TL;DR: To say “X is bad”, you need to be real clear about what you mean by “good”. If you do set up a straw man but don’t set up your definition of “good”, you will get to a real bad argument. In this case, the bad argument is set up around the claim that devouring nonfiction at a high speed is bad. As with everything, it probably varies a lot given context and individuals’ goals. Spaced repetition is good at exactly the narrow thing it’s good for.

First I’ll confess that in my private wiki I have a long-ass thing written out summarizing and criticizing Andy Matuschak’s thing on books, which seems to be of a kind with the thing linked here. I haven’t posted that response because it makes me too irritated to engage with every time I think about editing it1. The post I’d like to talk about today has less hubris in its central claims and is thus less infuriating.

Here is a list of supporting details from the piece that I am first intentionally not putting into the author’s argument’s framework. All of these are reasonable, factual, or beyond argument. Many of them are interesting.

  • Some people optimize their reading/listening habits for the rate at which they can expose themselves to information.
  • Some people do a lot of reading/learning without ever acting on it.
  • “[S]hoving information into your mind can create the illusion of knowledge, especially when you rush it”
  • There is a theory of learning that holds that the teacher/text takes the learning they have within them and metaphorically pushes it into the learner’s mind through the learner’s eyes and ears.
  • A study suggested students don’t pay attention to a lecture for more than 10 minutes at a time.
  • The author retained a lot from a college course where they had written reflections and small group discussions.
  • “[D]ates and key terms are bolded in textbooks and also the subject of exams at the end of the semester. But knowing the name of something without also understanding the context behind it isn’t knowledge”
  • There have been efforts to standardize instruction in primary and secondary education that have taken some of the agency away from teachers.
  • The author found travel more valuable when spending longer in a given place and trying to not act like a tourist.
  • To learn to ride a bike, you have to go try it, not read about it.
  • “Language learning provides another example. You can pick up trivial knowledge by banging your head against a classroom wall for 10 years. Or, you can immerse yourself for a year and walk away fluent.”

Okay, you got all that? Sounds like it probably fits into something reasonable, right?

Here is the real structure of the argument:

  • It is bad to consume a lot of stuff at high speed.
    • People who do this must be doing it because they think that this is the most “effective” way to learn from what they are consuming. (unsupported)
      • People must think this is a good way of learning because classes have lectures. (unsupported)
        • Education must commonly be structured with texts to read and lectures because teachers are supposedly stupid enough to “assume that learning is inevitable if students read enough books and spend enough time in the classroom.” (unsupported)
    • Instead, they should slow down and spend more time on each thing.
      • Essays and syllabuses force you to do this.
    • “Learning is most effective when we implement spaced repetition” (supported only relative to the implicit definition of learning as memorization)
      • Writing an essay must be useful because it’s like spaced repetition.
    • Direct experience is also necessary. “If you want to learn, you can’t just take things in. You have to figure things out for yourself.”
      • This is true for riding a bike and learning to speak a language.
    • Listening to things at 3x speed doesn’t make sense because different languages evidence similar averages of rates of information/second. (supported, but not any less of a non sequitur than it looks like here)

A list of objections, in approximately ascending order of nitpickiness:

  • The definition of “effective learning” throughout the piece has the unstated weakness of relying on what can be empirically evaluated, and therefore boiling down to ~”efficient preparation for recall testing”. Most secondary or post-secondary education is meant to happen at a somewhat deeper level than what can be neatly measured in this way. This gets big and complicated and controversial so I’m not going to try to dig into it – I myself am a big fan of spaced repetition!! – but if you’re not even flagging the complexity there, you probably shouldn’t be trying to cite studies like they give your very general arguments weight.
  • Despite the author’s plaintive “I’m not constructing a strawman here. I know many people like Mike, and I used to be just like him”, the piece suffers from its framing around Mike The Straw Man. What is Mike listening to all those audiobooks for? We spend time deriding him for never having actually “built anything”, but, uh, what does he want to build? “XYZ is bad because I made up a guy who’s doing it for no reason I’m going to disclose, and we know it’s not working for him based on the fact that I’m imagining him lacking results in his goals that I’m not going to explain” is a bananas framework for a written argument, just bananas.
  • Speaking of Mike’s goals: listening to a lot of stuff real fast is fun. There is an entertainment factor to people’s consumption. Not only am I willing to call that an inherent good, but I’ll say that that shouldn’t be discounted in terms of keeping people motivated. Calling it “instant gratification” is misunderstanding what people are getting out of it, especially when you’re abstracting the whole point of everything to the abstract good of “learning”.
  • “Systems are easy to scale when they use this reductive, cookie-cutter mindset.” I would say this was a fuck-you to educators throughout history, except that the transmissionist/passive learning construct itself is somewhat of a straw man to the point where I can’t even really identify what educators it’s meant to refer to. Look, I’m willing to believe that maybe someone in the 1600s or whatever wrote something down that really does look like the argument all of the constructivism/active learning people are taking a baseball bat to… but it smells like a straw man. It smells like something that people came up with via a teleological explanation for the forms of books and lectures while omitting the developed cultural practices surrounding those things as they were actually developed and used. (Hint: “As late as the 1700s, historian Robert Darnton writes, ‘For the common people in early modern Europe, reading was a social activity. It took place in workshops, barns, and taverns[.’]”)2
  • Not everything is worth equal amounts of consideration. Not everything that is inherently worth certain consideration is worth equal attention relative to an individual’s goals. Maybe this would all be more obvious if framed in terms of reading alone: “It is never worthwhile to skim something rather than read it thoroughly while taking notes and engaging in reflective discussions” is a claim with enough clarity to expose its own weakness. Straw Man Mike pursues everything with a need for speed, but actual humans deploy various strategies contextually. Evaluation of the strategy requires consideration of the context.
  • People often learn from being exposed to material that makes untrue claims. It is often quite important to learn other things from something than what a teacher wishes it to convey. If your discussion of learning can’t accommodate that, even if you pay lip service to the idea that transmissionism is bad, you’re likely still working from the idea that someone other than the learner needs to be designating what the learner should take away from the experience. This may be fine (if I take a survey course of sociology, it’s probably likely motivated by wanting someone to boil down what they think is important for me and my limited time to touch on) but can lead to some real contradictions when you’re not careful.
  • Following my mention above: quite often a syllabus is meant to facilitate someone focusing on a topic for less time than they would if they attacked it on their own. The author’s assumption is probably deeply indicative of something if we dig into it more, but let’s not.
  • Writing an essay is not like spaced repetition. Even writing an essay slowly over a period of time (you know, the way all students do, definitely never right before deadlines) is not like spaced repetition. If you think the point of engaging with material in the way that essay-writing demands is to engage with it longer rather than in a qualitatively different way from e.g. memorization… then I wish you would not write blog posts that aim to sound authoritative because people are going to listen to you.
  • Trying to abstract over all things that may be learned is useful in some cases but really dumb in others. “Direct experience” is real inapplicable for some things and essential for others. Using the example of riding a bike to suggest that we should all be trying for more “direct experience” re: any thing we might want to learn is dirty poker.
  • I love spaced repetition, but it is proven effective for exactly one thing: memorization. There is amazing data regarding its gold standard status at that and at that alone. Figuring out how to fit memorization into a larger system of learning is complicated and not itself empirically provable to be “effective”. Spaced repetition is not the opposite of transmissionism – it is the empirically honed method by which one may drill little atomic bits into one’s squishy brain, a goal indistinguishable from that of any catechism’s questions.
  • No one has ever defended the rate at which they listen to audiobooks by saying that some languages are spoken with a higher rate of syllables per second. I do not believe a single person has ever done it ever. A piece which pretends that is the case in order to justify bringing up an interesting graph is written in such bad faith as to merit pelting the author with tomatoes3.
  • It is startup culture brain poisoning to scorn Straw Man Mike for being “one of those white-collar workers who’s always on the brink of quitting his job.”
  • There is zero connection of ~”sometimes learners stay in the preliminary phase without moving forward to implementation to further their goals” with, like, anything in the piece. This is a bummer because I don’t often see that issue considered abstracted from certain domains, and there’s probably good discussion to be had there of what principles can animate one.
  • There is zero connection of ~”trivia is bad and not real knowledge” with anything else in the piece, which makes sense, because really getting at the difference between what is and isn’t trivial would require digging into learners’ goals, which this piece seems to be sworn not to do.

As frustrated as I am with the Western project of liberal education and its stalwart resistance to considering the material circumstances of its victims, judging from the stuff Hacker News keeps tossing up, I would rather questions of learning and education be left to the most ivory-towered than become chum for the disruption-minded.

  1. Since I have a real clear view of my own unimportance regarding how many people read this site/blog, struggling on stuff that makes me miserable with little sense of how it might impact anyone positively is… not my priority. ↩

  2. “Everyone did things so obviously stupid as to be counter-intuitive until the Enlightenment, and now we understand how things really work” is a form of argument that should raise some red flags, n’est-ce pas? ↩

  3. I am joking about this. Probably. ↩