Abstract
We contribute to the unresolved debate about why clusters evolve and transit across their life cycle. We differentiate types of
knowledge generated by combining endogenous and exogenous change processes in clusters. These drivers generate synergistic
multiplier effects and a cumulative causation in the process creating specific forms of new knowledge which determine a clusters
particular evolution. The paper presents a clear-cut distinction between sustaining knowledge based on incremental (exploitative
learning) innovation and evolution caused by disrupted (exploratory learning) turns from radical innovation in clusters. The
creation of sustaining knowledge by existing firms is exploitative in nature, competence-enhancing and occurs mainly in the
context of strong ties; and it is usually initiated by incumbent anchor tenants or technological gatekeepers (TGs). In contrast, the
creation of disruptive knowledge is mainly initiated by new entrepreneurial firms, is exploratory in nature, competence-destroying
and requires a context of extensive weak ties in order to grasp distant (to the cluster and technology dominance) knowledge. Each
type of knowledge determines a different shift in a clusters evolution and its technological life cycle. All in all, the paper presents
an undestanding about how capabilities are formed in clusters, how the cluster organizational routines are formed and evolve. In
general, it is necessary to understand that the cluster capabilities evolve, become redundant or are recombined to be transformed
into new capabilities which ensure advantage and returns. An explorative and logitudinal case study covering 40 years and
analyzing three major innovation journeys is presented using longitudinal data on two world-class clusters in the ceramic industry