Joi Ito's Web

Joi Ito's conversation with the living web.

Old school knowledge

I have some amazing friends who tell me that when they were young, they read the dictionary from cover to cover. Other friends of mine have read the entire Encyclopedia Britannica.

My sister calls me an "interest driven learner." I think that's code for "short attention span" or "not a good long term planner" or something like that. I can't imagine being able to read the dictionary from cover to cover. In fact, I don't think most people could sit down and read the dictionary from cover to cover.

Although reading the dictionary and the encyclopedia from cover to cover may seem a bit extreme, it often feels like that's what we're asking kids to do who go through formal education.

Courses are organized, sequenced in a very structured way as student scurry from class to class sitting through lectures and expected to pay attention as instructors go on and on about calculus, history and grammar.

Students with the ability to focus and motivate themselves either through the need to achieve good grades or through understanding the long term benefits of a good education are able to succeed.

Personally, I find the dictionary, the encyclopedia and videos online as excellent resources when I need to learn something. I find the need to learn things every day in the course of pursuing interests, preparing for meetings and interacting with exciting people. I'm extremely motivated to learn and I learn a lot.

I love the videos of professors, amateurs and instructors putting their courseware online. They are a great resource for interest driven learners like me. However, I wonder whether we should be structuring the future of learning as online universities where you are asked to do the equivalent of reading the encyclopedia from cover to cover online. Shouldn't we be looking at the Internet as an amazing network enabling "The Power of Pull" and be empowering kids to learn through building things together rather than assessing their ability to complete courses and produce the right "answers"?


Joi, you've hit on the principles of unschooling - or interest-driven learning - and you're right. In an age where we have access to information, examples, specialist practitioners, and communities of like-interested people in anything - the 'education' of it just needs to get out of the way.

It's a tough transition for the institution of education, but it's how we as humans, best learn and grown. And increasingly as we're seeing in works like Power of Pull and Drive or the talks by Sir Ken Robinson and even Brene Brown, it's a transition that's essential to our future.

Cheers to curiosity, interest, and capacities to explore them.

Absolutely right!!! Online course lead to nowhere but gives you a mere degree in hand!
Same goes with books too!
Books/Internet can be just maps to give you possible directions. Get out build & explore! Only way to discover new worlds out of these maps!

We've been over this before, Joi, maybe in conversation with Doc and his (and my) admiration for John Gatto. At this point, the social investment in the institutional-ity of education (buildings, endowments, the out-sourcing of commercial R&D to academic labs, credentialling processes) virtually guarantees that something resembling the present system will endure indefinitely — but the future belongs to people who perceive what you (and Doc, and Gatto, and others) recognise on the horizon and arrange their institutions’ resources so as to anticipate and meet in advance the next iteration of higher educational systems.

I think the right answer is Both. Initially, risks from unknowns will favor the push model (traditional factory-style education) over pull model (interest-driven).

Eventually, as unknowns are revealed and differences become clear, I think companies will choose from both.

Another factor pointing to Both being the right answer is that some students simply perform better in the push model. When left alone, they don't do anything.

I was having this very discussion with my brother a few days ago. Our current education systems often slow down the learning process. What I would love to see is a body of knowledge for students to go through moving from the most basic concepts to the most advanced, all at the student's pace. We need to rethink the concepts of what should be required knowledge as well. Much of the current curriculum is completely unnecessary.

Joi, why do you think online courses are in tension with "empowering kids to learn through building things?" I'm developing online courses inside a traditional university as a way to transcend it, as are many others.

Sure, there are many crappy examples of replicating the failings of the physical institutions online. But that's not "structuring the future of learning" any more than all the 1990s e-commerce sites that had people wandering around in virtual malls were "structuring the future of business."

There are a few issues here. One is credentialing, which will eventually be solved through a combination of badges, analytics, and greed on the part of faculty and universities.

The second is the education shouldn't always just teach you what you think you need to know. I'm not advocating a traditional liberal arts education necessarily, just something to counteract the natural tendency toward homophily. This is more true for some learners than others, of course, but that's a general issue.

The third issue is scale. Industrial institutions do it very well, but in a certain way. Education is no example. Online courses can and should scale in ways physical ones don't. Every higher-ed course everywhere, for example, should be made freely available online. But that's not the key scarcity.

Remember the story from Steve Jobs' Stanford commencement speech about how the Mac has great typography because he audited a calligraphy course after dropping out of Reed? That's all about distributed pull, but ask yourself why there was a course there for Steve to sit in on?

We can't fully escape the tension between push and pull. It's the same issue as the blogosphere vs. mainstream media. Citizen journalists would be impoverished if all the world's great newspapers folded, but at the same time, distributed collaboration can be superior to a huge percentage of what traditional media institutions once did.

I expect this latter problem will eventually be solved by the Li & Fung's of education, which don't yet exist. Such orchestrators are a key part of the pull economy, as Hagel and JSB emphasize, but there's nothing like them in education today.

It's too early. But it's really exciting.

This idea of 'cherry picking' one's topics of study works as long as the basics are covered. I'm one of those people who read dictionaries and textbooks for fun as a kid, but there are still areas where I'm seriously lacking.

The older I get the more I find my self counting on my fingers because my brain just can't effectively compute simple math, and yet I have no problem grasping concept or formula-driven math like algebra or trig. There are brilliant people who are incapable of stringing two sentences together correctly, but who can manually calculate the correct trajectory to 'slingshot' a probe past a planet's gravity and launch it into outer space.

Another factor is a topic you find initially unattractive that may turn out to be something you return to time after time and develop a deep fascination for, but you would never have bothered to investigate it yourself if it hadn't been a required course.

So while it may be fun to pick and choose what one is interested in, it's also important to cover the fundamentals, as unglamorous as they may seem, and to branch out in order to provide balance and make sure there are no glaring gaps in one's pool of knowledge.

"learn through building things" is the motto that schools should endorse and I wholeheartedly agree with. Nothing makes you better at something than actually using it in practice. Someone who reads a few Python books cover to cover will not be as good a programmer as someone who only read the basics and went on to practice it and learn more by a combination of trial-and-error, looking up online documentation and searching for answers on, say,

But that kind of learning looses out on rigor and intuition; and thus is good only after you've built sufficient foundation in the topic of study (interest).

Naman - I think you can be very rigorous and build a foundation without formal education. Intuition, I would argue, is in many ways much more likely to develop in a less structured setting.

Interest-driven learning is marvelous, and I think it's the best way for adults, and *some* kids, to learn. I'm afraid, however, that not all kids (not all adults, either) have the maturity and insight necessary to know what it is they'll need to learn in order to achieve their goals. Those kids would do well to trust those who tell them that such and such a pursuit may seem as dull as reading the dictionary, but it will benefit them down the line (and the benefit could be just the pleasure that comes with knowing things).

As Edith Hamilton wrote: "It has always seemed strange to me that in our endless discussions about education so little stress is laid on the pleasure of becoming an educated person, the enormous interest it adds to life. To be able to be caught up into the world of thought -- that is to be educated."

David, wouldn't it be better to spend the early years equipping kids with the skills they need to learn what they need to learn when they need to learn it? I mean I agree that reading, writing, arithmetic and civics are necessary for all of us (one reason our government doesn't work is largely because many of us don't know how it should work in the first place and operate as though they were unaware that "taxes" maintain roads and schools and infrastructure, or how to evaluate political candidates and break down emotional appeals) but in one of John Barnes' series of novels, childhood education had been set up so that kids were taught how to think logically, ask questions and get answers, use the internet and other tools to get information as they needed to learn it, and work together in groups to solve problems as needed, and that this, not memorising things, was the main thing they spent time on.

An educated person is not a person who can cough up huge chunks of information upon command.

An educated person is a person who knows how to find the information they need when they need it, knows how to evaluate the quality and sources of the information they find (for instance, being able to recognise that a long essay which restates the writer's feelings about a topic multiple times in increasingly inflammatory and emotional language without ever citing a source is fact-free), knows how to take a problem or system apart and recognise where things aren't working, and also a person who understands that cultural, scientific and personal information sets, theories and beliefs are the ones that have worked best so far, and that when they are actually observing something that doesn't match up with what they've been told, the thing to do is figure out whether the information they have is wrong/doesn't apply under these conditions, or whether they're not observing the whole process, rather than just deciding that those dinosaur bones must be 5,999 years old or that people consciously choose their sexual orientation all the time just because they were told that when they were younger.

One has to watch out for people's 'has to be like me' or 'like my experience' reactions. Systems that worked well for one person, or one generation, may not serve as well over longer hauls. Approaches that work well for one person may not be scalable.

The present systems in the US and the UK offer mostly a one-size-fits-all approach to students and subjects (and testing in the UK, increasingly in the US). It's hard for me to imagine that this educational assembly line serves anyone well except politicians, bureaucrats, and budget-makers.

I wrote a blog post a while back thinking about the role of hackerspaces in regards to an article I read that discussed the mechanisms by which phd allocations were engineered by higher learning institutions. Might be relevant to the thought process here.

I agree completely with Joi's commentary. I like to think of it as kinetic learning. Some folks need to see something break before they can learn how it truly operates. Others can just read the book and follow the recipes that are laid out before them. Different learning styles that's all it is. Many roads to many goals often intersecting.


Not disagreeing at all, but the format of educational institutions is at least partly driven by institutional needs. The need to educate huge numbers of people to some baseline standard as cheaply as possible, for example.

How does this interest-based learning work for people who aren't natural auto-didacts? Who aren't inclined to grab what resources are available to them and learn from them as and when needed?

Well, they need something like learning coaches rather than teachers, for one thing -- folks whose job is to inspire and coordinate rather than lecture and test.

Teachers need support to conduct their classes this way though. In my interactions with educational institutions, I mostly see those institutions demanding proof of standards compliance, not learner-driven processes.

The core problem of this approach can be found simply by taking a look at most modern politicians. We often chide people like Santorum or Palin for their ignorance, however these are actually pretty intelligent people (for the record, it really hurt me to write that). However, they choose to pursue their own interests. So the basics of science, economics, and global issues get lost in favor of fundamentalist Bible studies and free-market mantras.

Where I think the idea of interest based learning could be effective is in re-structuring the typical school day. We could easily provide more opportunities for students to pursue their own interests in a class structured on free learning. Often when I catch students not working on an assignment, they are playing a math game or 'words with friends.'

As a substitute teacher I've found that as long as the class is based on a foundation of creating interest and excitement, students tend to respond. The same student who is an angel in the algebra class where the charismatic teacher usually takes them outside to learn creatively can be an absolute terror in a computer course the very next period where the disinterested instructor sits behind a desk occasionally mumbling instructions. An important element of interest based learning, from a teacher's perspective, should be teaching students that your subject can be interesting.

not everyone wants to be a techie, some kids really like ancient history, or
geography, or reading Paradise Lost.
Some people like memorizing Byron or Tennyson.

Some kids hate group learning.

Some kids like abstraction---enjoy reading and working through an 800 page geometry text.


//and be empowering kids to learn through building things together rather than assessing their ability to complete courses and produce the right "answers"?//

Interest-driven curriculum could be used in any classroom. John Dewey wrote about it in 1900. The idea is that we begin using criteria and problem solving rather than memorization and instructor-convenience models. This kind if instruction can be achieved with traditional psychometrics, and is. Many skilled teachers create curriculum where the focus is on learning path to skill and knowledge acquisition. In many cases however, this approach is used to dismiss standardized assessment. What is troubling about this dismissal is the fallacy that learning is all relative and contextual, and that participating in the process is enough. The role of school is not only to provide surveys of information, tradition, and fact, but also to crystallize experiential learning into complex concepts. . . concepts that encapsulate a lot of information with a single word, like isometric, or genre, or base.

The idea that we should just learn what we want to does not work--especially with young people and republicans. Often, people only want to learn more about what they already know and reinforce what they already believe. Sometimes, people have buck up and learn something new and difficult. This is what is meant as a liberal education: the ability to develop and consider multiple hypotheses for explaining or describing a problem. Most people do not want to do this, because they are waiting for the sequel, extending what they already know and like.

However, this distaste for cognitive complexity and the Randy Moss approach to education "I learn when I want to learn" can still work with individual learning plans and criteria-driven assessments.

I just posted an essay on this myself recently.

David, why would anyone design an entire education system around the few students who "need" coercion?

And forcing people to learn something is the surest way of killing their pleasure in knowing things.

It needs to be both, though, ideally, separately. We need to know how to answer questions without having to reinvent the wheel each time, but we also need to be able to figure out which questions to ask in the first place.

The former is really best structured ad-hoc and as a life-long process. New tool/thinking pattern/whatever comes out – take a look to see if/for what it is useful and master it if and when you need. The latter is more difficult to both teach and measure, and, I suspect benefits from starting to learn early.

Modern day educational institutions, especially at the late-secondary and undergraduate level (it tends to be better both before and after), tend to muddle the two and not do a particularly good job with either.

Let's not overlook the existence of teachers and schools that are focused on "empowering kids to learn through building things together." Rather than trashing the institution as a whole, how might we share stories of innovation and support its growth? If the innovators walk away from public schools, that leaves plenty of room for those who are interested only in the stories told by high-stakes testing.

I few years ago I lived in a town called Dandong on the border of China and Nth Korea. I had taken a year off school and was learning Chinese.

One day, a young Chinese man that I had seen many times hanging around the foreign students dormitory found cause to approach me. He introduced himself and begun to speak to me in English. It seemed a reasonable endeavor given the lack of native English speakers in that part of the world. However, I was a reluctant listener. Other students had warned me about this guy and the way he would bail you up and practice his English for as long as you could stand. Those students were annoyed that he would never speak Chinese with them as some had made a pact not to speak English during their time in China. I am naturally suspicious of people who make pacts limiting communication and they were kind of douschy anyhow. Eventually, they all surrendered to speak-easy English and left China with as little Chinese as they had respect for the classroom fascists that made them swear the pact in the first place.

Between broken Chinese from douschy Westerners and broken English from a Chinese guy, I picked the later. I was not afraid of his English imperative. I listened with a friendly smile as I decoded what he said. It was not the usual brand of Chinese/English, all the little words that make a sentence work were missing, or in the wrong place, or muddled up, or something…it is hard to remember what you never quite understood in the first place. But there were big words – the long silly words that we seldom use to good effect – dotted all about. I could barely understand a thing that he was saying. Often he would pause, mid-sentence, after a word like ‘profligate’ or ‘quixotic’ and ask, ‘Is that how you use that word?’ Unsure that I understood the beginning of the sentence and unable to venture a guess at its ending, I would pause and smile and say, ‘Yes, I think you are right.’ He was always satisfied with my response and would announce, ‘This is the first time I have used this word. Is it the right word?’ And I would respond and around we would go.

Being talked at by this young man was breath taking. He was the embodiment of word salad. Articles, auxiliaries, pronouns, all absent, discarded in favor of bigger words. He would sausage-string his latest and most important words together indefatigable aeolist. Witzelsucht autohagiographer he would cacophony blandiloquent until I would lean in, searching for a point of reference, then he would announce, ‘Is that how you use this word?’ And I would allow myself to breath, finally back in command of the conversation.

I told him that I had come to China to learn Chinese, I said it in Chinese using all the small words I knew. He seemed less than impressed and said, ‘Inquisition (eyebrow lift like I already knew what he was talking about) you know how many Chinese characters?’ Stumped, I replied, ‘I have no idea…maybe 1000?’

‘You must know how many characters you know, how many exactly?’

‘I am really not sure, we don’t count them. probably about 1000?’

‘Count them and quantitative statistics how good is your Chinese!’ Definite sense of urgency in his voice.

At this point he announced to me that he spoke the best English in the whole town. I had my doubts. But he was adamant, and he could prove it. He pulled a tattered English-Chinese dictionary out of his jacket pocket and begun to leaf through the pages. In the past, you could tell how good a foreigner’s Chinese language was based on the sweat, oil and coal dust smeared about the pages of their standard issue English-Chinese-English pocket dictionary. Looking up a Chinese character used to be hard work. By any measure, his dictionary belonged to a master of the English language. The printing on the plastic cover had worn away, the edges were black and greasy, the elastic band that held it all together was tied in a knot in two places.

‘I know over 15 400 English words!’ Victorious cachinnation, he laughed maniacally.

I expressed surprise. I was nervous, that was a lot of words, imagine if he decided use them.

‘I know all the words in this dictionary…and more! No one in this town knows as many English words as I do. How many English words do you know?’

‘Ahhh ummmm?!’ (Not even words)

‘What’s your number English words you know?’

‘Hmmm, at least 16000.’

He seemed satisfied with my response. I imagine he was content with the notion that there was still work to be done. Those last 600 words would surely be the most difficult to pin down.

That was five years ago. He must have found the last 600 words by now, I am sure. Or maybe, he found more.

I always imagined that the idea of knowing all words in the dictionary was a great comfort to him. It is a little harder to imagine today though. My trusted grimy pocket dictionary has been replaced by an iphone and a translation app. If I am really getting my translate on I will use my expensive and flashy electronic dictionary. It would be difficult to count the number of words in either.

Joi! You need to get connected to Sudbury Valley School if you haven't already - the premise of their school is exactly what you are talking about. I definitely feel the "diplomatic schools" based on the Sudburry Valley School model are the way of the future of education. I've been researching these schools in Japan and would love to try to start up an English speaking one in Tokyo - Wanna help?