Fork the Academy

I recently wrapped up a series on GitHub. Throughout the series I highlighted what I thought were some of the most powerful innovations that software developers and writers can take advantage of in GitHub. In particular I looked at two of its collaborative features, the ability to "fork" repositories of text that retain a connection to the original and the issuing of "pull requests" as a way to enable outside contributions in an decentralized environment which leaves everyone with full control over the texts they work on.

The social and collaborative potentials that GitHub provides makes it easier than ever for anyone to contribute to an open source project or adopt and adapt a repository for their own needs and pursue their own directions. If something like this caught on in the academic world, if we could fork the academy, we might move beyond merely referring to the work of others through citations methods and directly improve, adapt, and re-use scholarship in a way that offers full attribution to the original author.

My last posting on the limitations of GitHub for writers focused on some of the technical reasons why GitHub itself is not an ideal platform for writers. Mark Sample has also voiced concerns about GitHub, specifically when it comes to sharing course syllabi. A number of these issues can be addressed as alternatives to GitHub emerge that are more suited for writers. I’ll be looking at some of them in the weeks ahead.

Perhaps more important than the technical challenges posed by a writing platform or the formats it uses, however, are the very serious disciplinary obstacles that stand in the way of any attempt to "fork the academy" and enable the radically collaborative authorship possibilities that something resembling GitHub might create.

Imagine, for a minute, what academic scholarship might look like as more than just open access, but fully open source, creating and developing their work with some of the practices and a GitHub-like platform that are in widespread use. A fairly traditional academic project within history might look something like this:

A historian, Leo, commits some notes on a collection of recently read and analyzed sources to a public distributed writing repository. Whatever sources are not protected by copyright he also commits to the repository. Leo starts the text of an article and separately a file with a shorter presentation version and a third version summarizing his findings for a blog entry. He makes "commit" updates to the article as he goes along, making changes and deletions, exposing his developing ideas to anyone who views the repository history. A scholar Leo has never met but who has been following his work, Amanda, issues a pull request with a few extra sources for him to consider, along with her notes on them. She also issues pull requests with some small edits to his early blog posting draft, and posted a few comments on his developing outline for the article.

Leo publishes his article (or rather, a snapshot of his article) in a peer reviewed open access journal that links back to the repository (or shows a special formatted version of its contents). The journal lists Leo as the primary author, but notes that it includes contributions (the ones Leo accepted) by Amanda. After publication, Leo starts getting much more feedback on the repository. Leo accepts some of the direct "pull request" proposed changes, and makes a few other changes in response to comments, attributing them as he goes along. Leo moves onto another topic and the project goes quiet for a while. Then Leo’s old friend Yuki tells him he has looked at the same sources and was thinking of going in a different direction. Yuki forks Leo’s article, adds more of his own material, uses some of Leo’s text directly and publishes a snapshot of the new version, with a new argument and direction. A history of the repository shows where Leo’s and Amanda’s original wording is, and perhaps some other attribution mechanism such as footnotes or other metadata also allows Yuki to indicate where his work was inspired or borrowed ideas. If the platform supported more complex inheritance, Yuki’s article might be a fork of several works with this clearly indicated in the work even as the new work is a product of remixing, editing, and adding as needed.

Some of you reading this may find this scenario horrifying. Others will simply judge it a fantasy that can never come to be. Together, the reasons behind these two types of response constitute the disciplinary, and the more personal obstacles to forking the academy. I have been talking informally with friends and colleagues about this for a few months now and I have seen a fair degree of enthusiasm, but also a recurring series of objections, which I will summarize below:

• The Economy of Citations: As scholars today, we operate within, and if anything are increasingly beholden to, an economy of citations. We are judged by the number of our publications, and each of those publications weighed by their publisher. Just as importantly, however, we are communicating across time and within our community of scholars through the act of citation, which serves both the cited, in terms of prestige, and the citer, in terms of demonstrating a familiarity with previous work. Ideas travel through citations, attribution is given through citations, and our publication formats are constrained by the need to maintain a system of citations. Forking the academy is a fundamental challenge to this as the dominant means of using and attributing ideas. To become possible, not only must the economy at the very least accept another form of currency that recognizes the varying possibilities of levels of collaboration, but also recognize and give credit for a wider variety of kinds of contributions to things like academic publications.
• The Taboo of Plagiarism: In forking the academy attribution is not eliminated, but brought to a whole new level as I hope is clear in the series on GitHub over the past few weeks and some of the alternatives we will look at going forward. However, this doesn’t change the fact that "forking" someone’s repository of writing may "feel" like plagiarism to some. When we hear debates of "libre" as well as "gratis" open access, with scholarship released under, for example, a Creative Commons Attribution license, there is a recurring theme that "someone will steal my publication" and publish it under their name. This often misunderstands the way that attribution licenses function and, in a GitHub-like writing platform, the way in which every word and line is (when not manipulated) tracked to its origin.
• The Cult of Originality: If we fork the academy, some might cry, the author will disappear as some kind of beastly MOOPs (massively open online publications) take over. Others have expressed concern that, beyond the issue of credit for an idea, there is the question of how a more singular voice is preserved. At one level this will be another arena in which our cult of originality will be debated and the value of creativity defended. The answer should not be a coercive environment in which all publications and writing must be thrust into a world of public repositories. Forking the academy cannot go very far without greater scholarly consensus involving real variety of choices and a diversity of approaches. On another level this concern is often more of a technical issue: GitHub and wikis do a poor job of giving you easy at-a-glance idea of who the "main" author of a text or its main inspiration. There is no reason, however, why we cannot think of new and creative ways to address this problem in a world of forked and collaboratively authored texts.
• The Brick in the Wall: The kind of collaborative writing I described in my example and which is common in the open source world may open the risk of a kind of "brick in a wall" approach to scholarship – a positivist approach which sees us all contributing to an ever greater, more complete body of knowledge about X. If it becomes more mainstream to reuse, adapt, and expand on existing work, will the relative value of starting fresh, beginning anew, and using completely new language become less? Will the illusion that we are all building a single tower together, rather than a multitude of towers, become more tempting? The experience of the open source world, again, suggests not, but some may be concerned that it will be relatively easier to build upon someone else’s work directly than to risk a completely new approach and thus forking the academy could, ironically, be seen as stifling innovation.
• The Fear of Transparency: When you compose your work in a GitHub-like versioning system, the development of your ideas, your misteps, your errors, your complete reversals are all exposed to the world. The hours you work, the speed of your progress, all of this is generally information available to anyone who looks at your repository. GitHub has the chilling "punchcard" tool that gives you an immediate overview of the "working hours" of an author. This kind of transparency is horrifying to some I have spoken to and while its severity depends a lot on how one designs or uses a GitHub-like platform, will be one part of the debate on its adoption.
• The Incompleteness Aversion: As PhD students, then as scholars, we like to drag our work out, fiddle with it, improve it, and sit on our work. Then in a moment of triumph, or more often, a whimper as we subject ourselves to a merciless deadline, we declare our work complete. In a forked academy, the "release early, release often" approach is far more natural. You compose your work in public, you share it in some public forum when you have reached some "milestone" in your progress, and you may create "branches" appropriate for different avenues. You may develop a text further and update it, producing new editions or "milestones" based on feedback received. The question, "when is your article done?" becomes a question local to a particular context, rather than referring to a project that has a fixed date of completion. This alternate approach has its own strengths but, as any advisor would tell a graduate student, also brings with it huge challenges.
• The Stolen Idea: We like to pretend that we are an altruistic bunch, sharing our knowledge with each other and working together to improve humanity’s understanding of whatever obscure corner of our field we work on. And yet, compared to a growing number of commercial computer developers who use and contribute to open source even as they work in a field where millions of dollars change hands and companies rise and fall every day, we are sometimes far more protective of our ideas and our writing. Many of us hide that wonderful document we brought back from the archives, or that excel spreadsheet with the meticulously compiled statistics, or that ironic twist on a standard historical narrative that we want to develop. We keep it close "until we are ready" or "until the article is done" or "until the book is out." Very often, we do it not because we don’t want to share it with the whole world, but because we recognize how vulnerable we are in an economy of citations, in a job market where that ironic twist may get us hired or get us closer to tenure. What if Yuki or Amanda had forked Leo’s project just as he was getting under way, and before he had fleshed out even his most important new ideas? What if they had quickly wrapped up an article or a blog post and Leo thus only gets credit for a few skeleton outline bullet points or his archival notes? Most of us have a few anecdotes about a professor stealing a grad student’s idea, or a scholar deploying the core ideas of another scholar’s conference presentation and becoming known for the main idea. If we are to fork the academy, the culture of secrecy must give way to one of sharing but this must be accompanied with incentive schemes that reward collaborative production rather than the heroic and solitary discovery. There must be better and more diverse ways for us to be evaluated and appreciated as scholars too that recognize the labor and importance of many tasks beyond the "final writeup" (this is already a topic of much discussion within the digital humanities community). Finally, if we are to release early and share boldly, then for several reasons we must get in the habit of declaring our aims, goals, and the scope of our project both to make it easier for people to find and help you in your work, but also make it more difficult for someone else to ignore you.
• The Fear of Misuse: Several of those I spoke to about forking the academy were deeply concerned about the loss of control when you allow anyone to fork and adapt your work. Whereas citing a work for nefarious purposes is not unheard of, even more distasteful would be the prospect of your scholarship being taken up by someone and directly used to the end of arguments you find offensive. If your name is there deep in the history of a forked repository, an author might feel like they are more directly contributing to the misuse of their work. This issue is closely tied to contending normative ideas about liberty. It has already been an important issue within the open source community itself. This comes out very clearly in the wonderful work by Gabriella Coleman, Coding Freedom as she reviews the debates over the terms of open source licensing within the Debian programming community. If we are to fork the academy, we must also have an open and inclusive conversation about concerns like these.

These are the main issues in the disciplinary and personal realm that I have found so far that may stand in the way of any attempt to fork the academy on a large scale, opening up the kind of collaborative social coding that has taken the open source world by storm over just the past few years. Many of these objections are on the level of basic principles, while others are practical problems deeply entrenched in the structure of our academic communities.

Do any of these objections resonate strongly with you? Are there other kinds of objections or challenges that I have overlooked? Which do you see as the most serious or difficult to overcome?

Image Creative Commons licensed by Konrad M. Lawson.