Self-organization

Computer scientists Tom De Wolf and Tom Holvoet (2005) proposed the following "working definition" of self-organization:

Self-organisation is a dynamical and adaptive process where systems acquire and maintain structure themselves, without external control.

Systems in this context consist of bounded collections of component parts, few or many different types, few or many tokens (copies) of each type, the parts interrelated and interacting, the boundary compartmentalizing the system in an external environment, its universe, the system 'open' in the thermodynamic sense that the boundary permits exchange of information with its universe in the form of matter and energy, the system as whole more less far from a state of equilibrium where stability arises through lack of change &mdash; what physical scientists call an open thermodynamic system far from equilibrium.

Structure refers to ordered or patterned configurations of spatial, temporal, and functional character (think of the market economy or of stellar galaxies), dynamical to processes neither static nor at equilibrium (think of crystal growth or cloud formation), and adaptive to processes that maintain structural viability, through change if necessary, in response perturbations from within or without the system (think of a biological cell). Without external control implies a system boundary, such that input to the system from beyond the boundary does not control or direct the processes that originate and maintain the system's organization.

Self-organization in biology
In biology, self-organization refers to the process whereby structural and behavioral order, or pattern, arises spontaneously at more and more global levels in a living system, a consequence of the interactions among more local, lower-level components of the system, those lower-level interactions the result of local physicochemical processes (Camazine et al. 2001). Some biologists consider self-organization as the fundamental basis of the order that emerges in living systems (Kauffman 1993, 1995).

In living systems, self-organization gives rise to a variety of global patterns, both hierarchically within the system and for the system as a whole, as well as among living systems, a variety upon which natural selection can operate to assess fitness of the system and its subsystems to environmental conditions &mdash; the interaction of self-organization and natural selection reciprocal in nature and, through evolution, determining of the global pattern of the biological world (Depew and Weber 1995; Batten et al. 2008).

To summarize, we propose a view according to which the relation between self-organization and natural selection can be divided into three stages: (1) during the first stage, the emergence of organization is a necessary requirement for natural selection to occur. Without organization, behaviors that can be selected upon are statistically so unlikely that the process cannot even start. In this situation, self-organization constrains selection. In other words, organization proposes what selection might dispose. (2) In the second stage, primordial forms of organization provide the material for natural selection to act upon, which in turn results in the development of increasingly complex forms of organization. Here natural selection provides specific constraints on self-organization. Selection passes on contexts for further self-organization. (3) In the third stage, the relationship (both competition and collaboration) between increasingly complex forms of organization and their impact on the environment affects the context within which natural selection can act (e.g., the current cultural selection occurring in human society could not be conceived in the context of the early primordial soup), making natural selection and self-organization complementary aspects of the same prcess. (Batten et al. 2008)

Order and pattern can emerge through self-organization in the inanimate world as well as the animate world, in particular in far-from-equilibrium open thermodynamic systems with physicochemically interacting components, systems such as galaxies and tornados, cloud formation and crystal growth (Haken 2008) and perhaps the cosmos itself (Smolin 1998). Computer scientists have developed methodologies and applications for software with self-organizing properties (Brueckner et al. 2005).