Everything has an edge of validity. Nothing is valid everywhere.
In today's New York Times op ed section Jonathan David Farley writes about the difficulty in using a mathematical analysis of a network "The N.S.A.'s Math Problem" without deeper understanding of the context of the data or the actors/players in the network.
He states "For example, Jafar Adibi, an information scientist at the University of Southern California, analyzed e-mail traffic among Enron employees before the company collapsed. He found that if you naïvely analyzed the resulting graph, you could conclude that one of the "central" players was Ken Lay's ... secretary."
I'm not so sure I would throw her out of the loop so fast as there have been other famous secretaries - Rose Mary Woods, who was Nixon's gatekeeper attributed with erasing 18 and a half minutes of Watergate history! Gatekeepers are to be reckoned with especially in highly political closed systems like Enron. Granted they are not the lead player but they are a critical point of entry to a network.
In a recent series of ONA Prac comments one group member wanted to know if there was software to identify the network surrounding "C" level or key players in an organization as a penetration strategy for marketing. Very relevant request.
Farley illuminates what we need to ask in a network analysis is do the the players share the same characteristics and group them together as a node revealing a big picture or "concept lattice". This term comes from the field of "formal concept analysis".
Often in my work with organizational surveys we would group the data results by demographics and find that these groupings were the more significant indicators than in what department a person worked. The demographic network became more relevant than the formal.
Farley also emphasizes that "using some common sense and knowledge of terrorists" led Gordon Woo, a mathematician and risk-assessment consultant to forecast that London was ripe for an attack.
It is the combination of math and the deep implicit knowledge of a system one is studying that makes for a break through in understanding.
This in fact is how the Enigma code was broken that he refers to in his closing remark. It was determined that weather ships (much easier to penetrate) carried the German code books as well as the war ships. The knowledge of the relationship weather ships played with war ships was critical to penetrating the network and ultimately cracking the code.
Weather ships and secretaries may not figure large in the network analysis maps but depending on the context and what you are trying to achieve may have deeper validity.
For those of us who are practioners helping organizations navigate in a complex networked world continuing to balance the implicit knowledge of the organization with explicit data is critical for success.
~ Victoria G. Axelrod