Whatever court sits in judgement over the realm of knowledge, it is increasingly clear that the rules cyberspace follows continue to be of its own. A key property of the knowledge economy is its lack of formal systems for anything: its rules, it follows, are likely to be informal. So also is the process of defining them, and this would probably not be democratic.

Democracy is egalitarian all right, but surprisingly incompatible with the knowledge economy in several other aspects. First, it implies a political structure that is hierarchical, where a few people are needed to govern the rest. Knowledge, however, tends to flatten things out, and a political system for the new economy would be more level in structure. Second, and more importantly, democracy is a political system that is painfully formalised. Like any formal system, this has its advantages - deviations from the correct path are clearly defined and easily monitored, for one. But the disadvantage of formal systems is their inflexibility. If, by definition, they have to follow a fixed form, then they can't possibly be very amenable to change.

Moreover, democracy's formalities make it impractical in a knowledge economy, especially as it can be very inefficient. The alternative, of efficiently authoritarian (though perhaps tyrannical) control, is ruled out, because cyberspace is far too dispersed to let it happen. If governance is not forced upon people by a dictator or a majority then it must be something informal and ambiguous - rule by consensus.

One thing to note about the organisation of cyberspace is that it is formed by thousands of communities. Small communities, or groups of people, often make decisions based on what is felt, by common consent, to be the common good. Decisions are not imposed, but consented to, and dissenters are left to themselves.

In a flat organisation, groups come together to form larger ones only when they absolutely need to cooperate in joint action. It is in their interests to form a consensus - as if they were just another small community - in order to ensure one another's cooperation. This makes it possible to choose with some objectivity, to make decisions that are technically or otherwise superior, and not necessarily to the advantage of any one group, majority or not.

The Internet Engineering Task Force (IETF) has always acted in this way to set the most basic rules of the Internet, without which cyberspace would fall apart - the network's communications protocol. Decisions have never been imposed, and indeed they can't be. The authority of the IETF - which is open to everyone - rests in the simple fact that those who disagree cannot cooperate. The community exists precisely because people do want cooperation, and the IETF is precisely for those who want to cooperate. Dissenters can ignore the consensus - and, in turn, be ignored.

The informal, loosely structured rule by consensus of the knowledge economy has parallels in more mundane matters - such as the evolution of language. With rare exceptions (the Academie Francaise) languages grow outside any formal processes. Words and their usage form locally, to spread by general agreement as to their meaning, without which communication would be impossible. Dissenters can hold out with archaisms, but will either be isolated or have to adapt. And formalised imposition doesn't overcome consensus here either - the French language-controllers are still stuck with "le week-end".