Geography Reference
In-Depth Information
chemico-physical) models can be applied across numerous dif erent i elds, including the
economic.
In Perona's view, the ef orts of Arthur (and, we would add, Krugman) to construct
a complexity economics are of this model-based ('theoretic', or 'scientii c-ontological')
kind. Thus we i nd both authors postulating an abstract economic landscape that is
viewed as an adaptive system, consisting of many agents who interact continuously
among themselves, while being allocated to a pre-given geometry of locations and
immersed in an overall 'evolutionary' environment (Arthur, 1994; Arthur et al., 1997;
Krugman, 1994, 1996), all represented by some system of non-linear equations, power
law functions and the like. In these models, agents' interactions and location decisions
are usually described by means of (predetermined non-linear) rules, which are repeated
through several steps, thus accounting for the system's 'dynamic' nature and spatial
development. Complexity in these treatments is not a characteristic of how causal proc-
esses are connected to each other, but a characteristic of the behaviour of the spatial-
temporal series of a particular variable, typically represented not by actual data but by
the solution sequence of a computational or simulation model (see not only Arthur,
1994; Arthur et al., 1997; and Krugman, 1997; but also Rosser, 2004). According to
Viskovatof (2000), although this approach to complexity economics (and complexity
'economic geography' as formulated by Arthur, Krugman and other economists) may
well abandon some of the artii cial assumptions of neoclassical economics (such as the
rational behaviour of agents, and an inherent tendency to a single, unique equilibrium),
it remains committed to deductive theorising, in this case as formulated by non-linear
models (and, it should be added, to the idea of equilibrium, even if the latter is now no
longer unique but dependent on the 'initial conditions' or 'starting point' of the eco-
nomic system). In ef ect, the complex economy becomes identii ed with the simulation
model and solution sequence used to represent it.
The problem is that there are far fewer examples of an ontological ('ontic') perspective.
One of the most concerted ef orts in this direction is that of Potts (2000), for whom complex
systems theory is the most suitable basis for constructing an evolutionary economics:
The hypothesis of evolution towards complexity is a conjecture to the ef ect that a balance
between order and chaos, between stasis and change, is the ultimate principle underlying all
evolutionary processes. Where equilibrium is the expression of 'balance' in an inert, mechanical
world of point-like existence, complexity is the expression, the structural signature, of balance
in a world of interacting dynamic systems. The hypothesis of evolution towards complexity is
the logical principle that interlinks the geometry of all economic systems. (Potts, 2000, p. 91)
Following Kauf man's (1993) contention that there is no strong reason to attribute the
emergence of order in biological systems solely to the force of selection, Potts argues
that much of the order and coordination in an economic system may not be the result of
'market selection' at all, but a spontaneous order of self-organised systems. According
to Potts, the concept of complexity - and more specii cally the hypothesis of evolution
towards complexity - contains within its meaning a number of high-level connecting
principles that are prominent in heterodox economic theories: 'evolutionary economics
is an eclectic rubric centred around the paradigm of the complexity of open systems proc-
esses, and its basic substance is both more encompassing and more protean than a simple
transferral of metaphor' (2000, p. 186).
