Some folks may take issue with this statement. Am I advocating getting rid of books? Am I saying that there is no use for materials in libraries? Has the collection ceased to matter? No. Collections and artifacts are very important, just secondary to the real goal: knowledge. Since knowledge is created through conversation, the primary “thing” libraries should be facilitating (collecting adding value to, providing access to) are conversations.
The end result of a conversation (or even the result of parts of a conversation that is ongoing) may well be an artifact. So a scientist may be studying a phenomenon. He is engaged in knowledge creation whereby he has internal conversation, talks with colleagues, and reads the literature (the medium of a long and ongoing conversation within the profession). Along the way, as he comes to conclusions (agreements) he may well publish an article (or book, or blog post). That publication is an artifact of the conversation. Yet, the conversation still continues.
Why does it matter…is this a small point? Well let us take that publication. The author understands that artifact in a given context like its relation to the field, plus a whole host of other factors like: will this will help define a domain; this will mark out a territory I am studying; this will help me get tenure; this will grow my reputation. Someone else reading that article may place it into another context (perhaps a radically different context like “this is an example of good scientific writing”, or “this work on astronomy explains dinosaur extinction”). The point is the artifact hasn’t changed, and it hasn’t brought its context with it. So the same artifact is being applied to very different understandings of the world (knowledge).
Now both the author and the reader need the artifact. It is NOT unimportant. It is, however, only part of the knowledge of the author and reader. Why care? Because the author and the reader may describe the artifact very differently, place it into a larger collection (classify it) very differently, etc. If libraries focus on the artifact, and provide their own contexts (for after all, what is Dewey or LCSH if not simply a shared context) without regard to the contexts of the individual member/patron/user, they will create a disconnect between how a user finds something, and what it means. This in turn creates pressure on the part of the user to utilize their own way of organizing things, which our systems only provide in a small and crude way (tagging, lists).
By the way, since much of what makes up a user’s context is socially derived (through the media, education, peer groups, etc), the contexts within a given user population may be very close. That’s why Dewey and LCSH have worked as well as they have for the library community. That’s also explains why aggregating things like tags and links to pages (like Google Page Rank) work so well. It is not that everyone thinks the same, or that the artifacts are similar, it is that groups of users have a great deal of overlap in their understandings. Participatory systems should build on the overlap for discovery, but allow for the individual context in the individual.