Why Forgetting Can Be Good

People often ask me how they can guarantee they won’t forget anything they’ve learned. But I think forgetting isn’t such a bad thing and that trying to avoid it completely is a loser’s strategy.

Trying for a perfect memory is a defensive strategy, it’s protecting the stockpile of knowledge you possess against atrophy. In theory it sounds good, who wouldn’t want to forget less and remember more? The problem is opportunity cost. The time you spend memorizing and making sure you never forget is time you could have spent learning something new.

Even within a specialization, the knowledge you possess is probably only a tiny fraction of all useful things you could know. Given that you know so little, playing defense seems like a poor strategy.

Learning vs. Relearning

Relearning is generally much easier than learning for the first time. Knowledge fades but it is rarely completely deleted, and I’d wager even for topics which have faded completely from conscious access there would still be a learning advantage for relearners.

The exception to this is for knowledge which was never really learned the first time. I often hear students complain about knowledge they’ve completely forgotten from prior classes, but my guess is that they never really understood those ideas.

Is Spaced Repetition Worth It?

A technique that’s been getting a lot of press is the idea of spaced repetition. It’s the idea that the best way to form lasting memories is to remind yourself of an idea right before you would forget it. Spaced repetition systems like Anki have been popular for flashcard style review.

I’m skeptical of the value of an SRS for most domains of knowledge. The problem is simply that just because you’re reminding yourself of an idea doesn’t make it useful. Useful and important ideas recur frequently, so spaced repetition is naturally built into the process of learning aggressively.

People have asked me if I have a mechanism to review material from courses I completed earlier in the MIT Challenge. Initially, I had considered creating one, but going through the classes has shown me that it isn’t necessary.

For any given class, some ideas will be very useful and important, others less so. If an idea is useful, it shows up in more than just one class. Huffman coding has probably shown up in 4-5 classes I’ve done, so I’m getting tons of repetition even though I never use a formal system to remind myself.

Even languages, the favorite child of SRS seem amenable to this approach. If you spend most your time actually communicating, the words and phrases you memorize are precisely the ones that come up most frequently. Perhaps just going out and speaking a lot is the best kind of spaced repetition.

The same is true in non-academic learning. When I read a book, I try to deeply understand it, but I don’t make any system to guarantee that knowledge is perfectly preserved. The reason is that I know if the ideas are useful and important, they will show up as themes in other books.

The solution to spaced repetition, therefore, probably isn’t looking back and reviewing, but going forward aggressively. Expanding your knowledge instead of fortifying it.

Aggressive Learning

My strategy is aggressive learning. This is the idea that my knowledge is far less than 1% of the total human knowledge available, and probably less than 10% for the things I’ve chosen to specialize. Given these enormous gaps in my understanding, it makes more sense to learn aggressively than worry about forgetting.

If the brain truly has a maximum capacity, I doubt if any person has ever reached it. Knowledge seems to be stored on top of each other, so even if some ideas are buried under new ones, they still form part of the foundation.

Aggressive learning doesn’t mean superficial learning. By all means, when you learn an idea, learn it well. My biggest takeaway after doing this MIT Challenge was the importance of writing exams and practice problems to learn ideas deeply, it’s too easy to skip the tough stuff and never really get the ideas you’re trying to learn.

When Should You Look Back?

Often the paths and trails into unexplored territory circle back to familiar ground. This presents an opportunity to relearn those old ideas or to view them in a new way and get a better connected understanding of them.

At the beginning of the MIT Challenge I took a class which introduced the Laplace Transform and, honestly, I didn’t get it. I could manipulate them well enough to pass a test, but I couldn’t see what they were or why they were useful.

However, because they were an important idea, they showed up again and again. Each time I circled back I got another chance to learn them more deeply. Had they not been important, I might never have seen them again and they would have faded from memory, but if they aren’t too important, why would that matter?

Aggressive learning works because it circles back and reminds you of precisely the ideas that are more prolific, useful and important.

Why Don’t People Learn Aggressively?

The attitude of aggressive learning and curiosity is forced out of students by education. Because almost all classes have a closed curriculum, where only a certain set of knowledge is tested, it rewards being able to get really good coverage of some arbitrary subsection.

But life doesn’t have a finite set of readings or ideas to understand. Success in the real world means knowing lots of things and being able to learn the answers to questions that won’t show up on any test.

  • Kevin Post

    Scott, thank you so much for sharing this article with the world. I very much appreciated it.

  • Olle Linge

    Although I agree with what you say in general, I don’t agree with the part about language learning. Speaking (and listening) your way to better language skills only works up to a certain, basic level. Once you pass that, words become too rare to retain simply by hearing them in context.

    If that’s the case, how do native speakers do it? They are completely immersed in their language and spend years and years in school and society, reading and listening. Few native speakers use SRS for their own language. However, this doesn’t mean that SRS isn’t good, it just means that it’s possible to learn through immersion, although it takes a lot of time. Indeed, I think SRS would be excellent for native speakers to learn more advanced vocabulary, perhaps in writing class in high school.

    For the rest of us, the use of SRS to learn languages is about efficiency. Most people cannot immerse themselves completely in a language for several years. If you can only study for a few hours a week, simply exposing yourself to the target language will be far from enough.

    So, in essence, if you’re talking about acquiring basic communication skills in a foreign language, you’re right, but as soon as you leave the basic domain, SRS starts being an essential tool. In a way, it can be seen as a substitute for the immersion you don’t have time for. If you do have that time, I still think (wisely used) SRS can be beneficial.

  • CeppuiqMX

    I agree with Olle Linge. Currently, I learn foreign words of my own language with the program Anki. If I just tried to relearn it when it appears in a text at some time in the future, I would not make much progress. Sometimes it’s crucial to repeat one word, or a number, or statistical records over and over. Because in your daily routine it rarely appears, although it is important. And this rare knowledge might make you better than others.

  • Jonathan

    WOW…just wow!

    I’ve been recently thinking, and have already reverted to using software that utilizes spaced repitition. I’ve been trying to preserve knowledge, but I didn’t take into consideration the opportunity cost. I think I will just use the holistic learning strategies that I’m comfortable with, and learn as aggressively as possible, then make a concept list which can be used for review later on. As you mentioned, if I can’t remember a certain topic, relearning should be a breeze.

    This is probably one of the best articles I’ve read on this website, maybe because it pertains to me (and it seems as if you have read my mind.)

    Scott, you make me feel stupid sometimes. Honestly. When I read some of the excellent ideas you present in these articles, I wonder why I couldn’t think it up myself. I don’t know if you have any articles on thinking with this kind of creativity? If you understand what I mean…

  • Stephan R

    Good step forward to learn new stuff rather than looking back at what you may or may have not learned.

    Essentially what you are saying is that as you learn aggressively the really important themes/ideas will recur more than once. That way you are able to draw connections you previously didn’t have about an idea. In this manner, important ideas will stick as you understand them deeply enough from different angles.

    It seems to me though that the number of important “truthful” ideas that count in the end is somehow limited as they keep popping up again.

  • Victor

    Awesome post.

  • daniel

    Good post, though sometimes it is worth to notice that one particular treatment of a topic excels in some way or has a unique or new idea and it might really be worth not forgetting it if one wants to innovate later on.
    One can forget even this kind of idea. For 99% of learning I agree with what you are saying, but the whole thing about if it’s important I’ll run across it again assumes the world of ideas to be static, which it mostly is, especially in the MIT challenge or other forms of formal education, but in the “real” world, taken to include academia, this is not really true. The thing is that you don’t know in advance which ideas you really should remember in detail, where “should remember” is taken to be mean “will help you to innovate later on” which I assume is one of the main reason people pursue an MIT-like education. How well aggressive learning works as a good enough substitute for repitition depends a lot on how far you wander away from what you have previously learned as you move on to other topics.

  • Thomas Carmichael


    Thanks again for another article that challenges me to reexamine the ways I view learning. I’ve been applying your techniques and have noticed an improvement in learning efficiency. But, perhaps more importantly, the holistic style of learning you teach allows me make learning fun through metaphor and comparison. Sort of like reaction coupling in Biology, I have learned to compare otherwise uninteresting information to information that I already find interesting (usually biology or health).

    Keep up the great work. I hope to save up some money and buy your book.



  • Dan

    Great post. Would you also relate Aggressive Learning to learning new languages?

  • Erik Frimann

    Well written, but you Americans are really odd, however likeable. Why this energy spent on learning a lot, while a more laid back attitude of studiying learning what matters yields more practical results in less time and certainly with less effort? How do you read a book? From cover to cover, taking notes?
    I do that, too. But only for pleasure. Normally I read the intro, the conclusion, the index and perhabs a short review. Then I decide if it is of any interest, and if not, I move on to another book, retaining a skoleton understanding of the book in hand. If there seems to be something useful, I dip to extract it, picking out promising chapters.
    Then I let it store itself, and try hard to forget the whole thing, trusting that it will connect itself synaptically with whatever makes it important.
    Well, you. I am old hand, so I know a lot, but forget where I stored it. Young people remembers everything and anything, byt then again: They don’t really know much. So we can live very happily together.

  • Oliver Powell

    Completely agree with you here, I have come to the realization that it really doesn’t matter if I forget knowledge I have learned, in actual fact , its inevitable. You will generally only remember well the concepts and idea’s which you work with on a constant basis or that happen to come up regularly. I think university is actually slowing down my learning in many ways, I feel the best way to learn something is to dive right in and force yourself to swim. The courses I am doing tend to be way too theoretical with not enough application. Some theory is needed, but the only way to truly understand something is to get your hands dirty. It has been said this is one of the reasons Germany’s economy is still doing fine, is because of the strong apprenticeship based system there, especially in engineering.

  • daniel

    My earlier comment did not really show that I feel that there is a *very* valuable piece of advice in your post. Say you’ve studied electromagnetism and feel you are forgetting what you have learned. Then I agree that you should try to learn more about electromagnetism and then go back to basics that you have either forgotten or never really learned as soon as something becomes an issue. I definitely agree with your advice and find it very good if applied to *one* subject. Without this restriction, your advice seems really geared toward the MIT challenge: Of course it is better to move on and see what concepts crop up again in the MIT CS curriculum if the goal is not learning electromagnetism but getting the best overall result in the MIT challenge (or managing to do this at all, which I find very impressive). Great post though, it really gave me something.

  • Hicaro

    Perfect. That’s exactly the point of hack learning. I have a superficial example: suppose I have a hard book to read for learn its subject. It has 10 chapters and I can read one chapter each three days. So I would take one month to read. Let’s simulate some rate of content retention in percentage of the whole book.

    Chapter 1: 1%
    Chapter 2: 2%
    …(and so on)
    Chapter 10: 10%

    Then, (1+2+3+4+5+6+7+8+9+10)% = 55% of content retention in one month. Now, suppose that I didn’t take this approach and I’m concerned in retain the whole book in my mind. Yes, 100%. So, I spend three days in the first chapter, and get 1%. Spend more 3 days in chapter one (rereading) and retain now 1% (again). More 3 days, 2% (use some Fibbonacci function or whatever, this is only a intuitive example). Simulating:

    3 days in chapter 1: 1%
    +3 days in chapter 1: 1%
    +3 days in chapter 1: 2%
    +3 days in chapter 1: 3%
    +3 days in chapter 1: 5%
    +3 days in chapter 1: 8%
    +3 days in chapter 1: 10%

    I took 18 days to learn 10% of the whole book. If I repeat the same behavior for the chapter 2, I would take one month to learn 20% of the book.

    However, there is the aspect of priorities of knowledge. How important is the information (converted in knowledge)? I think we can ignore this. Why? Because what is important, eventually, will emerge. As Scott said, important issues appear in more than one source. Quantity over quality? I don’t think so. People learn in different ways. But sometimes we had to go through.

    Okay. That’s the point: learning in a conscious way is hard. So gurus are starting to talk about “having fun” when learning. It makes sense, because it works. But it is not always flowers in learning. Sometimes we have to increase the intellectual stamina by working hard.

    I recommend the book: Mastery: The Keys to Success and Long-Term Fulfillment by George Leonard and the article: Teach Yourself Programming in Ten Years by Peter Norvig.

    Indeed, I wish to be wrong, as well my methods. That’s the only way to really learn: be wrong. Realize and fix it. Learn is not only about to put bricks together, but to break all and rewire. Make new connections. And life goes on.

  • Phil

    I really disagree with this

    I think you’re right that when you learn agressively certain ideas or approaches that you learn just naturally get a space repition effect because you keep reincountering them

    what you seem to be arguing is that if an idea or an approach is really important, you’ll see it again, therefore formal spaced repition is unneccesary

    I would argue almost the opposite, the ideas that you keep reincountering are the ideas that everyone else keeps retrying, or the well explored solution spaces,

    if you are going to truly do something breakthrough, its probably going to be exploring one of the ideas that don’t naturally get reinforced

    also our minds are set up to naturally reinforce ideas we get easily, I’m reminded of this qoute:

    “Well, the first rule is that you’ve got to have multiple models—because if you just have one or two that you’re using, the nature of human psychology is such that you’ll torture reality so that it fits your models, or at least you’ll think it does. You become the equivalent of a chiropractor who, of course, is the great boob in medicine.

    It’s like the old saying, “To the man with only a hammer, every problem looks like a nail.” And of course, that’s the way the chiropractor goes about practicing medicine. But that’s a perfectly disastrous way to think and a perfectly disastrous way to operate in the world. So you’ve got to have multiple models. “


    I think it probably pays off to try and manually break ourselves out of grasping onto ideas that make sense for us, and then using that hammer for every task thereafter

    using spaced repition to reinforce exception cases, seems like a good way to do that

    (obviously you have to be smart about your use of it, I think you want to remind yourself of different models to solve problems, not remind yourself of simple fact that are a 5 second google search away)

  • Christian Kleineidam

    I think you underrate spaced repetition.

    Even if the time I spent with Anki didn’t produced valuable knowledge the 30 minutes I spent per day with Anki are productive.
    It’s good cognitive exercise.

    If you are in a classroom lecture or you are reading a book you don’t get feedback that punishes you for being mentally absent.
    If you get mentally absent while doing SRS you notice it immediately. SRS is deliberate practice for retrieving information from your brain as fast as possible.

    Through the minimum information principle SRS teaches you to ask yourself: “What are the core ideas of the book that I’m reading?”
    At the beginning you are likely to put facts into Anki that aren’t central. With time you however get better at SRS. SRS trains you to separate those ideas that are important from those ideas that are worthless.

    The cards that I created a year ago were bad. The weren’t clear. Anki shows them to me again and so I can revise them to make them more clear.
    If you want to get better at concise writing doing SRS helps.

    Take a sentence from your article. “Relearning is generally much easier than learning for the first time.”
    The word *generally* has no use. It doesn’t help your reader to understand better. Every time you see fluff in your SRS cards you suffer a bit. The suffering helps you to get better at eliminating fluff.

    SRS forces you into the tough question: “What ideas in this text do I really need and which are fluff.

    You advocate the exercise of holding an imaginary talk that expresses an idea at the simplest level. Writing SRS cards is a similar exercise.
    Writing cards you are more accountable. If your cards contain more information than the minimum that’s possible you suffer for it.

    On a meta level Anki gives you another benefit.
    For each day it gives you a stack of cards that are due that day. If you skipped that day without good reason, you feel that you failed. If you complete your daily stack of cards you feel success.
    Outside of academic learning it gives you are structure for your learning. It’s like the exams at university. It focus your learning for a purpose.

  • Agota

    Anki and spaced repetition is great for learning vocabulary.

    Yes, if you just go out and speak a language, you will learn most frequently used words, but it’s an inefficient approach to learn them.

    I mean, vocabulary is the core of the language, if you know meaning of the words, you can guess the meaning of the words surrounding them and the meaning of the sentence.

    I assume you don’t know Russian. Imagine you go out to the street in Moscow and try to speak to people in Russian. You couldn’t express yourself even at a minimum level (I mean express yourself in Russian, not in English or gestures or drawings) and you couldn’t understand them. Why? You don’t have vocabulary.

    Now, if you put let’s say 2000 or 5000 most frequently used Russian words on Anki (well you have to know alphabet to read them in this case) and learn them, you will learn them relatively fast (..and definitely faster than trying to just “go out and speak” with your non existent Russian).

    Then, if you know the most frequently used 5000 words (especially when 2000 words usually make up what 90% of all spoken language?), you will be able to express yourself and to understand people. You might not know any grammar and say things like “Me to want to eat food where close restaurant”, but people will understand you. You will also understand them, because without grammar, you will still be able to guess the meaning of sentences when you know most words in it. Yes, it’s not smart to start learning a language by learning most obscure words in Anki, but you can get a frequency dictionary for a language and put it into Anki, and then you will learn the core vocabulary very fast.

    Obviously, just knowing a lot of words won’t make you fluent, for that you have to learn grammar and go out and talk to people, but it will give you a great start.

  • Scott Young


    I’m not enough of a linguist to have a strong opinion of SRS with respect to languages, I merely suggested that conversation seems to have a natural spaced-repetition effect.

    Everyone explains systematically how SRS seems to promote fluency, but I know of no person who became fluent through extensive flashcard systems but many people who became fluent through extensive conversation, so the real-world evidence at least suggests the possibility that SRS is designed to make you feel productive, but because the knowledge tends to lack context it is somewhat less effective.

    That said, this is just devil’s advocate, as I’m not really sure either way for the moment.


    I don’t see how anything in aggressive learning contradicts multiple models. In many ways it reinforces it, since by exploring the knowledge landscape more intensely, you’re bound to encounter more models. I would say Munger would have been an advocate of the voracious multi-disciplinary learning I’m suggesting.


    Canadian, so not American. I think you’re mistaking my analytical view of the learning process for being incredibly serious about learning itself. I read what interests me and read for enjoyment, but that doesn’t mean I can’t introspect as to my own subconscious strategies and try to retweak them to further my goals. Self-reflective self-improvement is the mission of this very website.


  • Jen

    I’m not sure how effective using an SRS is at improving fluency, but it can certainly help with understanding, and although the most common words that you need to know at a lower level are repeated often enough that as long as you expose yourself to the language enough you should be able to learn them naturally, once you know the commonly used words (or most of them), words which you don’t know are likely to come up less and less. I am fluent in Japanese, but I use an SRS to help me to remember words which I don’t know already, as I know that at this stage, if I don’t know it already then the chances are that it’s not very commonly used, but in most cases is still a word that most native speakers would be able to understand and use without problems. Thus it makes sense to try to remember it. Using an SRS for this, especially if you use a sentence with the word in for context, is just like constantly hearing or reading the word in lots of places, so it is useful. I don’t think anyone can become fluent without actually doing a lot of speaking practice, but it certainly helps you to reach higher levels of fluency much quicker.

    I think that the difference really is that language learning involves learning a LOT of vocabulary, and words don’t necessarily have that much logic to them, so you just have to remember this = this. There isn’t much understanding the idea of the word to be done, just plain memorisation. For anything where this is important, then using an SRS seems like a good idea to me. For anything where understanding ideas is important, then maybe it’s not such a useful tool. I personally find that when I need to learn ideas, as long as I can understand the idea then I don’t find it particularly difficult to remember anyway.

  • Phil


    I think to truly have a model you have to more than encounter it, you have to review it and use it

    Spaced Repetition provides a means to systematically review and use models you encounter

    Aggressive Learning provides a means to haphazardly review and use models you encounter (maybe you don’t like that characterization, that’s how I read your essay)

    maybe you’re the exception, the science that has examined this sort of thing indicates that most people don’t retain something that they hear once and never again for very long, (and what we enconter twice and then never again for a little longer, but in the grand scheme of things not that long)

    so, as I see it, the question is, is the stuff you’re going to see once and then never again important enough to review systematically? Especially considering that time spent reviewing, can’t be spent learning something new (that’s what you seem to be wrestling with in this essay)

    obviously alot of it isn’t, there are all sorts of random facts and trivia, that aren’t worth slowing down to capture

    to dismiss all the stuff that will be forgotten as not worth slowing down to capture is I think the wrong extreme to take that to

  • vvurdsmyth

    Yah, to me, in re-discovering, the way something works, or why (for example), can lead to a new, higher understanding and innovation.

  • Stephen Hill

    Scott, I’m a bit late to the party, but I have to agree that you’re giving short shrift to SRS. For example, I’m currently studying Italian. Each day, I try to read and watch some Euronews.com stories in Italian. Out of the words I don’t know, I pick the ones that seem most salient and add them to an Anki deck. It might be a few days or weeks before I see “disoccupazione” again, but thanks to Anki I’m virtually guaranteed to know that it means “unemployment.” Anki provides more opportunities to encounter the word, and that makes me more likely to recognize or remember the word when I need to. As others have already noted, in a non-immersion environment, this kind of repetition is essential.

    FWIW, I find Anki even more useful for reading Greek and Latin, where using the language means usually means reading it. If I read a page of Plato in Greek, I’ll find many words I don’t know. Reviewing those words in Anki before I continue means that the next time I see them, I don’t have to look them up, which saves a non-trivial amount of time.

  • tunesmith

    It seems your arguments against SRS are basically arguments against reviewing knowledge in general.

    I can see the point up to a point. For instance, I was really into tracking all of my todo items in todo software for a while. It eventually was a huge drag on my productivity, and discouraging as well. I’ve since learned that the truly urgent tasks are going to present themselves. When you look at the todo items that are truly urgent but easily forgotten, it’s a small subset of what tends to be put on todo lists. For that, stickies are sufficient, and I’ll use actual planning software to plan out complicated sequences of events (the kind of thing todo software isn’t good for). So in that case, investing energy on reviewing all my various todo tasks wasn’t really an efficient use of time.

    But I think learning is a different matter. I generally know *how* to do my todo items. But with knowledge and learning, you either need a set of knowledge or you don’t, and if you need it, it makes sense to want to retain it. And if you can’t rely on being in an environment that will force you to review it (not all of us are in high-impact university environments) then you’re going to forget – review is the only way to retain it.

    I think the criticisms are more relevant towards responsible deck management. For instance, if you have a deck that asks 100 questions that all drill the same concept, you’ll probably eventually want to pare down that deck, because why spend time reviewing 100 questions that drill the same skill? We can all make our own determinations for how important retaining vs relearning is for any given subject, but that can be accomplished just from suspending decks for a period of time.

    For me, I’m sticking to a pattern of reviewing around 200 cards in around 45 minutes, every day. If I feel like I’m not able to learn new subjects quickly enough, I’ll either suspend a deck in favor of a new one or increase my study time. So I think smart triage is a better approach then just questioning the value of active reviewing.

  • Mizter

    “This is the idea that my knowledge is far less than 1% of the total human knowledge available”
    Our total knowledge is probably less than 0.1% than that

  • dag

    It’s hard to agree or disagree with you. Personally, I think both starting points are fine, I mean, both aggresive and defensive learning, as you call them (weird names, though descriptive!).

    From my point of view, the appropiate learning is that you are comfortable at. What is more, perhaps all of us should choose an option inbetween. I prefer the aggresive one, but I think it’s connected with my personality. I’ve met people who practise the opposite with very good results, so I don’t see the point in seeing it in a detrimental way. They’re simply different. And as they are equally fine, everyone who practises a particular one has something to learn from the other way in order to find a balance. At the very least, that’s what I think.

  • Terrance

    I understand your point of view, and I believe it is limited. You describe a good way to go about learning information that could be beneficial. I walk around wondering how the things I look at work. I read up on it, understand it, and then move on. These abstract learning methods require basic knowledge that is repeated across many aspects of our lives. And we learn them for the reasons that you describe. (e.g. gravity, motion, pressure) Even without the knowledge of their nomenclature, you could still understand them and move forward. You would not be able to communicate your ideas to others. This is key. We must be able to communicate our ideas to others.
    Particular fields of study, medicine, programming, geology, etc. require specific knowledge. We don’t have the luxury of learning and forgetting. The knowledge must be known when it is needed. You may not need to know a particular drug interaction with a rare disease 99% of the time. But that 1% error is unacceptable (not that it doesn’t happen). Misinterpreting the ratio of Ca++ to Mg++ in a mineral will put it in a totally different spectrum of heat and pressure.

    The point is that in a modern world, sometimes described by the chaos theory, it is important to be learning in both ways. It is no longer sufficient to be great at 1 particular thing. We must have vast broad range knowledge acquired in ways that you describe, as well as specific knowledge acquired via srs to allow us to compete in a particular field.
    It is well known that the modern dynamic economy changes rapidly, that most people will not go through life having 1 job for 50 years like my grandfather, but that we will need to be successful in several areas. We must be able to beat the competitors in out particular field and at the same time be ready to beat the competitors in another field if the situation arises from changes in technology, popular trends, market forces, etc.
    Successful businesses today utilize similar methods. They must be great with their particular product of the moment while taking steps and leaps in different directions (R&D) constantly seeking new fields of success.

    I am not suggesting one method over the other, but both. You are right, and wrong. I think people will understand this. Formally educated book smart people may not be successful. Common knowledge “street smart” people may not be successful either. Success requires a combination of both. So do both.

  • Malcolm

    I agree that SRS is good for aggressive learning, and I’d like to point out one way in which it’s better than just immersion:

    Immersing yourself in a language will expose you to the words at the frequencies at which they’re used. Sure. But Anki will expose you to words at the frequencies at which you need to hear them to remember them. It STOPS repeating words you’ve already learned (as often).

    Consider a musician learning a song. If they keep making a mistake at bar 40, should they return to the beginning and play through the first 39 bars every time before trying to get bar 40 right? I mean, people do this, but I hope you’ll agree that it’s not effective. Effective is to play just bars 39-41, etc, until that’s working, then maybe 36-44, etc. Similarly, SRS allows you to focus on the 90% of stuff you don’t know, rather than the 10% of stuff you do.

    Speaking of measuring things by percentage, you probably know that in the case of languages especially, the ~1000 most common words make up 80% of the words people say/write, so if you already know those words then most of your immersion-conversation isn’t even teaching you any new vocab (grammar, idioms and more abstract stuff, sure).

    The following articles make really good cases for SRS.

  • Andrew

    SRS has its advantages and its major faults. You can target short questions and answers you want to memorize. From little used knowledge to everyday knowledge.

    Learning Japanese kanji characters is one. Mostly all Japanese writing has kanji in it, aside from children’s book. You need to learn kanji to read Japanese. Combined with mnemonics, you can learn the 2000+ standard kanji’s signs in months, rather than years. The effort to do it without any sort of organization is crazy.

    Some information is probably better to read in context. Anki was made in the mind for information you need to brute force to get. Formulas, information, ingredients etc. It can also be tested for other stuff. But it’s less ideal.

    Bottom line is, SRS is a valid option to normal learning methods. It is also very good for fixed info you want to learn. This is very fitting for the school systems and tests we have. If that’s a “good” way to learn is another matter :p

  • Roman Bolshakov

    I’ve easily learned about 1600 kanji in a year just by extensive reading. SRS is good only for meaningless information. Language is not a meaningless set of words