Geography Reference
In-Depth Information
acceptable modus operandi. Translation in this case is motivated by a search for plausi-
bility rather than accuracy. This ensures that the necessary adjustments can take place,
while preserving the central structure of the social system (Tilley, 2005). Actors may also
use existing language and texts (in form of written documents, visual displays, speech
acts, and the like) when engaging in discourse with others. Discourse is a mechanism for
structuring not only interaction but also perception through the use of linguistic rules,
symbols, and analogies (Gibson, 2005). Through discourse, actors draw on, transform,
and reproduce their social reality. For example, the academic discourse about clusters
has produced a wide range of ideas concerning l exible specialization, cooperation, col-
lective learning, and the like. These ideas are in a constant struggle for survival, as they
compete for human attention and are retained and passed on in form of, for example,
best practice and policy initiatives.
The concern for translation mechanisms rel ects a version of agency that is more
socially engaged, with more discretionary, evaluative, and transformative possibilities
than what is of ered in many evolutionary accounts at more aggregate levels. While
social transmission is dif erent from genetic transmission, because of the capacity of
the mind to reinterpret and reapply the information passed on, information process-
ing behavior is also subject to the possibility of mutation. The relative probability of
further replication in a changing environment af ects the speed and direction in which
the relevant population evolves. Idea-based information varies, is selected, and is copied,
through a variety of mechanisms and with more or less imperfection, and this is all that
is needed for evolution to occur.
In research practice, it may be very dii cult to determine which particular mechanism
is operating in a given empirical instance. The analysis of mechanisms is complicated by
the fact that they exist in a nested hierarchy. Higher level mechanisms may be driven by
mechanisms operating at lower levels. For example, the argument that knowledge circu-
lates in a cluster through the interi rm mobility of employees (Agrawal et al., 2006) needs
further specii cation because mobility may be motivated by a variety of cognitive and
social logics. To what extent causal regression to lower level mechanisms is analytically
required, is a question of debate not only in the philosophy of science (Craver, 2001). It
is also a concern for cluster policy makers who grapple with the question of the level of
action at which intervention is most likely to be successful.
The study of mechanisms forces attention to the actual processes connecting causes
and ef ects. This improves theoretical understanding, for example by not mistaking
spurious associations for real causal relationships, which is a danger especially when
the goal is to explain macro-level phenomena. A systematic concern for mechanisms
also helps avoid the proliferation of seemingly disparate theoretical concepts if the same
mechanisms work in dif erent cluster processes. The problem is that in socio-economic
systems as complex as clusters several mechanisms will probably operate simultaneously,
and they may counteract one another in their ef ects on particular outcomes. Although it
is dii cult to disentangle multiple mechanisms empirically, cluster theorists ought to i nd
the notion of mechanism appealing and worth pursuing in their research.
4. Concludingremarks
The impermanence and l uidity of regional clusters calls for theoretical explanations
that appreciate the various ways in which human agency is embedded in those aspects of
