Finally, although our network and interview data have helped to advance our knowledge
of this area, there is no substitute for micro level event histories of actual network
systems in action. An event-by-network review of a ‘‘year in the life’’ of a network manager would go a long way toward better specifying how network managers do what they do (for a working model of this form of research see Barley [1986, 1990]). Besides being expensive, this form of research is bound to be intrusive and thus difficult to get access to the desired participants. Nevertheless, it may be essential to advancing theory in this area.