I was listening to NPR yesterday, and heard a "Sound Clip" - a short segment of audio provided by a listener that "collects the sounds that fascinate our listeners." For some reason, that started me thinking about preservation.

I think I had always thought about the preservation of digital content items like Sounds Clips as a part of the process of preserving our culture - a snapshot, if you will. But the time that I have spent thinking about interactive media troubles that conception. Increasingly, we will live in an interconnected skein of rich multimedia, interactive services, and fully participative interfaces. People now, on the net, leap from semantically sensitive text to connected video to connected data display or analysis to writing, responding via Twitter, blogs, and community spaces like Facebook and MySpace. This has to mean something for digital preservation.

As a teenager said in the New York Magazine article, "Save everything ,"

If I don’t delete it, I’m still gonna be there. My generation is going to have all this history; we can document anything so easily.

What I have realized is that our conceptions of preservation are as outdated as our traditional conceptions of libraries themselves. Preservation cannot succeed if it is about capturing "things" - digital objects, or web sites. That's because we live in a sea of data, and what we need to capture, if anything, is something within the network: not objects, per se, but the flow, the pooling of the water that our data are on the network. This is a very different conception of digital objects - because it takes us away from objects and into the interactions and linkages between data - and I have frankly not thought through all that it should suggest for our efforts in preservation. Preservation must be different when we recognize that digital production is constantly happening, bound in networks, and made real through use.

Whatever preservation is, it might mean more of a focus on these things:

• Flow the preservation with the content, don’t dam it up to save it. What's worth preserving is not "the web site" or the "archive" but the interactions that people build among them.
• Collaborate to save aspects of the whole, but recognize you'll never have the whole.
• Don’t worry about fidelity; it shifts. All of the web is interactive. The "true" version will never exist. (It never really did; it always existed in multiple editions.)
• Engage with production. We can only preserve the living, and the best time and place to preserve is when the connections exist and work and demonstrate the utility of their social value.
• Recognize the value of the latent preservation increasingly inherent in the web.

Dave Winer speculated with a few of us the other day about what happens when a Flickr user dies; what happens to the account? A valid question indeed. But I think increasingly we need to realize that the data on the net is not going to vanish; like pools of water it might join other larger bodies, or might get cut off from network flow and slowly stagnate, but it will increasingly rarely vanish. Why? Because storage is cheap, and the value of having the data in the network will only increase.

These are incomplete and raw thoughts, but I think we need to shift our thinking about preservation to adjust with the way that our networked lives work, and move beyond focussing on preserving a static web more characteristic of our understandings and use 10 years ago.

 [P.S. Presentation I gave on these topics at Berkeley.]