Wrestling with Jefferson’s Moose

Some months back, I had the pleasure of hearing an old friend give a talk at Harvard Law School. I met David Post a dozen years back, when his research at Temple Law School brough him to Tripod, the internet startup where I was working, to talk with us about his theories of internet governance. It was fascinating to hear him talk about his new book and find myself transported back to the kitchen of our old office space (the top floor of a plastics factory in Williamstown, MA), talking with David about the idea that online communities should govern themselves.

Post’s book is titled “In Search of Jefferson’s Moose: Notes on the State of Cyberspace“. The search for the moose doesn’t take long. The first half of the book, which is a celebration of different facets of Jefferson’s genuis, introduces the moose in the prologue. The massive animal, skinned, prepared and shipped by a political ally in the Colonies to Jefferson in France, graced the entrance hall of the Hotel de Langeac in Paris, where Jefferson was serving as ambassador. The search for the moose isn’t a search for the massive animal, but for the reason that Jefferson, extremely busy with negotiations critical to survival of the new nation, should have time, interest and reason to coordinate the killing, preservation, shipping and reassembly of a moose.

The reason for the moose, we learn, is to settle a scientific dispute with French naturalist George-Louis Leclerc de Buffon. Leclerc argued that the animals of the New World were bound to be smaller than those of the Old World, and that animals from Europe brought to the new continent were bound to suffer and degrade, because America was cooler and wetter than Europe, which would be detrimental to quadruped development. Turns out that nearly every subclause of that explanation is untrue, and Jefferson may have seen the scientific dispute as having certain political overtones. The presence of a massive moose – larger than any European quadruped – in his entrance hallway must have been an elegant rebuke to Leclerc’s argument.

Of course, this is a book about the Internet, not about the natural sciences. Well, not so fast, Post warns. He draws an elegant analogy between the Internet and the rivers of North America. Before the advent of railroad travel, travel over land was significantly slower and more fraught than river travel. Communities connected by rivers were knit together into an interconnected economy and society to a degree that frontier communities were not. Post frames the Louisiana Purchase in terms of connectivity – by controlling the Mississippi, Pittsburgh (and dozens of other frontier cities) was suddenly connected to Boston via a combination of river and ocean travel, and able to experience the “network effects” of that connection. (If this doesn’t make sense, glance at this ugly but helpful GIF, which shows just how tiny the original colonies are to the vast, interconnected watershed that drains to the Mississippi.)

So what of the moose? He’s a metaphor for the challenges of scale. Post postulates that Jefferson’s unique genius was in seeing the possibility of American representative democracy scaling to a vastly larger nation – in geographic and population terms – than the original colonies. The purchase of vast wilderness looked not just like an expensive proposition to many observers (and was challenged by Jefferson’s lifelong rival, Alexander Hamilton, amongst others) but an invitation to make a massive new nation entirely ungovernable. Post celebrates a decentralized, “loosely joined” model of governance that Jefferson anticipated growth, new state formation and the massive expansion of the nation under representative government.

In this, he sees a parallel to the internet’s massive growth, and the scaling of one of the net’s key architectures, the domain name system. He explains how the internet’s domain name system delegates levels of control to different authorities, allowing Harvard University’s sysadmins control over all domain.harvard.edu names, but giving control of resolving harvard.edu to a more central domain name registrar. It’s an elegant explanation of the current DNS system, and proceeds to a useful discussion of wider internet governance issues, including the move to IPv6. But the DNS explanation includes a seed of what I think is one of two central problems of the Post’s argument: an overly romantic view of the contemporary internet, and a failure to acknowledge corporate points of control.

Ultimately, “Jefferson’s Moose” is an argument that the internet might be able to be governed not by nations, but by the online communities themselves. Post outlines an argument between internet exceptionalists and unexceptionalists. Unexceptionalists believe that law on the internet works more or less like it does off the internet – if it’s illegal to sell Nazi artifacts in France, France can ban Yahoo from selling those artifacts there. (That’s not a theoretical – LICRA vs. Yahoo! is probably the most discussed case regarding Internet jurisdiction issues, at least around a place like Berkman.)

Post believes the unexceptionalist position is untenable, pointing out that companies take actions that are illegal in some jurisdictions and then simply avoid having assets in those countries that can be seized. He suggests that the absurdity of that situation suggests need for a new type of law that recognizes the internet as an exceptional space, governed by a different jurisdiction of law. Post doesn’t offer much about what this governance might look like, suggesting that it needs to develop in a bottom-up fashion, not a top-down one. He points to the emergence of banking systems in Second Life and postulates that it’s better to attempt to govern such systems under the laws of those communities, rather than in the mishmash of jurisdictions that might otherwise apply.

And here’s where we loop back to the conversation Post and I had a dozen years ago. Observing that Tripod had roughly 15 million members, making it – at that point – the eight largest website in the world in terms of traffic, Post wondered whether it was time for our users to have a “constitutional moment” and determine how their spaces were to be governed. As only a snotty 24-year old can respond, I said something to the effect of, “We just had a constitutional moment, Professor Post. I sat in my office, wrote a new terms of service and posted it. If our users don’t want to be governed by it, they can go somewhere else.”

Obnoxious as that response was, I don’t think I was wrong. Our users weren’t citizens, they were customers. We could – and did, better than many of our competitiors – listen to our users, act on their feedback and try to make our service more useful to them. But the moment we turned governance of the community over to the users, we’d be running the risk of being sued by the people who actually did have rights – the shareholders who’d capitalized our company. In a conflict between users demanding ad-free webhosting and shareholders demanding we make an effort to turn a profit, the shareholders win, otherwise we get sued in decidedly enforceable jurisdictions, like the State of Delaware, where we were incorporated.

This isn’t to say that there aren’t large web communities capable of self-governance. Wikimedia comes pretty close, and it’s a huge and important site. You can even argue that it engages in the most controversial decision a community needs to face – taxation, which it handles through regular fund drives. But it’s worth noting that it’s a nonprofit foundation, not a company trying to please venture capital backers or public markets. (And it’s worth noting that when faced with real-world legal action, say a libel suit, Wikipedia’s deliberative decisionmaking can turn into rapid unitary action.)

Yes, communities like Facebook are experimenting with ways to make users feel more like citizens. My colleague Jonathan Zittrain looks at a recent effort on Facebook to allow users to vote between terms of service: “It calls to mind the age-old trick of asking the children whether they’d like to wear their red or green pajamas to bed – with no choice about when bedtime actually is. Facebook still holds the quill and frames the choice.” Facebook set up a vote so that it would only be binding if 30% of the community – an enormous number of people – participated. The Register notes that Facebook changed the turnout figure upwards, possibly worried that an earlier threshhold would be met. They needn’t have worried – less than 1% of members participated, possibly because they thought the choices were a sham, possibly because they didn’t care, possibly because they didn’t know about the vote.

I don’t think the response is to declare Facebook an unrepresentative, facist state. They’re a company that realizes they depend on user goodwill for their continued survival, and so they’re experimenting with ways to enable user input, while retaining the control they need to attempt to run a profitable business… and stay within the letter of the laws of the jurisdictions in which they operate. While I think experiments will take place pushing the limits of what’s possible in terms of internet self-governance, I don’t expect to see them take place within large companies.

This matters because most people participate on the internet on spaces controlled by large corporations. In the 80s and into the early 90s, the internet looked more like the frontier Post celebrates. Many people ran their own mailservers and DNS servers, hosted their own websites. Your ISP might well have been a local or regional company where you knew the sysadmin, or a university where you could walk across campus and work through problems with the geek with the reins. That doesn’t happen these days, at least for customers using consumer broadband connections. And most people don’t run their own webservers – they use Blogger.com or YouTube or Facebook, and they’re subject to their TOSs, which are very hard to dispute or change. Those companies – the ISPs and the hosting companies – have enormous amounts of power, and they’re unlikely to institute consumer-driven democracy any time soon… and if they did, their shareholders would likely rebel.

And Post knows this. There’s an excellent section of the book that talks about the change in how connections are formed on the internet. In years past, we saw machines building a modest number of connections with other machines. Now we see a distribution where most machines are connected to one other machine (the upstream router of their ISP) and a few machines are connected to millions (large webservers that serve communities like Facebook or Twitter.) To me, this distribution suggests that any analysis of power and governance has to focus on the corporations first. I don’t know if Post doesn’t follow this line of logic because it’s not one Jefferson followed, or because it’s inconsistent with his hopes for internet governance, but it strikes me as a critical weakness in the arguments.

All that said, there’s much that’s valuable and insightful about this book and about the connections between Jeffersonian thought and the contemporary internet. And, as is my response to all cyberutopians, prophecy isn’t prediction – it’s a challenge to effect change in the world. The changes Post would like to see happen would be dramatic and fascinating, and likely quite positive, even if they look somewhat impossible to me.