Chris Kelty: The Cultural Significance of Free Software

Anthropologist Christopher Kelty is visiting Harvard from Rice, discussing his book “Two Bits: The Cultural Significance of Free Software” at the Berkman Center today. The book is an ethnography and “analytic history” of free softare, focusing on the cultural importance of the free software movement. He’s interested, in particular, in the ways that free software has “modulated” and informed other sectors, like citizen journalism.

Framing the talk, he asks, “Why do geeks look alike?” We do, as it turns out – he offers two photos, of geek legends Ken Thompson and Dennis Ritchie, who could have used each other’s driver’s licenses. He offers this as a special case of the idea that geeks seem to be able to find each other across national, language and cultural barriers. And he observes that a fascination with free software seems to link together geeks of all cultures.

His ethnographic study of geeks begins in Boston with MedCommons, a company that focuses on medical information using open source tools and models, and moved on to a hacker group in Berlin, a hyper-networked tech innovator in Bangalore and finally the Connexions open education project at Rice. This final study is especially interesting, as it demonstrates how the logic of open software “modulates” and influences another field – educational materials.

Kelty offers five key components of the free software world:
– Fomenting a movement (evangelizing for free models)
– Sharing source code
– Defining an open infrastructure
– Writing copyleft licenses
– Co-ordinating collaboration

Discussing these components, Kelty explores some critical moments in the history of free software. He suggests that the fact that Unix was neither an academic or commercial project means that it was very difficult to decide how the code should be owned. Once the Lions’ commentary on the UNIX 6th edition – basically a literary criticism of source code of the operating system – began circulating in the computer science community, it was essentially impossible for AT&T to restrict the circulation of the code. Sharing that code, and the resistance to restriction helped create the culture of the code. Unfortunately, it also made it possible for Unix to fracture and create so many incompatibilities that, in turn, it set the stage for Microsoft to introduce Windows NT, pledging to end the confusion of the Unix market.

The book looks at the standards battle between the OSI Reference model and TCP/IP. While OSI managed to get acceptance as an international standard (which is a very, very hard thing to accomplish), TCP/IP outcompeted it in the marketplace, largely because it was central to Unix, freely available and, eventually, installed with every modern machine, whether the user knew it or not. In the end, this created an open infrastructure that allowed a wide range of new applications to take place.

A discussion of copyleft licenses focuses on the early history of the GPL, which Kelty identifies as an Ur-license that nearly every other free software license is descended from. He outlines the battle Richard Stallman fought over the future of Emacs, which included the strange case of Stallman being accused of violating the copyright on something he invented.

For examples of coordinated collaboration, Kelty points to the debate between organizational models for Apache and Linux. Linux has, historically, functioned as a benevolent dictatorship with Linus Torvalds coordinating contributions to the kernel. Apache has worked more as a meritocracy, with an oligarchy admitting people to a core team on the basis of contributions. He points out that these communities are anything but self-organizing – if anything, they’re organized around code management systems which create a technical structure for contributions and releases of a free software project.

The same basic principles that have given such vitality to the open source movement are being adopted and experimented with in other fields. Kelty points out that there’s now an attempt to build open source biological systems, through projects like BiOS and the Registry of Standard Biological Parts. The source code in biology are the publications and ideas of the field. The open infrastructure is the lab protocols that allow experiments to be replicated. Coordinating collaboration happens through tools like the Public Library of Science. Kelty observes that free collaboration in these systems is more common outside of academe than within, which seems counterintuitive, until one considers the incredible importance of taking credit within the academic world as a way to promote your career.

One of the more interesting ideas I gleaned from Kelty’s talk was the phrase “recursive publics“. The concept is a response to Habermasian visions of the public sphere – a space that’s voluntary, rational, independent of power, and forces accountability. (It won’t surprise you that many academics argue that these spaces no longer exist, and may never have existed.) Technology is important in the creation of these spaces, at least in the modern age, and Kelty suggests that part of what’s interesting about free software is the way that it’s sought to make free each aspect of its existence. Once you free the code, it raises the importance of freeing the language… then freeing the operating system and the protocol designs. Follow this logic to its end and you understand why free software folks are advocating for net neutrality or trying to build open, alternative networks via mesh. The goal, ultimately, is to create a public space that enables the existence of free software… built via the principles and code of open software.