I thought about attaching this to the AJATT thread, but whatever. A while ago while still working out my own theory of how I learn best, I found this site to be a very lucid and balanced overview of different theories of second-language acquisition: http://www.timothyjpmason.com/WebPages/L...Review.htm

This is an addendum I guess to my earlier comment: http://forum.koohii.com/showthread.php?p...7#pid38977 (I think for SRSers, this is the most important link there: http://www.supermemo.com/articles/stability.htm)

I'll throw in this overview on Baddeley as well (who updated his model in 2000 or so to include the 'episodic buffer'): http://en.wikipedia.org/wiki/Baddeley's_model_of_working_memory

There's also Pinker's The Language Instinct and its opposite, Sampson's Educating Eve.

I wrote about Krashen here.

Quote:Acquisition-Learning hypothesis
The Acquisition-Learning hypothesis basically states that there's a difference between learning a language and acquiring a language. Learning a language is to learn about the language, to be aware of grammar rules and consciously think about the language. Whereas language acquisition is a subconscious process similar to the way children learn and requires learning in the form of meaningful communication. Krashen argues that acquisition is more important that learning.

Monitor hypothesis
The Monitor hypothesis tells us that what is learned 'about' the language will only be useful when the learner has time to carefully edit and parse what they've said or written. In a sense, what's learned because they're 'monitor' when it can. Krashen notes two different types of learners, those who are "under-users" and those who are "over-users".

Natural Order hypothesis
The Natural Order hypothesis states that there is a 'natural' order in which grammar structures are acquired. That means despite how grammar is introduced to the learner, they're only going to be acquired in their 'natural' order (remember that there is a difference between acquisition and learning).

Input hypothesis
The Input hypothesis concerns itself purely with the acquisition process. It states that the learner will only learn 'i+1'. That is, if the learner is at level 'i', they will only learn when they are exposed to comprehensible input that is slightly beyond they're level.

Affective Filter hypothesis
The Affective Filter basically says that negative emotions such as self-doubt, anxiety, boredom, all serve to get between the learner and the language. When the learner is plagued with negative feelings, they tend to either prevent effective learning, or prevent time spent with the language.

Thanks alyks, a useful and smoothly written summary. However, I think it's too subjective and uncritical, at least for the purposes of providing an overview to theories on language learning. I'm well aware of many blogs posting their thoughts on Krashen, but few of them seem to parallel his similarities and differences to other theories and in an objective way that isn't arguing about the Krashen revolution, et cetera.

One important thing that the your summary is overlooking: a learned language develops into an acquired language with use, and reaching the goal is faster when coupled with learning. It is making a false dichotomy. We don't have to JUST acquire language or JUST learn language. Most things in life work better with a balance.

A lot of people make a straw-man argument out of traditional book-learning. "I don't think about English grammar rules in my head when I speak (native) English, so why should I bother studying Japanese grammar rules?" The answer is that once you learn grammar rules and USE them, they become internalized and you don't think about them when you speak. It's the same thing as RTK. We learn the kanji using stories and mnemonics, but we don't keep using them forever.

The only difference with learning purely from exposure (sentence method etc) is that you need to figure out the rules for yourself (and likely miss out on exceptions and nuances). You are just making more work for yourself when something is already laid out for you. Maybe it just gives the appearance of being more effective since while you're getting those sentences you are applying your Japanese. Again, a false dichotomy. If you book-learn grammar it doesn't preclude you from reading at the same time.

Natural order hypothesis goes against my personal experience with second language acquisition, but it is a known phenomenon with babies learning their first language. I don't think it applies much to second language learners since we don't have to learn the advanced concept the grammar describes, we only need to relate it to our native language.

Jarvik7 Wrote:A lot of people make a straw-man argument out of traditional book-learning. "I don't think about English grammar rules in my head when I speak (native) English, so why should I bother studying Japanese grammar rules?" The answer is that once you learn grammar rules and USE them, they become internalized and you don't think about them when you speak. It's the same thing as RTK. We learn the kanji using stories and mnemonics, but we don't keep using them forever.

The way I see it, there's no point to learning grammar rules because you might as well go straight to acquiring them from meaningful input. Which leads me to the second point.

Jarvik7 Wrote:The only difference with learning purely from exposure (sentence method etc) is that you need to figure out the rules for yourself (and likely miss out on exceptions and nuances). You are just making more work for yourself when something is already laid out for you.

The point isn't to try and "figure out the rules" the point is to try and "figure out what people say". There's a huge difference. If you try and read a lot and figure out that 全然 only goes with a negative, then I can understand your arguments. But when you try and figure out what somebody is saying when you hear something like "全然好きじゃない!" in a story, then it's different. The reason he argues output before input is because of this, you imitate what you've seen used a million times and you don't have to worry about exceptions and such. It's that you learn from imitation and communication and output based situation and context.

Example, would you find it easy to list of random words you know in your native language off the top of your head without any kind of stimulus? Kinda hard. But if I told you to list words started with the topic of computers, it would be much easier. "Computer, has a chassis, motherboard, CPU, central processing unit, center, keyboard, keys, keyhole, doorknob, room, house, ..." simply because in our native language words and sentence patterns are associated with context and real life.

I agree with both Krashen and his critics. I don't think there's any real difference once one gets down to effective learning, individuals will decide the merit of dividing or maintaining 'learning' and 'acquisition'.

I think the models and general structure of these theories are important only in attempting to apply them in schools and whatnot, as well as in critical analysis that needs empirical support rather than just hypotheses to support these models and applications.

As self-studiers, we have the luxury of just experimenting for ourselves depending on our time and energy. Personally, in terms of finding a common set of advice for language autodidacts, I feel that a wiki or something would be better than a collection of comments and posts that simply advocate one way or another.

I actually wrote an essay about my own method and theory, but realized it's kind of pointless to share. ;p

Thanks Nestor. I appreciate overviews.

alyks Wrote:The way I see it, there's no point to learning grammar rules because you might as well go straight to acquiring them from meaningful input.

No point? The idea is that while adults can choose to learn like an infant through immersion only, many believe that adult brains are able to combine learning and acquisition in ways that infant brains cannot. So the point is that adult learners will learn faster and will have a better understanding at earlier stages by combining natural input with grammar knowledge. I think Jarvik is suggested there's little benefit to an either-or approach. Why not take advantage of the full power of your brain?

The idea is not to memorize detailed grammar rules and then tackle real stuff. Learning the basics first (as you did), will obviously make subsequent input more enjoyable and more "comprehensible" (to use Krashen speak). And learning about other grammar constructions as one's encounters them won't harm language acquisition, but will speed it up. Think of it as a little comprehension hints whispered in your ear. Krashen also advocates this. (I'm talking about basic explanations and examples, not heavy duty linguistic analysis)

Quote:The point isn't to try and "figure out the rules" the point is to try and "figure out what people say". There's a huge difference.

I took Jarvik's "figure out the rules" to mean that if one has learned to speak through input alone, then it is implied that one has internalized the rules. It doesn't mean that the purpose was to learn what the rules are or that one will be capable of reciting those rules. Again, the rules will eventually be internalized either way (learned ones fall away as Jarvik says). Learning about them just speeds up the process.

Another point of learning grammar pertains to exceptions. It takes a few seconds learn an permissible exception (allowed despite contravention of a rule) whether through grammar or input. But the input method is not able to teach impermissible exceptions (not allowed despite being in accordance with a rule), because you won't encounter it. So your output will contain these types of errors.

There's stuff on this buried in Nestor's links. Alyks, you are a frequent advocate of Krashen, but I get the sense you apply his general theories a bit narrowly.

alyks Wrote:The way I see it, there's no point to learning grammar rules because you might as well go straight to acquiring them from meaningful input. Which leads me to the second point.

But grammar provides you the foundation to make that input meaningful. Grammar allows you to destruct a sentence and make sense of it. Even in English a language which i 'acquired', i still have to deconstruct my own sentences (in essays and what not) when i type them out, when i check them before submission, it is a painful process. But I rarely have to do this when i speak.

I think the reason most people quit on learning 'properly' (through grammar) is because it requires more mental attention than raw sentence bashing does. You have to arrange things in your head, think very hard do all this other stuff.

The same thing applies with learning ANYTHING. Mathematics. Perhaps the single most pure and sensitive subject when it comes to 'grammar' and making sense of things.

alyks Wrote:Example, would you find it easy to list of random words you know in your native language off the top of your head without any kind of stimulus? Kinda hard. But if I told you to list words started with the topic of computers, it would be much easier. "Computer, has a chassis, motherboard, CPU, central processing unit, center, keyboard, keys, keyhole, doorknob, room, house, ..." simply because in our native language words and sentence patterns are associated with context and real life.

I don't see how this at all relates to acquisition, this just proves that we associate words in context. This has nothing at all to do with learning grammar/bashing sentences.

Build yourself a castle made from pure sand grains, rather than one made from random blocks of anything you find on the street. My castle will take longer to build at the start, but after a while, it won't be grain by grain, rather block by block, then section by section and will build exponentially faster than yours. Because you will have to keep filling up gaps at the bottom. But after I use strong cement for the foundation of my castle i won't even need to think about the cement anymore, rather I will walk across it while your castle will have random bits of wood, cement, metal nailed all over the place with holes and gaps that you will have to keep filling up "marge get out the ACME cement~~"

alyks Wrote:But when you try and figure out what somebody is saying when you hear something like "全然好きじゃない!" in a story, then it's different. The reason he argues output before input is because of this, you imitate what you've seen used a million times and you don't have to worry about exceptions and such. It's that you learn from imitation and communication and output based situation and context.

If you're just learning utterances, I see that as a massive waste of time. From that same phrase you could have learned what 全然, 好き, じゃ, and ない meant and be able to apply them in novel sentences, instead of an utterance of limited use (in this example, a Japanese person would rarely outright say "I don't like it at all", instead opting for something softer like あまり好きじゃない). If you ARE learning what all those elements mean over the course of many sentences, then you are figuring out usage rules in your head, if not deliberately. This is the same thing children do when they learn their first language. At first they can only repeat utterances and then eventually they figure out the rules and can make novel constructs. The problem with this is that it takes a long time compared to exposure supplemented with book learning. Silly example: my girlfriend's (Japanese) sister gave birth to a baby around the same time that I started learning Japanese. I speak better Japanese than the kid Once the kid starts going to school and supplements the everyday exposure with book learning (国語 class) she'll probably pass me, because she will be putting a lot more time into it than I do.

While I think a balance of exposure based learning and proper studying is best, having a technical view of Japanese (as in grammar rules and the names and categories of things) is definitely an asset. It really makes learning classical Japanese, dialects, and things which seem irregular (but actually are not) easier. Plus, as part of my scientific nature, I like to know how things work, which in turn gives a better understanding as to the meaning and nuance.

Ex off the top of my head:
五段動詞 「行く」in 未然形 (which is 行か)
＋ negative 助動詞 「ず」in 已然形 (which is ね)
+ 接続助動詞 「ば」
=行かねば (must go)

Now you could have just learned that 行かねば (ならない) is a more formal form of 行かなきゃ(ならない), but then you'd have to just relearn a whole new phrase instead of understanding how to make phrases more formal in general. You also wouldn't know why one is more formal than the other.

Nothing in that example was obscure Japanese. You should know the negative ず from such simple constructs as 役立たず/せず (する in old 未然形 + negative aux. verb ず), and ば and 已然形 from such things as "行けば分かる". You should know 未然形 as the negative form, but it actually has nothing to do with negative other than the fact that negative auxiliary verbs can be attached to it (it actually means something more like not-yet-realized form).

Another problem with the example I gave is that many people just learn it as a set chunk that means "must X" and not as the causal relationship that it is. This is because every textbook I've seen presents it as a chunk, the same way you'd get through simple exposure. Not knowing that it's a causal relationship means you're not going to be making novel utterances like 今行かねば売り切れて買えません (If I don't go now they'll sell out and I won't be able to buy it). You'll just be saying "I have to go".

For a more simplistic example: do you really need to memorize 青い、青くない、青くなかった、
青かった、青ければ、青くなければ、 etc individually (repeat for every 形容詞), or can you just learn once how to conjugate い adjectives?

Knowing this kind of stuff makes it easy to figure out grammar you're never seen before by puzzling it out in much the same way you can puzzle out a word you've never seen before written with kanji that you know. You can also say correct Japanese that you've never been exposed to before.

hehe Prof Jarvik: You may have just scared off a bunch of people. And classical Japanese as a dangling carrot? hmm -you might need to get out of your library a bit more.

Jarvik7 Wrote:For a more simplistic example: do you really need to memorize 青い、青くない、青くなかった、
青かった、青ければ、青くなければ、 etc individually (repeat for every 形容詞), or can you just learn once how to conjugate い adjectives?

In the time it would take to internalize the conjugations, less energy would have to be applied to seeing them all in the course of entertainment from which enough repetition would give you the ability to say them all without thinking.

Again, I don't see the point of grammar. You could learn to conjugate them all first, then expose yourself and internalize them. Or you could just expose yourself and the same thing would happen making the need to learn how to conjugate first them nonexistent (the conjugation is learned unconsciously).

Quote:Now you could have just learned that 行かねば (ならない) is a more formal form of 行かなきゃ(ならない), but then you'd have to just relearn a whole new phrase instead of understanding how to make phrases more formal in general. You also wouldn't know why one is more formal than the other.

My experience has been to the opposite. 行かなきゃ is a good example, because I learned this quite easily and understand it entirely without ever reading any kind of explanation. It's also quite easy to apply it to other verbs after only learning a few. Patterns are pretty easy to pick up.

Quote:If you're just learning utterances, I see that as a massive waste of time.

So yeah, it's not. I've been doing it for most of my time learning Japanese and I can tell you with certainty that it's not as bad as you say, and a whole of fun.

Also, I don't like to think of grammar when reading/speaking Japanese. (sorry, had to say it )

I mean, there's nothing wrong with reading through the first couple of sections in Tae Kim's guide or something. I would even recommend it for the absolute beginner (I myself went through a bit of UBJB by just importing the sentences into Anki and learning what they mean). But only for understanding. Consciously trying to remember the rules is what I don't like because with enough exposure it becomes automatic really quickly anyway. This isn't just speculation, it's what I've been doing for the past 5 months and I can conjugate really well.

The hypothetical universe you guys live in with hypothetical people using hypothetical methods to learn what seems to be a hypothetical language is a weird one.

QuackingShoe Wrote:The hypothetical universe you guys live in with hypothetical people using hypothetical methods to learn what seems to be a hypothetical language is a weird one.

That is why test subjects must be kidnapped.
http://www.alljapaneseallthetime.com/blo...no-grammar
This article came at a suspisciously coincidental time. Does he read these forums?

Not so coincidental, because people don't actually ever stop talking about this.

alyks Wrote:

Jarvik7 Wrote:For a more simplistic example: do you really need to memorize 青い、青くない、青くなかった、
青かった、青ければ、青くなければ、 etc individually (repeat for every 形容詞), or can you just learn once how to conjugate い adjectives?

In the time it would take to internalize the conjugations, less energy would have to be applied to seeing them all in the course of entertainment from which enough repetition would give you the ability to say them all without thinking.

Again, I don't see the point of grammar. You could learn to conjugate them all first, then expose yourself and internalize them. Or you could just expose yourself and the same thing would happen making the need to learn how to conjugate first them nonexistent (the conjugation is learned unconsciously).

It shouldn't take more than 30 minutes to memorize the conjugation table for い adjectives, less than the length of a single anime episode. I can't imagine you managed to get exposed to & isolate, figure out, and internalize all of the conjugations completely in anything less than a few weeks. Then again I'm betting you learned adjective conjugation long before you started sentence method

Quote:My experience has been to the opposite. 行かなきゃ is a good example, because I learned this quite easily and understand it entirely without ever reading any kind of explanation. It's also quite easy to apply it to other verbs after only learning a few. Patterns are pretty easy to pick up.

I think you missed what I was saying. Assuming you know 行かなきゃ and are exposed to 行かねば-->

Did you also pick up on the causal relationship? (行かなきゃよかった, a slangy form of "I wish I never went").

Thora Wrote:hehe Prof Jarvik: You may have just scared off a bunch of people. And classical Japanese as a dangling carrot? hmm -you might need to get out of your library a bit more.

Having access to the second largest library in Canada (#1 largest Japanese library) is too tempting

Really though classical Japanese is mostly of use for filling out your knowledge of Japanese. A lot of classical still finds use in modern speech, and a lot of stuff that seems weird in modern Japanese makes total sense when you look at how it evolved from classical. I don't really care about reading Genji Monogatari in the original. (although JP folktales are awesome)

Thats the same reason why I'm also studying Chinese (plus it's so easy if you know kanji). It gives still more background knowledge. To continue the example above, the 2nd "行" in 行かなきゃ行けない uses the Chinese meaning that isn't written in most kanji dictionaries I've found --> "OK, alright". It's something that has kind of fossilized from the days of 漢文. Also seen in "それは行けませんね". It's not a slangy "That won't go".

It's not just an exercise in academia, I come across a lot of it in translations I'm doing to get certified. Japanese people love wordplay.

<insert complaint about the awful definition of 行 in kanjidic. It manages to have two definitions, one right, the other wrong, and the rest of the meanings of 行 missing, including the main meaning as seen in 行く>

In response to the ajatt link.. Like most stuff khatz writes it's TLDR with an annoying writing style that attempts to be funny, but I'll negate his entire argument right here (it's not the first time I've heard it). There is such a thing as grammar. There is no such thing as a CORRECT grammar (it's subjective). Just because grammar is (usually) descriptive instead of prescriptive doesn't mean it doesn't exist. Just because it evolves doesn't mean that it doesn't exist. Just because goed isn't a word doesn't mean that grammar doesn't exist. You might as well just say that nothing exists and achieve enlightenment.

Just because a native speaker doesn't think of grammar when they speak and can't instantly produce a rule describing something or say if a word is transitive or not doesn't mean that it doesn't exist. It just means it is internalized. As an analogy think of this: when we move our arm we just move it without thinking of actuating individual muscles. There are rules that our brain follows in deciding how to actuate that muscle and how much force to apply to it and when to stop moving it to pick up that glass of water. We don't think of any of that when we pick up the glass of water. We just think "pick up that glass of water". Does this mean that muscles do not exist because we do not consciously interact with them?

I think khatz should try studying up on linguistics, first language acquisition, language disorders (there are some amazingly weird language disorders that mess up some people's grammar "centers" so stuff comes out garbled but in a regular way, but they think they are speaking normally and understand when spoken to normally. For example people with Broca's Aphasia leave out all function words. Conditions such as these are pretty much proof of grammar existing in our heads), second language acquisition, etc instead of being a basement theorist describing his personal experience as the universal truth in the form of a revelation, with everything else being (obviously) false. Or at least that is how he comes across to me and a lot of other people I know.

Jarvik7 Wrote:It shouldn't take more than 30 minutes to memorize the conjugation table for い adjectives, less than the length of a single anime episode.

Yes, it's easy to say that. "Sure, it won't take long to learn this, it's only a little bit of grammar..." So you learn the adj chart. Then you think "verbs could be useful" learn the formal past/present pos/neg tenses for verbs. Then you decide that the って form is important. All of this is very innocent.

But eventually it gets to where you're a bum on the street begging for a JLPT grammar list and trying to figure out sentences like "みんなで力を合わせて、平和で住みやすい社会を作ろうではないか。", forcing yourself to memorize conjugation charts/usage and reading stupid children books that you hate but think you have to because they're at the right level. Where does it stop?

You let study become your main method and that's what Japanese becomes, study study study.

Learning a language by learning all the rules is... scary. Sounds incredibly boring, headache causing and much more difficult then is has to be. Even if it was hypothetically faster to do it your way, I still wouldn't do it. Talk about sucking the fun out of everything.

Besides, without studying grammar and focusing on exposure only, I still got to the point where I understand most of the grammar points for JLPT 2 quite easily.

Jarvik7 Wrote:Then again I'm betting you learned adjective conjugation long before you started sentence method

I knew how to recognize them, but that's it.

Jarvik7 Wrote:Did you also pick up on the causal relationship? (行かなきゃよかった, a slangy form of "I wish I never went").

Yes.

No one is saying "just study rules". If you just study rules all day but never actually use it, wtf was the point? I feel sorry for anyone who just studied rules for four years before they ever thought to read something. Study & usage go hand in hand. Just because it's possible through exposure alone (which no one is debating) doesn't mean it's the best/fastest way.

Quote:But eventually it gets to where you're a bum on the street begging for a JLPT grammar list and trying to figure out sentences like "みんなで力を合わせて、平和で住みやすい社会を作ろうではないか。", forcing yourself to memorize conjugation charts/usage and reading stupid children books that you hate but think you have to because they're at the right level. Where does it stop?

Never met anyone like that who actually studied properly and used their Japanese. There is nothing wrong with asking for a JLPT grammar list if you intend on taking JLPT though. You pass standardized tests by studying for the test. Although I didn't even study grammar when I took JLPT last year since after looking at the list I realized that I already had known everything on it for a long time.

Quote:Then you think "verbs could be useful" learn the formal past/present pos/neg tenses for verbs.

Verbs in Japanese only have 5 forms (hence the 5 in godan verbs, the category that has a vowel change for each form). Everything else is done with vocabulary. You might think that there are more, but that is because you learned from the top down and not the bottom up so you never learned the underlying simplicity. Five tenses is again, not an insurmountable obstacle and can be done in less than a day. Also, these are the same forms that are used with adjective conjugation.

Learning core Japanese grammar like this should take less than a week.

I think Thora nailed it with the bit about applying the adult intellect, rather than infantilizing yourself in some kind of twisted Norsk experiment that treats grammar like the bogeyman. It's not an either/or situation, like folks who take their gurus too seriously seem to believe. ;p

Also, I don't know my foot from an adverb (not in English either), nor did I memorize conjugations. There's all kinds of ways to incorporate 'grammar' to turbo boost your learning and develop a foundation.

Aklys -- do you avoid dictionaries as well? If not, then what's the difference between looking up the meaning of a word and looking up the meaning of a word pattern?

pm215 Wrote:Aklys -- do you avoid dictionaries as well? If not, then what's the difference between looking up the meaning of a word and looking up the meaning of a word pattern?

It's 'Alyks' not 'Aklys'. Quite often I do, but not always. To put it a better way, I'm going to read even if I don't look up words. In my experience grammar is often very easy to figure out in context. Words can take a bit of time before context will allow you to get them, and the number of words needed to understand is far greater than the grammar needed. Pretty much every time I have trouble understanding what's going on in a story is simply because there are too many unknown words. Besides, I don't know how you would look up grammar as simply or quickly as you would a word, I can press a single keystroke to look a word up.

Also, my argument here was against memorizing grammar as your language learning. Grammar will be internalized and take care of itself on its own. Memorizing verb charts or word patterns and what not is a waste of time. The equivalent of this for words would be to memorize a dictionary and force them into all into active memory as you learn them before internalizing their usage through context.

Except there are only 5 verb forms and you're going to use every single one of them right from the beginning. It's nothing like memorizing an entire dictionary which has >100,000 words, most of which you will never use.

Didn't you do RTK? Didn't you invent the movie method to memorize every kanji reading before learning vocab? Those are both magnitudes more pre-study before application than verbs are.

alyks Wrote:It's 'Alyks' not 'Aklys'.

Oops, sorry. I guess I play too much nethack...

Quote:Besides, I don't know how you would look up grammar as simply or quickly as you would a word, I can press a single keystroke to look a word up.

Surely this is an argument *for* pre-study of grammar? If you can look up a word easily you can just do it as you go along where you need to. If it's something you can't look up then having learned it in advance is the only way... (Anyway, there are grammar dictionaries, although I don't know of any serious electronic ones offhand.)

pm215 wields an Alyks.
Alyks welds itself to pm215's hand.

~J

Jarvik7 Wrote:One important thing that the your summary is overlooking: a learned language develops into an acquired language with use, and reaching the goal is faster when coupled with learning. It is making a false dichotomy. We don't have to JUST acquire language or JUST learn language. Most things in life work better with a balance.

Thank you. Too many people forget this.

alyks Wrote:The way I see it, there's no point to learning grammar rules because you might as well go straight to acquiring them from meaningful input. Which leads me to the second point.

By that logic, there's no point in learning English rules of grammar, either... And yet every single one of us was required to do so in school.

Language learning (rather than acquisition) plays an important role in learning the language faster. Imagine if you were learning to read. You've never seen a book before and all you know is that language is someone recorded on the pages. You would have to fully formulate all the concepts involved: letters, words, groups of letters making certain sounds, etc, etc. And then you'd have to study all the letters, letter groups, and words to figure out what they mean. Given enough time, you could figure it out.

But add in structured learning, and suddenly learning to read is a -lot- shorter process. It doesn't replace actual reading, it just jumpstarts and then supplements it.

Language learning is the same way. You jumpstart with some vocab so that you can start learning pieces here and there from actual language. Then you supplement with more vocab and some grammar, and you continue your high-speed trek through the language. There comes a point when learning vocab and grammar from books becomes pointless, of course. But that point is definitely not at the beginning.

alyks Wrote:Memorizing verb charts or word patterns and what not is a waste of time.

Actually, it's a huge timesaver. You can memorize all the conjugations needed in Japanese in a day. Doing the same in context would take weeks. Instead of having to guess what ご飯を食べれば？ means, you would know it instantly.

Structured learning owns. Which is why it owns to learn kanji before you start reading.