy?' which all of us, myself included, should have taken a more seriously at the time. Don't go drinking the kool-aide.

With regards to deleting sentences, do you delete them after they are easy enough for you to understand? What about production?

They are many easy sentences out there that I just am unable to produce in actual conversation. Does it make sense to delete an "easy" sentence if you are unable to produce it(for sake of simplicity, let's say the sentence is very helpful and would be to your advantage to memorize).

Recognizing and understandいんg the meaning of a sentence is a lot easier than outputting that sentence when talking to people.

By production, I'm more referring to the general grammatical structure of the setnence and any relevant vocabulary (rather than saying the sentence word-for-word).

mafried Wrote:

IceCream Wrote:and at that point, isn't it better to use the srs only for words?
i mean, if i wanted to remember some difficult words in english, i don't think i'd enter them into anki in a sentence at all. it'd be superfluous, cos i'd already know how to use them in a sentence, i'd only need to know their meaning...
so, yeah, unless theres something about the sentence you dont really get, or can't use on your own, there's no point in having it in your srs, i think...

BINGO!

I disagree. I enter English differently from how I enter Japanese, admittedly, but for example, here are my cards for "boundedness":

[latex]Boundedness.\\

What does it mean for a sequence $a_n$ to possess this property?[/latex]
[latex]$$\exists M \geq 0 | \forall n \in N, |a_n| \leq M$$\\

What is this property of some sequence $a_n$ called?[/latex]

Even though the actual word is isolated (mostly to facilitate use as both an answer and a question), the second part of the prompt implicitly puts it in context—it's not just "boundedness" and nothing else, it's a property that can be possessed by a sequence. I see nothing but harm in removing that part—I'd still have the right idea in a very general sense, but I'd no longer know that I'm talking about the formal mathematical definition in relation to sequences.

The other half of what I see your argument being is that you're arguing that "important" is logically equivalent to "frequently encountered", which I decline to let stand without defense.

~J

That's not English, that's technical vocabulary specialized to a particular field of mathematics. If you were instead testing your knowledge of English, not math, "the right idea in a very general sense" is exactly what you'd be looking for.

To the second point, I'm not equating "important" to "frequently encountered." In fact, those rare words can sometimes be the ones which carry the critical meaning of a text. But when you consistently know at least 4 out of 5 words on a page, you can usually guess the meaning of the rest by context, even those "important" ones which carry most of the meaning. (EDIT: This is especially true of kanji-based languages, for obvious reasons.) Studying the simple, frequently occurring words will get you to the point of knowing 4 out of 5, at which time you can take the rest from context. My point is that you need only make a systematic study and review of the approx. 5,000 words it takes to get to that level, then switch to learning in context and stop your reviews. Those 5,000 words you will see often enough that reviewing them in an SRS is pointless. The remaining you will pick up from context (or when absolutely necessary a dictionary) and those which are important and necessary to remember will naturally select themselves into your long-term or active memory.

You might be tempted to speed this process along through systematic study and an SRS, as many people on this forum have, but in fact that would not be wise because it neglects the fact that we're dealing with a very long, and very imprecise tail (statistically speaking). Put differently, you will have 100,000+ words to choose from and no clear indication of which will actually be important until you encounter them in real-life usage. But even so you shouldn't learn every new word you encounter, because a good majority at this level are words you are unlikely to ever see again. Yes, there's still another 5,000 words to commit to active memory, and another 10,000 to passive memory, but which 15,000 words depends entirely on the contexts in which you choose or find yourself using the language, and is so personal that we cannot, say, tell you in advance which words you will need beyond that initial 5,000. Active, real-life usage is itself not just a good way to study and review (it is) but also a much better selector of the vocabulary and idioms you need to learn than SRS+your intuition could ever be. Don't trust your intuition on this.

I'm starting to ramble... which is part of why I haven't written that post yet. I'm seeing massive results (in French and German--Japanese is no longer important to me), and although I understand the how, it's quite hard to put into words exactly the why's.

mafried Wrote:That's not English, that's technical vocabulary specialized to a particular field of mathematics. If you were instead testing your knowledge of English, not math, "the right idea in a very general sense" is exactly what you'd be looking for.

I'm not sure I agree; it's certainly true that the context means that precision is more important, but if it's not important to know the difference between, say, "damp" and "moist" (and "the right idea in a very general sense" most certainly won't tell you) then why on earth have both words? Subtleties are maintained because they're important.

I reply only on this peripheral bit because it's late. More tomorrow, maybe.

~J