Geography Reference
In-Depth Information
have dif erent interpretations of what constitutes good or novel ideas. Ideas are impli-
cated in valuation and analysis. Some ideas are implicitly held, forming the taken-for-
granted backdrop of interpretation. Other ideas are explicitly expressed, constituting
the substance of communication and negotiation. Some ideas produce consequences in
the background, others work in the foreground. If ideas are oriented to outcomes, they
have cognitive content, such as ideas about the likely ef ectiveness of certain policy initia-
tives. Normative ideas consist of values and are typically not outcome oriented, such as
notions about the distributional fairness of a particular governance system.
Some scholars have criticized theories of institutions and organizations for neglecting
how ideas and meanings are implicated in change (Blyth, 1997; Ingram and Clay, 2000).
The key point, they argue, is that what people believe (ideas) is just as important as what
they want (interests), and what they want is often determined by what they believe. In
some i elds, such as political science and political sociology, there has been renewed
interest in 'how ideas matter' (Campbell, 1998; see also the contributions in Goodin and
Tilley, 2006), picking up where Talcott Parsons (1938) left of long ago. Economic geog-
raphers also refer to ideas as input factors in 'innovation systems', as the raw material of
'creative clusters', or as outcomes of strategic deliberations, but they have generally not
problematized ideas as a subject of inquiry in their own right. The concern for ideas by
cluster researchers who draw inspiration from population ecology, organizational evolu-
tionary theory, and complexity theory has, at best, been latent. Ecologists studying popu-
lation dynamics have generally looked at the outcomes of competition and cooperation,
such as i rm survival, but they have typically not studied directly the cognitive structure
of a population of actors to answer questions about how the actors choose between dif-
ferent ideas. They have usually taken the existence of resource niches as given and have
not asked how actors actually construct their world in these niches. When researchers in
the i eld of organizational evolution invoke arguments from institutional theory, they
generally refer to those that explain convergence and reproduction, while downplaying
the older versions of institutional theory that take a more l uid perspective on struggles
over resources and ideas (Hirsch and Lounsbury, 1997). Complexity theorists typically
of er no direct account of the social constructions involved in the spontaneous process
by which order is created. How do actors frame their actions vis-à-vis one another? What
are the ingredients of the 'self-referencing framework' that guides decisions in a way
consistent with the system's accumulated history? What exactly goes into actors' choices
in constructing what are essentially contestable social structures? In other words, we
need to know something about the substrate on which decisions, routines, competencies,
and so forth depend. In the social and cultural domain of clusters, ideas would seem to
qualify as such a substrate, with sui cient variation to make evolution possible and with
sui cient longevity, i delity, and fecundity to permit replication (Dawkins, 1976).
Ideas may be thought of as constituting the micro-foundation for the entities that con-
tribute to cluster evolution. Being a foundational unit means that they are distinct from
the entity of which they are part, such as rules and routines. The evolutionary question
is whether or not an idea is sui ciently discrete to function as a unit of selection (see the
debates in Aunger, 2000). If ideas are to function as replicators (Dawkins, 1976), in the
sense that they have generative powers and can be transmitted in their entirety, they must
be analytically divisible. Just as clusters, organizations, or routines are not established in
a fully specii ed form and do not evolve
in toto
, individuals do not acquire information
