- The Critical Idiom of Postmodernity and Its Contributions to an Understanding of Complexity
Paul Cilliers’s Complexity and Postmodernism: Understanding Complex Systems attempts to bring together developments in neuroscience, linguistics, logic, computer science, the philosophy of science, and poststructural theory in an effort to locate unifying themes in these exciting fields. Cilliers seizes on “complexity,” a term used to describe large-scale, non-linear interaction of nodes or agents in a dynamic environment, as a way to discuss possible structural resonances among the brain, natural language, artificial intelligence, deconstruction, and the legitimation of knowledge in contemporary society. By means of this ambitiously interdisciplinary approach, Cilliers hopes to overcome certain persistent simplifications in the thinking of both representation and organization.
Cilliers introduces the terms “distributed representation” and “self-organization” (or “self-organized criticality”) to improve upon the standard analytical and rule-based methods of understanding complexity. He takes up the “connectionism” attributed to neural networks as a model for the contingency and dynamism of complex systems such as those of the brain or of natural language. Connectionism treats the interactions of the nodes within it as a dynamic whole, each individual node working in concert with all other nodes of the network to adapt continually to environmental changes. This is in stark contrast to the rule-based descriptions of complexity which, imposing the rigidity of principled behavior on the nodes, cannot account for the contingency of environmental conditions and localized adaptations. Through distributed representation, Cilliers circumvents the shortcomings of the rule-based understanding of complexity because he is able to demonstrate that distributed representation is not representation at all, but rather the recognition of localized contingency. Each node interacts in concert with the other nodes of a neural or language network because each node acts and reacts as a system, not individually. This interaction is further explained through self-organization. A complex system, able to organize its individual nodes or agents through concerted action, does not have a central organization center but has the capacity to self-organize at local sites where environmental changes are detected.
Cilliers, following Saussure and Derrida, recognizes the complexity of natural language in terms of both its stability and its evolutionary capacity. Discussing natural language’s ability to instantiate meaning through a system of phonetic or graphical differences, he claims that while language users are bound to certain language rules, they are nonetheless free to adjust those rules and hence to influence the evolution of the language. This seemingly contradictory statement finds its theoretical underpinnings in Saussure’s concepts of the signifier and signified, where signification involves mental representation and the enactment of this representation through the utilization of the signifier in either spoken or written language. A language user has to choose among a host of socially sanctioned signifiers to represent a mental state. As Cilliers observes, “The system of language transcends the choices of any individual user, and therefore has stability” (39). But while he recognizes the constraints of social conditioning and common culture that temper any “free play” of language, Cilliers conceptualizes language as less the closed system described by Saussure than the open one of Derrida. Derrida, by denying the metaphysics of presence, claims that meaning cannot be generated outside of language and hence “where there is meaning there is already language” (43). Drawing in particular on the Derridean notions of différance and trace, Cilliers tries to show that natural language is a complex system which adapts dynamically over time and across multiple environments through a system of phonetic and graphic difference. Because language is constituted by nothing more than relationships, there are traces of other signs inherent in every sign. Language, through difference and deferral (hence différance), self-organizes signs through distributed representation.
Cilliers uses his discussion of natural language as a segue into a consideration of artificial intelligence as a complex system. In a chapter entitled “John Searle Befuddles,” Cilliers asserts that Searle’s contention that artificial intelligence does not possess intentionality and hence cannot be called intelligence at all is untenable. Cilliers briefly summarizes Searle’s views on artificial intelligence through a description of...