Mushi Wrote:

JimmySeal Wrote:"Let's meet me at my house tomorrow."

As an aside (as this doesn't affect the validity of anyone's position in this discussion), is the above even an error in grammar? "Let us meet me" looks gramatically correct to me, just impossible, or requiring a cloning device.

In English 'meet' can be transitive (vt) or intransitive (vi), but not both at the same time. "Meet me" is the transitive (has a direct object) use, and "Let's meet" is the intransitive (no direct object) use. That's why it's grammatically incorrect.

I'm on much shakier ground with Japanese grammar, but as I understand it Japanese verbs are either transitive or intransitive, but not both. (That's why you see lists of (vt) and (vi) pairs, i.e. 直す naosu 'fix' and 直る naoru 'be fixed'.) 会う is a (vi) and is always used with the indirect object particle に (or と but let's not go there), never with the direct object particle を. This is confusing because "I met Mushi" is a (vt) sentence in English, but 私は虫さんに会った is a (vi) sentence in Japanese.

Of course, once cloning devices are commonplace this grammar point will be obsolete.

Anna B Wrote:In English 'meet' can be transitive (vt) or intransitive (vi), but not both at the same time. "Meet me" is the transitive (has a direct object) use, and "Let's meet" is the intransitive (no direct object) use. That's why it's grammatically incorrect.

So would you say this is incorrect?

"Do you want to go see the new neighbors tonight?"
"No, I'm too tired tonight. Let's meet them tomorrow."

"Let's meet <somebody>" is still transitive. "Let's meet me" in particular just sounds nonsensical because what does it mean to meet yourself?

Incidentally, does anyone know of a good unabridged E-J dictionary? Preferably a big, heavy thing that defines at least half a million English words?

Being from a Japanese family, I always had access to English-to-Japanese dictionaries. Although my Japanese reading ability has never been good, I found that those dictionaries were nevertheless nice to have. The problem with them though, is that they not particularly complete.

Few things bug me more than an abridged English dictionary. They always seem to contain all the words you already know, and none of the words you don't know, but would like to look up.

yudantaiteki Wrote:I only use J-J if I feel like the J-E dictionary isn't giving me the full story, or the word isn't in the J-E.

Yep, me too. Mostly I start with J-E and only look in the J-J if the J-E senses look a bit all over the place and like there might be an underlying commonality in the J-J definition.

IceCream: are you really SRSing every word you ever look up, even the ones which you only looked up to check a reading or where you were kind of half-sure what the word was and wanted to confirm it? If so, you could probably stop doing that...

(OTOH I bet you're way ahead of me in vocab already so take my advice with a pinch of salt :-))

Blank Wrote:

Anna B Wrote:In English 'meet' can be transitive (vt) or intransitive (vi), but not both at the same time. "Meet me" is the transitive (has a direct object) use, and "Let's meet" is the intransitive (no direct object) use. That's why it's grammatically incorrect.

So would you say this is incorrect?

"Do you want to go see the new neighbors tonight?"
"No, I'm too tired tonight. Let's meet them tomorrow."

"Let's meet <somebody>" is still transitive. "Let's meet me" in particular just sounds nonsensical because what does it mean to meet yourself?

You're saying that a nonsensical sentence can be grammatically correct? If so, I would say that your point is nonsensical.

@IceCream: Let's meet together would be redundant but would make the most sense, and would also be grammatically correct, since 'together' is an adverb, not a direct object.

Of course a nonsensical sentence can be grammatically correct. Grammar is about structure, not meaning.

IceCream Wrote:@Jimmy Seal... just found this post... do you still learn like this? did you change at any time? What's your perspective on non-dictionary learning after 3 years?

The other day I read an interesting article which put a case for bilingual in preference to monolingual dictionaries. It also had some other suggestions about dictionary use. Here is the link:

http://learnalanguageortwo.blogspot.com/...ngual.html

The article comprises extracts from the book "The Art and Science of Learning Languages" by Amorey Gethin (AG) and Erik V. Gunnemark.

A more detailed summary of this book is given at http://www.lingua.org.uk/voc.html plus a critical appraisal of it here: http://www.lingua.org.uk/vocdb.html

Two of the article's recommendations are:
- to not make word lists (does that imply throwing out SRS too?)
- to use a dictionary as little as possible, trying instead to infer meaning from context - a view also expressed in 'Polyglot' by Kato Lomb.

However, others such as Rob Waring in 'The Inescapable Case for Extensive Reading' (http://www.robwaring.org/er/what_and_why..._vital.htm) and Jan-Arjen Mondria in 'Myths about language acquisition' (http://www.babylonia-ti.ch/BABY207/PDF/mondria.pdf) point out that, in order to infer meaning, a learner needs to know 95% or more of the surrounding words which presumably makes word lists and dictionaries necessary.

One way to increase the chance of inferring meaning from context is to concentrate on input from one subject or one author. See 'The case for narrow reading' by Krashen (http://www.sdkrashen.com/articles/narrow/all.html)

My apologies if you're already familiar with these articles. I know they've been referenced in other posts. I'm not going to offer advice on which way to go here as I'm too much of beginner myself. The articles are interesting, contradict each other to some extent and may just leave you confused!

Anna B Wrote:You're saying that a nonsensical sentence can be grammatically correct?

Yep!
http://en.wikipedia.org/wiki/Colorless_g..._furiously

Edit: oops, missed yudantaiteki's post. I think that's what I'm trying to say

Learning words from context without verifying them is very dangerous. I've learned both from my own experience and my teaching that it's very easy to come up with plausible meanings of sentences that are totally wrong. You may not realize how much you can misinterpret and still be able to have the feeling that you are following the main context and reading along. That's why I think that translation can be a useful exercise from time to time; that's a great way to tell what parts you don't really understand.

I've seen this in fan translations -- while many of them are fine, there are some where the person obviously has no clue what they're doing but they're able to construct plausible English sentences that are grammatically correct and have something to do with the action on screen. I still remember one where the translator obviously did not understand the Verbがいい（よい） structure (i.e. 見るがいい) -- it's a very blunt request/command, but the translator obviously thought it meant the same thing as 見ることがいい (i.e. "the act of looking is a good thing").

yudantaiteki Wrote:I've seen this in fan translations -- while many of them are fine, there are some where the person obviously has no clue what they're doing but they're able to construct plausible English sentences that are grammatically correct and have something to do with the action on screen. I still remember one where the translator obviously did not understand the Verbがいい（よい） structure (i.e. 見るがいい) -- it's a very blunt request/command, but the translator obviously thought it meant the same thing as 見ることがいい (i.e. "the act of looking is a good thing").

I intend on responding to Ice Cream's original question tomorrow but I thought I should respond to this now.

The most revealing point of your anectdote is the part I've marked in bold. The problem here is not that learning from context is ineffectual, but that the people you're describing, well, obviously had no clue what they were doing. People who do things haphazardly will fail at what they are doing. This is true no matter what they are basing their approach off of.

Anyone who has seen Verbがいい more than twice and actually paid attention to the context would know what it means. The issue here is that these "translators" were letting the words pass in through their eyes or ears, and out through their fingers, unhindered by rational thought.

As David Bond says:

Quote:24. Only use language you have seen or heard

My own advice to a language-learner, from the very beginning, would be extremely simple and straightforward. Never use language that you have not seen or heard and which, as far as is humanly possible, you therefore know to be correct. This is at first an extremely difficult discipline – but only, I would emphasise, at first; like anything else it will rapidly become second-nature if you stick to it. The student that learns this way – ultra-patient, some might think it – will perhaps proceed a shade slower at first but I have no doubt at all that they will learn the language a great deal more satisfactorily and satisfyingly in the end as a result. With specific respect to dictionary-use, I would advise learners only ever to use a dictionary to look up things they already know, either to check spelling or to find a word (in the language learnt) that is on the tip of the tongue or simply to satisfy your curiosity as to what the dictionary has to say about it. Never use a dictionary as a means of learning anything. It isn’t any such thing.

The above pertains to production in L2, but it applies as well to translating text from L2. Learning from context does not mean assuming the meaning of words and phrases on first sight. It means observing, recalling, re-evaluating the meanings until you are sure you understand their nuances. The person you're describing clearly did not do that.

Well yes, but I find it hard to read something and correctly identify everything I don't know and determine the meaning from context. I would be surprised if anyone, at any level, is able to correctly identify and learn 100% of what they do not know, correctly, without any reference works.

(I'm not saying that you must always use a dictionary to look up all words you don't know or that you should never try to learn anything from context, just that you should always be aware that you may figure out things wrong and that you may want to sometimes verify your hunches, especially if you are somewhat uncertain.)

IceCream Wrote:@Jimmy Seal... just found this post... do you still learn like this? did you change at any time? What's your perspective on non-dictionary learning after 3 years?

Hello Ice Cream,

Yes, I do still stand by everything I said in this thread 3 years ago, and I have applied the principle to reading in Japanese, Chinese, French, Italian, and Spanish with a lot of success.

There are conflicting arguments about whether learning vocabulary and grammar in terms of L1 does long-term damage to one's ability to learn the language. I'm still of the camp that tends to think that it does, and think that at best you will wind up having to unlearn these associations later on. I think it's better to naturally learn the feel of the language and its vocabulary from the start, and avoid that unlearning process.

Regardless of what the true answer to the above dilemma is, I stand by the argument that reliance on dictionaries (including monolingual dictionaries) and textbooks does hinder language learning, if only because of the additional effort and burden it places on the learner. Reliance on dictionaries convinces learners that they cannot read texts unless they understand most or all of the words. Thus they wind up doing one or both of the following:

1. They spend time trudging through the text they want to read, while looking up almost every word they are unsure about. This places a burden on the learner because they have to drop what they are doing and look up a word every few seconds (extra burden points if it's a paper dictionary). And it's frustrating as well, as learners find that they continually forget the words they look up, and have to look them up again and again. This quickly saps the fun out of language learning.
2. They spend time reading through boring textbooks and/or asinine children's material trying to reach a point where they understand almost all of the words in the texts they actually want to read. They assume that until they reach the magical 95% mark, the stuff they want to read is just too hard. This quickly saps the fun out of language learning.


Regarding the 95% figure that people like to bandy around in defense of textbooks and dictionaries, I am not going to argue with the research that went into producing this number, but I will say that it is very misleading if taken at face value.

It may be the case that you need to understand 95% of the words in a passage to infer the meaning of the rest, but this only means that need to know 95% of a given section, not the whole text. Probability dictates that as long as the level of the text isn't way too high, there will be areas with a high concentration of words you know, and areas with lower concentrations. If you focus on the easier-to-read parts, and gloss over the harder ones, you will accumulate vocabulary, and the effect will snowball until the unapproachable parts are now approachable.

This is why I advocate starting off with novels you've already read in a language you understand, and comic books.

If a picture is worth a thousand words, and you're reading a comic book panel with 10 written words, that means you already understand 1000/1010 = 99% of the words in the panel. I'm just being rhetorical here, but the concept is valid. If you see the word 箭 in or around 4 comic book panels with an arrow in it, chances are it has something to do with arrows.

If you start off reading translations of novels that you've already read in your own language, you can gloss over the hard parts while only recognizing words here and there and still follow along with the story. Once you reach a spot with a few words you already know, you can use this to infer the nuance of one or two more new words (knowing the story also helps with this). If you go through the book for a second pass (which I recommend for the first 2 or 3 books), you'll be amazed at how easy it is the second time through.

Don't get misled by the myth of the 95% comprehensible input requirement. Remember that if infants had to already know 95% of the words they hear before they could understand any new ones, they'd never get past their first word.

JimmySeal Wrote:If you see the word 箭 in or around 4 comic book panels with an arrow in it, chances are it has something to do with arrows.

You're talking about a concrete object, though.

It still seems to me like you're taking the usual path of "Some learners misuse X and X has the potential to cause difficulty, therefore X should be completely avoided no matter what benefits it might have." It's not a choice between "Look up every word and don't move on until you understand everything" and "Don't use dictionaries at all".

Of course there are pitfalls and dangers in using dictionaries, but your context suggestion has pitfalls and dangers as well. You seem to be constructing a perfect scenario where someone reads through a book and learns every word purely through context, without any misunderstandings or mistakes. But I could equally well say that with a dictionary, someone will look up every word and perfectly understand everything and not be misled by any of the definitions because they'll see from the context what the correct one is.

I am a proud dictionary user; I used them learning, I still use them now. I don't think they've caused any harm to my learning. I suppose if someone else wants to do it without dictionaries and is able to that's fine. I have no problem with the idea that you shouldn't look up every word or that you should sometimes try reading without dictionaries, but the idea that you should never, at any point in your learning, use a dictionary seems untenable to me for most learners.

JimmySeal Wrote:Anyone who has seen Verbがいい more than twice

Most verbs and nouns appear only once or twice in a typical light novel series:

Nouns:
Bakemonogatari 0.58 → 58% of the nouns appear only once or twice in the whole series
Denpateki_na_Kanojo 0.61
Gosick 0.53
Kanokon 0.58
Kara_no_Kyoukai 0.60
Kino_no_Tabi 0.61
NHK_ni_Youkoso 0.71
Spice_and_Wolf 0.50
Suzumiya_Haruhi 0.60
Toradora 0.60
Zero_no_Tsukaima 0.54

suru-verbs:
Bakemonogatari 0.52
Denpateki_na_Kanojo 0.60
Gosick 0.57
Kanokon 0.53
Kara_no_Kyoukai 0.56
Kino_no_Tabi 0.62
NHK_ni_Youkoso 0.74
Spice_and_Wolf 0.45
Suzumiya_Haruhi 0.50
Toradora 0.54
Zero_no_Tsukaima 0.49

I doubt there's enough context to figure out what they mean (with any confidence).

yudantaiteki Wrote:Learning words from context without verifying them is very dangerous.

One example of this I see a lot, is with perfectly well educated native speakers who mistakenly believe that "penultimate" means something like "the best of its kind", when it actually means "next to last". Any dictionary can tell you that, and it's not exactly an uncommon word, but still...

I imagine that they've heard or read the word in contexts like, "'The Empire Strikes Back', the penultimate film of the trilogy, is a tour-de-force of special effects and storytelling..." You really can't make a good guess based on contexts like that.

I've never felt like the monolingual dictionaries help much in those cases. EDICT is problematic because of the lack of examples, but Koujien's definitions don't really help me with those above words any better than EDICT does. This is where you really need help from the context of where you're seeing these words.

So many Kojien definitions of Chinese compounds are just the kun-yomi of the kanji -- for instance, 照合 is 照らし合わせ確かめること。 Personally that's no better to me than "collation; comparison". Now, the example of 指紋を照合する is helpful but my J-E has helpful examples as well. 統合 is 二つ以上のものを一つに統べ合わせること。 For me, if I don't know 照合 I doubt I know 照らし合わせる, and if I don't know 統合 I doubt I know 統べ合わせる.

I've noticed an interesting phenomenon recently with native speakers of English (of which I'm one). If they hear or see a word they don't know twice within the space of the same conversation/article/5 minutes/whatever they tend to ask what it means or look it up.

I've seen it happen a lot recently. So even natives use dictionaries/ask what words mean. They can also learn from context obviously but just pointing out that it's not the sole way they learn. I know I do both.

As for the 95% thing... I've revised it to 99.95% for 1 reason. If you know 95% of the vocab in a novel then on average you're looking up 5 words in every 100 and there's roughly 250 words on a page. So within 4 pages there's 50 words you don't know. It's too much missing information to actually read with flow and good comprehension. Even 99% is still 2 words per page at that rate...

I think for the most part learners are easily deceived by using percentages to represent understanding in language.

mezbup Wrote:As for the 95% thing... I've revised it to 99.95% for 1 reason. If you know 95% of the vocab in a novel then on average you're looking up 5 words in every 100 and there's roughly 250 words on a page. So within 4 pages there's 50 words you don't know. It's too much missing information to actually read with flow and good comprehension. Even 99% is still 2 words per page at that rate...

I would disagree with that, actually.

Here's a couple of paragraphs out of 私の男 by 桜庭一樹, chosen because it's about at that 95% level for me.
こんなにこの場所に、思い出に近づいたら、また養父にあってしまう、とおそろしく思った。不安で、心がざわめいて、気が重いのになぜかどんどん足早になった。銀の夢荘にたどりつくと、わたしたちの部屋だったところのドアが、少し開いたままになっていた。思い切って外階段を上がった。ハイヒールが甲高い音を立てる。カンッ、カンッ、カンッ、カンッ。。。部屋の前にたち、おそるおそるドアノブを握りしめた。
一気に開く。
六畳間の開け放された窓から、強い西陽が射してはげしく目をくらませた。わたしは瞬きをして一瞬、立ちつくした。カーテンがなくなっているのだと気付いた。ゆっくりと靴を脱いで、部屋に上がった。机もなかった。冷蔵庫も、食器棚も、ふるいタンスも、なにもかもなくなっていた。部屋は文字通りのもぬけのからで、ずっとタンスが置かれていたところだけ、畳の色が若くて、ついさいきんまで誰かがここに住んでいたことを示していた。

Words I did not know on my first go-through:
思い切って
西陽
くらむ（くらませる）
食器棚
もぬけのから

銀の夢荘 has already been established as the name of their apartment building.

That's five, and I'm averaging somewhere between 2 and 6 words per page.

But, does it slow down my flow? Not really. 思い切って - "Okay, she gets back to her old apartment, she's been told by the landlord to clear out the old stuff (but where has her foster father gone?), and the door's partly open. How do you THINK she climbs the stairs?"
西陽　I may not know that as a word but I can at least tell from the kanji that it's sunlight from the west.
くらむ -- "Afternoon sunlight is doing something to her eyes -- so surely it's blinding her or dazzling her or making her squint, something like that."
食器棚 -- I know 食器　and 棚, and if the fridge and the chest of drawers are gone, then surely it's something like them -- it's a shelf for 食器.
もぬけのからI couldn't have gotten from context -- it means completely empty, like the shed skin of a snake or insect -- but this entire passage is hammering home the idea that the room is completely empty, so I wouldn't have missed the meaning of the passage as a whole even if I hadn't looked that up.

Now, if I don't look all of these words up, am I going to learn their correct meanings? No, not necessarily. However, if I'm approaching this as "I'm enjoying reading a novel," yes, I'm perfectly capable of enjoying reading a novel while at about 95% comprehension. And reading a whole novel at a fairly fast clip is REALLY helpful for internalizing grammar stuff and vocabulary that you already know but isn't quite second nature to you. And aside from the language-study aspects I'm happy to have read things by contemporary authors whom I like, and I'm happy to have read things by people like Mori Ogai even if I understood about half of what was going on!

Meikyo sounds pretty good and I think our library has a number of copies but I can hardly stand using paper dictionaries anymore...

If I had to understand 99.5% of anything without a dictionary to make use of it, I would have been kicked out of graduate school long ago. I don't think there's any fixed percentage of a text you have to understand; it depends on exactly what the text is, and what's in the percentage you don't understand.

IceCream Wrote:@JimmySeal: it certainly sounds like a decent way of going about things to me... but... do you use anki or anything? and if so, how do you use it? The reason i'm asking is that it just seems like, its not always possible to remember the last context you saw something in. In fact, i find that particularly difficult. So, don't you forget what it meant by the time you next come across it?

Yes, I use Anki for practicing kanji readings and kanji production. I have cards with individual words in kanji and the hiragana reading on the answer side. I have other cards with short sentences that have a word underlined and spelled in katakana, with the kanji on the answer side. Most of these kanji production cards are taken from kanken practice books.
If I encounter a sentence in my reading that causes an Aha! moment for me, sometimes I'll write it down and put it in Anki and maybe just underline one word for producing the kanji reading, or just put a dash on the answer side so I can read the sentence and pass it. But I will only do this once in a great while. Anything more would be over-burdensome.

yudantaiteki Wrote:

JimmySeal Wrote:If you see the word 箭 in or around 4 comic book panels with an arrow in it, chances are it has something to do with arrows.

You're talking about a concrete object, though.

Ok, how about if you see someone running in several panels coinciding with the word 跑? There's a good chance it means "run" (it does). If you see several panels of someone with a surprised look on their face and the word 驚嚇 you can bet this has something to do with surprise. Incidentally, these are actual examples of words I figured out this way.
And this isn't limited to comic books. On the previous page of this thread I pointed out that I figured out the meanings of "tuer," "lorsque," and "tout le monde" (all abstract concepts), from reading Harry Potter.

yudantaiteki Wrote:It's not a choice between "Look up every word and don't move on until you understand everything" and "Don't use dictionaries at all".

You're right. It's not just the two extremes, but those who rely on dictionaries also tend to over-rely on them, and wind up finding themselves on a slippery slope. The maintainer of Tadoku seems to agree, and just look at poor mezbup a few posts up. He thinks you can't read something without looking up every word and need to know 99.5% of a text's content to attempt reading it. Balderdash.

yudantaiteki Wrote:You seem to be constructing a perfect scenario where someone reads through a book and learns every word purely through context, without any misunderstandings or mistakes.

If you think the scenario I'm describing is that someone will pick up every word they see and without any mistakes, that's obviously not what I'm saying. I think I said clear enough that learning from context requires reevaluating the guesses you've made and adjusting them as necessary.
And your statement that looking up words will provide perfect understanding is just not true. Even with one piece of context, a dictionary can still be misleading, and I've already said, time consuming to boot.

And to be clear, I'm not opposed to occasionally finding out the meanings of words through methods other than inference. Asking a Japanese person what a word means will give you a two-way exchange about the word's meaning and nuances. Looking up a word on Google images will often give you a clear image of what a word represents without representing it in your native language or some overwrought definition.

IceCream Wrote:I dunno how many words that is, but it's less than half a page, and there's 7 words i either have never come across or have srsed at some point and forgotten. If Mezbup's calculations are right, that's more than 95%.

A word count and some calculations brings it to 97.4%... Exactly what I'm talking about 7 words unknown out of 273 and that's on half of 1 page... out of (how many??)

I think there are two ways to go about it... either you go for a very long term approach where you look everything up and srs it or you just read and skip over everything you don't know until you see words that come up repeatedly and thus u learn based on frequency naturally.

No reason one can't do a bit of both.

My main point is for native-level ability you need to know a shitload. Way more than you ever realise when you first begin learning the language.

Well, all I can say is that I made heavy use of dictionaries and it worked out fine for me. It has not hampered my long term progress.

mezbup: But as usual, what is "native level ability"? As usual, I don't really care whether some arbitrary label can get applied to my Japanese.

It seems like it's a really common argument dodge on RTK to respond to a criticism of a method or statement with "Well, that's fine unless you want native-level ability (or "fluency") in which case you have to agree with me."

mezbup Wrote:

IceCream Wrote:I dunno how many words that is, but it's less than half a page, and there's 7 words i either have never come across or have srsed at some point and forgotten. If Mezbup's calculations are right, that's more than 95%.

A word count and some calculations brings it to 97.4%... Exactly what I'm talking about 7 words unknown out of 273 and that's on half of 1 page... out of (how many??)

I think there are two ways to go about it... either you go for a very long term approach where you look everything up and srs it or you just read and skip over everything you don't know until you see words that come up repeatedly and thus u learn based on frequency naturally.

No reason one can't do a bit of both.

My main point is for native-level ability you need to know a shitload. Way more than you ever realise when you first begin learning the language.

Definitely need to know a lot in terms of kanji/vocab, how about 4000+ kanji?

JimmySeal Wrote:and just look at poor mezbup a few posts up. He thinks you can't read something without looking up every word and need to know 99.5% of a text's content to attempt reading it. Balderdash.

lolz what I was talking about wasn't really what you need to know in order to "attempt" to read it but rather to actually read it or be able to read it for pleasure. It was about over-all reading ability.

In terms of where I think you ought to be when you attempt reading something well that really depends on what it is and where you're at. It's kinda obvious that you start small and work your way up. If you needed to know 99.95% in order to start reading you'd be up shit creek without a paddle.

I generally look everything up out of personal preference. Coupled with a strategy of once I've finished a larger "text" re-read it a second time knowing 100% of the vocab. I've made wicked progress this year cos of that.

But you don't even need that to read for pleasure. I don't understand 99.5% or even 95% of the material I use in my dissertation research.