Gas station without pumps

2014 November 12

Autodidacts (against and for)

Filed under: Uncategorized — gasstationwithoutpumps @ 22:05
Tags: , , , ,

Lately I’ve seen a lot of blog posts talking about autodidacts (people who learn things without teachers) as if they were some strange breed of alien being. For example, there is the post Ed tech promoters need to understand how most of us learn | The Hechinger Report, which includes the following paragraphs:

This is a very particular take on learning: the autodidact’s take. We shouldn’t mistake it for most people’s reality. Productive learning without guidance and support from others is rare. A pair of eminent researchers has gone so far as to call the very notion of self-directed learning “an urban legend in education.”

In a paper published in Educational Psychologist last year, Paul A. Kirschner of the Open University of the Netherlands and Jeroen J.G. van Merriënboer of Maastricht University challenge the popular assumption “that it is the learner who knows best and that she or he should be the controlling force in her or his learning.”

There are three problems with this premise, Kirschner and van Merriënboer write. The first is that novices, by definition, don’t yet know much about the subject they’re learning, and so are ill equipped to make effective choices about what and how to learn next. The second problem is that learners “often choose what they prefer, but what they prefer is not always what is best for them;” that is, they practice tasks that they enjoy or are already proficient at, instead of tackling the more difficult tasks that would actually enhance their expertise. And third, although learners like having some options, unlimited choices quickly become frustrating—as well as mentally taxing, constraining the very learning such freedom was supposed to liberate.

And yet, to paraphrase the economist Larry Summers: There are autodidacts. Look around. We all know at least one successfully self-taught expert, and the tech world is teeming with them. How’d they get that way?

While I do see a benefit to teaching (or I wouldn’t spend so much of my time teaching), I don’t think that the autodidacticism should be dismissed as “an urban legend in education”. In fact, the end goal of all my teaching is to turn out students who can continue to learn on their own, without needing the continuing crutch of having a teacher lead them. I’m not sure how successful I’ve been in a lot of cases—I see students for a 10-week class and then they disappear, giving me no clue whether they have developed new ways of learning that stay with them or they have just managed to fake it through my course and relapsed to expecting to be spoonfed immediately afterwards.

I think that Annie Murphy Paul has it wrong when she claims that few people can be autodidacts—she seems to be assuming that it is some sort of innate gift that one is born with (Carol Dweck’s hated “fixed mindset”). I am convinced that becoming an autodidact is something that most people are capable of. I recently read an account of one student who turned herself into an autodidact, and what prompted her to do it—How to become a programmer, or the art of Googling well | okepi:

He was the very picture of the competent hacker I held in my head, that I nursed a secret crush for. But most extraordinary, he threw something together using tools that he’d never used before. Yes, he did spend more time on Google than he did coding, but through sheer force of googling and a prior, general picture knowledge of how these things worked, he’d roped together a pretty sophisticated and working app. He knew where Twilio belonged in the grand hierarchy of things, knew exactly where to apply it, and so, even without knowledge prior, was able to figure things out.

And I despaired. How do you get so good that you can build something out of nothing?

The rest of the semester passed glumly, and without incident. Come winter, I began to panic again. Driven by the need to become employable, I tried my hand at a couple Code Academy website tutorials. Hm. Not bad. I made an attempt at my first website—pretty terrible, just one, static page full of boxes and awful colors, but it was something. Something I realized. Just like my code-god compatriot, when I didn’t understand something, all I needed to do … was google it.

To a large extent, the difference between the autodidact and the ordinary student is not one of competence, but of confidence. It is Carol Dweck’s “growth mindset”—the conviction that you can learn the material and are not doomed forever to learn only what someone predigests for you.  There are tremendous resources now available to everyone that can turn them into autodidacts: Wikipedia, for example, has thousands of excellent articles in all sorts of sciences (and the science articles suffer much less from point-of-view problems and vandalism than pop culture articles).  And, as “okepi” says, Google can find all sorts of answers for you (she goes on to much larger accomplishments later in her post).

I learn a lot of stuff on my own by reading Wikipedia articles, reading survey articles, reading research papers, googling stuff in StackExchange, going to weekly research seminars, even (sometimes) taking classes.  [The astute reader will have noticed that I did not include MOOCs or videos in that list—despite the claim that MOOCs are a godsend for autodidacts, I have found them profoundly unmotivating, and videos as a learning tool are just too bloody slow for my taste—I fall asleep before anything has been conveyed.]

There are some things for which teachers are essential—it is very hard to learn a foreign language well on your own, without a native (or near-native) speaker to help you hear the differences between what you say and how a native speaker would say it.  Theater is hard to do on your own (though a group of autodidacts could get together to learn to act).  Feedback on writing is very valuable, as is having an audience for public speaking. And there are times when it is useful to have the structure of a scheduled course to help with time management—to keep you on task to meet an external deadline when there are dozens of other things to do. But in a lot of cases, a textbook is all the structure that is needed, or an on-line tutorial document, or even just a particular problem that needs to be solved shaping what needs to be learned.  I learned those skills decades ago, and I think that my son learned them well by the time he was halfway through high school.

So I know how to be an autodidact, but how do I teach it to others?  That is a question I have no easy answers for. I try giving open-ended assignments, I try scaffolding by having students search for answers to specific questions, I try deliberately leaving material out of a lecture or a lab handout and telling students to go read about it in Wikipedia, and I try whatever else I can think of that will get students to learn on their own.  For some students something clicks, and they start doing more learning on their own—sometimes a lot more. For others, I’ve not found a secret sauce.

I particularly despair of those students who take copious notes in class and want to record my lectures (I have two of them this quarter)—they seem to have developed the attitude that I am the sole source of knowledge, and that if they just cram everything I say into their memories, they’re golden. But I’m not interested in hearing my words echo back to me—if I wanted that, I’d lecture to an empty classroom.  I’d much rather the students wrote down two or three keywords from my lecture, so that they could find what others had to say on the topic using Google and Wikipedia—or even looked up the topics I’m covering in the textbook (which does have an index). I’d rather that they thought about how to derive the algorithms we are learning in class, rather than trying to memorize what are really fairly arbitrary recursive definitions (and ones that are more easily derived than memorized).

Does anyone have any good techniques for converting note-takers into autodidacts?  Those are the techniques I need to learn (and I didn’t really see anything in Teach like a Champion that would help).



  1. Lots of good observations here.

    On this issue, I like to start off with the idea that no one can force an idea to stick in someone else’s head — there has to be a brain in there that is cooperating to some extent — in other words, in the end it is we who do our own learning. An analogy I like is to how optimum music learning is done during a week: there are lessons, there is a lot of private practice, there are payoff experiences with groups, and sometimes in front of audiences. All musicians knows that it is they who have to learn how to play, but that getting tips and critiques of various kinds are key (even on learning how to auto-learn via practice and exploring). Sometimes these will come from books, but often it’s very useful to have experts in the loop as well.

    So I think of “autodidacts worth commenting about” is those who drive their own learning, and use every resource possible to get fluent. The ones “not worth commenting about” are those who fail to get fluent enough for anyone to say that their learning efforts had real and/or interesting effect.

    I think it is a main job of teachers at all levels to help students become such self-driven learners, including having a sense of how to assess themselves and find ways to be assessed.

    It seems there are many parallels with good parenting, and also the different strategies needed for the diversity of outlooks that genetics and happenstance provides.

    It also seems that these points of view are grounds for deep criticisms of how general education in the US is set up and conducts itself.

    Comment by alanone1 — 2014 November 13 @ 05:25 | Reply

    • I’m basically in agreement with you, Alan. Teaching is a lot like parenting, and roughly the same range of strategies to help people learn are available. As a home-schooling parent, I found the analogies even more direct. A big chunk of my work as a parent was providing the right level and kind of support so that my son could grow into a competent and successful adult. I didn’t always get it right, but he seems to have turned out fairly well despite that. And he is quite capable of teaching himself stuff—most of his computer science and computer engineering knowledge is learned from the web, honed by repeated practice at solving problems that he’s set himself. (For the past year, most of these problems have been product-design problems from his start-up company, and have involved things I never learned in school, like finding manufacturers for prototyping projects and doing careful optimization to minimize parts and assembly cost while keeping functionality.)

      For a lot of what he wants to learn having the context of a particular problem to solve or the structure of a course to help him manage his time is still valuable, so he is well served by being in college, where there are professors to point out interesting problems for him to work on, and to teach him techniques he hasn’t previously needed (like generating functions, for example).

      Comment by gasstationwithoutpumps — 2014 November 14 @ 21:04 | Reply

      • An autodidact in a good university is a very fruitful combination!

        Comment by alanone1 — 2014 November 15 @ 04:08 | Reply

  2. IMHO autodidacts => immediate synthesis with newly acquired knowledge

    Two remarks;

    1) (I’m sure I am preaching to the choir here…) Those students who write class notes verbatim are not reviewing and understanding the material before class. As a teacher, I would take those students aside and try to convey to them the importance of being ahead of the class lectures instead of behind. Class is a time to gain understanding, to ask questions to clarify those concepts that remain fuzzy from reading the material.

    2) what seems to be unstated in this conversation is that autodidacts synthesize with what they have just learned. It is insufficient to merely absorb the material, one must immediately synthesize with it. A remarkably documented instance of this is in the book “In Code: A Mathematical Journey.” Here a young woman learns number theory by reading (and discussing with her father) but then immediately used what she just learned to program Mathematica to illustrate each little bit of knowledge, i.e programming Mathematica to find Greatest Common Factors, finding relative primes, etc. Once she had the basics firmly entrenched (because she synthesized little bit she learned) she was able to do “great thing” with what she had just learned.

    I’d be curious to receive other peoples opinions…

    Comment by Rich — 2014 November 13 @ 11:11 | Reply

    • I think that autodidacts vary in the way they process new material as much as other learners do—I’m not sure that I could characterize them all as immediately synthesizing with what they have just learned.

      I do find it easier to retain material that I use, but I’m more often using the material first, then learning it, rather than learning it, then using it. That is, when I’m in the middle of doing something and I realize I need more knowledge or skills, I go looking for something to help, and I often blindly apply something that I don’t fully understand. If it works, I then dive into the subject so that I can understand what I have been doing—often improving on what I did blindly at first. So in my case, the application often comes before the real learning, rather than the learning followed by synthesis.

      Some of the labs in my applied circuits course are structured this way, but many students do not take the opportunity to dive in and really understand what they have done, so they are left with rather flimsy partial understanding that evaporates before they can use it again.

      Comment by gasstationwithoutpumps — 2014 November 14 @ 21:10 | Reply

  3. i agree that learning can really only happen when the student is curious and even desperate for knowledge. This is the native state of the autodidact. The student sitting in an obligatory class is in a much more difficult position….that of trying to stay alert and curious when they may not be.

    Comment by tashadepp — 2014 November 22 @ 19:39 | Reply

  4. […] via Autodidacts against and for | Gas station without pumps. […]

    Pingback by How do we create more autodidacts? | Computing Education Blog — 2014 December 22 @ 05:38 | Reply

  5. As with parenting, what is appropriate for a toddler and for a teenager are different. So, we obviously offer more guidance for first year students and then less and less as they progress.

    Currently, I’m a big fan of flipping the classroom (with pop quizzes to be sure students stay current and know their syntax), but I don’t know how well this would work with huge classes.

    For upper level (junior/senior) students I require student membership in the ACM ($19/year – a bargain!) and assign readings from Books 24×7. Where possible I give alternate sources (e.g. “Here are 3 sets of tutorials; choose the one you like and work through it.”) I am also very explicit about my goal of making them autodidacts, though I frame it as “the transition from student to professional.” (Even in more elementary courses I provide a list of alternate references – both explanations and “cheat sheets”.)

    Next, I model explicitly(and often) how I go about learning new material (read several books at once, one elementary, one more advanced, etc.) and about developing good sources (what do you think of the various O’Reilly series? of the other series of professional books? Which ones should you start with? Which ones do you use for reference? etc.). Among the “strongly agree -to-strong disagree” sentences on my end of semester course evaluations are “This course improved my ability to learn from a variety of sources.” and “This course improved my ability to learn new material on my own.”

    Finally, this year the last group project I assigned in an upper division course had a list of 4 or 5 things we had studied in the semester which had to be used and a list of about 8 things which they hadn’t studied. The project had to include all on the first list and at least 3 things from the second list (no references provided my the instructor) with links to the documentation and resources they had used for the new items. There was also an exhortation that the project should be “fun”. The project generated a tremendous amount of excitement. In fact one colleague remarked that he had never heard so much laughter coming from a lab.

    Comment by Margaret Menzin — 2014 December 23 @ 06:53 | Reply

    • I’ve never been a member of ACM, though I’ve been a member of IEEE for at least a couple of decades. The library has all the ACM and IEEE publications, so I’ve never really needed personal subscriptions to the journals. I carried IEEE membership largely for the term life insurance, which was reasonably priced even with the cost of membership added, though I used to go to some IEEE conferences, and the member discounts were worth it then.

      While I do model for students how I learn, I’ve never asked them about what sources they use (other than citations on research papers, of course). I’ve also never explicitly given them a list of things that they have to learn on their own.

      Comment by gasstationwithoutpumps — 2014 December 23 @ 08:23 | Reply

      • True, most university libraries have a good collection of books, but once students graduate they need other sources. So, I like to start them down that path. (Many schools also limit the number of users who can read a particular e-book at once —- very inconvenient if you’ve assigned some reading.) Of course, using professional books (as opposed to textbooks) is a help, wherever they read them. Still, my students like the ACM membership and when I point out how many books they don’t have to buy, they are quite happy about the annual fee.

        The ACM Student membership provides other benefits – a career newsletter, mentornet, etc. and once you have a course’s worth of students in the ACM then you have enough students for a student chapter (leadership opportunities.)

        Further, students should put ACM membership on their resumes – shows they are serious about their profession.

        Should we give students a list of what they need to learn on their own? Why not? Isn’t that what we expect them to do after they graduate? If we provide a scaffolded set of experiences then we help them on the way to becoming independent. And, as I think we all agree, if they can’t learn on their own then their CS degrees will be out-dated in a few years.

        Comment by Margaret Menzin — 2014 December 23 @ 09:22 | Reply

RSS feed for comments on this post. TrackBack URI

Leave a Reply

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

You are commenting using your 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

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: