Geography Reference
In-Depth Information
and the interdependencies between agents are 'global' in the sense of operating through-
out the trading system (Batten, 2001). For example, Batten (2001) argues that medieval
Europe in the Carolingian era was a weakly interactive system in which villages and
groups of villages were autarkic and self-contained. However, as long-distance trade and
the geographical circulation of goods and merchants intensii ed, this weakly interactive
economy was transformed into a highly interactive one, and the European economy
self-organised. Does this mean, however, that all modern economies are self-organising?
What are the critical transitions and thresholds in dif erent types of interaction and circu-
lation that dei ne whether a region or city is strongly rather than weakly interactive?
A dei ning feature of a complex system is that it is composed of interacting sub-systems
and hierarchical levels, and on this basis we might argue that the national economy can
be divided into smaller territorial sub-systems such as regions, cities and localities. Batty
et al. (2004), for example, claim that cities are the quintessential example of complex-
ity. But if complex systems dei ne their own scales and can shift scales unpredictably,
then can we assume a certain geographical scale can always be dei ned as a complex
system? The boundaries of complex systems are never pre-given but are a matter of
framing, which is partly about the purpose of study and research strategy, and partly
about the constraints arising from the operations of the system itself (Cilliers, 2005a,
2005b). Dividing a complex system into territorial sub-systems is bound to be at the cost
of analytically fragmenting and simplifying its complexity, but it can be justii ed if the
constituent l ows and connections are producing some forms of identii able boundary
and system integrity. Clearly, identifying spatial economic sub-systems does not imply
that they are closed. Rather the boundaries of complex systems are neither fully open
nor fully closed. Moreover, Cilliers (2001) also argues that when components are richly
interconnected, there is only a short route to the outside of the system. There is no safe
inside; the system boundary is folded in so that in a richly interconnected system we are
never far away from the boundary. Boundaries are functional constitutive components
that do not separate but connect systems with their environment. Sub-systems can be
identii ed within complex systems if the sub-system becomes responsible for maintain-
ing its own coni guration, and it can then be said to show 'organisational closure'. In
this state the organisation is determined purely internally even though the sub-system
(a region or city, say) exchanges energy and matter (that is, l ows of goods, services,
knowledge, capital, money and people) with its environment (other regions and cities).
The interactions between these sub-systems (regions, cities, clusters, etc.) may also deter-
mine systems at a higher hierarchical (and spatial) level so that a 'boxes within boxes'
architecture emerges.
The subsequent question is whether we can justii ably argue that spatial entities such
as regional economies and clusters are economic sub-systems that show emergent prop-
erties. Do they emerge from unplanned and unintentional localised interactions between
agents and display relatively ordered paths of change? What is not clear here is whether
such units (clusters, cities, regions) show 'organisational closure' such that they retain
their organisational identities and coni guration despite l uctuations and perturbations
in their external environments. Is it feasible that regions, clusters and cities develop key
internal interactions that provide them with some organisational 'attractor' and continu-
ity in the face of environmental changes? Given that market-based economic systems are
always importing knowledge and information then there is no simple distinction between
Search WWH ::




Custom Search