I suppose I shouldn’t blog while down about the state of humanity. But I do many things I shouldn’t do.
I was just reading through some news reports of the continued misunderstanding of faculty on what Wikipedia is, how it works, and why it matters.
I doubt that you’d be able to get a teaching post somewhere without understanding how to do library research, or get tenure track position while confused about why endnotes mattered to an argument. You don’t get to teach science without understanding what the scientific method is. But you can walk into any college in the U.S. and demonstrate rank ignorance about the process underlying the most consulted reference work in the world, and you get to teach students, no problem.
My daughter was told at her high school — a top high school in the state — that students in college who consult Wikipedia in college (not plagiarize, not cite, but CONSULT) get kicked out. We’ve dealt with that; she goes to a new school now.
But the problem stands. On the most pressing issue of our age — how we advance knowledge in the world of the read/write web — it’s perfectly fine to be ignorant and teach. No one will stop you, no one will supply professional training to help you, no one will guide you, no one will correct you. We ask why these technologies, which have such potential to do such good, have not had the impact they might have. But in this case it is certainly a case of education holding us back.
One thing I’ve learned from my deep dive into wiki is that wiki is most powerful when seen as a collection of *ideas*. Those ideas might be stories, examples, software patterns, chord progressions, whatever. But when treated as a repository of ideas instead of a collection of publications wiki gains a certain type of power.
Ward demonstrates this nicely this morning in his federated wiki running journal (metaphorically called forage.ward.fed.wiki.org). He starts his day of reading this article on net neutrality and net regulation:
It’s a multi-page treatment of the relationship of law to the internet which argues that in fact we already have many other legal tools at our disposal. But Ward doesn’t summarize it, exactly. He mines it for ideas he can name and connect. He finds one, and adds it to his journal:
David Reed says, “Not all laws come from governments. There is a whole body of “common law” that is generally accepted, transcending government. One such law is that you cannot steal a package that you’ve agreed to transport from point (a) to point (b). That is true whether or not there is a “contract”. It’s just not done, and courts in any jurisdiction, no matter what the government, will hold to that principle.” webpage
He makes a compelling case that the “inter” part of the internet works pretty well without governance by ITU or FCC or anyone else for that matter.
And he gives the idea a name: Steal the Package.
I fork the page, not necessarily because I agree (although I do, in this case) but because this is a useful concept to think with. At some later point I’ll connect that thought to an idea of mine. The whole process functions in some ways like the creation of sub-disciplinary jargon to express ideas quickly and succinctly, but it does it in a way that makes these terms accessible to anyone. In a wiki, each idea gets a page. Complex thoughts are formed by connecting pages.
That idea can quickly flow through a network, maybe even change the debate. As it flows through the network it can be extended, qualified, annotated, connected.
This is a different activity than forwarding a link via Twitter, and it’s different than writing up a response in WordPress. It’s a form of analogical, metaphorical thinking that we have barely tapped into as educators. Yet it embraces the core of education — you collect and curate a collection of ideas that will serve you well later. By chunking those ideas into terms you develop the ability to construct and express complex thoughts quickly.
We’ve come out of a decade of using wiki as a glorified book report publishing engine. We have barely tapped its educational potential at all.
Did this this morning, it was fun. Higher Ed policy wonk Bryce McKibben makes statement on Twitter. We have an exchange, I write it up in fedwiki and link him to it for review.
Here’s a screenshot of the fedwiki page:
Bryce replies yes, it’s fairly accurate, and suggests an additional link to a think tank report. I review it and add it.
Resulting page is here: http://journal14.hapgood.net:3000/view/welcome-visitors/view/net-college-price-and-tax-credits
From now on, when others post for or against the College Board analysis on Twitter, I can link them to this page and ask what they would add. This page is also passed into my feed for forking and extension by others in the federated wiki network.
The idea is not to kill StreamMode, but to redirect it, when appropriate, to more recursive and expository tools.
In getting ready to present federated wiki at OpenEd and I’ve edited the “Arthur Clarke” scenario down to seven minutes and fifteen seconds, and crunched it all together.
The set up for this video is this:
Arthur Clarke has some insights in 1950 about global positioning systems. He doesn’t realize that other people are working on this and might benefit from his commentary. But it doesn’t matter, because in this world he uses federated wiki and his ideas auto-magically flow to and are extended by other smart people.
Yes, it’s a world where wiki and the Web exist in 1950. Roll with it.
What to notice about this process:
1. It flows seamlessly from an Evernote-ish journal stage, to a Twitter-like/Tumblr-like stage, to a wiki like stage. This encourages contribution while still moving people towards reuse and extension.
2. As pages move through the network, they are transformed by the nodes they pass through, and they also end up connecting previously unconnected authors to one another. This stuff happens in other platforms, but here the dials are turned up to eleven.
3. It manages to move information through a network without the benefit of a central server (ala Twitter, Tumblr, Blogger, wiki). Everyone owns their own server, but the system protocols allow people to easily plug into a fluid, user-defined network.
4. It replaces wiki wars with an evolutionary model — information people fork to their own site survives and propagates. Information people aren’t willing to host dies out.
If you already get what’s wrong with the Internet, I think you’ll get why this is a worthwhile initiative. But if you need more rhetoric, this post from last week goes into depth on why these changes matter, and how they could make the Internet a better (and less toxic) place. But it is 5,000+ words and this post is 302.
by Mike Caulfield.
Keynote delivered at NWACC 11/6/2014.
Part 1: Sputnik
I’m going to start this keynote by stealing a story from Steven Johnson, a historian of technology.
Johnson uses the invention of GPS as a case study in how innovation happens. It’s his favorite story and he’s told it everywhere from a TED Talk to Science Friday, so apologies if you’ve heard it before. But he repeats it for a reason — it’s just an excellent example of what innovation and progress actually look like.
The way the story goes is this. Sputnik launches in 1957, and America freaks out. The Russians are in space! It’s like Ebola, ISIS, and Gamergate all rolled up in one.
Meanwhile some physicists at Johns Hopkins are hanging out, and wondering — would it be possible to listen to this satellite? It’s sending out a signal, mostly to just show it’s still up there. Kind of like a Space Age Machine That Goes Ping.
So a couple of these guys go and fiddle with microwave receiving equipment — they just want to get this ping on tape. But when they record it, the pings are not the same. They vary slightly. And they realize that this is classic doppler effect stuff — the ping is compressed as the satellite approaches and stretched as it moves away.
And we get really lucky here, because one or another of them says, hey let’s turn this project up to eleven – let’s use the variation of these pings to plot the trajectory of Sputnik around the earth using a mathematical model of the Doppler Effect.
And here’s the part that that interests both me and Johnson, because a couple weeks after this, their boss calls them in. And he’s heard about their project, and he has a question.
He says, if you could calculate the position of the satellite from a set position on the ground, could you calculate your position on the ground from knowing the set position of a satellite?
Because there was this problem he was tasked with — we needed to know the position of nuclear submarines, and astrolabe and a chronometer weren’t going to cut it.
This sequence of events leads to the development of the Transit positioning system by the APL lab at Johns Hopkins and by a little known new entity called DARPA.
You know the rest of the story — we start by using GPS to plot positions of nuclear subs in 1960, but in the 80s and 90s we commercialize the system. We move through a series of inventions to this point in time where the cell phone in your pocket is plotting your position right now, ready to plot your run or clue you into the nearest parking, gas station, or bar.
So that’s Johnson’s story. But here’s another piece of the story that Johnson doesn’t cover.
This is a letter from Arthur C. Clarke, the famous sci-fi author. And besides suggesting communications satellites, he notes that you could put three satellites up in the air and use them to plot your position anywhere on earth using something the size of a watch.
It’s from 1956, a year before Sputnik went up. And it turns out that Clarke had proposed a version of this in 1945.
Now the model is somewhat different than the APL version as we’ll see later — in APL they were using doppler shift with Low Earth Orbit (LEO) satellites. Clarke’s idea involves geostationary satellites in a high earth orbit.
So I’m not saying that Clarke had the full idea here.
Still, nowhere in the history can I find any indication that the people working on this had heard of Clarke’s idea — an idea he had had since WWII.
And ultimately, even if they had, this is just one idea among thousands that never saw the light of day. This is the one where we have a record. Who knows how many other ideas about satellite aided GPS were never captured anywhere?
So one thing we can do is celebrate the environment that actually led to this invention at Johns Hopkins, as Johnson so rightly does.
But what I’m obsessed with (and really, what Johnson is really obsessed with too, in a sense) is how people who wanted to position nuclear submarines were not familiar with Clarke’s proposal.
And my sense is that this sort of thing happens almost every day — someone somewhere has the information or insight you need but you don’t have access to it. Ten years from now you’ll solve the problem you’re working on and tell me about the solution and I’ll tell you — Geez, I could have told you that 10 years ago.
How does this happen? Why does communication break?
One answer to that is right in front of us. This is a letter, addressed to one person who might find it interesting. Clarke couldn’t have addressed it to the folks at APL because he didn’t know they would be interested.
And this is why this concept of “openness” has become the most important concept in the digital world.
You don’t know who can benefit from your information, the modern solution to that is to not even try to guess. Unless there is a compelling reason you should always publish it as openly as possible. If you don’t, and nuclear war breaks out, it’s on you.
Publish company information to everyone in the company. Publish non-confidential information to the entire world.
This is the lesson I think most of you already know. But I think we often stop there, with openness. And I don’t think that’s enough.
We need to look more deeply into this because this is THE problem of our century.
I fervently believe that amazing solutions to so many of our major problems — renewable energy, education, disease — exist out there somewhere, but they are in pieces. You have a piece of the solution and someone in Bangalore has another piece of the solution. And if those ideas find each other in ten years, we’ll save thousands of lives, but if we can help those ideas find each other in ten months, we’ll save millions.
So I want to celebrate our advances in this area, but I also want to critique them. Because it’s worth the effort to do better.
Part 2: The Broken Social Media Spiral
Clarke says something interesting in 2003 about the GPS idea he had had a half century before. He says it was obvious. Anybody in their right mind could have seen it. He didn’t think it was that special an insight.
Carol Goman calls this phenomenon “Unconscious Competence”. You don’t know the value of what you know. It’s not just that Clarke didn’t send his letter to the right people. It’s that Clarke didn’t think there was that much of interest to tell. He sent out that letter, but for the ten years before that that he had had that idea, he didn’t send letters to anyone.
So, we start from this point — knowledge capture — and move forward. The biggest problems of the information age is how we make the most of the massive amount of information we collectively have.
And to make use of it, to really make use of it, a few things have to happen. We have to:
record it somewhere
route it to the right people
extend , organize, localize it
pass it back into the streamfor the next iteration
There’s a broad feeling that social media has solved this problem. I think it’s solved a lot of it. But as I think we’ll see, there’s a lot left to improve.
So let’s start with the writing down piece of this, the record, part of what knowledge management people would call externalization. So Jim Groom, for example, was here two years ago giving the keynote, right? Now, Jim and I go back to 2007. We’ve been working and thinking in an area you can call EDUPUNK Connectivism for seven years.
I talk to Jim a lot — and we do it through comments, Twitter, and blog posts that reply to one another.
So what the what the web has done, and blog-like technologies in particular, is move these individual exchanges really quickly to externalization. Ninety percent of what Jim and I have talked about over the years is online, in public, where it’s findable, searchable. Others can benefit from it. Openness combined with these blog-like products — Twitter, Instagram, Pinterest, Tumblr, whatever, makes externalization the default.
That’s progress. To be frank, that’s a TON of progress. But there’s a couple things that makes this approach less than ideal for broad dissemination of ideas.
First, let’s look at the externalization problem. The first problem is that social media tends to get only a certain kind of idea down. Remember Clarke with that letter — it wasn’t just that he didn’t publish it broadly. it was also that he didn’t know it was worth publishing.
These platforms are conversational which makes us overly concerned with publishing interesting stuff.
But here’s the problem — I’m embedded within a pretty advanced group of people in educational technology. Ideas that we think are common might be revolutionary for others. But we’ll never produce posts or tweets about them because everyone in our clan already knows them. And the stuff that we do produce assumes you share our background, so it’s not always readable outside our clan.
And it works the other way too. We’re dumb about a lot of stuff that other folks could teach us a thing or two about. But the chance of that “unconscious competence” reaching us is pretty close to zero.
On the routing stage, I actually think routing goes pretty well on these platforms. I’m amazed at what finds it’s way to me via twitter and blogging.
But the extension piece, that’s an issue. When a blog post from another twitter subculture finds me it’s been routed through all these other nodes. So let’s say an economist publishes something on the String Quartet problem, which is a classic productivity problem that ends up affecting higher ed. I get *some* annotation, right? Someone posts in their stream a link and says something like “This relates to higher ed too”. That’s a helpful prompt.
And maybe that’s enough annotation for me to grok how this obscure economics post relates to my work. Maybe.
But for a nontrivial set of things if information is going to useful to the circles it moves to it is going to need to be recontextualized and reframed. And in a perfect world it would actually be re-edited, wiki style, to foreground the parts that most apply to higher ed and eliminate the pieces that don’t.
A world of compentent extenders would also be a world where we don’t treat these posts like the exhaust of our thought process, thoughts to be expelled as we think them and never returned to. Ideally, when we learn more about an idea we posted several months ago we’d go back and update that post. If I think of a new thing it’s connected to, I’m going to want to write in that new connection.
Extension is where things like wiki have excelled, where communities have worked to extend and connect ideas rather than just retweeting them.
So there’s a composition teacher buried in me — it’s what I did before educational technology. And looking at this I can’t help see the “Kinneavy triangle”.
Kinneavy took the speech triangle from lingustics (speaker, listener, referent) and used it to explain composition. You had narrative (I), dialogue/persuasion (focussed on the you), and exposition (focussed on the “it” — what we’re talking about).
And his idea was that you move students through these modes of writing. But I’m interested in this as a sort of lifecycle of information. An idea starts out with what it means to you, the “I” in this situation. Then it pings around a social network and is discussed (the “you” phase). And then in the final phase it sort of transcends that conversation, and becomes more expository, more timeless, less personal, more accessible to conversational outsiders.
When I look at this triangle, it seems to me that different technologies excel at different stages. Things like Evernote and Delicious or Diigo excel at that “I” part. Here you just take notes on what means something to you. And you don’t want it to be dialogic necessarily, because that ends up limiting what you capture. Start out by just caring about yourself, and you’ll actually capture more.
Twitter and blogging, on the other hand, excel at the dialogic and persuasive functions. Ideas ping around and reach unexpected people. Sometimes you even learn something.
For the expository phase, it’s wiki that excels. By cracking open ideas and co-editing them, we turn these time-bound, person-bound comments into something more expansive and timeless. We get something bigger than the single point of view, smarter than any single person.
So one thing I’m interested is how we create a system that allows information to flow in this way. One way might be to link up Evernote, Twitter, RSS Feeds and Wiki in a certain way.
Another way is to start at the end technology — in this case wiki — and look at what it would take to make it work better in the other stages, the I and the You, the personal and the dialogic.
So that’s what I’m going to do today. I’m going to demonstrate a newer technology called federated wiki which allows the sort of communal wiki experience, but also supports those earlier stages of the knowledge life cycle.
Part 3: Kate Middleton’s Dress
So here’s the problem with using wiki for those first two stages. Wiki, as it currently stands, is a consensus *engine*. And while that’s great in the later stages of an idea, it can be deadly in those first stages.
As an example of this, how many of you have heard of the Kate Middleton Wedding Dress Wikipedia fiasco?
OK, so this is fascinating. Here’s how it went down.
There was all this talk before the Kate/William Royal Wedding about the royal wedding dress, which is apparently a big deal. It’s historic.
There’s a whole fascinating history of wedding dresses and monarchs. The white wedding dress that’s such a staple of weddings nowadays? It goes back to Queen Victoria’s wedding dress. Diana’s dress, if you’re old enough to remember those overhead shots of the wedding, had a 25 foot train, which apparently made it really difficult for her dad to sit in the carriage with her on the way to the wedding. Kate Middleton’s dress had more modest nine foot train.
You can see the world in a grain of sand according to William Blake. And wedding dresses are not my thing, but for some people they’re that grain of sand.
So some enterprising person went to Wikipedia and started a page on Kate Middleton’s dress. And she and some others started to build it out.
About sixteen minutes later, someone – and in this case it probably matters that is was a dude – came and marked the page for deletion as trivial, or as they put it “A non-notable article incapable of being expanded beyond a stub”
So all hell breaks loose. You get the pie fight of all pie fights. Here’s the smallest sample of comments on that talk page. These are all just about whether the page has a right to exist.
Dress defenders say this is a dress of historical importance. Dress attackers say there’s no other wedding dress pages up. Dress defenders say – are you honestly saying that Wikipedia is defined by what’s NOT in it? Wouldn’t that, taken to its logical conclusion, mean that you couldn’t add anything?
Finally founder Jimmy Wales shows up on the talk page and to his credit says:
Strong keep – I hope someone will create lots of articles about lots of famous dresses. I believe that our systemic bias caused by being a predominantly male geek community is worth some reflection in this context. Consider Category:Linux distribution stubs – we have nearly 90 articles about Linux distrubtions, counting only the stubs. With the major distros included, we’re well over a hundred. One hundred different Linux distributions. One hundred. I think we can have an article about this dress. We should have articles about one hundred famous dresses.–Jimbo Wales (talk) 08:58, 30 April 2011 (UTC)
You can imagine this as a movie moment. A Jimmy Stewart, Frank Capra moment where everyone on Wikipedia looks sheepishly at one another and comes to their senses, right?
Except of course, no. This is Wikipedia, so a bit after this someome replies — to Jimmy Wales, co-founder of Wikipedia — basically, stop being a bleeding heart social justice warrior. You’re driving away all us male geek editors with your “activism”. Ugh.
And if you want to know why wiki never got the traction, I think that’s the reason right there.
If you find a community online big enough to be socially interesting it comes with this baggage. You have an idea, you hear a fact, you learn a technique you want to share.
You go online to share it and you’re teleported past the personal and dialogic and suddenly find yourself having to defend the inclusion of this fact or this edit FOR ALL TIME. In many cases, you’re arguing with *pedants*, and even where the conversation stays amicable, is this really how you want to start your day?
And it gets worse, because if you lose that battle (notability, accuracy, citations, linked ideas — whatever the battle is) your contribution disappears. It’s easy to say that it’s all in the revision history, but in practice what Google can’t see does not exist, and Google can’t see that revision history.
Part 4: Introducing Federation
Wiki is a relentless consensus engine. That’s useful.
But here’s the thing. You want the consensus engine, eventually. But you don’t want it at first.
It’s funny, I was looking over this keynote last night, and I saw this line and realized — this is the simplest explanation of federated wiki.
You want the consensus engine, eventually. But you don’t want it at first.
This is a problem. It’s pretty easy to build a system that starts with consensus and then fragments into personal opinion and individual statements. In fact, we build a lot of systems like this accidentally. Entropy can be added rather easily to any system.
It’s harder to build something that starts fragmented and personal, but then organically becomes a shared communal space. You’re looking for a system that produces what Polanyi called “spontaneous order”.
So back in February I saw a video of Ward Cunningham, the initial inventor of both wiki and really of wiki culture — he was the first wiki admin as well, and many of the conventions of wiki come out of his unique take on how collaboration fails and how it succeeds.
And what he says in this presentation is we end up being pushed to consensus in these systems because we’ve got the technology upside down. Here’s a slide to show what he means by that.
In a traditional wiki, you have multiple people sharing a single server, and the server is the ultimate arbiter of what’s on the wiki. In a federated wiki, everyone has their own server which stores the records associated with them. But the meaning is made in your browser. Your browser pulls wiki records from all over the internet, and makes them look like they exist on a single server.
If you have a background in network theory, I think you’ll see immediately at how this inversion creates a sort of evolutionary ecosystem where we reach consensus not through arguing who gets to control a page on a specific server, but by seeing which versions of a page spread to where.
If you don’t have a background in network theory, I want you to forget this slide, and just watch what I’m going to do next. In my entire time talking about this I’ve found almost no one gets the massiveness of the architectural shift at first — I don’t want you to understand it, I just want you to know it’s there.
Ultimately, the theory is that this thing here is the “spontaneous order” engine that can move from the fragmented to the unified.
Either that hypothesis is right or wrong. My job here is not to argue a side of that hypothesis — it’s to convince you this is an hypothesis worth testing.
Again, if you don’t completely get this idea, don’t worry. The view from the user side is much simpler.
Part 5: Arthur C. Clarke Uses Federated Wiki
So I want to sketch out what life looks like if Arthur Clarke had federated wiki in say 1950, which we’ll guess at as a time when this GPS idea came about.
Incidentally, if I happen to be wrong and APL did know about Clarke’s idea, you can replace references here to GPS with communications relays, satellite phones, home computers, or any of the dozen major predictions Clarke envisioned years before they happened.
In our scenario, Clarke keeps a journal, and one day he thinks:
If three geosynchronous satellites were in orbit they could ping you the time. By measuring the different delays of the pings you could calculate how far you were from each satellite and triangulate your position.
The following video shows how that might look.
Important: The video contains the next two minutes of the presentation — the presentation won’t make sense without it.
So that’s the capture and linking side of things. We jot down idea, write reactions to readings, organize our own thoughts. Over time, we connect them, like a personal Memex.
The next stage is that routing we talked about — how do ideas spread? In this video we show you how Clarke’s idea spreads through copies in a decentralized way that requires no central service.
Important: The video contains the next two minutes of the presentation — the presentation won’t make sense without it.
And then, finally, *really* neat things happen. Clarke has recorded his ideas and linked them. Readers like Maria have propagated those ideas by follow a fork-to-like convention. When they finally reach a physicist working on a similar issue we move to dialogue and eventually the extension of the resource. Again, the network here does not just route the pages; the pages are expanded and extended by the nodes they touch.
Important: The video contains the next three minutes of the presentation — the presentation won’t make sense without it.
And there we are. We’ve moved from the personal, to the dialogic, to the expository. We’re working on resources and ideas together rather than thumbs-up or thumbs-downing. Kind of lovely, right?
Part 6: Academic Uses
(Here we walked through class examples and personal examples, mostly demonstrating how students move through a similar progression, from the I to the You to the It. It was unplanned and driven by audience questions, and I don’t have a transcription.)
I don’t have much of a conclusion here, actually. If I’ve done my job, you should be either excited about this, or absolutely terrified. Either is fine, actually.
As long as you’re not complacent.
I’m hoping we can talk about this. Some possible applications. I can share where the fedwiki team is with it.
I’m hoping I get some allies. Maybe you’re all allies.
But even if you’re not, I hope this can open up an honest discussion about the ways in which social media is not serving our needs as it currently stands. As advocates we’re so often put in a situation where we have to defend the very idea that social media *is* an information sharing solution that we don’t often get to think about what a better solution for collaboration would look like. Because there are problems with the way social media works now.
My hypothesis is that this federated scheme solves many of the problems. I might be right.
But what I *know* I’m right about is that these problems exist, and they are serious.
Minority voices are squelched, flame wars abound. We spend hours at a time as rats hitting the Skinner-esque levers of Twitter and Tumblr, hoping for new treats — and this might be OK if we actually then built off these things, but we don’t.
We’re stuck in an attention economy feedback loop that doesn’t allow us silent spaces to reflect on issues without news pegs, and in which many of our areas of collaboration have become toxic, or worse, a toxic bureaucracy.
We’re stuck in an attention economy feedback loop where we react to the reactions of reactions (while fearing further reactions), and then we wonder why we’re stuck with groupthink and ideological gridlock.
We’re bigger than this and we can envision new systems that acknowledge that bigness.
We can build systems that return to the the vision of the forefathers of the web. The augmentation of human intellect. The facilitation of collaboration. The intertwingling of all things.
This is one such proposal. Maybe you have others.
I should be prepping for the NWACC keynote. — it’s in a couple hours. But of course in going over my notes and reading some recent posts (particularly this one by Bonnie Powers) I suddenly doubt the route I’ve chosen into my subject (which is, of course, federated wiki).
Why? Because I talk about the societal implications of using the current breed of software, how our news-pegged attention economy fixation causes us to miss important ideas and shuts out opposing voices and minority opinion. And that’s stuff that’s really important to me.
But the real reason that I go out and plug federated wiki — the idea-too-big-to-get — is that using it has given me my brain back. I’m engaged with ideas again, not just personalities. Writing in my fedwiki journal gives me the space I need to think without worrying about how interesting I’m being, whether I contributing something new to the conversation. It gets my head out of the stream for a bit. It feels nice, like a personal library of slightly musty books on a beautiful rainy afternoon.
Maybe if a million people were using Federated wiki that feeling would disappear. Maybe I’d get addicted to forked pages, extensions, the like. Maybe having a thousand people on my feed would recreate the self-consciousness that exhausts my introvert self.
Maybe. But if there’s even a chance we could make the future less of the conversational pigpile that forms Twitter or the personal exhibitionism of Facebook and Instagram, we should pursue it. Federated wiki provides the routing and discovery architectures of current social media. But it also has a place for quietness. It allows one to attempt to break out of time, to see rather than react.
If I could somehow transfer that experience to people, these presentations would be a breeze. As it is I’ll just be talking about smaller things, like saving society from nuclear war and finding cures for cancer.
Keynote given at NWeLearn, 10/23/14. Originally titled “Taking Education Out of Airplane Mode”
Speaker’s note: I write unique pieces for presentations I give. I’ve not yet learned to economize and give set presentations to multiple audiences. And I approach presentation as articles, revising them mercilessly over months and months. So word of warning, the early drafts of this were titled “Taking Education out of Airplane Mode”. Somewhere about a month ago, a better title would have been “Design Patterns and the Coming Revolution in Course Design”. And then yesterday I was looking at this and thinking the real title should be “How to Fight Big Design Without Becoming a Design Anarchist.”
It’s all the same idea. As Tom Stoppard would say, it’s “two sides of the same coin, or the same side of two coins”. It’s all intertwingled. But I thought you should know.
This presentation is also fresh. I’ve danced around these questions on my blog the past few months while I wrote this, but kept this larger work under wraps. So this is my first comprehensive presentation of this idea anywhere, and I’m really interested in what you think.
One thing not mentioned in my bio is that I’m a songwriter. Here’s a slide pertaining to the third most bizarre incident of my life, when a song I had released under the name The Russian Apartments shot to 26 on the Latvian Airplay charts.
You’ll see that I beat out Jennifer Lopez, the Foo Fighters, and Depeche Mode, but couldn’t quite take down Katy Perry.
Though what you’re probably looking at here is number 36. Chris Brown is still putting out music? Who knew. The magic of Latvian radio. Anything can happen.
I mention my songwriting not because I think there may be some Latvians in the crowd I will impress, but because songwriting for me is a way to think about creation, and what we are in this room are creators.
Songwriting is like most other creation. You start with inspiration: a snatch of melody or a killer progression, but that’s when the real work begins. You start to add pieces — a bridge, a pre-chorus, a keyboard hook, a bass line. When you work solo on multi-track, as I do, and you are scoring 5 to 20 instruments you become acutely aware that each decision you make constrains you further. As you progress, there really are limited ways these things can fit together.
This should feel increasingly claustrophobic, but it feels quite the opposite, at least for a bit. There is a feeling in the middle of songwriting of acceleration, a feeling that you are not so much creating something as you are chipping away the stone to reveal a song that is already there.
This “emergent” coherence to this process feels like the peak of being human, and it’s not for nothing that when it works — when a song we listen to fits together like something living — we feel human ourselves. It doesn’t matter that “Waterloo Sunset” by the Kinks is a fairly lightweight pop or that “Debaser” by the Pixies is sophomore surrealism — the songs progress in a way that almost seems inevitable. They move from being a sequence of vowels and rhythms to words and melodies and beats.
Now a lot of people think this coherence is an abstract alignment of part to whole. But it’s much more than that.
It’s the feeling that the parts of a song are *re-acting* to *one another*. When we talk about what has the “spark” and what doesn’t, that’s the secret. And it’s built into the design process.
The same is true of great courses – the elements of a great course exist in a creative tension, and drive us forward into unexpected but seemingly inevitable directions.
Yet when I look at how we create courses and how we make songs, I’m struck by the difference. Each process is equally analytical. But the songwriting process generates serendipity, creative tension, and spark. And the course design process – the official course design process — all too often works against those things.
Take away that course design process though, and chaos reigns. It really does. Years as an instructional designer have taught me that some people can create a complex tech-mediated course from their gut, but most can’t. The scattershot nature of much of our “un-designed” offerings is a result of the assumption that most teachers can be the John Lennon of course design.
So I became obsessed with this question of whether there is a third way. Can we get coherence and creativity? Data-driven design and spark? And I started looking at other models in other disciplines. It got really interesting, and that’s what I want to talk about today.
The Coherence of the New England Downtown
This is the town where I spent the last 10 years of my life, before I came out to the Pacific Northwest.
I don’t know the sequence it was put together, but I do know these buildings have different dates on them.
You’ll have to forgive the wideness of Main St., a historical mistake they’ve done their best to work around. The rest of the street grew up around something that began as a square but ended a thoroughfare. The church was one of the first things in, early 1700s or so, and things sort of arranged themselves around that.
And weirdly, although these buildings span from the early 1700s to the mid-20th century (and although Main St. is far too wide) the effect of the downtown is one of coherence. And Keene had its issues, but I miss that downtown badly.
Here’s where I live now. Well, OK, not this. I live in a smaller twelve unit project, but along the same lines. It’s part of a development, and around it there are shopping centers. When we think about sprawl, we tend to think that it was due to a lack of design, but often the opposite is true. The development in this picture was fully planned out before a single building went up. Every last detail was controlled.
So here’s a question: how is it that the experience that was designed by hundreds of different people over hundreds of years feels more coherent and alive than this development which was designed by a single person?
The key, I’ve found, is emergent structure.
The Keene downtown was never built to be finished. If you walk down towards the end of Main Street, you can see the loose ends of the street. If Keene is smart, when new buildings go up at those loose ends they will emerge out of the constraints of the surrounding architecture and environment.
Like a song, the new pieces react to and mesh with the pieces in place. Unlike a song, they leave loose ends for the next expansion to pick up.
The development I live in is different. It was designed to be finished on delivery. Self-contained, with all threads wrapped up. Things are connected, but they don’t *react* to one another. And so, architecturally at least, we get this sort of still-born community. A structure that turns in on itself. Inside are wonderful people making the best of the situation, but they are doing that in an architecture that works against conviviality.
Does that in-turning structure sound familiar to those who have been watching the direction of online course design?
If not, you’re lucky. If so, you’re a bit ahead of me here, though trust me, we will get there soon.
Alexander and Beyond
Observations about emergent structure aren’t unique to me. They were made in much better prose with much better precision by architect Christopher Alexander a half a century ago. And by others too: Jane Jacobs, for instance. By a lot of people. But Alexander, in particular, invented a whole new way of design, built around this concept. In Alexander’s methodology, referred to as “a pattern language”, designs are generated the way one might generate a sentence — with local context and needs pinging around the rules of grammar and producing infinite sets of expressive meaningful sentences. And this forms what Alexander called living structure.
Here’s how Alexander’s system worked. You’d have a pattern and it’d look like this — “Light from Two Sides”.
It’d be based on empirical study of how people react to different rooms. Alexander says that given two rooms, people will congregate in the room that has windows on two walls.
He’d have his students go out and measure this stuff. The patterns were data-driven. And for him, this is sort of a cognitive constant that the builder has to design around.
And that’s important too — it’s not an aesthetic to Alexander. Many aesthetics could emerge from this. It’s a cognitive and physical constraint, like people being most comfortable when the temperature is between 60 and 80, or people finding white noise more peaceful than sporadic explosions.
Have you ever wondered why people are so excited about the corner office? Well, here you go: Light on Two Sides.
But what makes this observation special is the way it feeds forward into other design details.
How do you create these multi-windowed rooms? You’ll find a reference here to higher level design concepts like the Long Thin House. And you’ll find reference to lower level concepts like Deep Reveal, and Windows Looking out on Life. One pattern leads you to another. Want Light from Two Sides? You can do Pattern Long Thin House to maximize the number of rooms with that light pattern. You can use courtyards, Cascade of Roofs, staggered outside boundaries.
This is not how we approach design right now. We approach design much more like the architect of my housing development in Washington. “OK, we’re going to need roads, and houses, so here are the plots, and here’s where the houses will go…”
We have a top-down design process that creates sterile communities, communities are the result of top-down constraint.
Patterns as they were developed by Alexander were supposed to help mitigate this. They’d make visible the values and knowledge of a community of users. They’d form a sort of contract as well.
A good example of this is the University of Oregon, which has 70 or so patterns in its campus plan which they have had since 1970. The idea of the pattern approach was to move out of the self-reflexive, feature complete design, but provide a template that made the campus cohesive. Here are the principles that Alexander defined for that project:
- Organic Order: the design emerges through a process, not from an initial blueprint.
- Incremental Growth: development occurs by small increments.
- Patterns: a pattern language guides the planning process.
- Diagnosis: development is guided by an analysis of the problems with the current design.
- Participation: user involvement must prevail throughout the planning process
- Coordination: working together benefits the product as a whole entity.
You’ll notice that the patterns are just a piece of this. But it’s the mechanism which holds the rest together.
Design patterns are the mechanism by which the communal, long-term vision meshes with the participatory incrementalism of the short-term process.
So here’s a pattern from OU called arcades. Keeping people outside is important to your community values and purpose, but local weather sometimes precludes that. Those two things — the constraint and the community value — ping off one another and suggest arcades as a pattern.
But what about research and accountability? The research is built into the pattern. You see that line up there about the data in Pattern Language? Halfway down the page? We call this thing a name and then we can collect data about it. We only adopt it as a community if it works in a broad variety of contexts, it supports community values, and the data tells us it works.
Once we’ve established the pattern as a community sanctioned one, the community of users can go to the architects and say we want an arcade on this. And if the architects say no, there better be a darn good reason.
You’ve given design to the people by giving them design patterns with the RESEARCH BAKED IN.
Here’s a different kind of pattern called Department Hearth. And there’s a community value here that we want to facilitate discussion among people working together. And that’s run through an empirical finding that space matters — people talk to each other more when we engineer chance meeting. So buildings have to accommodate that.
Do all buildings have a common hearth and an arcade? No, not exactly. But the process of design requires that the architects explicitly consider each pattern, and if they reject it they must provide a justification why. Power to the People.
And the things the users end up specifying from the patterns — they make it easier for the architect to build the building.
It’s like a template, but it’s also the opposite of a template. The template constrains, but not in a generative way. The template doesn’t set your creativity buzzing.
This? This does.
Could this work with learning design?
I think it could.
Arcades and Department Hearths provide a way for a community to develop buildings in a decentralized way while creating the sense of a coherent whole. And maybe this approach could provide us a way to work with faculty in the driver’s seat, and to rid ourselves of so much of the dead design that plagues our higher education.
Maybe it could give faculty the tools to create their own vibrant spaces while engaging in good design and producing a coherent experience.
Programming’s Sea Change
Where the story of design patterns is going to lead you next is software, because it is in software that it has had the most influence.
In the mid-80s Ward Cunningham, Kent Beck and others were frustrated by software design. This was an age of scientific design and management. You’d determine your goal. You’d write up the product design. You’d produce your timeline/waterfall chart, sketch out your module dependencies then write a bunch of modules from scratch. In a short twelve months you’d have your prototype, show it to the customer, find out they hated it and then you’d go back to the drawing board.
Cunningham and others were fascinated by design patterns, and a lot of the work they did was developing libraries of design patterns for code.
They captured community solutions to software problems the way Oregon captured community solutions to architectural problems. And they were somewhat successful with that and somewhat unsuccessful. I think you can say that the idea of software patterns had some influence on every piece of software in your pocket right now, but only a fraction of them were built using something like a patterns repository.
But the thing that they were proposing when then pushed design patterns — well, that’s a different story.
Because the other piece of what they proposed was something called “agile programming”. And what the advocates of agile programming (or eXtreme Programming) said was radical at the time. They said, roughly, that the design emerges from the construction of programs, rather than the construction from the design. You build something to find out what it is you really want to build. And you do it not so much based on a step by step plan as through rearranging ideas that have worked before and seeing what happens.
In other words, even though programming involved an awful lot of people working together, it could and should be less like that waterfall chart, and more like writing music, or composing an essay, or fishing, or cooking by taste.
It’s difficult to grasp how crazy this sounded to people in corporate software in the 80s and the 90s. Sure, a process like that could produce a small piece of software. Individual programmers had kinda sorta been doing this with smaller project from time immemorial.
But to bring such a thing into the enterprise? And to hold the work together with patterns and pair-programming and other organic methods? It was nuts. It was sloppy. It was a method fit only for individual hackers, and spaghetti coders — not for highly paid teams at IBM.
Top-down design tells us that the abstract concepts give rise to the instantiation of those concepts. There’s a hierarchy. Theory begets practice. Specification begets code.
Agile programming said your specification *was* the code. It said your common architectural designs — if considered at the right level of abstraction and empirically verified — could guide you through construction better than theory, needs analysis, or top-down design.
That sounds abstract, but this part is not. People don’t write 18 month prototypes anymore. There may be some companies that do waterfall charts of programming projects, but you be hard pressed to find the most talented programmers there.
In fact, the way software is designed today looks a lot like the way Oregon designed their campus:
- Organic Order: the design emerges through a process, not from an initial blueprint.
- Incremental Growth: development occurs by small increments.
- Patterns: a pattern language guides the planning process.
- Diagnosis: development is guided by an analysis of the problems with the current design.
- Participation: user involvement must prevail throughout the planning process
- Coordination: working together benefits the product as a whole entity.
And this is my thesis — as Instructional Designers we’re sort of in what the dark days of the 80s were to programming.
In many places we’ve become implementers of the spec.
We are, in many places, pushing faculty into ADDIE inspired processes — processes that were created to build instruction for military personnel going into very defined jobs. We’re using those same processes to teach students who are going to switch jobs a dozen times in their careers.
Here’s a final thing about Ward and others in that agile software movement.
They had this crazy “agile” idea, but to their minds they were less scientists inventing the idea than anthropologists discovering the system under the system. Big Design was in full force, but if you pulled back the curtain, you’d find that what good programmers were doing, in actual practice, was more organic than that. They had hints of agile programming in their process — they were just in a system that made them feel guilty for that.
I don’t know if this resonates with anyone out there. But it resonates with me. I don’t want to work in a system that makes me pretend things I know about process are not true. I don’t think you do either.
I’d started this presentation saying that we were designing courses wrongly, and were being hindered by that. I’m hoping by now you can see there are approaches to design that we have not truly considered as a field.
But how do they relate to learning design?
Here’s my take. What we currently have in our institutions in the face-to-face space is a sort of design anarchy. Every professor chooses his or her own approach. And while that allows the occasional Jim Groom or Michael Wesch to flourish, for the most part it produces something architecturally similar to this:
Many face-to-face classes are good, but so many are not. There are an awful lot of classes which have no empirical basis for the way they are designed whatsoever. And even where they are good, the experience can be so disconnected that the most flexible of students feels a bit overwhelmed.
At the same time, the online world is moving in. This is a world where many design methodologies came out of the military, a world comfortable talking about courses as reproducible experiences, suitable for mass manufacture at a zero marginal cost. And they are bringing in something that looks like this:
And I think a lot of administrators are frankly relieved about this, because as more and more education moves online, the idea is that we can bulldoze our strip-mall exburban eyesore and replace it with something centrally managed and controlled. And courses will be delivered as these closed, feature complete products, designed by the experts — us, the instructional designers.
And the feeling I get is in this massive battle that’s about to happen you are expected to be either for design anarchy or for Big Design’s waterfall process.
When I try in my job to move away from Big Design, I feel I always have to fight the impression that I’m being unserious, or unstructured.
I want to say to people, look, try running a blog as your class site, putting all your communications and materials into a single place. One which also aggregates student work. Seriously. It will change the nature of your class.
Well, hold on a second. What learning objective does that serve? What letter of ADDIE is that? People can get a bit freaked.
And when I talk to people on the design anarchy side, it’s worse — I’m making this too complex, they say. Let everyone do what they want, and it all works out.
Well, has it? Do most of our students feel like they had a coherent, interconnected school experience? Do most of our classes engage in empirically supported methods?
In a way, of course, the design anarchists are the ones who should listen up the most.
Because if you’ve watched the K-8 space, you know who wins the anarchy vs. Big Design battle. It’s Big Design and it’s not even close.
We are moving to a future where every minute of every day of a grade school teacher is precisely scripted by a textbook company in collaboration with a district. I have zero doubt that unless something changes this will be the fate of much of higher education as well.
If you’re an instructional designer, then maybe you get to write that script. Again, yay you.
Instructional Design Patterns
But what if we could do better than that? What if we could transcend the ADDIE mindset that has plagued us all these years?
What if we could push good design not by learning designers scripting classes, but by us creating the learning design patterns and the enabling architecture for faculty to implement them?
What if instead of saying hey, you need a course blog, and we can do that because of ACADEMIC FREE-ANCE!!!111! — what if we agreed as institutions that courses need a “Learning Hearth”?
LEARNING HEARTH (1):
Begin with a preposition:
When each student’s digital learning space is separate from other students learning spaces, and there is no peripheral awareness of what other students are doing, a student’s experience becomes disconnected, losing the SYMPATHY OF NUMBERS (34) provided in a physical space. Under these conditions there is little chance for a community to develop; students become disengaged, unmotivated, and less likely to learn.
THEREFORE, in every digital space create a LEARNING HEARTH, a space in which students are forced to bump into each other’s work. This space should combine announcements, course materials, and important links with feeds of student created content, linked and excerpted to entice the student to engage with them. Links should be composed of RIVERS OF WORK (48) and class information should be NEXT TASKS FOREGROUNDED (12).
This leads us to RIVERS OF WORK, and NEXT TASK FOREGROUNDED.
Here’s RIVERS OF WORK.
RIVERS OF WORK (48):
Begin with a preposition:
When students do tasks in isolation, they can often misunderstand tasks, or apply less effort than when they complete these tasks in the presence of others. Additionally, when courses consist of online assignments, students can feel disconnected from other students in the class. Theories of social facilitation also suggest that a public-ness to work can increase motivation and engagement.
THEREFORE, create RIVERS OF WORK which show students the latest work submitted by their peers, and encourage them to react to that work either by comment, their own work, or a remix of the presented work. Attention should be given to river size as small rivers do not generate forward class momentum, and large rivers can be overwhelming (see data).
RIVERS OF WORK are a crucial part of the higher-level pattern LEARNING HEARTHS, and maybe curated into ASSIGNMENT EXEMPLARS.
An important thing to see here is that looking at RIVERS OF WORK will lead you to the LEARNING HEARTH pattern, and the LEARNING HEARTH pattern leads you down to RIVERS OF WORK. And RIVERS OF WORK suggests ASSIGNMENT EXEMPLARS.
And thinking about RIVERS OF WORK makes you think of what those rivers would consist of. Reading response? Article summaries? Updates to a crowd-sourced project?
You do this already, I’m sure. “What would this class look as a wiki? A blog?”
But again, we’re not supposed to think like this. We’re supposed to define the Student Learning Outcome and choose the best possible assignment for that SLO.
We’re supposed to take the plot of land, draw the roads, divide it into plots, and determine the shape of the house by the plot, and the layout of the rooms by the shape of the house.
Eventually, at some level, where you put your dining room table is determined by the needs of the road system. And that’s bad, because the roads have to be aware of the needs of your dining room table as well.
Top-down gives you consistency, but it doesn’t give you anything living structure. The magic actually comes about when you think — what does a RIVER OF WORK look like in Psychology 101?
It’s in the collision of the design pattern with the local need. We’re allowed to think through the pattern.
What this generates with WordPress in a Philosophy class for a general education class will of course be different than what it might generate with Canvas in a Physics class. But there will be a coherence all the same to the student experience. And while we are not enforcing checklists of standards on faculty, we are, in fact getting the sort of consistency and quality that matters.
It gets better. The work and testing you do with LEARNING HEARTHS at your institution — and the things you learn — transfer easily to work at my institution, because we have a common language.
And if we’re not afraid of the “f” word — feelings — we could start to look at what a learning hearth should FEEL like. Alexander says, quite rightly, that during design holding on to what something should feel like is the best way of keeping on course, because you know when you’ve taken a wrong turn right away. Select three examples of Learning Hearths for the faculty member to look at. Have them pay attention to how it feels.
Sound like hippie nonsense? Read Kahneman, read cognitive science on thin-slicing.
Again, I think this stuff is, in fact, what many of us are already doing, but just not in a formalized or institutionally approved way. And not in a way for which we have a language. And not in a way that our claims can be empirically tested, or formally analyzed to guide improvement.
And I want to say quite clearly that idea of learning design patterns is not new. As I’ve gotten into this I’ve found that people have been kicking this idea around since at least the mid-90s, and that in the past few years various efforts have made some significant progress. Here’s a book I just recently came upon from 2010, for example:
And I would bet you that after this keynote four or five of you will come up to me and tell me your own learning design pattern project. That’s exactly what I want out of this talk.
As for the rest of you, some of this, no doubt, will seem confusing to you. It probably is confused, a bit.
But I return to the software example.
Years ago you had the maverick hackers and the institutional plodders. If you worked in a group you were probably following the approved process feeling miserable, or not following it and making other people’s lives miserable. And life sucked for programmers until the industry realized that you could have fluidity and structure at the same time if you made that a priority. And what I feel now, instinctively, is that we are in those dark ages.
I started out writing this cheerily, working on it in my free time, seeing where it led. And somewhere in the middle — not of this piece, mind you, but of the writing process — it got very dark.
We are in dark days. Everywhere I look I see the exuberance of the mid-aughts being crushed by processes and requirements and specs that don’t fit. And where I see signs of hope, it’s all cowboys and mavericks, stuff that will never mesh with our institutions, that sets itself up as in conflict with the very institutions in which it is trying to root.
I’ve got nothing against cowboys. As the great philosopher William Nelson said, all my heroes are cowboys. You look at the stuff I’m doing on my own campus, and it’s being run off of off-campus servers. You look at where I’ve been successful and it’s grass-roots under-the-radar stuff.
The dark piece of that is that it can’t take root in the organization in a way that transcends the individuals involved.
Here’s a story for you. I came to Washington State University, gung ho about my new job, and wondered why we were behind on the use of things like blogs and wikis. I started building stuff, because there was nothing there.
It wasn’t until I was a year into it that someone said to me — this is great! It reminds us of the old MediaWiki we used to have.
Turns out WSU was a *leader* in this stuff. Papers, keynotes, conferences. Early — and impressive — use of a university wiki. The use of worldware for ePortfolios. Student blogging and course blogs in 2006 or so.
Leadership changed. People were moved, rearranged, let go. Servers were shut down, and student work deleted. And this experimentation was obliterated so fully that I only came on it by accident.
This is not a unique story. I have a friend who presented to the board of a prestigious college about Networked Learning. And after talking about blogs, wikis, and the like he was told that they’d had all of that – in 2008. It’s gone now.
If I opened up the floor right here, I’m sure many of you could tell similar stories.
It’s easy to shake our fists at some imaginary villain here. But if we want better learning design, we can’t get that, strangely enough, by simply producing better learning design. Or we can’t get it permanently at any rate.
Instead, we have to look at the process by which we produce classes, and come to new community and professional understandings of what that process should look like.
We have to articulate that process publicly, so that the people following new processes are not mistaken for cowboys or illiterates.
And we have to make that new process comprehensible to faculty, and give them the tools to participate more fully while still preserving the integrity of the process.
We have to root deeply into the Way Things Happen. Or even more than that – we have to change the soil so these ideas can take root.
I said it got dark in the middle of writing this. It did. It kept getting darker, until I was talking to someone involved in the Agile Software movement. And what they said is, look, we launched this thing in the 80s, we didn’t get everything we wanted, but programming is a much better job today than it was even ten years ago. The way companies have programmers program now — almost everywhere — is much more in line with how we naturally think about these things.
In programming, they found ways for people to bring that organic experience of programming alone into programming with people. In programming, they found a place in between 1970s exburb strip mall loops and the sterility of the 1980s mall.
And that’s promising. It excites me. Because I can imagine us looking at the way we work now in 10 years and thinking, my God, how did we put up with that?
Maybe the learning design approach I’ve specified here is not where we’ll go. But if we can open up a honest conversation about process, perhaps these experiments can take root, and this Groundhog Day could end.
Perhaps. And perhaps, for me, is a word of hope.
Keene Downtown Shots
Outer Strip w/ Car Dealerships:
OU Pattern Language (arcades, department hearths):
from The Oregon Experiment. http://bit.ly/1pxB7Gr
Pattern Language (Light from two sides):
Planet of the Apes, Lady Liberty