Author Archives: adam

About adam

Autodidactic Polymath

Progress Report, Next Steps and JLPT N3

Added another 300+ sentences including tae kim basic.  I believe that’s 22/day and I have another ~500 to add.  So if I keep this pace, I can catch sentences up with vocab in under a month.

After that:

I re-read nukemarine’s suggested guide and realized that if I do another 500 tae kim sentences and 600 kanji, I’ll be at around the JLPT N3 level.  I’m not sure about the Kanji yet, but I believe I will progress further into Tae Kim Grammar once I catch up with my core sentences.  I might even filter some Tae Kim in along with the core sentences or at least pre-read the lessons.

As far as kanji goes , I hesitate to spend time studying kanji because it won’t make me any better at speaking or listening.  However, I have begun to understand that once I get into studying native material, being able to read kanji will allow me to study a much larger variety of Japanese media.  Tentatively I am considering at least memorizing the 200 some odd primitives to make kanji easier once I get back to it.

And speaking of native material, I am planning on practicing on NHK Easy News rather soon.  I just went over there and realized that I can’t follow along too well.  But I think it would be a good exercise to spend some time looking up all the unknown vocab and practice reading until I can understand a whole article and the video.  I think it’s important to keep challenging myself and move on from sentences to paragraphs.  It’ll also be good to do something a little more fun and useful.

Also:

I’m also considering playing around with a few different mnemonic methods to see if I can increase my speed a little.  I’ve heard some good things about memrise.com, memory palaces and also keywords.  Considering giving one or all a try for a month each and see if there’s any noticeable difference.

1975 Study: 35% better vocabulary retention using keyword method.

Mnemotechnics in second-language learning.

ABSTRACT Discusses the author’s keyword method, a mnemonic aid for vocabulary learning. According to this method, a foreign word is connected to its English translation by a chain of 2 links-similarity in sound (acoustic link) and a mental image of the interaction between the 2 words (imagery link). Experiments using Russian and Spanish vocabularies are described to illustrate that the keyword method produces significantly greater vocabulary recall than the rote-rehearsal method of second-language learning. Results also suggest that (a) providing the keywords for the S is better than having him generate his own keywords, (b) imagery instructions have a significant advantage over sentence-generation instructions when the keyword method is used, (c) the keyword method does not affect word retrieval times, and (d) the keyword method produces significantly better backward associations than the rote-rehearsal method. Problems in the application of the method are considered.

http://www.researchgate.net/publication/232565697_Mnemotechnics_in_second-language_learning

New Paper suggests singing can facilitate foreign language learning

A new paper in Memory and Cognition suggests Singing can facilitate foreign language learning

Abstract

This study presents the first experimental evidence that singing can facilitate short-term paired-associate phrase learning in an unfamiliar language (Hungarian). Sixty adult participants were randomly assigned to one of three “listen-and-repeat” learning conditions: speaking, rhythmic speaking, or singing. Participants in the singing condition showed superior overall performance on a collection of Hungarian language tests after a 15-min learning period, as compared with participants in the speaking and rhythmic speaking conditions. This superior performance was statistically significant (p < .05) for the two tests that required participants to recall and produce spoken Hungarian phrases. The differences in performance were not explained by potentially influencing factors such as age, gender, mood, phonological working memory ability, or musical ability and training. These results suggest that a “listen-and-sing” learning method can facilitate verbatim memory for spoken foreign language phrases.

Review: All Japanese All the Time (AJATT)

A lot of people have strong opinions about AJATT.  Most either love it or hate it.  I’m kind of in the middle.  On one hand there’s khatzumoto’s prose style and the sheer brute force nature of his method.  On the other hand, khatz is is clearly an intelligent, well read person and there is some actual insight even if you don’t go in for the full AJATT method.

Executive Summery:

  • The AJATT website is verbose and hard to navigate so it take a while to find what his method actually is, so you’ll spend a lot of time reading english instead of learning Japanese.
  • Khatzumoto’s bombastic personality turns a lot of people off.
  • Spending every moment of your life doing things in Japanese is probably effective, but not appropriate for everybody’s lifestyle.
  • His method might be the most efficient for becoming fluent, but you will go months/years before you can even use the language if you aren’t studying all day, every day.
  • The AJATT method is probably best for people who want to learn as fast as possible and don’t mind putting the rest of their life on hold to do it.
  • If you can parse the website, he has some helpful insights everybody can use.

First off is the website.  Maybe you should just click on it and go there to see what I’m talking about and then come back for my explanation.  It’s impossible to find anything explaining his method.  I guess people complained about this so now he has an overview, a table of contents, how to use this website, and “nutshell” trying to explain what the website is all about and it’s still hard to figure it out.  If it wasn’t such a popular website, I might have gone there once and never came back.

Keep reading and eventually you’ll either leave in disgust or find something that actually makes a lot of sense.  I can’t say what will make sense to you because a lot of what he says is controversial (like not studying grammar or vocabulary) but I found the not studying grammar part interesting, so I went with it.  And that’s how I would recommend approaching this site.  I believe it is worthwhile clicking around and reading and if you don’t agree with some things he says, nobody says you have to do it.  But eventually, you might read something that changes what way you think about language learning or motivates you or whatever.

Key points of his method are:

  • Always have Japanese going in the background (tv, music, talk shows…) for passive immersion.
  • Learn all 2000+ general use kanji first using RTK and Anki.
  • SRS
  • Learn kana after kanji
  • Don’t learn kanji readings
  • Mine 10,000 sentences, learn and practice reading them
  • Don’t learn grammar rules
  • Don’t learn individual words – learn sentences
  • Don’t translate Japanese sentences into English (use native translations or monolingual J-J dictionary)
  • Start here and here if you want to get a taste of what AJATT is all about and draw your own conclusions.

 

At the heart of AJATT, as the name implies is having something Japanese going “all the time”.  In other words, passive immersion such as having Japanese TV, music, podcasts or whatever on in the background pretty much 24/7.  People can split hairs about how well this works, but I can attest that it will work, but not very well.  I have learned a few Japanese words by listening to Japanese music before I started studying he language and without even trying.  Once I started studying, hearing occasional words that I know float past my consciousness served to reinforce the word in my memory.   However, I only learned a few words over 100s of hours of listening which isn’t that efficient.  But the point is that something is better than nothing if there’s no cost.

My feeling about this for my life is that there actually is a cost to me because I like to listen to podcasts and music and I get more enjoyment out of it than a bunch of what amounts to gibberish at this point in my study.  Eventually when I get further along in my study however, I do plan on listening to more Japanese media and I expect I will learn more and it will be more enjoyable – changing the equation.  However, if I was moving to Japan in a year and wanted to learn as much as possible, I would probably start listening immediately and often.

Another significant part of the method is learning all 2000+ kanji before anything else.  Again, I believe this is prudent advice for some people, but not others.  Since I only have about an hour per day to study, it would take me about a year to learn them all and I’m not sure if I would have the motivation to study an entire your without something that I could put to use sooner.  Also, I am more interested in speaking and listening asap and feel that reading can wait even if that order is not the most efficient.  However, I know that Japanese is very much related to kanji and knowing kanji from the get-go makes everything easier.  So, again for the person who is moving to Japan in a year, learning kanji before anything else is the way to go and that’s not controversial.

One thing that IS controversial is not to study vocabulary or grammar.  If you’ve read anything about how I am going about things, I’m all about vocabulary but I’m on board with the not studying grammar.  I believe the reason for not studying vocabulary is that in khatz’s method, you are using the kanji as a key to the meaning of a word and you have readings as part of the answer section of your flashcard.  In the context of the AJATT method, this makes a certain amount of sense but since I’m going about things in a different order, it’s hard for me to comment on this.  I can say that I’ve been able to read some things based on the 400 or so kanji that I know with more ease than I would expect, however I also know a lot of vocabulary.  I really can’t see how explicitly studying vocabulary would be harmful or inefficient.

Learning grammar by reading rather than studying grammar explicitly is something that I can comment on and actually it’s one of my strongest convictions.  I’m not saying not to study grammar at all, don’t start out studying grammar.  Rather do a bunch of reading and get a feel for how sentences are put together.  It’s rather easy to understand simple sentences with a noun, verb and a particle and requires no explicit study.  Later, when you get into sub-clauses and more compile grammar read the sentence a few times and then look up the grammar structures that are a bit confusing as needed.  AJATT makes similar suggestions (if you can find them).

The bottom Line:

I imagine for some people, AJATT is more or less what they should do if they want to learn Japanese fast and don’t mind giving every moment of every day for 18 months over to Japanese.    I, however have a life and I’m not in such a hurry to learn Japanese.  I prefer to learn at a slow steady pace and have a good time doing it.  I prefer to learn vocabulary immediately so that I can use it immediately and kanji can wait(even though I know that it’s probably more efficient to learn first).  I would get discouraged and quit if I spent a year studying kanji and couldn’t understand a single Japanese word much less sentence.  Even if you’re not ready to go full AJATT there’s probably something for everybody on the site if you don’t mid digging a bit and can get past the author’s tone.  Personally I go there on occasion and read some of the most popular articles to get ideas and mix up my thinking a bit and that’s probably the best recommendation I can make.

 

 

Beyond vocabulary

It’s 2:30am and I should be sleeping but I’m going to write my goals for the next few months. I just realized that it’s about time to start writing in Japanese. Maybe not today or tomorrow, but soon I will start writing a journal in Japanese. Probably just a sentence or two per day at first, but hopefully more later. After that I will probably make an italki account to get someone to write to.

I’ve started adding tae Kim grammar cards to the mix. Added 150 last week so that does it for basic grammar. I rescheduled them between 1 and 15 days so they don’t bog me down but I still see them fairly often. On the grammar cards I’m really trying to be strict with the the grading because the cards are pretty literal translations. I’ll review those for a week or so and then probably start adding essential grammar cards.

I’m still adding core sentences too. I thought I only had 500 more left to catch up, it’s more like 1000. This should keep me busy for a while.

Progress Report: [491d]::[524hr]::[2342vocab]::[320kanji]::[997sentences]

The obvious change from last report are all those sentences.  I think it’s over 500 new sentences in the last month or so.  I’ve given up adding any cards other than sentence cards until sentences catch up with vocabulary.  Since I know the vocabulary in the sentences I’m making short work of them.  About 800 of the vocabulary are not from core, but from another deck, I am really only at about 1500 core vocabulary.  So that looks like I have about 500 more sentences until I’m caught up.

I’ve noticed that I can do dictionary lookups within anki(I think it’s an iOS function – just press and hold and you get a context menu).  That’s really helping with sentences because I can remind myself of words that I forgot and I’m even keeping cards that I don’t know all the words because I’m learning hem from context.

My deck is a little hard to comprehend right now because I integrated it into the optimized deck and there is a lot of inconsistency.   So it’s hard to figure out which cards still need to be unsuspended.  Also, I’m suspending sentence cards if they at all confuse me, so it’s not clear I’ll be able to get to 1500 sentences without a lot of figuring things out.

Things have gotten a lot easier

For some reason, and I’m not sure exactly why, my accuracy has gotten a LOT better in the past month.  From about 90% to over 95%.  I think it has a lot to do with adding the sentences because I have pretty good accuracy with the sentences and I’m not being to strict with them.  I also think it is because I’m getting a lot more vocabulary practice just reading the sentences because there’s around 3-8 words in each sentence so it ups my vocabulary accuracy too.  I’ve also stopped adding vocab cards, so the cards are getting more mature.  Whatever the reason, it’s a very welcome development.  I wish I would have started earlier with the sentences – probably around 400 in the core vocab.

What’s next?

From here I will continue adding sentences until I catch up with the vocab cards.  That probably shouldn’t take more than a few weeks.  Then I will start adding both vocab and sentence cards together for a while and see how that goes.  There are a few grammar structures that I don’t completely understand, and I’ve been making them, so I will probably look those up.  I will probably browse through tae kim again and start adding some tae kim grammar cards.

Sentences are Easy!

I’ve been adding sentences again for the past several weeks and I’m happy to report that sentences are really easy!  I guess it comes as no surprise that if you know all the words, reading japanese sentence is really no sweat.  Filing in the cloze deleted word is a little more challenging, but still nothing to slow me down as much as a fresh vocabulary word.

Of course I’m stacking the deck where any card that contains a word I don’t know is suspended.  However, I’ll unsuspend them in a few months and hopefully I’ll know all of the words.  I’m even noticing sentence cards that I’ve suspended before because I couldn’t understand them, and now I understand them quite well.

So far I don’t have a count of the new cards which I’ve gone through.  I’ll update the count and forecast a timeline once I get a few more weeks for sentence study data to play with.  If you haven’t already, consider reading my recent post on adding Anki cards without getting bogged down.

 

Lessons learned – what’s working.

I’ve written most of this before, but I thought I’d put it in one place.

What’s working so far:

  • learning kana first
  • skipping kanji
  • finishing reviews every single day
  • focusing on vocabulary
  • using anki
  • suspending katakana vocab
  • studying new vocab at the beginning of he day

What I’d do differently:

  • start with the optimized core deck from the very beginning
  • start sentences sooner(around 500-100 vocab)
  • don’t bother with other methods (pimsleur, rosetta stone, actual classes..)
  • skip kanji completely (I learned ~400.  I could have learned 400 more vocab instead)
  • don’t cram too much
  • prioritize accuracy over speed