Note: This is a revised and updated version of a post from my old blog. It is one of the last posts I’ll carry over to Steemit, because I’ve been focusing less and less on language learning, but I think it is applicable to learning other things besides a foreign language.
This post takes a look at what it would take to retain a new word solely through reading. I think about this kind of thing as I’m learning new concepts in web development and cryptocurrency, and recently my approach has defaulted to an “if it’s really important it will come up again” type of thing. What do you think is the most sensible way of retaining something?
For this post I want to focus on the question: “Is reading enough?” Or more specifically: “If you just keep reading (or watching TV, or getting some sort of comprehensible input), will you retain what you look up or figure out from context, and progress at a comfortable rate? Is it worth worrying about anything else (like an SRS)?” I’ll just jump right in.
I think remembering a word boils down to two main factors: frequency and memory strength. The frequency is how often you’ll see it; how many pages you need to read before you’ll come across it again. The memory strength is how well you know the word; how long you can remember it without seeing it. The question is: Is your memory strength strong enough that you’ll still remember the word when you see it again?
A Closer Look at Frequency
It’s actually pretty straightforward to produce some specific numbers for the frequency part of this, so to get a clearer idea of what it might look like, I went through and calculated how frequently you’ll see a word in English based on this frequency list. The list was created using sources from spoken English and written English, but I used it to estimate how often you’ll see a word while reading.
If you’d like to get a better idea of how I calculated everything — or to check my math (I’m pretty rusty) — there is more information below.
The column “Pages” has the number of pages of English that you would need to read to come across the word, with the assumption that a page has 250 words. I had to go with ranks that were multiples of seven because I was working off of an incomplete list (that’s why it’s rank 252, 504, 1001, etc.).
| Rank | Word | Pages |
|---|---|---|
| 7 | to | 1 |
| 252 | far | 13 |
| 504 | model | 22 |
| 1001 | somebody | 44 |
| 2002 | bomb | 99 |
| 3003 | shade | 169 |
| 4004 | steep | 259 |
| 5005 | bizarre | 369 |
| 6006 | postwar | 472 |
| 7000 | one-time | 634 |
| 8001 | breach | 792 |
| 9002 | lest | 987 |
| 10003 | pediatrician | 1157 |
The chart tells us that you’ll see the 252nd most common word about once in every 13 pages, the 504th most common word about once in every 22 pages, the 1,001st most common word about once in every 44 pages, etc.
As a more specific example, let’s take a look at the word “shade”.
You will only see the word “shade” once in every 169 pages or so. Using an assumption of 60 pages a week, this means that you’ll only see it about once every 3 weeks.
So with shade, you can assume that you’ll be able to retain just from reading if your memory strength is around 3 weeks or more. At that point you’ll come across it before you forget it.
A Closer Look at Memory Strength
This part of the equation is much harder to estimate. For a general idea, I’m going to use the intervals from this post:
Say if you learn a new word in a foreign language, you’d want to practice it again a few minutes after hearing it, then a few hours, then the next day, then in 2 days, then 5 days, then 10 days, 3 weeks, 6 weeks, 3 months, 8 months, etc. After a while it’s basically permanently memorized with a rare reminder.
In other words, it would take 6 reviews spread over 19 days for memory strength to reach the 3 week mark. I’ll call this point, when frequency equals memory strength, the “natural retention point.”
This is just a rough estimate, but it helps to have an understanding of what 3 weeks of memory strength might take to reach.
So What?
While progressing, the main factors will be retaining and deepening your understanding of what you know, and learning new words (or sentence structures, etc.). After the natural retention point for each piece of information, the retaining and strengthening can both be taken care of just by reading. But what about before it?
I’ve been getting the feeling more and more that I started SRSing with Japanese well before I really needed to. And once I gave into my frustrations with it, I was already comfortable enough with my level that I didn’t really feel the need to continue.
However, something keeps bringing me back to trying supplementary tools like flashcards and notebooks, and I think it’s the knowledge that there’s this whole period of time before a word reaches the natural retention point that is more efficiently managed by something other than reading.
But at the same time, even before the natural retention point, reading is probably enough in most cases. It will involve a lot more forgetting and looking up, but with bits and pieces sticking here or there, you’ll continue to move forward.
I’m going to keep trying out different things and searching for the magic silver bullet, but my biggest take away after thinking through all of this is that I’d like to focus on the best way to bring an unknown word to the natural retention point without a focus on forgetting, which is what I think the forgetting curve tends to do. I’d like to focus on building up enough memory strength momentum that I can drop the word into the current of my L2 input and forget about it.
What is the quickest way to take the word “shade” to 3 weeks? Is it the SRS-like approach above, or some other combination of focused study? Maybe a mnemonic and 12 minutes of review? Maybe drawing a picture of what it means and whistling about it for 5 minutes? Maybe it is just to keep reading and stop worrying about what the most efficient method is.
To be honest, I went around in circles with this post for awhile and I think I’ve managed to overcomplicate a fairly simple question here (I’m so good at that! :P). More than anything, I wanted to illuminate that reading is enough after a certain point. There is a specific point where the natural flow of input through your brain will allow you to retain words, but before that it’s less straightforward.
What do you think, though? Is this even worth worrying about? What’s your favorite approach? For awhile I’ve been focusing almost entirely on reading and it has been going really well. Either way, I think it’s time to get back to my book. 🙂
Calculation Notes
I’m going to work through the 3003rd most common word (shade) in the table to show how I came up with these numbers.
From the frequency list I used, I took the rank, word, and frequency. The frequency is the number of times that the word shows up in the 450,000,000 word corpus. For the 3003rd ranked word, this number is 10,677.
Then I found the proportion of the corpus that the word represents, which for “shade” is 10,677 / 450,000,000 = 0.000023726666667. I used this to represent roughly what percentage of the written English language is made up of the word shade. So if we multiply this number times some number of words, we’ll get the number of times that we’ll see the word shade within that number of words.
With this, we can assume that the number of pages you read, times 250 (the assumed number of words per page), times the proportion the word makes up of the language, is equal to the number of times you’ll see the word in that number of pages. If we set that number equal to one, we can solve for pages and then get the rough number of pages that one would need to read to see the word one time. In other words:
pages x 250 x proportion = 1 → pages = 1 / (250 x proportion)
For shade this works out like this: Pages = 1 / (250 * 0.000023726666667) = 168.586681652149, which we can round up to 169.
