Developing Super Memory Pt1 Retrievals vs Recognition

まだ早いだな。。

 

 

 

There is an important, probably THE most important distinction to make when looking at the “goal” of memorization. For any goal, we must establish criteria that satisfy the terms of what we are trying to do. What is memorization defined as? Generally, memorization  is possessing the ability to recall your target without reference, quickly (a few seconds) in 1-3 seconds.  Meaning, if i’m trying to remember the word for “by chance”, or “accidentally”, which is 偶然(ぐうぜん)guuzen, the fact that I typed that word from memory immediately means I have it memorized.

The process of remembering the word with no reference is called a retrieval. Meaning, I had to “retrieve” the word from my mind. If I were to see the word, for 偶然 in reading, I’d be looking at what I call the word shape. The goal of learning a new language is really just a very complex approach to memorization of words or word shapes, patterns and phonetics. You work to “retrieve” these words, word shapes and patterns and memorize the mimicking of the pronunciation of the new language. This is all data. But it is a LOT of data, hence the tendency for people to stress over “difficulty”. But really, your brain’s capacity for storage is relatively endless and when we look at our task (relative to our true ability), especially having mastered a language to fluency already (your native tongue) there is no need to focus on limitation in that regard. What you will experience is significant challenge, but know this: everyone who gets the results you want has done the same thing. 

Retrieval Is Everything

Retrieval is the true gauge of where you are. You can trick yourself into thinking you have something “memorized” because you can read it, or after seeing it in print you can “somewhat” recall what it is. This is recognition based, which has obvious benefits, but massive perceptual disadvantages if you use this as your only guage for memorization. Remember you want to be able to eventually produce what you see and read with reasonably ability. This is a retrieval activity. You can speak your native tongue fluently because you have made millions of phonetic retrievals in your life. It took you a few thousand to get the hang of it. Only by training to bring things from memory, do we truly memorize a target word, phrase or construct. So to define memorization as being a full memorization, it is a word, phrase or construct we can both recall & recognize. 

Hold on to this idea of “full memorization”  as we continue throughout this article.

Retrieval and Priority

The brain prioritizes what to memorize by exposure frequency and a certain kind of relative stress. Seeing words over and over in Anki, or trying to read hundreds of sentences per day forces your brain to “retrieve” what you have stored into memory, which will eventually go into long term memory. This means that your process of retrieval must be as frequent and efficient as possible. As i’ve said on this blog before, we don’t have the luxury of a child who has 10-12 years to learn their language in small doses. We must utilize methods to create very high level of exposure to set the stage for memorization, but also do so in a way that we do not get completely overwhelmed or burn out. Let’s look at how this applies to our “approach mechanics” for efficient tasks.

The way I like to look at it, is this:

Can I design what i’m doing in such a way that I can get (a) VERY high exposure rates with (b) as LOW a time per data unit as possible.

By “data unit”I mean, word, phrase, etc. So I want to have VERY high exposure and also VERY low time per item i’m trying to memorize.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This isn’t my typical day, this just happened to be yesterday. For my deck “Master of Vocabulary” I was averaging 9 seconds per card, for almost 900 cards. this means i’ve had 900 retrievals of my target information. Let’s say, I do 900 retrievals, and then read 500 sentences. I’m crossing the 1,000 retrieval mark per day. What i’m doing is pushing the limit of what I train to pull from raw memory each day. But to do these numbers requires unusually efficient approaches. Let’s take a look at why:

Retrieval and Approach Problems

Most people assume you just dive in and start trying to read a sentence with your target word. In the learner’s mind, they think they are just focused on the target word, but all sentences are composed of words we may or may not know. We must have certain rules in place to not limit ourselves. When the brain sees too much data it does not understand, it freezes up. This “brain freeze” is what causes us to stop, which we must avoid. A child learns the English alphabet, then small words, then very simple sentences, then more complex ones. This is a “progressive load”, like a weightlifter adding more weight as he or she goes through a routine. Always remember that your memorization is also progressive. Your brain can handle hundreds of short bits of information, but will freeze if the data is too large (too early).

It’s like giving a baby an X-men comic to read. They’d look at the pictures and lose interest because they can’t understand anything that characters are saying. Think of yourself. Are you trying to read an X-men comic at a baby level of understanding (and approach)? I’ll give you an example of this:

Our target word is 役割(やくわり)yakuwari – role, part someone plays.

Here is an example sentence:

人はちゃんと自分の役割を果たしていたもの。hito wa chanto jibun no yakuwari wo hatashiteita mono.  –  People carried their own weight then.

Just visually, this is very noisy if you can’t read basic Japanese. Secondly, if you don’t know Kanji you can’t read “人” ”自分”、”役割”、”果たして”。Also when looking at sentences, you can’t predict the level of what you are reading. For example the expression:

果たして (はたして)hatashite「as expected, just as one thought] is an N1 grammar pattern (a very advanced pattern). You also need to know the meaning of ちゃんと(earnestly), and how いたもの works in this sentence. 果たす means “to fulfill” so the sentence literally translates to:

人はちゃんと自分の役割を果たしていたもの。The people earnestly all of their roles fulfilled.

So this sentences does NOT use the “果たして”, form, but the “果たす”form of this Kanji.

I picked this word (and sentence randomly) to illustrate a point. Just by trying to learn this ONE WORD, in this ONE sentence, we must learn several other words, and high level grammar (that could be quite confusing).

Now multiply this instance by 4,000 and you can see why people get into monstrous problems and say these things are impossible. I’m showing you this sentence because this is the type of sentence that often a complete beginner will try to learn. Even if you are “advanced” this is a tough sentence to try and process if your goal is word memorization. 

This process fails miserably (at scale) because in trying to read this sentence you are not doing two things: (a) training the retrieval of your target word with speed (b) keeping your time with this target word significantly low to gain value from the exercise.

A person could spend 25 minutes to an hour just breaking down the components of this sentence, for one word! 

Retrieval and Improved Approach

If you are still with me, understand that we must follow these rules blindly. We must set things up so that we can (a) see our target word OFTEN, keeping the frequency high and we must (b) ensure that this process is not clunky and slow. It must be fast to be effective. The good thing, is that this is where things get very cool. A few years ago I exploded my research and understanding of memorization through a website called Vocabulary Labs. It is the first place I really started to get a different sense of how to look at memorization relative to the brain, word segmentation and the idea of retrieval based training. There I saw a brand new format of flashcard, designed to ensure we hit our two targets. I did this while training my German words, and have since adapted it to Japanese.

If i can read 役割 I am technically not “retrieving” it from memory. But you say to yourself? Isn’t the goal to be able to read these words? The answer is yes. BUT what’s to stop me from confusing yakuwari 役割with words like yakusoku  約束(promise), yoyaku 予約 (anticipate) . With enough exposure these are never ‘confusing’, in the same way you don’t confuse the word “promise” with “guarantee” with “assurance”. Only by using these words, or seeing the proper usage a certain number of times do we reach ‘full memorization’. 

Why This is Important

This is important because it sets the stage for what we actually have to do. Not only are we memorizing these words, but we must design a way to use these words. Fortunately, the word “usage” has broad definitions, and i don’t want anyone here to panic. Remember, by “retrieving”, or “recalling” a word from memory, you start to store it into longer and longer term memory. We can retrieve/recall a word a few ways:

  1. we can retrieve the word by reading it alone
  2. we can retrieve by reading a very short sentence
  3. we can retrieve it in a longer contextual sentence
  4. we can retrieve/recall the word by itself (consciously)
  5. we can recognize it in spoken speech

When you are trying to do things at scale, this type of understanding is crucial. You want to figure out what you can do the most, with the most benefit without it becoming unusually stressful because it is so much data. But as we’ve established, you want to keep exposures super high and time per exposure as low as possible. From this list, the quickest things to do are either: read a single word, or read a very short sentence. The latter three all take time you can’t gauge necessarily. Reading a long sentence filled with words and grammar I don’t know can be a serious waste of time. You might “feel like” you are getting somewhere, but you could have spend the 25 minutes from that one sentence doing hundreds of revisions at a few seconds per word. Recongnizing a brand new word in spoken speech is a matter of memorized phonetics, which depending on the speaker, what you are watching and how clear it is, can be noisy or clear as day. Phonetic memorization comes only after several hundred hours of exposure to your target language. This means that trying to recognize a brand new word in speech is a waste of time until  you get a handle on the way the language sounds. You generally know you’ve reached this point when: you can hear everything a person is saying clearly, even if you don’t know what they are saying. 

So only by training these things at high volume in short chunks that allows the brain to process it, do we allow for “elite training”, a way to really put pressure on our goal.

 

Why This is Important Pt.2

We cannot replace time. Most of us do not understand that it is the quality of our revision, not just methodology that really gives the best results. But really you must know that if your method becomes harder and harder and harder at scale, then your stress will exponentially increase, which means the method is going to break you eventually.

It is imperative that you can do an activity with high repetitions that is only as stressful as the effort YOU put in. I know the demands of learning 100-200 words per day (very high) and I also know the demands of 50 per day (not as high) and 25 (very low, almost easy).

Let’s assume you can read 2,136 characters already and want to start diving into learning vocabulary. If you want to try and memorize the word 特徴(personal characteristic, trait). Ask yourself? Do I get more benefit trying to read this word in a long, complicated sentence? Or do I simply get benefit seeing it over and over and over and over? We both know the answer to that. But this means what we ‘see’ must also be efficiently designed

Below explains how to make this process super efficient.

The Secret To Efficient Retrieval/Recall

First we put our target word in a format that forces us to retrieve a part of it from memory. Look at these two potential flashcards:

FRONT   彼の役割  BACK 役割 やくわり his role

FRONT  彼の[role, part to play]   BACK 役割 やくわり

Notice in the second card, I am now forced to remember the second Kanji and its reading. I am forced to “retrieve” that piece. By “retrieving” that piece, I am giving my mind some work to do. By just reading the word, I have no guarantee that I can pull it from memory. I guarantee that I can recognize it, but recognition alone does not equal memorization. 

Notice also how short these sentences are. Here i’m avoiding ‘bloat’. It gives my mind room to very quickly try to retrieve this data. If I can’t remember it, I just set it to return in a minute in Anki and within that minute I could see another 30 words. See where this is going? I also get to use the word  a lot and build repetition, which eventually builds expectancy.

彼の[role, part to play]  his role

僕の役[role, part to play] my role

彼女の役[role, part to play] her role

市民の役[role, part to play] the citizen’s role

I can hit these sentences in rapid succession and train my memorization in seconds. The more I forget (strategic failure) is the more my brain works to retrieve the “piece” it needs to learn. Remember, the brain will be lazy if you make things easy. If you can just read it, the brain isn’t working, which means you probably won’t be able to quickly pull this word from memory in conversation, or if you are doing self-practice.

You: Uh, why is this important again?

Everything is nice when you go through your first 100 words and feel excited that you are “learning a new language”. That ballgame changes when you realize you really need around 5,000 words to make the gains you seek. 100 words is only 2% of this goal. Whatever you do, must work AT SCALE.

Words are not always inherently interesting. Many feel useless, but need to be known. How often do you use the word ‘predicament’? Probably never. But you’ll never forget it. Or the word ‘essential’. I can’t remember the last time i’ve said it out loud, but its necessary to know. Many of the words you learn are like these and there is no way around learning them. There is no “hack” or “shortcut”. You may disagree with this, but look at it this way:

Learning 1000 high frequency words does nothing for your skill level, since you will need to learn about 5,000 anyways to truly have command of your language. Meaning, I see no point in spending months and months with 1000 word vocabulary, when I could develop a much more robust 5000 word vocabulary. The guy who learns 1000 words and speaks ‘fluently’ is deluding himself, as he can barely read anything he sees, can barely understand most of what he hears beyond extreme basics and cannot say that much! He or she may be able to say very basic phrases and words with proper pronunciation, but would be lost otherwise. (Trust me, i’ve been there).

With an efficient approach I can get through the muck of running through hundreds and hundreds of these type of sentences very quickly and “train” the words. Some words you brain remembers easily, others it will keep forgetting and you just have to keep seeing the words and reading them to make them stick.

Training words isn’t always “fun” but unavoidable. This is why I think faster is better. Enjoy a bit of pressure, get saturated, turn it into a kind of fun battle without getting too frustrated. Last night I watched a few minutes of an Anime and saw in the opening sequence a character said something about  精神の戦い seishin no tatakai.

which is basically a ‘war of the mind’. I only learned 精神 maybe two days ago. It is only by learning these words and being exposed can we hear them in speech or retrieve them. 

Remember i’m writing these posts in the midst of everything i’m learning and gaining. I know it will take more time before my brain processes “everything”, but the process is starting.

I’ve written about Brain Power X. Where the brain ‘does its own thing’ once you feed it enough data. I don’t know why 精神 せいしん is a word that my brain immediately picked up instead of others, but that’s just the process. As you keep going the brain will hold on to all the data you prioritize, especially when you start prioritizing retrievals, meaning you make an effort to say certain words and phrases.

But how do we balance it all? Now that we know we have this large task ahead, how do we organize all the data.

That’s in the next post. Developing a Super Memory Pt. 2

 

 

About marcusbird

Writer, Designer, Filmmaker
This entry was posted in Uncategorized. Bookmark the permalink.

1 Response to Developing Super Memory Pt1 Retrievals vs Recognition

  1. 救世 says:

    This was helpful. Used to burn myself out trying to do about 200 words everyday. I started immersing with anime, games, and books last year and it has helped me out quite a bit. I noticed that after I tried doing less words per day (50 words) I could do more in other places. Like reading 3-4 chapters instead of struggling through 1 chapter each week.

    It’s like a battery or memory in a computer. If you do strenuous things all the time the device (your brain) won’t function or it will run slower.

    In short, your post is a good reminder and I’m lookin forward to Pt. 2.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s