Knowledge and anonymity

By David Weinberger

A debate is wanting to rage. It's about digital identities and the fate of anonymity on the Internet. The argument is boiling under the surface, but it needs to happen in the open, everywhere, for much of what we value about the Web is at stake.

I am highly biased (what a surprise!), so let me tell you flat out where I stand and then how I see this affecting KM.

We currently have a wide spectrum of ways in which we let others know on the Internet who we are, ranging from a username and password to simple verification checks ("Mother's maiden name?") to dongles on our PCs to biometric scans. Each of these ID procedures has been designed to meet a real and specific need: a password to log onto our corporate network, a biometric body scan to update national security records. And each works well enough for us to do what we need to do over the Net.

The impetus for instituting a standardized and universal digital ID scheme comes not from the bottom up and not from some need that's been irking us. No, it comes top down from those who have an interest in enforcing restrictions on what we can do digitally. The main impetus is coming from Hollywood, via Congress and Microsoft, in order to enable vendors to control how we use the digital content we've purchased from them. The degree of control they want far surpasses anything possible in the real world.

So, on the one hand we have the bottom-up acceptance of ID schemes that have emerged in response to real needs. On the other we have top-down schemes that would impose unreasonable restrictions on how we use digital content. Care to guess where I stand?How does this apply to KM? If the "strong ID" folks win, anonymity will be a special case on the Web, not the default. And we will become stupider because of it. Anonymity lets us be wrong without shame.

KM, as I think we all agree at this point, comes in two basic flavors. The first is noun-based. It wants to find and make accessible all the certified, authenticated, reliable knowledge in an organization. The second is verb-based. It wants to make an organization smarter by enabling the sorts of activities and processes by which knowledge nouns emerge.

Obviously, many KM systems do both. But identity and anonymity play different roles in them. In a noun-based system, many chains of authentication ultimately end with a known person: You can trust these data because they came out of Jim the Brainiac's research lab; you can trust that turning the widget to the left instead of the right will fix the spigot because Betty the Master Mechanic says so. That type of identity is strong enough that if the data has been cooked or the widget breaks when you turn it to the left, Real World Jim and Real World Betty will be looking for new jobs.

But anonymity is crucial in verb-based KM systems. Anonymity lets us get over the human foible of thinking less of people for being wrong. If we're identified in an online discussion as BabyGator but there's no link back to our real-world identity, we are free to suggest half-baked ideas and, more important, to ask dumb questions. And nothing makes an organization smarter than half-baked ideas and dumb questions.

Yes, of course, in an anonymous conversation, people are also free to be flaming jerks. But we have means for dealing with such people, starting with sharp replies, moving to silence, and ending with bozo filters. And, yes, there are times when a company may need to track an anonymous contributor back to a real-world person. But those occasions should be rare and in reaction to such outrageous behavior that it's necessary to risk damaging the conversational milieu that is a company's greatest asset.

Anonymity helps makes organizations smarter. I hope that in our zeal for Accountability and Responsibility, we don't lose the benefits of being nameless