Seite - (000199) - in Knowledge and Networks

193 The Theory of Path Dependence and Path-Breaking and the Role of Platforming For decades regional clusters (e.g., Bathelt, 2005; Lazzeretti, Sedita, & Caloffi, 2014; Martin & Sunley, 2003; Porter, 2000) have been seen as hot spots of knowl- edge creation and transfer or, more broadly, learning (e.g., Bathelt, Malmberg, & Maskell, 2004; Cooke, 2001; Ibert, 2007; Malmberg & Maskell, 2002; Maskell, 2001; Morgan, 1997). This view cumulated over the years into the conception of clusters as the institution for knowledge- based economic development, although negative path dependencies and even the dangers of lock-ins have also been recorded with respect to clusters for some time (Grabher, 1993; Hassink, 2005, 2010; Henning, Stam, & Wenting, 2013; Lagerholm & Malmberg, 2009; Martin, 2010; Martin & Sunley, 2006). The path dependence of knowledge development has been noted for regions and geographical network trajectories (Glückler, 2007) as well as for firms and other types of organizations and their knowledge management prac- tices (Coombs & Hull, 1998; Nooteboom, 1997). In regional clusters this phenom- enon is, hence, likely to have its roots not only in technological but also institutional and organizational path dependencies (Sydow, Lerch, & Staber, 2010). Path dependence is mostly used with the broad meaning of “history matters,” less often in a much more specific, analytical sense alluding to the seminal works of Paul David (1985) and W. Brian Arthur (1994). In this chapter we adopt this latter under- standing, which is currently also conquering the analysis of regional development processes and might eventually be used for additional cumulative knowledge pro- duction on cluster formation and transformation processes in economic geography (Bathelt & Boggs, 2003; Boschma & Fornahl, 2011; Henning et al., 2013; Li, Bathelt & Wang, 2012; Tödtling & Trippl, 2013; Wolfe & Gertler, 2006). What is more, we try to clarify the relationships between this particular understanding of path dependence and the challenging task of path-breaking—and the role that plat- forming may have therein. Such clarification is necessary if actors wish to have strategic influence on the development of a regional cluster. According to David (1985), who developed this understanding based on his research on QWERTY,1 path dependence is understood as a tapering process trig- gered by a small event and leading, at least potentially, into a lock-in. For three reasons, this understanding is more specific than the general argument that history matters. First, a small event (which in the case of QWERTY is still debated; see Kay, 2013) triggers the tapering process of becoming more dependent on the course of action embarked on after this event. Second, from a certain point in time often retrospectively called a “critical juncture” (Collier & Collier, 1991), positive feed- back mechanisms take over and make it increasingly difficult to leave the given course of action, or path. These mechanisms make the once-chosen path increas- ingly attractive. Economists typically refer to this phenomenon as “increasing 1 QWERTY refers to the most common keyboard layout for Latin script. The name derives from the sequence of the first six keys in the upper left row of letters when read from left to right. 10 Platforming for Path-Breaking