Last Fall, the NSF sponsored a workshop in Ann Arbor on the History and Theory of Infrastructure: Lessons for New Scientific Cyberinfrastructures. The summary report is now available.
Beginning with the premise that development of electronic media to support scientific research may be informed by studying other infrastructures, participants explored issues of path dependence, 1st and 2nd-order distributional impacts, and the underlying (political) philosophies that "ought" to be central to the design of a cyberinfrastructure. These last include broad participation and collaboration, both within and across disciplinary boundaries.
My own sense is that we're in something of an "eCambrian era"- we've known for some time that traditional intellectual divisions are sub-optimal, but are only beginning to realize how specialized "noetic cells" may combine to form whole new "knowledge species."
Hold onto your hat - over the next few decades, we'll experience a profound shift in how knowledge is acquired, assessed, organized, and communicated.
H/T to William H Dutton (Oxford) for the link.