Networking Imagination: Main Article

networking imagination: Creativity and Collaboration in the Digital Age

by Jordan Wirfs-Brock

In January 2010, a group of five writers, one programmer and one facilitator met in Berlin. They had two words, a timeline and a goal: “collaborative futures,” five days, write a book.

On the first day, they outlined the subject matter. They plastered the walls with colored sticky notes. They discussed free culture, where collaboration breaks, the role of egos, and countless other topics until midnight.

On the second day, they wrote – no judgment, no editing – for nearly eleven hours.

For the next three days, they continued to write, edit, structure and re-structure. They made sushi. They rearranged things. Interested collaborators dropped in digitally and in-person to help. They rearranged things some more.

By 10 p.m. on the fifth day, they published a 33,000-word book called Collaborative Futures.

This process, called a book sprint, makes National Novel Writing Month seem sluggish. It’s a feat that would be difficult, if not impossible, without digital tools. And it’s just one way that people are creatively collaborating online. Research has shown, over and over, that the more people connect, the more they create. When doors open and cubicles come down, offices are more productive.

When people live in densely populated cities, they generate more ideas per capita. As the Internet works its way into every aspect of our lives, it brings new ways for people to connect. Whether it’s a group of individuals, such as the Collaborative Futures authors, an individual operating creatively within a digital network, or a crowd that takes on a collective voice to solve problems and answer questions, our collaborative creative processes are adapting to the digital age. In many ways, the Internet is allowing the natural creative processes we already have to work better than ever. But the key to effective creative collaboration, whether it’s an individual leveraging social connections or a crowd creating something together, is the strength of the social structures that underlie digital communication. Understanding how the Internet influences collaborative creativity isn’t just important for artists and inventors who make a living off of new ideas. Almost everything people do is either a collaborative act or a creative one. Usually it’s both. There is no such thing as an idea in a vacuum.

Scribophile: Individuals authors writing within a digital network

As intriguing as the group authorship of Collaborative Futures is, most works are still written by a single author sitting in front of a single keyboard, even if that author is contributing to a broader group work. Writing’s reputation as a solitary activity is so deeply ingrained in society that the reclusive writer has become an archetype: Emily Dickinson. J.D. Salinger. Thomas Pynchon. Harper Lee. Cormac McCarthy.

But it’s precisely the solitary nature of writing that makes writing groups so important. Because writers often work in isolation, their connections with other writers can become a creative lifeline. Writers often rely on networks of fellow writers and readers. And those networks are moving to the web.

Alex Cabal, a programmer with a love of literature, founded the website Scribophile in 2007 as a way for writers to workshop each other’s stories. Scribophile is a 21st-century version of the salons and coffee shops that fostered writers’ groups in earlier centuries.

“You can’t sit down with four people in a room and put a good novel together,” Cabal says. “But once you are done with what you are writing, it’s really important to get it out into other writers’ eyes.”

Because it's housed in a virtual space, Scribophile can connect far more people than could ever squeeze into even the hippest offline literary hangout. The site has between 7,000 and 8,000 users (not all are active) who’ve posted 8,800 works of writing and 68,300 critiques. The day before I spoke with Cabal, Scribophile users completed 111 critiques, a number Cabal says is actually rather low.

Exposing an idea – or story, or painting, or software program – to other eyes and minds is essential to the creative process. Some creativity researchers even claim it’s a defining part of creation. In a chapter called, “Creating Creativity: User Interfaces for Supporting Innovation” in Human-Computer Interaction in the New Millenium, Ben Shneiderman writes, “Creative work is not complete until it is disseminated.” Something hasn’t truly been created until it has been shared.

Providing an outlet for sharing creative works is one of the Internet’s great strengths. Niki Kittur, an assistant professor at Carnegie Mellon University’s Human-Computer Interaction Institute, says “It’s much more compelling to have two million people watch your YouTube video than to share it with your friends and your mom.”

Sharing is about more than stroking egos and getting page views. It’s about the feedback the audience provides.

“The audience is not only your audience, but can be your partners in the creative process,” Kittur says.

Through a system of social incentives (which he calls karma) and clever content organizing, Cabal has ensured that every writer who posts an article, poem, story or essay on Scribophile will get at least three critiques. The feedback covers everything from tone to grammar to structure:

possibly a colon here (I'm horrible at grammar so have someone else look at it.)

I like the description but feel the last piece is unnecessary and makes me think the writer is trying too hard to convince me of her hate

She's wearing his sneakers?

‘caress my lips’ - this would flow better if it were: to let my lips caress against …

Cabal tries to make sure that the critiques are, well, good. He highlights particularly well-done critiques and offers guides for Scribophile users who are new or nervous about critiquing.

“Getting it out in a social environment is good for writers, especially if they’ve been cooped up in their writer’s attic,” says Cabal. “If you write the great American novel and it’s taken you ten years, you are the only person who has been looking at it, so that skews the flaws that might be there.”

But on Scribophile, writers don’t just post drafts near completion: They post story ideas, pitches and early drafts. In forums, they discuss their ideas and the writing process. They aren’t waiting until the great American novel is fully formed before exposing it to community scrutiny.

Creativity researchers call the kind of feedback Scribophile members leave on each other’s work “evaluation” and the generation of ideas “ideation.” Psychologists and creativity researchers used to think the best way to foster creativity was to separate ideation and evaluation: First ideation, then evaluation. Brainstorm, then critique. Negative remarks from others during the ideation process could discourage new ideas. If writers get disapproving comments on early drafts, they could give up altogether.

But Raymond McCall, an associate professor of planning and design at the University of Colorado who studies software, observed software design and noticed that when ideation and evaluation are intertwined, creativity actually flourishes.

McCall – who published his work in the May 2010 issue of Human Technology – says that feedback, even if it was negative, stimulated the designers to come up with new ideas. This feedback came in the form of discussions (or arguments) about the ideas and in implementing the code itself. McCall writes that while sometimes software designers argued against disapproving feedback, “often the response was to accept the criticism and propose a new or modified position.” Disagreements and arguments, if presented in the right way, provide a creative opportunity. Even if the code is messy, or the client doesn’t like the approach, forcing ideas into the world for evaluation breeds more ideas.

To put this in the context of Scribophile, when writers post their work for critique, the feedback they receive doesn’t simply help them refine that piece. It could spur ideas that lead to new work.

Evaluation and feedback aren’t one-way streets: Commenting on the work of others helps Scribophile users become better writers. User P. Doughbury writes, “I enjoy doing critiques because thinking critically about the work of others helps me focus more critically on my own writing.”

When Cabal created Scribophile, he wanted to ensure the diversity of the community, so he didn’t require members to “audition” to join like some writers’ websites do. The same things that make close-knit writers’ groups beneficial – intimate knowledge of each member’s style, strengths and weaknesses – also limit them. “If you participate in the same group for a long time, it’s going to be people you know, people who are already familiar with your writing,” says Cabal. “That’s not always ideal for someone who is interested in getting a more varied perspective.”

But just because people have the tools to connect with diverse collaborators doesn’t mean they will. The Internet makes it easier for people who share views and interests to interact in self-imposed silos.

“People are attracted to others who are like them,” says Marc Smith, a sociologist who studies online communities. In any given population, you are “like” – in terms of gender, age, economic status, hobbies – two percent of other people. It’s a human trait, “to form these little insular in groups,” says Smith. “And the Internet makes that possible.”

This is the type of phenomena Smith’s company, the Connected Action Consultant Group, studies.

“In fact,” his excitement is evident even over the phone, “do you want to see a picture?”

Smith guides me to a screenshot posted on his Flickr page. On the left is a spreadsheet, on the right a graphic that looks like a spider web: an mass of black lines connect a rat’s nest of red, blue and green nodes. Although the overall picture is a mess, there are clear clusters of blue nodes at the bottom and red nodes at the top. They look like two spiders lurking on opposite sides of the web.
Smith tells me, “What you’re looking at is a picture of NodeXL,” the software his company designed to visualize and analyze social networks.

To make this image, Smith queried Twitter for all messages with the #GOP hashtag from September 5, 2010. On the image, about 1,000 tweets are displayed as nodes. Each line represents a connection such as a follow, re-tweet or reply.

Smith analyzed the content of each tweet to see if authors were supporting or criticizing the Republican Party. “The red dots are people that are GOP-a-OK people,” Smith says, and the blue dots, “are GOP-not-so-much people.”

The groups are as tightly knit as they are distinct. A few stray lines connect the red and blue blobs, but it’s nothing like the knot of lines engulfing each blob.

“The blue people don’t even listen to the red people. The red people do not follow, do not reply to, do not even mention the blue people,” Smith says. “Even though they are talking about the same thing.”

This pattern – called homophily, the birds of a feather effect – isn’t unique to politics. It happens with knitting, gluten-free cooking, amateur astronomy and writing.

“We are going to fall into these little puddles of association,” Smith says. But online, those puddles aren’t so small: Smith likes to say that if you are one in a million, then by sheer statistics there must be at least 1,600 of you on the Internet.

Although Scribophile’s users are geographically diverse and range from college students to retirees, they still exemplify homophily. Because they all consider themselves writers, they’re birds of a literary feather. For Cabal, this is an example of how homiphily can be more beneficial to the creative process when it occurs online rather than in person. Instead of slicing by geography, the group is sliced by interest, and that makes it a greater resource for the individual operating within a broad digital social network.

Scribophile helps writers break out of their solitary minds to improve their individual creative work. But something else is happening within the digital community: Users are coalescing into groups that produce collaborative projects. They are getting together to publish literary e-magazines such as Gnarrative and Ruthless Peoples Magazine. The line between individual creativity and collective creativity is easily crossed, and group creation naturally flows out of the Scribophile network.

Collaborative Futures: Finite group authorship

Scientists, journalists and others have been co-writing articles since long before the Internet. But FLOSS Manuals, the publishing company with a focus on open-source software guides that published Collaborative Futures, has leveraged digital tools to take group authorship further through book sprints.

FLOSS Manuals has been using book sprints since 2008 to collaboratively produce books on everything from the One Laptop Per Child Program to graphic design principles. Adam Hyde, founder of FLOSS Manuals, adapted the book sprint process from a similar one used by software programmers to rapidly write code.

In a tone that is at times self-referential and irreverent (an advisory at the beginning of the book warns: “Note: We offer no warranty if you follow this manual and something goes wrong. So be careful!”), Collective Futures addresses topics such as crowd-funding, anonymity, peer-to-peer sharing, social creativity, and the differences between design and creation.

But the texts created are only a small part of collaborative book sprints: The experience of writing and the communities they generate are just as valuable. Michael Mandiberg, a co-author of Collaborative Futures, calls book sprints “writing-based endurance performance art.” They are both productive and performative.

Collaborative book sprints have as much in common with the performative arts of jazz music and improv theater as they do with solitary writing. Keith Sawyer, a psychology professor at Washington University in St. Louis, began studying group creativity by observing – and participating in – jazz and improv. He notes that groups don’t necessarily have a vision of what the performance will be when they start. Instead, improv theater groups start with prompts from the audience, jazz groups start with a key and time signature, and the authors of Collaborative Futures started with two words. The art emerges without any single member of a group knowing what it is until it happens.

The fact that it’s emerging from multiple minds is what makes it interesting. “Creative ideas are always a combination of existing ideas,” says Sawyer. And with groups, you have a bigger pool of ideas to draw from and thus a better chance of producing a new, exciting juxtaposition of existing ideas. “You have those ideas coming in, building on each other, in a way that enhances or accelerates the creativity beyond what the individual members could do alone,” Sawyer says.

The co-authors of Collaborative Futures wanted to document their own emergent creativity. “We were aware that what we were doing was an event and a process, as much as it was a final book,” says Mandiberg, an artist and professor at the College of Staten Island/CUNY. “And even that final book was just a intermediate resting point in a longer process.”

The book has since been released in a second edition, written during a second sprint, and continues to be edited and updated remotely. Collaborative Futures invites its readers to share, re-mix and translate. It even has a section titled, “Write this Book.”

The book was written using an open-source software called Booki, which stores edit-histories and allows collaborators to write simultaneously. To emphasize the performative nature of the writing process, the authors of Collaborative Futures included excerpts from their electronic chats in the appendix. Some of their interactions are about the nuts and bolts of putting together the book. Others reflect the constructive argumentation McCall observes in software design. In some cases, disagreements among the authors lead to new ideas:

kanarinka: hey adam - we were just talking about the "This BOok is Useless" section
...
kanarinka: and i'm not REALLY asserting that the book is useless
kanarinka: more like a rhetorical device
adam hyde: yeah
adam hyde: but i think the point itself needs more weight
adam hyde: its a bit of an easy shot at the moment, needs more substance and nuance...at the moment, i see what it is trying to say but it doesnt capture my imagination like i think it should
kanarinka: ok - tell me more
adam hyde: the book doesnt assert that we never collaborate. and i think this is what this chapter is implying
adam hyde: [it] doesnt seem to grasp the fact that we are talking about particular kinds of collaboration
adam hyde: possibly limited to the free culture digital media zone
adam hyde: and that is a limit for sure
kanarinka: well the idea of making a book about collaboration imagines that there is something such as not collaborating
adam hyde: does it?

The chapter, “This Book Might Be Useless” incorporates elements of this exchange. It begins:

… this book might possibly be useless. Because collaboration is everywhere. To imagine that we could write a book about collaboration is to imagine there is such a thing as not collaborating. And to imagine a long history of not collaborating with each other. And the ability as “individuals” (Western liberal subjects) to operate separately from the “others” and the world, environment, context around them. And all of that is false! How can we even begin?

At times Collaborative Futures uses a first-person narrative, but usually it’s a collective voice. As Mandiberg describes, “One or two people wrote the major chunks of most chapters, and then those chunks were churned through, chopped up, and reformed by at least two more authors. I know that I touched almost all of the chapters, but that is partly because I took on somewhat of an editor role at the end, trying to achieve a more consistent voice.”

When the book was re-written during the second sprint, that collective voice emerged even more as an entity that’s distinct from the sum of the individual voices.

Researchers have seen this same process happen with problem solving in groups. The role of collaboration isn’t just some intangible quality: It can be quantified. When it comes to problem solving, the whole is not the same as the sum of its parts. Most times, it’s greater.

A team led by Anita Williams Woolley, assistant professor of organizational behavior and theory at Carnegie Mellon’s Tepper School of Business, devised a way to measure group intelligence based on the group’s ability to accomplish a range of cognitive tasks, a kind of group IQ test.

Woolley’s team put people into random groups of two to five and gave them challenges that tested different modes of thinking. They solved verbal and spatial puzzles. They made moral decisions about a hypothetical college basketball player who bribed a professor. They planned a shopping trip. They estimated the median age of the U.S. Population. They played checkers against a computer. They brainstormed the possible uses for a brick.

With individuals, a person who solves one mental problem well tends to also be good at solving other types of mental problems. Although measuring intelligence is murky, this is the largest single piece of supporting evidence that individual intelligence is a quantifiable trait. Woolley's team found that the same holds true for groups: The intelligence of each group is an independent, reproducible property.

What’s even more astonishing is that the relationship between a group’s intelligence and the intelligence of its individual members is weak. But other factors play a bigger part in predicting the group’s success.

The strongest predictor of group intelligence is social sensitivity – the ability to perceive the emotions of others. Woolley’s team measured social sensitivity by showing people photographs of eyes and having them interpret the implied emotions. Groups with higher average social sensitivity tend to have higher group intelligence. Groups with more women perform better, too, which the researchers think is because women generally score higher on social sensitivity. Another factor that contributes to group intelligence is how equally each member contributes. “Groups where a few people dominated the conversation were less collectively intelligent,” write the authors in the article, published in the October 29, 2010 issue of Science.

Woolley compares groups of humans to social groups of animals: flocks of birds, schools of fish, ant colonies. “They aren’t cognitively complex,” as individuals, Woolley says, “but they are hard-wired for this kind of collaboration.” Woolley says social sensitivity is the human version of that instinctual hard-wiring.

And in the same way that an ant colony can accomplish things that would be impossible for an individual ant, human groups typically (though not always) perform better than individuals on complex tasks.

The larger the group, Woolley says, the better they perform, to a point. Anyone who has ever tried organizing a crowd knows there are upper limits to this. But that doesn’t mean collaboration with large groups isn’t impossible. It just means that social factors become even more important. This also applies to groups connecting digitally.

“Social sensitivity ability absolutely translates to the internet,” says Woolley. “The ability is broader than the way that it is tested.” Meaning that the eye test only measures one aspect of social sensitivity. Other aspects include the ability to read body language and textual cues. And when the cues aren’t built in, like on the Internet, Woolley says, people adapt and create new ones: subtext in e-mail, emoticons or new verbal expressions.

As groups get larger, they need more structure to collaborate successfully. For the authors of Collaborative Futures, that structure came digitally. Mandiberg says that during the first book sprint, which included only a small group of authors, a lot of work and discussion was done face-to-face. But the second sprint, which had more collaborators, was done mostly online through Booki.

Large groups can work together collaboratively, group creativity expert Sawyer says, but “There’s a dialectic between structure and creativity.”

With structure, some of the improvisational elements of creativity can be lost.

Wikipedia: Crowd-authorship with a social grounding

The Wikipedia entry for “Collaboration” was created on August 24, 2003. Its most recent revision is from November 24, 2010. In the seven intervening years, 446 people have contributed. The most prolific contributor to the page has made 146 additions and edits. But the entry, despite its myriad authors and edits, appears seamless to the reader. The reason, according to creativity and human-computer interaction experts, are the social, ethical and stylistic codes that Wikipedia editors have created and agreed to.
“It starts out with less structure,” Sawyer says of open-source communities like Wikipedia, “and then it evolves.”

The type of authorship that appears on Wikipedia may seem to have little in common with the writers posting work on Scribophile and the FLOSS Manuals book sprints. On Wikipedia, the collaborators don’t necessarily know each other, and you can’t even list them all on a single page – or five pages. Yet Kittur, who studies digital collaboration on sites like Wikipedia at Carnegie Mellon, says there are more similarities than most people realize.

In person, collaborators get to know each other, learning their strengths and weaknesses and building trust about each other’s judgments. On Wikipedia, that same bond isn’t automatically there, Kittur says, and as a result he’s observed counter-productive behavior. “You can have people who are much more involved in the page feeling territorial about their work,” Kittur says, and they don’t always take kindly to newcomers. “People who are heavily invested in a page are more likely to revert others who kind of come and make small edits to their pages.” This back-and-forth can be seen on page histories and editor forums.

But social structures and digital leadership emerged within the Wikipedia community. “It really takes social intelligence so that you can have people work together without getting into flame wars,” Kittur says. “The people who are heavy contributors tend to know each other, work with each other, help each other out. Some of that social lubrication in the real world seems to also be working here.”

Kittur studied Wikipedia’s collaborative structures and found that increasing the number of contributors to a page tends to have a negative effect on the page’s quality. Unless, that is, social structures are in place. That structure doesn’t arise from the crowd as a whole: it comes from an individual acting as a leader within the crowd. If someone takes control and guides the efforts of others, adding contributors makes a page better. Social structures and strong leadership are most important when a page is brand new.
With the right social structures, Kittur’s research shows, even people who passionately disagree can collaborate. The Wikipedia pages on abortion and evolution, for example, are surprisingly civil. “Even though it’s such a controversial issue,” Kittur says of the entry for evolution, “the creation of the article wasn’t controversial.” The page links to an outline, for example, of how the creationism-evolution controversy is playing out in society.

Scribophile users echo this, saying that the reason the site works when other writing communities turn hostile is because of the underlying mutual respect inherent in the karma Cabal uses. I started a forum on Scribophile asking users what they like about the site. The response was overwhelming and nearly instantaneous.

User Trish compares Scribophile to a writing site she had to audition for, “You'd think this made it a place you'd really like to be a part of. You really might think that. Don’t. The level of snark was high, the quality of the [critiques] low, and the writing on there was just as good or bad as here.”

“The writers here seem to have more grace than other sites,” user P. Doughbury writes. “It is easier for me to critique someone’s work knowing that she already paid her dues in karma to get her work on the site.”

“[T]he whole Scribophile group becomes one,” writes user Frances Pawley. “We don't all see eye to eye, but there haven’t been any battles since I joined; unlike a previous writing site I was on. Where the norm was to chuck a few insults around daily.”
The social structures holding Scribophile together didn’t happen spontaneously. Cabal has been modifying them since the site launched, taking the lead within the community just like the main editors on successful Wikipedia pages and Mandiberg in Collaborative Futures.

Sawyer points out that digital tools – such the forums and wikis used by Scribophile, FLOSS Manuals and Wikipedia – don’t foster group creativity on their own. A wiki is, “a big whiteboard that happens to be on the computer,” Sawyer says. “It’s just putting people together and sort of hoping. But the environment isn’t designed in such a way that it helps people, or guides people in collaboration.”

Communities such as Scribophile, FLOSS Manuals and Wikipedia show, however, that if those creative structures aren’t built in, they can emerge from strong leadership.

While Sawyer doesn’t think true group creative flow can occur online, Anita Williams Woolley, who led the group intelligence study, has a different perspective. “Collective intelligence is observed in these groups that are not face-to-face, that are digitally connected,” she says. The key ingredient in group intelligence, social sensitivity, is a “basic ability [that] transcends face-to-face environments,” she says.

Collaboration extends into all aspects of life. “Almost nothing humans do, or can do, do they do alone,” says sociologist Marc Smith. “We are a collective animal.”

Nearly 80 percent of all adults and 93 percent of teenagers in the U.S. use the Internet, according to the Pew Internet and American Life Project. Which means more and more instinctual collaboration will be happening digitally, from mathematic proofs to poetry translation to map-making to journalism.

It’s nearly impossible to find an activity that isn’t collaborative. “You didn’t put yourself on this earth by yourself,” Smith says. “We are highly interdependent, and we don’t really notice it.”

The same is true online, with one key difference: We leave behind digital records of nearly everything we do. “We are leaving these machine readable traces, and they are telling a story,” Smith says. “They are telling the story of how we weave our societies together.”

###