May 3, 2006 · Tim Wu
Back on March 30 I presented Who Controls the Internet at Ed Felten‘s Infotech lecture series at Princeton. The crowd was extremely sharp; the discussion was great, and I had the chance to meet Brian Kernighan, from whose book I learned C programming. I must say there is something uncanny about the enthusiasm for political theory and policy found in computer science departments today. Seems like everyone is a policy-geek — what ever happened to just being a geek? Maybe that’s what engineering department are for.
Anyhow, during the talk, someone asked an interesting question — what’s the difference between our book and Larry’s Code and Other Laws of Cyberspace?
It’s a perceptive question. From the outside the books look very different — ours is about international relations and nations, Larry’s is about code as a kind of law. But on the inside beat similar hearts, reflecting how much Jack and I learned from Larry, his scholarship and his book. How could it be otherwise?
Larry’s enduring contribution in Code is the concept of dueling regulatory modalities. This is the idea that a society at any time is being governed by multiple, and different types of rules from different sources. Larry didn’t just mean federal and state law. He meant something deeper: entirely different types of regulation, or as he puts it, that we are regulated by law, social norms, market forces, and what he called architecture or code.
Perhaps this idea was somewhere latent in the sociology and economic literature. But once presented by Larry it became an idea that once you get you cannot forget. (Similar to Larry’s translation theories of constitutional interpretation, which have a way of sticking in your mind and refusing to leave). So when you’re driving — what regulates you? The speeding laws, sure. But also norms that say, for example, no skipping the line to get on the highway. And while it sounds odd, tolls on the road regulate your behavior, making one road more expensive and another cheaper. But most profound of all was the idea that the architecture of the world regulates too, just like law. A speed bump is an alternative to a speeding law, and maybe more effective too. Conversely, you don’t need a speeding law for bicycles because by their nature they only go so fast.
Our book takes on a slightly different problem of competing sources of order. The question was, how would the world come to regulate the borderless or extra-territorial conduct found on the internet? In the 1990s there were three competing answers: (1) self-regulation or “private ordering,” (2) use of international law, or (3) national laws. While (3) was certainly not a fashionable answer at the time, our book is the story of how the use of (3) outgrew the rest, at least from the 1990s through now, and what might be said about that. (And maybe the most surprising thing has been the lack of use of international law tools to handle the extraterritorial problems created by the internet, with the exception of the Cybercrime treaty).
But at the heart of our theory you can find our debt to Larry. For in our studies of the techniques used by nation-states, the same theme emerges — the use of intermediary, or what legal scholars sometimes call “gatekeeper” controls. But if you really think about it, when acting through intermediaries the government is using a technique Larry described in his book — using law to shape the code. If the law were to tell carmakers that no car could go over 65 mph, that’s a form of intermediary control, but also what Larry was getting at.
Prompted by Code, we also talk about an almost too-theoretical question — you might call it the question of “what comes first.”
In short, is law, and its basic provisioning of public goods (like physical security) necessary to successful systems of norms, a free market and so on? This position, which is more or less Hobbesian, is one we approach in the book. I don’t think the argument is bulletproof — good cases can be made, perhaps, that norms are where societies start, leading to laws, leading to functioning markets — and so on. But having Larry’s ideas helped us think this through.
I don’t want to go on too much in this vein, and there isn’t enough space to detail all the ways our book has learned from Code. But suffice to say it and much of the other writing in this area all begin with that volume.