Author Archives: adam

About adam

Autodidactic Polymath

[Method] Listening – Reading

Listening – Reading (L-R) is a method of language acquisition written widely about by a polish polyglot who went by many different internet noms des plumes.  Phi-Staszeksiomotteikirunandemoii, and buonaparte are the few I know of, but there are most likely many more of her(him?).

If you click on the link, you’ll rapidly find out that the description is very confusing, but it needn’t be.  The core of Listening – Reading is as simple as reading in your native language while you listen in your target language.  But there are a few caveats, cautions, additional steps and guidelines below.  For the most part, I’ve used the buonaparte’s own words drawn from many different threads and condensed everything down to just the essentials and edited for formatting and readability.

Buonaparte has also graciously shared a comprehensive list of materials he has prepared for this method applied to Japanese.  I’ve also written a companion article summarizing buonaparte’s specific adaption of L-R for learning Japanese.  All thanks go to Buonaparte.

The Materials

  1. A recording performed by good actors or narrators in the language you want to learn
  2. The original text (of the recording)
  3. A translation into your own language or a language you understand
  4. The text(s) should be long: novels are best

The Listening – Reading Method

The order ought to be EXACTLY as follows:
1. You read the translation.
If it is a text you have read many times in your life, much better.  You must be passionately in love with the text you’re going to study and know it well.  You only remember well what you understand and what you feel is “yours” psychologically.  

2. You listen to the recording and look at the written text at the same time.
Because the flow of speech has no boundaries between words and the written text does, you will be able to separate each word in the speech flow and you will get used to the speed of talking of native speakers – at first it seems incredibly fast.  If you’ve ever tried to listen to native speakers of any language, you must have noticed that at first you do not know which groups of sounds form words and that the speakers speak as if they were machine guns.  The aim of STEP 2 is to cure these two small drawbacks, and at the same time to get some exposure to meaning, sounds, rhythm, intonation in L2.

*  If you already understand quite a bit of the text, can recognize the boundaries between words and the speed is no longer frightening you can skip to step 3.  If the speed is frightening you go on until it stops being so.

3. You look at the translation and listen to the text at the same time, from the beginning to the end of a story, usually three times is enough to understand almost everything.  If you’re a fast enough reader you can read much faster than people speak, so you’re able to know IN ADVANCE the meaning of what you’re going to listen to, and to be in a position to guess at least some meaning (with a good translation almost everything) of what you’re listening to.
This is the most important thing in the method, it is right AT THIS POINT that proper learning takes place.  If you’re not capable of doing it without stopping the audio, you might decide to read a page (or a paragraph) and listen to the passage once or twice and go on.

Because of the IDIOLECT of the author the first 10-20 pages might be a nightmare for some, but then it’s getting easier and easier, the longer the text the easier it becomes, but it’s still the same IDIOLECT, variation after variation on the same theme, more and more celestial music.

The aim of STEP 3 is obvious: MEANINGFUL EXPOSURE, INPUT, LISTENING COMPREHENSION.  And ultimately: NATURAL LISTENING – understanding completely new texts without any crutches, you only rely on your ears and what you already know. It basically means you are able to understand NEW recorded texts (usually slightly simpler than the ones you have “listened-read”) without using any written texts, neither the original nor a translation and without having read them in L1 before.  I might add here: garbage in, garbage out.

4. You repeat after the recording, you do it as many times as necessary to become fluent.  Of course, first you have to know how to pronounce the sounds of the language you’re learning. Do not try to speak until you’ve reached the stage of natural listening.  Blind shadowing (without understanding) is a waste of time and effort.  Repeat after the speaker what you only understand (the meaning) and can hear properly (phonemes, rhythm, etc).  Listen-repeat – if it’s correct: listen-repeat, listen-repeat.  If it’s not correct, do not repeat any more, only listen.

5. You translate the text from your own language into the language you’re learning.
You can do the translation both orally and in writing, that’s why the written texts should be placed in vertical columns side by side: you can cover one side and check using the other one.

Caveats

  • The layout of the texts to learn is very important.   The original text and the literary translation should be placed in parallel vertical columns side by side.
    Sensory memories – visulal (iconic) and auditory (echoic)- are very short and disappear within a second, so you get lost when you have to look for words, they should CONSTANTLY be within your eyes’ and ears’ reach.
  • The translation:  a) interlinear, word for word (3 to 5 hours of audio) (for beginners)  b) literary, but following the original text as closely as possible.  Then you can check almost instantly whether you understand or not.
  • Listening to a short text time and again does not mean new exposure, it is still the same mechanical repetition. It might have its merits as well: you’re exposed to sounds, rhythm and intonation, but that’s about it, nothing more.
  • L-R is NOT watching subtitled movies.  You CANNOT read subtitles in advance, they appear on the screen at the same time as the characters are speaking, you have no time to pay attention to what you’re (mis)hearing, you concentrate on what is going on in the movie. Quite often, subtitles in L1 have very little in common with what is actually being said in L2.  By the way, L-R (reading in L1 with an occasional glance at L2 and LISTENING to L2) works MUCH, MUCH, MUCH better than just reading in L2 and listening to L2. I know, I’ve done both, girls and boys and both.
  • L-R is NOT Assimil and suchlike.  What makes L-R different is: 1. using long novels right from the start in fully bilingual format, with bilingual etexts in vertical columns with matching cells, side by side on ONE page, recorded by professional actors.  2. Step 3 (= listening to the target language while reading in a language you understand.  3. Using self-explanatory texts (= knowing the content beforehand, both the meaning and emotionally)  4. speaking and writing only after the incubation period, that is after getting to the stage of natural listening.  5. the Assault (= massive exposure in a relatively short time)  6. taking into account all the subsystems: pronunciation, grammar, vocabulary and discourse (= how to produce texts), discourse in textbooks is artificial and often wrong).  7. And that’s true, it IS the cheapest way of learning a language, both in terms of money and time.

How it works

GRAMMAR AND VOCABULARY  are in the texts,  why should you bother with lengthy and often wrong explanations?  When L1 and L2 are not closely related, say English and Japanese or to a lesser extent Polish and Japanese, you might want to read some basic information about L2 grammar, but nothing more.

READING  When you’ve done the right amount of listening-reading with parallel texts, you don’t have to learn the skill separately.  With languages using a different script, say Japanese for Indo-Europeans (us, unlucky bastards), ‘listening-reading’ saves a lot of toil, thousands of hours compared with traditional methods using textbooks and flashcards.

WRITING  After the right amount of exposure to complicated texts with full and beautiful DISCOURSE, a little bit of written retranslation from L2 to L1 should be enough.  You don’t need to translate whole books, though, only the phrases or sentences you feel you wouldn’t be able to say or write yourself.

 ‘Listening-reading’ is a SYSTEM (a set of interdependent elements that mean something as a whole). If you skip or omit one element, the structure crumbles.  Let me be stubborn once more.  Listening-reading’ is a system and that’s its only advantage. Not its particular components, not even STEP 3.  You can incorporate some elements into your own learning, but to exploit L-R to the maximum it is much better to use it as a whole.

Progress Report: [841d]::[968hr]::[2329vocab]::[787kanji]::[2312sentences]

After I wrote this, I completely changed gears and didn’t do anything I planned out here.  Instead, I’m continuing studying RTK until I finish in about a month.  Then I will start on native media.

The first thing you should notice is that I’ve added 421 new kanji since my last report.  This is for 2 reasons.  The first reason is that I’ve been wanting to start reading and I don’t want to spend too much time finding material with complete furigana coverage or adding it.  The other reason is that I’ve decided to study for the JLPT N3 test in December.  I haven’t decided to actually take the test, but studying for it is forcing me to round out my stills into something more useful for actually using Japanese.

Overall, kanji went quite fast compared to vocabulary.  I was able to add about 10 kanji for every hour I studied instead of 6 or 7 vocabulary words.  The kanji I added were the ones tagged as N3 in my kanji deck.  I’ve also suspended any unstudied non-N3 vocabulary, so I have about 80 more vocabulary to study.

As of today, I’m starting 2 new cards – audio only vocab to start getting better at listening comprehension, and kanji only vocab to get better at reading without the aid of furigana.  I’ll continue to mature the old cards for now.

June Goals:

– Audio only vocab cards.
– Vocab cards without furigana.
– Add the last 87 N3 vocab words

July & beyond:
– NHK easy
– Finish Tae Kim
– Start taking practice jlpt tests
– Reading parallel texts
– Listening practice (podcasts, subs2srs, nhk easy, buonaparte’s listening-reading)

Analytics: The difficulty of finding leeches

When I first started thinking about leeches, I assumed cards that were easy or hard in the learning phase would stay easy or hard in the review phase.  I’ve noticed those problematic cards that I had a hard time getting out of learning and are still giving me trouble months later.  If I could just find the cards which were giving me trouble early on, I could just suspend them and learn just the easy cards.  Presumably I only remembered the cards that continued giving me trouble.  Because unfortunately, the next set of charts shows I was not a very good judge of what was actually going on.

The following charts show the relationship between how many reps it took to get each card to an interval of 7 days and how many reps to get the same card from interval 7 to interval 90.  I was fully expecting to see a nice relationship where difficult cards would stay difficult.  In a scatter chart, you would see a tight grouping of dots sloping from the lower left to the upper right.  Instead, what I got was the following set of charts where just as many easy cards became difficult as difficult cards became easy.  This exercise is making me think that it will be difficult to find leeches with any accuracy.

coresentence_ivl7v90 jfbp_ivl7v90corevocab_ivl7v90rtk_ivl7v90tk_ivl7v90

Anki Analytics: Card difficulty

In a follow up on my post on leeches, I graphed the amount of reps it took to get each card to an interval of 90 days.  Again, the chart for kanji is the odd man out with it’s plot being more linear than the others, suggesting that different types of memories behave differently.  But even with kanji, we see that the most difficult cards take many multiples the number of repetitions that the median card takes.

By my calculations, the easiest 80% of my core vocabulary cards takes roughly the same number of reps as the hardest 20%.  In other words, I could learn 4 easy cards in the same time it takes to learn 1 of the harder cards.  It sure would be nice to identify those difficult cards early somehow.

low high mean median
Tae Kim 5 20 8.386740331 8
RTK 3 98 37.78409091 34.5
Core sentences 2 59 9.30834753 7
Core vocab 2 194 30.17112299 19
JFBP 5 164 19.59797297 7

rtk_ivl90 tk_ivl90 jfbp_ivl90 corevocab_ivl90 coresentence_ivl90

Analytics: Leeches

This is a new series where I combine a few things that I am currently learning into a topic I have no business pretending to know anything about.  In addition to teaching myself Japanese, I am also attempting to teach myself programming and also data analysis.  Although it’s going very slowly, I am hoping to figure out a few things that will hopefully make the ankiing a little more efficient.

My first target is those damn leeches.  Leeches are what anki calls those cards that you keep forgetting over and over.  According to the supermemo site, around 50% of your time can be spent learning 2.5% of the material.  That 2.5% that is taking half of your time are leeches.  Depending on your goals, wouldn’t it be nice to be able to identify that 2.5% of material and spend that 50% of your time learning twice as much?  Personally, I would rather learn 97.5% of core twice as fast before spending the time to learn that last 2.5%.

Unfortunately we don’t know what those 2.5% hard vocab words are, and even worse, anki doesn’t give us nearly the tools to find them.  All that anki gives us is a setting that once you fail a card more than a set number of times (default is 7), anki will suspend that card.  The thinking being that you are more likely to learn a new card in less additional time than keep trying (and failing) to learn the one you’ve failed so many times already.  But I’ve always wondered what setting has you learning the most amount of material in the least amount of time?

This is the question I set out to answer.  I wrote a small program that counts the number of reps to either learn a card or become a leech.  I considered a card to be “learned” once it’s interval surpassed 4 months.  I did this for all cards, and averaging the reps to learn a card and the reps to become a leech for every card I’ve studied.  The result is the average number of reps it would take to learn a card assuming a given leech threshold in anki.
image (1)

The above graph shows the results for the 4 decks I’ve been studying.  The first thing to notice is that “core sentence”s and my” Japanese for busy people” decks are much easier than my “core vocabulary” and “kanji” decks.  The other thing to notice is that for all decks except for kanji, setting the leech threshold to the lowest setting results in learning the most number of cards in the fewest reps.  Kanji appears to be most efficient setting the leech threshold to 8, but any number higher than 4 appears to be just fine.  The final thing to notice is that all of the vocabulary and sentence decks appear to have a similar curve, and a very smooth one.  I take this to suggest that for all vocabulary decks I study, setting leeches to the lowest setting will result in learning the most amount of vocab words in the least amount of reps.  However this isn’t the only consideration.
image (3)

The second graph shows the ratio of learned cards to suspended leeches for each deck and each leech threshold.  As you can see with the “hard” vocab and kanji decks, at lower thresholds anki is suspending more cards than I would be learning learning.  In fact, setting the leech threshold to 1 for core vocab and kanji would result in learning only 18% of the vocab deck and 6% of the kanji deck.  This is hardly desirable, but finding a good balance between efficiency and completeness might make sense for some people.  For instance, setting the threshold to 9 for kanji and 6 for core vocab gets me in the 50-60% coverage range.  That still seems less than optimal to me, but something that I have to think about as there is no clear cut answer unfortunately.

That’s it for now.  Please put you thoughts, criticism, praise and especially suggestions in the comments as I’m happy to make this better with your help.

 

Progress Report: [806d]::[910hr]::[2329vocab]::[309kanji]::[2312sentences]

Stats:

  • 2312 sentences
  • 2329 vocab
  • 309 kanji
  • 910 hours (67.8 min/day)
  • 806 days
  • 264,871 reviews

Kanji

Less kanji then last report because I lowered my leech threshold to 4 and suspended anything less than that.  I started adding kanji a few weeks ago, so the number should go up again.

Sentences and Vocab

I stopped adding sentences at the same time as vocabulary.  Basically sentences are a lot easier than vocabulary getting hints from the context.  Consequently I don’t need to study sentences as frequently.  Since I stopped adding sentences, I’ve been able to add vocabulary cards much more quickly.  Once I get several hundred vocabulary, I’ll add sentences with an interval of 30 or so.  It worked so nicely when I did that before.

Understanding

Finally I feel like I’m getting somewhere practical where I could be able to actually use what I’ve learned some day.  I got some audio files of NHKeasy to play in the car.  Although I can’t follow along at speed, I feel like I recognize most of the words.  Same with some Japanese conversations but to a lesser degree.  Cramming vocabulary is looking to be less and less urgent now.  I need to figure out a way to start studying listening comprehension to be able to follow along at speed.  I feel like I could really make a quantum leap in understanding conversation if I study this effectively.

My story

  • Late 1980’s Took Spanish in High School.  I didn’t do too well in class, but I enjoyed learning the language.
  • 1996 Moved to a city where more people speak spanish then english.  Unfortunately I forgot most of the spanish I learned.   I didn’t pick up any spanish “from immersion”.
  • Summer 2003  Met a Japanese girl at a party.
  • Winter 2003   Listened to pimsleur while working.  Unfortunately I learned nothing from it.  Listened to pimsleur in the car…learned nothing useful for my upcoming trip, but learned a few vocabulary words.  Went to Japan for the first time with the girl.
  • 2004 kept listening to pimsleur on and off.  Mostly off.  Played around with rosetta stone.
  • 2005 Girlfriend got a job in Japan.  I wanted to try living in Japan, so started taking a class.  the textbook was “Japanese for busy people”.  Finished JFBP I.  The plan was for me to move to Japan and try finding a job there.
  • 2006 Started realizing that it would be impossible to work in my career in Japan.  Got busy with work and stopped going to class halfway through JFBP II.  The only good thing was that I visited Japan often.
  • 2007 Got married to the Japanese girl from the party and she moved back to Los Angeles.  I’m not studying Japanese at all now.
  • 2009 Got a smartphone and thought I could learn something while walking the dog around the block.  Downloaded a few japanese learning apps.  Started casually learning kana.  Started listening to pimsleur in the car again.  Listened to level 1 and 2.  Always frustrated that I didn’t have flashcards to supplement the CDs.  Frustrated that the school where I used to take classes seemed to close down.  Just generally frustrated buy not knowing what to do.  Eventually dropped the whole thing.
  • 2013  Got interested again and tried a few different free flashcard apps.  Reviewed my hiragana so that I could learn some Japanese vocabulary.
  • Feb 2013 Decided to buy the anki iOS app and downloaded the JFBP deck because I had the textbooks and wanted to review the vocab that I already halfway knew.  Officially started anki on Feb 20 2013.  Studied every day and realized that every day I studied made it that much harder to break the chain.  Finished JFBP I vocab.  Found nukemarine’s guide and realized that the resource I was designing in my head existed and it’s name was core.  Started learning core vocabulary and kanji (via an rtk deck).  I suspended katakana words because I wanted to learn japanese words.  Along the way, I started adding sentences.  Realized that after several months of study I still couldn’t understand anything on Japanese TV.   Decided sentences and kanji could wait and I needed as much vocab as I could handle, Unfortunately I decided that learning vocabulary rendered in kana instead of kanji was a good idea.
  • Spring 2014  I was getting impatient with my progress around core 2k and I still couldn’t understand much on TV.  Realized that I needed to learn grammar and exposure to inflected verbs, so I decided to start adding sentences again.  I raced through the sentences because I already know most of the vocab, suspending cards where the vocab prevented my understanding the sentence.  After about 2 weeks of sentences I realized that my vocabulary accuracy improved dramatically.
  • Summer 2014 Core sentences are caught up with vocabulary.  Decided to start Tae Kim.  Already being vaguely familiar with some grammar made for a lot of ah-ha moments.  Finished 2/3 of Tae Kim (Basic and Essential).  Started adding vocab and sentences in parallel.  Ran sentence gloss on my sentences so I can easily look up unknown words in the sentences.
  • 2015 Realized that reading kanji was important, so I started RTK in earnest.  Took me 8 months to finish.  I did recognition (kanji to keyword) because I had no interest in writing and writing was a significant burden to study.  In hindsight this was not ideal and if I did it again, I would probably learn the traditional radicals as an aid to learning the kanji via vocabulary.  RTK took me 8 months to finish.
  • 2016  RTK review hell.  Mostly because my memory sucks, but also because of inefficient study methods.  Also started studying song lyrics and japanese pod101 beginner dialogs.  Regarding jpod101 I pre-learned the vocabulary and then listened to the dialogs and read them too.  This was huge actually.  Longer texts are superior to single sentences.
  • Spring 2017  Continue with Jpod101.  Focus on getting down vocabulary reviews to make room for reading and listening.
  • future Probably get more into native material.

Set my leech threshold to 5

I noticed that the majority of my high interval cards have very few lapses.  The majority of my cards with several lapses have intervals under 30 days suggesting that the ones with lapses are destined to take a larger share of my time.  This is just a note (to myself mostly) that I just set my leech threshold to 5 from 7 and suspended all cards with 5 or more lapses.  I ended up suspending about 400 cards which breaks my heart, but my workload went down by 50 cards tomorrow, and 400 this month, so I’ll be able to replace the suspended cards with easier ones rather quickly.  I might even set my threshold lower if this works out as well as I think it will.

Edit:  My first day after suspending ~400 ‘leeches’ I had an extra half hour to learn new words.  Consequently, I was able to add 40 new cards the first day.  I spent a little more time time than normal yesterday, but I didn’t have to worry about getting killed with reviews today because my review total today was just average.

Some stats:

Cards with an interval over 1 month: 4740 total
0 lapses: 3242  <—-68.3% of my month+ interval cards have no lapses
1 lapse:  620
2 lapses: 328
3 lapses: 193
4 lapses: 133
5 lapses: 102
6 lapses: 93
>6 lapses: 29

all intervals: 6520 total
0 lapses: 3122
1 lapse:  881
2 lapses: 527
3 lapses: 346
4 lapses: 268
5 lapses: 197  <—these
6 lapses: 204  <— and these were suspended today
>6 lapses: 976 <—-these were already suspended leeches

 

Phonetic Radicals: How to guess a kanji’s reading

I was reading something today which mentioned phonetic radicals.  What’s that?

“The right side is usually what’s known as the “phonetic compound.” This portion has a specific reading attached to it. If you see this phonetic compound, you can sometimes guess the reading of the kanji. Sometimes by learning one phonetic compound’s reading you can know how to read six or seven other kanji that contain it.”

Apparently 67% of kanji have phonetic radicals, so it would be very useful to learn them.

Why are you learning Japanese?

People have different reasons for learning a language.  Some people just want to be more worldly and others like the way Japanese sounds.  Others want to move to Japan or at least travel there.  Personally I am married to a person from Japan and want to fully enjoy Japanese things with her and I want to be more independent when I am there.  I want to understand things as she understands things and not have to have them explained to me.

I’d love to hear your reasons for learning Japanese.  Please let me know in the comments.