Based on this article
http://joechip.net/extensivereading/2011...er-it-too/
I am trying to read articles taken from
http://news.kids.yahoo.co.jp/headline/
with my new acquired kindle.
I am having a great time, I copy and paste an article creating a .txt, and then I email to my @kindle.com email.
I try to follow these three principles of extensive reading:
1. Don’t look up words in the dictionary.
2. Skip over parts you don’t understand.
3. If you aren’t enjoying one book, toss it aside and get another.
These principles were created by Kunihide Sakai, a retired English professor who champions extensive reading, and loosely translated from his site tadoku.org.
However, I admit I am having a hard time avoiding the use of a dictionary. I feel that I am losing that word that I don't know forever.

With practice, it gets easier.

Also, it's not a hard rule. If I see a word many times and it's killing my understanding, I look it up. Sometimes, if I know it's a really important word, I'll look it up the first time. But I usually try to wait until I've seen it 3 or more times so that I've got some context for it.

Yeah, i think the no dictionary rule is mostly about discouraging obsessive compulsive over use that would make your reading grind to a halt. You want to balance comprehension and speed so that the text is still enjoyable.

edit: Even if you don't check a word, it can be a good idea to keep lists of unknown vocabulary as you read for later study. This way you can distill out from your reading precisely what you need to learn.

I think it's better to design/prepare authentic materials with various tools/strategies so you can get a more learner-adaptive and comprehensive language exposure (condensed reading = intensive + extensive), and it's also better to use, say, Rikaichan: RevTK Community Edition to look up words, which includes having audio for the multisensory component. There's tonnes of studies that show computer-assisted learning in this way is best.

Example: Review of Effects of Glosses on Incidental Vocabulary Learning and Reading Comprehension

“Many studies have confirmed that glosses are very useful for reading comprehension and incidental vocabulary learning. With the development of computer and multimedia technology, gloss formats are not limited in text only. This technologically advanced tool offers different modalities of gloss, i.e., pictures, animations, video, sound, etc., which cause immediate access, reader control and absence of interruption in reading...

... Many studies have confirmed that a gloss is more useful than no gloss for incidental
vocabulary learning.”

IMO both "Look up every word" and "Never use the dictionary" are extremes that you should avoid in general, although I think they're both worth doing sometimes. I used to read the short Tensei Jingo articles in the Asahi newspaper when I was at the intermediate stage, and I would read each one three times -- the first time with no dictionary, the second time looking up every word, and then the third time a day or two later to see how much I retained.

A very important skill to develop is to use a dictionary judiciously to improve your comprehension of a passage without slowing you down so much that you're not getting anywhere. This is something you have to learn over time, and avoiding the dictionary completely won't help you learn it.

I like this article when it comes to reading:
http://www.tofugu.com/2010/08/30/5-step-...-for-kids/


I feel like you should at least know the pronunciation of the word, if there's no furigana.

Also, if you like reading and you genuinely want to understand, feel free to use a dictionary. I've done so many times for Moshidora, I skip over stuff I don't understand but I also bust out the dictionary if it gets to a point where I don't understand the overall story and I try to only look up words in the parts that seem important.

By the way, extensive reading should definitely not be your primary learning method.

Here's a paper I just found that I'm disappointed I didn't discover sooner; it's by Paul Nation, who has been cited here before, re: vocabulary sizes...

The Four Strands

Abstract: The activities in a language course can be classified into the four strands of meaning-focused input, meaning-focused output, language-focused learning and fluency
development. In a well designed course there should be an even balance of these
strands with roughly equal amounts of time given to each strand. The research
evidence for the strands draws on the input hypothesis and learning from extensive
reading, the output hypothesis, research on form-focused instruction, and the
development of speaking and reading fluency. The paper concludes with 10 princi-
ples based largely on the four strands. The strands framework and the principles
provide a basis for managing innovation in language courses.

Edit: And I should mention ‘glossing’ (e.g. marginal) is typically differentiated from ‘dictionary’ searching on the basis of how the latter interrupts the reading process by requiring an external inquiry, so pop-up dictionaries fall in with glossing functions (e.g. studies using Lextutor for English and finding double the vocabulary growth when using a popup dictionary [and sound for accessing another modality, apparently, wow, they were on the ball: http://llt.msu.edu/vol9num2/pdf/horst.pdf])

I use dictionaries and Anki obsessively but after a while you just stop coming across many new words, so it's not that bad. Takes a while though.

I came across that extensive reading link that that OP posted, and also the tofugu article, a few days ago. My understanding is that extensive reading is reading a lot "at your level". So for a beginner, it would mean reading a lot of children's books and similar. As I was pondering this in relation to my own reading and Japanese learning strategy, I came across a problem with it IF one already knows the jouyou kanji, i.e. one has finished, or is close to finishing, RTK1. Everyday now the Asahi online news has its own tab on my browser, and everyday I try to read it, learning words like 原発, 水爆, 実験, 福島県. *headsmack* Sign of the times. Then one night I was wondering if I should drop my reading level and just read children's books. Well it didn't take long for me to crush that idea, because I realized while browsing the books that I know more kanji (through RTK1) than those children's books and fairy tales I looked at will give me. Not to mention that the things I read in Asahi are more interesting to me than the children's books.

This is just my anecdotal experience, and was not meant to dissuade OP from his/her chosen learning strategy. Do what works for you.

The right level for extensive reading is a level where you can understand the at least the basic gist of what you're reading without any dictionary lookup. (Could you narrate it back to a friend in your native language? Could you make logical guesses about what happens next? It depends on your level of comprehension, but it also depends on your frustration threshold -- some people are more frustrated than others by reading texts they only partially understand.)

Seamoby, you might find that novels written for young adults or adults (light contemporary fiction, not anything really dense and heavy) makes a good compromise between the difficulty of children's books and fairy tales, and the difficulty of Asahi news. Newspapers are densely packed with information in a way that requires a LOT of vocabulary, often very subject-specific vocabulary.

For English, the general recommendation is only 2-5% of the material contains unknown terms (and the difference between those levels seems high; I believe I read that in general for written English, without tailoring your materials to your level, this requires something like knowing 8000-9000 word families (~35,000 words). (For a learner.) The difference in that 3% is big, though, like thousands of word families.

Thought I'd quote this:

“Taken together, the research confirms that worthwhile vocabulary learning
does occur from reading. However, the pick-up rate is relatively low, and it
seems to be difficult to gain a productive level of mastery from just exposure.
Hill and Laufer (2003) estimate that, at the rates of incidental learning
reported in many studies, a L2 learner would have to read over 8 million
words of text, or about 420 novels to increase their vocabulary size by 2000
words. This is clearly a daunting prospect, and thus it is probably best not to
rely upon incidental learning as the primary source of the learning for new
words.

Rather, incidental learning seems to be better at enhancing know-
ledge of words which have already been met. This conclusion is congruent
with Waring and Takaki’s (2003) findings that reading graded readers does not
lead to the learning of many new words, but that is very useful in developing
and enriching partially known vocabulary. Studies with a variety of test types
have shown that exposure leads to improvements in multiple types of word
knowledge. Also, given that repetition is key to learning words, the benefits of
repeated exposures in different contexts for consolidating fragile initial learn-
ing and moving it along the path of incremental development cannot be
underestimated.”

- Instructed second language vocabulary learning (Norbert Schmitt)

That's mostly for ungraded materials without supplementation with glossing and frequency-related design and such. Overall for incidental/extensive without explicit, graded doesn't fare much better, but in terms of graded vs. ungraded extensive/incidental:

“... One way of incorporating incidental learning into a language program is to
organize an extensive reading component (Day & Bamford, 1998). Although
readers need to know 98–99% of the words in a text, many authentic texts will
still be suitable for more advanced learners, especially if teachers provide sup-
port for the more difficult vocabulary (see below). However, for developing
learners, the vocabulary load will probably be too high in authentic texts, and
so the use of graded readers is recommended, as the vocabulary load is both
fine-tuned for the learner’s level, and systematically recycled (Nation &
Wang, 1999; Al-Homoud, 2007).

Graded readers used to have a bad reputa-
tion for being boring and poorly written, but that is no longer the case, with
several major publishers providing a series of interesting and well-presented
readers. Most importantly, research shows that substantial vocabulary learn-
ing can be derived from graded readers. For example, Horst (2005) found
that her participants learned over half of the unfamiliar words they encoun-
tered in the graded readers they read.”

Another nice bit somewhat related to Nation's meaning-focused vs. language-focused Four Strands paper:

“... there are good reasons to believe that vocabulary requires a different approach
which incorporates explicit attention to learning the lexical items themselves:
• learners who understand the overall message often do not pay attention
to the precise meanings of individual words
• guessing from context is often unreliable, especially if the learner does
not know 98% of the words in the discourse
• words which are easily understood (guessed) from context may not
generate enough engagement to be learned and remembered
• new words which learners have met in discourse need to be met again
relatively quickly to avoid their being forgotten. In order for words to
be met 10 times in reading, learners would need to read 1–2 graded readers
per week. The typical learner simply does not read this much.
(Laufer, 2005)”

Of course, we have all sorts of tools to blend it all together now, especially with direct card creation. ^_^ Interestingly, they seem to find that post-task explicit learning of words encountered in reading is very good, so that works well with reading followed by SRSing words you added with Rikaisan. They also found that adding audio is better than reading without, and both of those are better than just listening.

Edit: It also seems that with shorter texts with fewer total unknown words (e.g. 750 running words [word count of 750]), 1 unknown in 15 running words rather than 1 in 50 seems to fare better. Again, this being contextual inference in English, without glosses in the margins, etc.

Up until now I've taken the 'look up and anki every single word' approach. Though, us folks here at kanji koohii with an intuitive sense of how to learn say: if it's not working, change what you're doing. It's worked fine (GREAT) up until now, but as I'm approaching 17k words I'm finding it to be a lot more work than I want it to be. Still do-able and not so much of a hassle but it's not developing the "fluency" of reading that I also crave. This thread comes at a perfect time as lately I've been thinking about the next step foward in my study method.

As my vocabulary has progressed from roughly 2.5k in Jan 2010 to nearly 17k mid 2011 I have gone from being able to read short articles with heavy use of a dictionary to extremely hard novels/books with still, heavy use of a dictionary. My point being, as my level has increased I've always moved onto harder reading material to keep the learning going. This has been a wonderful approach but ever since the 15k mark I've noticed the law of diminishing returns has reared its ugly head. The other day I read something, an article I guess you could say, which had absolutely nothing in it I didn't know and I could read it at a near native speed with no comprehension problems. Ah, brilliant! I thought. It felt good. I can definitely see the value in taking a step back from the vocab crusade and just reading a novel aimed at maybe 小学生 or 中学生... I don't even care if its full of furigana this time. It's about time to delve into reading a metric shit tonne of stuff that really is at a level I (should have by now???) mastered.

Also, another consideration of mine lately is when I hit 20k vocab to go more monolingual and to stop piling stuff into Anki (only stuff with odd readings I'll never remember). I can understand definitions in a monolingual dictionary well enough now and I think it really is an important final step. I feel like that in a few months when I hit 20k, the study method will have reached it's usefulness limit and that I'll be at that point where I'm naturally ready to take that step of learning words through a process of osmosis like I do in English. I'll start glossing lots more instead of hitting S all the time and saving it to anki. Anki will be reserved for words that are, unique, special, worth remember or something I won't encounter but MUST remember if I see it again. There's lots of stuff that you're not going to encounter again for ages and it's not THAT much worth remembering even if you do... so I figure just gloss over it and when if it comes up again, just look it up again.

Almost feels like taking a step backwards considering that's what I used to do when I first learned vocab by listening to music and looking up words in the songs I listened to back when I first started learning Japanese almost 3 years ago.

However, I have a goal to read 100 books and I think if I want to read a book a week, this is the only way I'll be able to get through all that! If I spend the next 2 years reading a book a week in Japanese, I feel confident I will pick up that last 10k I'm after AND really really realllllllly polish my Japanese to a near native level. The subsequent 5 years (which will most likely be spent in Japan) will bring me to NATIVEEEE. Haha.

Anyone who learns this language is crazy btw and thanks to everyone for all the great discussions, tools, tips, tricks and resources over the years.

Fillanzea Wrote:Seamoby, you might find that novels written for young adults or adults (light contemporary fiction, not anything really dense and heavy) makes a good compromise between the difficulty of children's books and fairy tales, and the difficulty of Asahi news. Newspapers are densely packed with information in a way that requires a LOT of vocabulary, often very subject-specific vocabulary.

Hi, yes, thanks for the suggestion. I do try to pick my Japanese reading battles, and try to use a variety of sources, not just Asahi. Obviously, not all articles in Asahi are accessible to me at my current level. For example, the 天声人語 column is hard for me. But there are other parts which are readable and are good for vocabulary mining, for example, some articles in トップ写真, ライフ, or トラベル. The good things about these articles are they tend to be short, are "disposable", are varied in subject, and vary from day to day. Also, for me, what make Asahi and other online newspapers harder to read are the proper nouns (names, places, etc.).