I have been thinking about knowledge and collaboration for a long time, for it greatly affects my own life as an scholar and researcher. The open source movement didn’t invent collaboration; academics were collaborating freely — both as in beer and as in freedom — before software engineers. After all, professional engineers work on products that are bought and sold, while academics (in principle, if not in practice) share their wisdom in return for society’s generosity in funding their exploration.

In practice, software engineers collaborate a lot more and a lot more freely than academics do. Wherever you look, the situation is better in industry with all the cut-throat competition than in academia, with its public charter. Some of it is because academia is actually a lot more cut-throat than industry — there are fewer jobs and there’s less money. Further, unlike an industry professional who can sell expensive widgets for a living, an academic only has their data and their content to flog to the world. The sociology and the economics of academia is well understood now and I will remain silent on this issue from now on; you can always read the Chronicle of Higher Education to see the daily lamentation.

Let me talk about a structural issue instead. You might have heard of the famous slogan: “the medium is the message.” In other words, the means through which you communicate influences the content of your communication. TV news is not the same as newspaper news. For the same reason, academic collaboration isn’t the same as software collaboration. Software collaboration — mostly done via version control systems — is real time, ongoing and continuous. The time cycle is in the order of hours, if not minutes. The technologies that support collaboration are more or less instantaneous: you run a git push origin master and your collaborator has your contribution in front of them.

Academic writing is a lot slower. Its collaboration technology is built around citations, responses and feedback that have a time cycle of months or more. Worked well in the seventeenth century; now, not so well. It’s true that you can write your scientific paper on a Google doc and see your collaborators’ response in real time. That’s missing the point — collaborating on an office document has none of the language and ritual of paper writing. Every element of a scientific article, from the abstract to the introduction, the citations, the data, the discussion, the conclusion and the references, is designed (unconsciously, as a result of a slow evolution over centuries) to address a single problem: how can I communicate my work to a community that lives far away from me and doesn’t have access to my mind or my lab? It’s that mental organization that has enabled a scholarly edifice, built on top of each other’s work. Unfortunately, that design has a half-life of months.

We now expect instant feedback from our communication systems — wherever you look from phone and Skype to SMS, Whatsapp and Facebook messenger, people are used to ongoing, real-time conversation across the world. When I first came to the US in the early nineties, I was still writing letters by hand to my friends and family. Most of them didn’t have a phone or a computer. I would write a letter, post it and then wait for a month or so before I received a reply. In a couple of years, we had all switched to email. It’s true that the handwritten letter had an emotional impact that an email can never have, but for most purposes we don’t need that handwritten note. Certainly not in an academic setting. Scholarly collaboration needs to reflect this new cognitive landscape. A revolution in knowledge needs a revolution in communication.