KSR and the Degrees of Separation Game.

A blawg entry, which originated as wry plan to turn the KSR v. Teleflex decision into a game structured around the notion of six degrees of separation, turned more thoughtful once my wry mind opened to the matter of how inventions originate from among those “having ordinary skill in the art.”

In the game scenario, participants would start with a patented invention, then deconstruct it, like a house of cards, into its elements, which were well-known.  If that could be done in six or less steps, then a point was awarded.  The fabulous board game so envisioned was tossed aside, once thought was given to the forms and flow of inventive development. 

At the ends of the spectrum are well-populated R&D staffs, and one nutty dude in his garage.  But both of those norms imply working alone, in secret from competitors or from all others.  One notion of obviousness is that an idea or element from a related field of art would have been obvious to combine with prior art within the skilled persons’ field.  It’s that notion of borrowing that may earmark what is invention and what is obvious. 

Those more skilled than I have studied in detail how similar, but unrelated artisans pass ideas around, which leads to one or more of them achieving an innovation.    Uzzi and Spiro, Collaboration and creativity: The small world problem, 111 Amer. J. Sociol. 447 (2005) analyzed collaboration among creators of past Broadway musicals.  
That study provoked a recent study of how c
ompanies, indirectly linked in a network of strategic alliance relationships with only a few degrees of separation, are more innovative.  Profs. Schilling (NYU) and Phelps (U. Wash), Interfirm Collaboration Networks: The Impact of Large-Scale Network Structure on Firm Innovation, MANAGEMENT SCIENCE, Vol. 53, No. 7, July 2007, pp. 1113-1126.  
It concludes that firms more connected to others are more creative, which influences the number of patented inventions they obtain.  Phelps and Schilling conclude that 
“companies that network and form strategic alliances are more creative and develop more patented inventions than those that don’t.”  Their study revolves around “clustering,” and “reach” as factors derived from “publicly reported contractual alliances.”  “The structure of alliance networks influences their potential for knowledge creation,” and so, patent output. 
Clustering is a measure of redundant paths between companies that represents their level of cooperation, and pathways for information to be passed along quickly and accurately.  Reach represents the likelihood, speed, and integrity of knowledge transfer between clustered companies.  The profs’ abstract for their study notes that “firms embedded in alliance networks that exhibit both high clustering and high reach (short average path lengths to a wide range of firms) will have greater innovative output than firms in networks that do not exhibit these characteristics.”

Once they had computed and assessed their longitudinal study of the patent performance of 1,106 firms in 11 industry-level alliance networks, the “results indicated support for [their] argument: the combination of clustering and reach was associated with significantly higher firm patenting.”  Much of what these Professors set out was way over my understanding of statistical relationships and analysis, but I accept the premise, and so accept the research that underlies their conclusions. 

Trouble is, that proof of collaboration, or more pointedly of cross-pollination, may prove the sort of obviousness, or the lack of inventive separation, that the KSR decision set forth as among the reasons to deny patents to what were believed to be inventions.  The decision does draw doubt toward the professors’ validation of their “dependent variable: patents” based on earlier research that “Patents provide a measure of novel inventions that is externally validated through the patent examination process.”