I will eventually write a post on what I call the Golden Number Theory but presently this post is for my records. From my previous experience with German, I believe it is possible to predict what I call a “comprehension explosion” once you’ve learned a certain number of words (3000+) and progressively listened to a certain number of hours of native speech. Interestingly, it appears this excludes the need for “comprehensible input” (as the input becomes progressively comprehensible on its own once your number of learned words gets large). From what i’ve seen, what needs to be “comprehensible” is the context, versus words, grammar and phrases.
Why No Comprehensible Input?
A child born anywhere on planet Earth will learn to speak the language of its people to fluency without fail. This means that our brain has a “biological software” as it were, that allows us, in a predictable and repeatable manner, to acquire a language. The process from this standpoint, is a certain period of listening (as children physically cannot speak until they are about 2), contextual acquisition of words and grammar and then reproduction of speech (ages 2-4). It is usually by age 4 that a child becomes fluent in their target language, and possesses an exposure to several thousand hours to the language and a vocabulary of about 1500 words (without the knowledge of reading or writing). This highlights that the two most important components of this natural process are words and listening time. Meaning, any child, anywhere, with the right amount of listening time and contextual word/grammar acquisition will learn the target language to fluency. As a child ages, this ability lessens as they learn more about ideas involving personal limitations and their brain grows larger and perceives reality differently.
However this means that within all of us, this software still exists and theoretically should work in the same manner. As an adult, one can accelerate aspects of this “natural processing” period, by consciously utilizing a large majority of time listening while progressively adding words. This is important, because I feel that many learners (as I have in the past) think the key is in just listening, which is not correct. A child has the luxury of infancy to blissfuly spend years slowly acquiring the data that will allow it to speak and spending their lives excitedly learning things bit by bit in context. An adult has no such luxury and cannot reproduce naturally the tens of thousands of interactions that produces contextual learning in a child. We as adults can however use present methodologies to “shortcut” aspects of this multi-year process, compressing it into just months, as many people have demonstrated in various ways.
This means that input in the beginning, does not necessarily have to strictly be “comprehensible” versus “context heavy”. I say this because upon learning a brand new language, a new learner will have the following disadvantages:
- a lack of knowledge of all language phonetics and how these chain to make words sentences and phrases.
- a lack of time spent recognizing these patterns sonically, in context, or with any frequency for any length of time.
- a lack of any and all vocabulary of the target language.
- Inability speak the language.
- Inability to read the language.
This means that even if I get content with transcripts, if I have no vocabulary to read the transcripts, they are useless to me until I learn the vocabulary. Also as I’m listening to these transcripts I have no experience with the phonetics of the language, so it will be mostly noise. With vocabulary or phonetics, I won’t be able to process what i’m listening to. This means that in the beginning, no matter how good the tools you have are, everything is incomprehensible.
With this in mind, let’s look at the fact that the level of exposure needed for your brain to start mastering your target language’s phonetics is quite extreme and also your vocabulary requirements. By “extreme” I mean, you need SO MUCH exposure and SO MUCH vocabulary that in the beginning, your efforts are better spent on the high end of acquisition of this data, so you can trigger your brains natural software.
Let me be clear: I am not using the perennially overused example “learn like a child” reading baby stories and trying to enjoy watching silly cartoons. My only comparison with children relative to what i’m researching is relative to the following:
The order of acquisition.
A child acquires sounds which the brains translates eventually to patterns we call speech, which it eventually is able to process as individual words and phrases.
The child then collects these words and phrases and begins to process reality relative to these objects and begins slowly producing these words and phrases as it learns to associate the muscles of the mouth with the production of what it has spend thousands of hours absorbing.
This absorption, at a specific, measurable point (about 2 years for the majority of children) then produces the ability to speak that rapidly accelerates within the next year to two years.
THIS is the research point here: What number of hours relative to vocabulary triggers the child’s ability to now understand more and also speak?
With this in mind, when thinking strategically, what presents the best use of time and resources relative to our natural biological software initially are items 1-3 from our list above.
My best results in any language learning attempt(German) has been when I did the following three things, which might seem counterintuitive.
1. I only studied vocabulary (not grammar) for the first month and a half.
2. I watched a lot of native media (without subs) in context rich media (shows i’d seen before), I wouldn’t call it comprehensible, but stuff I could follow.
3. I studied grammar after my brain started to naturally put things together.
4. After about 4.5 months I started to just “understand” things I couldn’t grasp just three weeks earlier, which indicated to me I had “triggered” what I call a ‘comprehension explosion’ with a specific overlap of listening time, word acquisition and then grammar learning.
At the time I got these results, I did not understand all the mechanics of what I was attempting, but after several failed attempts at multiple languages, I’d done enough research to know what didn’t work or what was sorely lacking in efficiency.
I came to the realization that no matter how quickly you learn the grammar of a language, to communicate effectively you will need a vocabulary of several thousand words. This varies by language, but generally three to five thousand is the minimum you need to really get a handle on things.
Secondly, I learned that even knowing all these words is worthless if you can’t recognize them in native speech. This only happens after what i’ve dubbed extreme exposure to the language. So what I’ve found is that in the race to speak quickly (which is the bedrock of a lot of Youtube channels) people don’t realize they will hit these massive hurdles at some point. There are shortcuts to speaking fluently (in a certain manner) very quickly, but there is no shortcut for listening time and no shortcut to massive vocabulary acquisition. By “shortcut” I mean, exposing yourself to and attempting to learn 3-5000 words for a dedicated learner can take anywhere from two to four months and this is minus the base requirement for listening that triggers your brain (which my research is pointing to a number around a 1000 listening hours) which a minimum of 8 hours per day will take a minimum of 125 days. About 24-40% of this also needs to be active watching, which means you’ll have to watch to build conscious context in your mind. The demands of all these things present the “true cost” of language learning.
Sure I can spend 30 days hammering out 10 hours a day practicing to speak ‘fluently’ with a vocabulary of 300 words, or 50 verbs, but I wouldn’t be able to really hold many conversations beyond extremely narrow and basic ones or understand any native media beyond elementary school level. I would “appear” to be fluent, but really have no skills in the language. When I lived in Japan I was in this exact position. I’ve always had very good pronunciation and I can speak pretty quickly if I need to say something, so I gave off the “illusion” of fluency when I spoke to Japanese people. I could handle a very “narrow band” of conversational situations, but if things got deep I had no speaking ability. I think this “narrow band” phenomenon is a representation of what many rapid learners label as “fluency” or “ability” when its really a pinhole looking into the room that is a language. So when I was in Japan, beyond the basics, I couldn’t get any ‘deeper’ into the language. I also did not possess my present mindset, where I didn’t truly believe I could learn another language as an adult. Despite epiphanies, seeing folks do it and getting inspired myself, I thought it was really only special people or geniuses who had the ability to acquire another (or multiple languages). But this is not the case. What these “special people” possess is an incredible fortitude of patience and unbreakable, well disciplined habits that take them through the drag of acquiring massive amounts of data over a staggered period of time. Language learning is the ultimate mental marathon, with one of the most enviable prizes once you reach the finish line, and the finish line does not need to take very long to reach either, with the right strategy, but what many fail to reveal with “strategy” is the absolute mind numbing commitment this takes and the constant careening and obstacle dodging you must learn to do to avoid yourself as your mind and body begin to heavily resist the load of the work you’ve put on yourself.
More on that later.
Now that I’ve put that all on paper, let’s look briefly at some basic tasks (that are unavoidable and have minimums of time relative to one’s acquisition goals)
Minimum Words Needed (5-10k)
5,000-10,000 words – irregardless of system used, this is at the minimum a 8-16 week process depending on how many words you are aiming for per day. At ten words per hour this is a minimum of five hours per day, clocking you in at a whopping 280 hours of study for this acquisition. I don’t believe it needs to take this long, but I guarantee the average person does not have the time, raw energy and inclination to go after 100 words a day. I did this with German, but it was all I did for about five weeks. There is no way to do 100 words per day and the revision required while trying to do speaking, reading and writing, there isn’t enough time in the day. So 5,000 words at 50 per day
5000/50 = 100 hundred days or 3.1 months
Minimum Listening To trigger the brain
General research points to the golden number of 1000 hours to gain an extremely sharp processing ability in the language. I believe this number can actually be much lower on that spectrum, at around 350-500 hours for a near 100% comprehension (with the acquisition of 3-5,000 words).
With German, I noticed my first “leap” at the 3000 word mark (week 5). Two things happened immediately: I was able to hear almost everything being said in films (that’s when I started watching stuff in German) and when reading cards as I able to figure out new words in context. This was with zero listening time. Over the next three months I doubt I hit more than 250 hours listening time before my first “quantum leap” when I started understanding teachers speaking in German about how to learn German in German, watching movies were pretty okay (even some German ones) and then i was also able to play some video games in German (role playing). So this was after 4.5 months.
So even though my situation might be a bit unusual, let’s say 250 hours put me in a position to start understanding a lot of native speech with about 75-90% comprehension. Even though that was happening , this was in late 2019 (stress) and even though i’d made gains the language STILL felt impossible. I remember trying to listen to a podcast and just giving up. I didn’t listen or touch German for about six months. Oddly, in early 2020 when all the madness began I decided to commit to listening and started listening to a German podcast. To my shock I found that I had no issues following the speakers and ended up listening to around 80 or 90 podcasts over the next 60 days. After this the language “opened” in another way, I could watch interviews, news and stuff that wasn’t too esoteric (meaning using heavy language I wasn’t exposed to). For those 60 days I think I average about 2.5 hours a day, so that’s 150. So let’s call my total for German 400 hours to reach that level (with low motivation to boot).
that’s 40% of the 1000 hour requirement and I was at about 90% of the word requirement at 4,700 words.
The difference with Japanese is the first barrier of entry which is obviously the written language. Without knowing the Kanji, hiragana and katakana you can’t read anything and by proxy cannot really “go after” thousands of words without getting a handle on this first. This is why I devised my current method to learn all the Kanji in 8-10 weeks, while now consciously doing a progressive load of listening and word acquisition.
This requires INCREDIBLE effort, which is why i’m ensuring to document the key components. Just the demand of the Kanji alone has been staggering and I am sure writing this post is an act of pure procrastination.
But i’m on track. I’m at around 1,600 fully memorized so not that many left. Once I wrap the Kanji, I’ll be able to focus heavily on new words. I find that the mental demands of memorizing Kanji and then doing 50 words a day is too much. My true intention is to do 100 words per day to race to 5000 as quickly as possible, which cannot be done while i’m learning the Kanji. So until I finish the Kanji i’ll add probably 10words per day and then jump to 50 then 100.
So the mission is to see if and when a “comprehension explosion happens”. Remember I said my first major one happen with German at the 3000 mark. Based on my anki deck:
I’m at 975 words learned from the deck (based on other words I know I’m assuming my base is around 1,200 total)
This is the only way really to track this type of goal. I have two decks i’ll be working with to aim for my 5K mark over the next two months.
So after 8 weeks this is where I am:
1,600+ Kanji memorized — 1,200 words — 182 hours of listening.
note: I started the “listening phase” a bit late, (didn’t have the epiphany yet). So i’ve been doing listening for 1 month and 9 days, and “intense listening” for just 10 days. By “intense” mean I set things up so I aim for 8-10 hours of listening per day (and usually hit 7.5-8).
I’m so ready to be done with the Kanji so I can focus on word acquisition. This phase is much easier with romance languages because there is no barrier of entry technically, just different phonetic pronunciation. This means that progress can be made much faster (as demonstrated with German). I think i’ll save my next benchmark post until the 250 mark. Hopefully by them i’m done with the Kanji and can go hardcore with words.
By the way none of this action leaves much room for speaking, reading none of that. In fact this stuff takes so much effort, as i’ve said before, I think chasing the ability to speak too early will cause this reality to backfire on certain learners as the “true cost” reveals itself. In other words, spending a few months in a more disorganized manner to learn how to “speak” versus locking in 1000 hours of listening time and 5,000 words (which guarantees a certain trigger of comprehension explosion) i’ll go into more of this later. Obviously there are people who master the language quickly, but trust me, these people have no life beyond language, to engage in this for 12-15 hours a day you have to be a little bit crazy (which is why some people get crazy results!) but for now i’m just focused on testing the raw science based on my theory. Even writing this post took about an hour, which could have been new words, or thirty new Kanji (get where i’m going with this?)