Educational Technology and the Sources of Innovation

Based on Jon Udell’s excellent post, I spent the weekend getting reacquainted with work of Eric von Hippel, the researcher who pioneered the study of user-driven innovation.

What’s interesting about von Hippel is that his research hits on the common themes of the open education movement, but does so in a slightly different key.

Briefly, there are a number of intersecting debates about MOOCs. There is what Reich frames as the Dewey/Thorndike debate about what learning is. There is the centralized/de-centralized debate about what the web does best. There is the debate about about whether MOOCs are disruptive or innovative or neither, and the discussion over how much ability to remix teachers need to make classroom learning work well (answer, probably, is quite a bit).

But people on both sides of the debates are often driven by a larger question that we are not naming directly enough: “What are the sources of innovation?”

User Innovation

This is the question that von Hippel has been investigating for over thirty years now. And if we see innovation not as something that has happened, but as something we want to continue to happen, this may be the most important question of all.

The traditional answer, says von Hippel, is that product industries (“suppliers”) are the innovators. In this view a company comes across a set of “sad users”, finds what their problems are, and designs (via research and development) a solution.

But is that really how things happen? Since the 1980s von Hippel has been looking at the history of “transforming” innovations in various industries. These are innovations which haven’t just offered a slightly better or slightly cheaper product, but ones that have radically altered what is possible in an industry. A great example of such a transforming innovation is the center pivot irrigation system, considered by many to be on par with the invention of the tractor in the history of agricultural technology:

Before the center pivot system, farmers had to draw water from a single well and then pipe that water throughout the farm. The fundamental insight of the system was that instead of piping the water all over the farm (with the resulting leakage) you could drop a well in the center of a section of crops, and then use a gigantic rotating sprinkler to irrigate a large section of crops from that well. If you’ve ever flown over the country, you’ve seen what such farms look like from the air:

What von Hippel points out is that major innovations like these almost always come not from suppliers, but from “lead users”, a set of highly motivated and skilled users for whom the current technology or practice is restrictive. In this case, for example, the first center pivot system was created by an individual in the 1950s who wasn’t initially looking to market it, but simply to solve the set of local problems he was facing:

The Valley Corporation then came in later and perfected it, allowing it to work more flawlessly, with less user intervention. They perfected it and prepared it for mass adoption. But the innovation was not theirs.

Look under between 75% to 80% of all major innovations, and this is the story you find again and again, from the first heart-lung machine, to the development of wider skateboards, to protein-enhanced hair conditioners. On the web, people were running makeshift blogs well before Blogger, net-sync’ed folders well before Dropbox, video + question sequences well before Coursera. What smart companies do, for the most part, is not “innovate” but find what “lead users” are hacking together and figure out how to make that simpler for the general population to tap into. Research often plays its most important role after the fact, not in producing designs, but allowing us to determine which lead-user designs work best in common scenarios, and to understand what, exactly, is making them work.

EDUPUNK and User Innovation

For many readers, this process may call to mind the EDUPUNK wave of 2008. The term was coined by Jim Groom in a conversation with Brian Lamb and subsequently extrapolated on by a number of edubloggers, eventually hitting the New York Times (if I remember correctly) as a word of the year.

What some may not remember is that the coining of the term was a reaction to the announcement of Blackboard that they were moving to a Learning 2.0 platform, one that would (supposedly) finally integrate the technologies they had worked so hard to keep out of education because they weren’t perceived as serious or safe.

Lead users like Jim had gone out and done their own thing, hacking together syndication feeds, wikis, and modded themes into a workable replacement for a learning management system that did far better at meeting the emerging needs of the open classroom. And when it was looking like they were out of Blackboard for good, Blackboard came up with this system of blogs and other “2.0” features which replicated much of the functionality, but at the cost of hackability. Here’s Jim in that piece:

Corporations are selling us back our ideas, innovations, and visions for an exorbitant price. I want them all back, and I want them now!

Enter stage left: EDUPUNK!

My next series of posts will be about what I think EDUPUNK is and the necessity for a communal vision of EdTech to fight capital’s will to power at the expense of community.

I’ve never fully gone for the “capital’s will to power” bit of that, although I know that piece remains important to Jim. But for me the piece that resonated — and still resonates — is the disturbing vision of an educational-technology-complex that is aligned against the communities of innovators that it supposedly serves.

While a company like Blackboard, which produces tools to create things, may seem qualitatively different than an irrigation system company, it’s not different in the respect that it codifies practice. To the farmer coming up with an irrigation plan the range of devices and options available to her are just as much building blocks in an overall design as is the Blackboard gradebook or discussion forum.

And as with other industries, most of the practice that Blackboard codifies (and the rudimentary architecture to support it) was developed outside of Blackboard by user innovators. And that’s fine. But the message Blackboard sent (and I think intentionally sent) over the years to skittish administrators was “Now that we’ve offered these innovations in the product itself, you can rein in all your experimenters and put them back in the box.”

As Jim so rightly points out, such actions and attitudes destroy innovation communities rather than foster them. And it’s not just Blackboard either. The entire education reform-industrial complex has often waged war on educational communities, based on the perception that questions of educational practice are mostly solved, and if we could get teachers to just teach using the centrally specified method (or foundation-approved test) we’d be set. Technology thought leaders even make bizarre claims that there is no innovation going on in education, outside, of course, the Silicon Valley entities here to save us.

People have termed this approach “a war on teachers”. It’s that, certainly. But since a subset of those teachers are where the innovations of the future are likely to come from, it’s a war on innovation as well.

The Sources of Educational Innovation

Once we see the question “What is the source of educational innovation?” as a core question of the debate, certain things become clearer. In fact, the answer an individual has to that question is probably highly predictive of what technologies they favor.

The current breed of xMOOCs emerged as a fluid hacking together of different educational elements in places like Stanford. In this environment, teachers using the system were encouraged to extend and supplement the product through both technological and pedagogical innovation.

But, as Bob Dylan would say, things have changed. As MOOCs have reoriented to see a significant piece of their customer base as providers of blended learning (rather than the students themselves) they have failed to invite that user base into the culture of innovation, presumably due to their erroneous belief that innovation begins at the top, then filters down to the masses. The licensing, technology, and content, and supporting community are all designed to preserve their innovation as shipped, in an effort to protect it from the users.

On the other hand, EDUPUNK technologies (varieties of cMOOCs, ds106, FemTechNet, Open Course Frameworks, P2PU) have continued to engage their users, asking the the users to experiment, remix, hack, and redistribute. They are, in the words of von Hippel, “user innovation toolkits” which encourage users to alter, and even subvert, given designs. Because they codify much practice in convention rather than code (see, for example, the use of tag-based RSS and the harnessing together of readily accessible technologies) they retain a fluidity that promotes experimentation. They are, in a word, so EDUPUNK.

You can look at either of these paradigms, and ask which one is more innovative, or which one fits with your model of education. We can ask which framework is more effective or more suited to various local conditions. But the key question for administrators and policy makers is not just which system is more effective today, but which framework will continue to grow and adapt in the future.

And on this question the historical record is fairly clear — open frameworks which allow lead users to hack are the systems that will produce long-term gains. As a case in point, take Lego Mindstorms, a project built over 7 years by LEGO engineers which was significantly improved by user hackers within three weeks of its release.

Rather than fight against those hackers, LEGO decided to embrace them. And maybe this is where I differ from Jim in this respect — I don’t think gutting user communities is necessary to for-profit enterprise. Counterexamples like the one below show that both the interests of investors and users can be aligned. In fact, given LEGO’s explosive growth in the face of a recession, one could see a more enlightened capitalism as a force for good:

I believe that this idea of fostering user innovation informs the rhetoric of Instructure around the Canvas LMS (the reality will emerge over time). It’s the business plan of Lumen Learning’s Candela OER Project, which acts as a publisher, polisher, and integrator of products produced and maintained by their user base. It’s something along the lines of what Alan Levine is proposing in his recent Shuttleworth grant proposal, and what Jim Groom and Tim Owens having been pitching under the Reclaim Project banner.

And at the same time, it is the antithesis of much of what we see out of Silicon Valley, which, not well versed enough to invent the wheel reinvents instead the tree trunk roller, and then mounts a campaign to get lead users to give up their makeshift wheel-and-axle systems as too ad hoc.

The situation is further complicated, because local knowledge is “sticky” in two major ways. First of all, many educators and educational technologists have extensive tacit knowledge of what works that is difficult to express to people who design products. As von Hippel points out, when such knowledge is sticky at the point of use (in this case the classroom), it makes sense to push design functions downstream.

Knowledge is also sticky in another way in education. It resists generalization. Despite what Udacity might tell you, there is no “magic formula”. Rather, there are dozens, perhaps hundreds of magic formulas: the success and applicability of which are determined by the subject and skills being taught, the specific capacities of the students, and the nature of the local learning environment. What works in one situation is not always applicable to other situations.

When knowledge is sticky in this way, the importance of hackability to innovation is even greater. Yet while industry moves more and more towards recognizing the importance of user-driven innovation the educational-reform-industrial complex still treats such innovation as a disease in need of a cure.

The Last Innovators

The truth is that Salman Khan, Sebastian Thrun, Andrew Ng and others know this at heart — they are all, in fact, former lead users who solved their own problems with technology and then took their solutions to a broader market. And that’s wonderful: we’ve benefited from their contributions.

But they are only a fraction of a fraction of user innovators out there. We can’t afford to regard these figures as the last innovators to ever walk the earth. If we wish to engage in ongoing innovation, we need to focus on generating conditions that foster more communities of more such people, not less. That means making sure that educational technology is as hackable as farm equipment, shampoo, and skateboards. That means choosing technology for your campus based on what your most creative and effective users need, so that they can advance your local practice, and steering away from lowest common denominator technology. It means looking to our practitioners to lead the way, and then asking industry to follow. And ultimately it requires that we cease to see innovation as a set-and-forget product we buy, and engage with it as a process and a culture we intend to foster.

Photo/Image Credits: Center pivot system: USDA, via Wikipedia; Kansas fields: U.S. satellite image via  Wikipedia; Center pivot prototype: T-L irrigation; Jim Groom as EDUPUNK: bavatuesdays; Tree-trunk roller: Jonnie Hughes.