The philosophy of Open Knowledge

4 · The digital commons4 min read

Free software, copyleft and the engineered commons

The digital chapter of this history begins with a printer.

The printer that would not share

In 1980, the artificial intelligence laboratory at MIT received a new laser printer from Xerox. It jammed, as printers do. Richard Stallman, a programmer there, went to do what programmers at the lab had always done: fix the driver so the machine would at least notify people of a jam. But this time there was no source code. The software was proprietary — a trade secret — and a colleague at another university who had the source had signed an agreement forbidding him to share it. Stallman later described that refusal as a betrayal of something he had taken for granted: that programmers, like scientists, help each other.

The frustration crystallized into a conviction: software, like mathematics, is knowledge. Knowledge you cannot inspect, learn from, fix or share makes you dependent on whoever controls it — and a society running on opaque software is a society that has outsourced part of its autonomy.

The four freedoms, and the trick that protects them

In 1983 Stallman announced the GNU Project — a complete operating system anyone could use, study, modify and share. In 1985 he founded the Free Software Foundation and articulated the four freedoms that define free software, numbered from zero, as programmers count:

  • Freedom 0 — run the program for any purpose. Yours to use; no one dictates what for.
  • Freedom 1 — study how it works and change it. This is the learner's freedom, and it requires source code: the human-readable text of the program.
  • Freedom 2 — redistribute copies. Helping your neighbor is not piracy.
  • Freedom 3 — distribute your modified versions, so the whole community benefits from improvements.

"Free as in freedom, not as in free beer": the point was never price but autonomy. And Stallman's masterstroke was legal, not technical. The GNU General Public License (1989) used copyright law against enclosure. Here is the mechanism, because it deserves to be understood: copyright gives an author near-total control over copying. The GPL says: you may copy, modify and distribute this work freely — on the condition that whatever you build from it carries these same freedoms. Break the condition and you lose the license, leaving you with ordinary copyright infringement. Enclosure of the commons thus becomes legally impossible: the gift protects itself. This mechanism is called copyleft, and it is one of the great legal inventions of the twentieth century — a hack, in the best sense, on the very law that made knowledge ownable.

The bazaar wins

In 1991 a Finnish student, Linus Torvalds, released a kernel called Linux under the GPL. Combined with the GNU tools, it became a complete free operating system — built not by a company but by thousands of volunteers coordinating over the early Internet. It should not have worked. It runs, today, most of the world's servers, all of its Android phones, and essentially every supercomputer on Earth.

Eric Raymond's essay The Cathedral and the Bazaar (1997) tried to explain why the chaotic, open "bazaar" style out-engineered the closed "cathedral": openness turns users into contributors; transparency means problems are found early ("given enough eyeballs, all bugs are shallow"); and motivation that comes from reputation and craft outlasts motivation that comes from assignment. In 1998 the term open source was coined — proposed by Christine Peterson — to present the model to business without the ambiguities of the word "free". Business listened; today the software commons underlies nearly every product you touch.

Ostrom: the commons was never a tragedy

One more thread matters here. Economists had long taught the "tragedy of the commons": shared resources get overexploited, so everything must be privatized or state-run. In 1990 the political scientist Elinor Ostrom published Governing the Commons, decades of field evidence — Spanish irrigation communities, Swiss alpine pastures, Japanese village forests — showing that communities routinely govern shared resources sustainably for centuries. What they need is not an owner but rules: clear boundaries, rules fitted to local conditions, participation of users in making them, monitoring, graduated sanctions, cheap conflict resolution. She was awarded the Nobel Memorial Prize in Economics in 2009 — just as Linux and Wikipedia were proving her point at planetary scale, with version control instead of village assemblies.

Free software supplied the missing piece of the open-knowledge tradition: a working legal and social technology for building commons on purpose. Everything in the next lesson stands on it.