Reviewed by Tommi Leung, United Arab Emirates University
Cognitive scientists have witnessed a paradigmatic shift in the theory of cognition from a classical symbol-based to an interactive ecological approach that focuses on the dynamics between mind, body, and external world. Robert D. Rupert’s Cognitive systems and the extended mind examines this theory of situated cognition. A number of variations on situated cognition are discussed: the extended view (cognition extends beyond the boundary of the organism), embedded mind (cognition depends on the organism’s use of external resources, while not extending into the environment), and embodied mind (no active role of the environment in cognition). R rejects the extended view of cognition and argues that the embedded and embodied views provide important understanding of cognition.
The book is divided into an introductory chapter and three major parts. Part 1, ‘The thinking organism’, presents arguments against the extended view based on problems of demarcation. A non-extended systems-based approach is proposed, claiming that a single cognitive process involves the interaction between various bodily concepts, whereas the external environment is not an integral part of cognition.
Part 2, ‘Arguments for the extended view’, examines previous arguments in support of the extended view. Ch. 5, ‘Functionalism and natural kinds’, argues that what is considered an extended cognitive process by proponents of extended mind can be accounted for by internal bodily concepts and mechanism. Ch. 6, ‘Developmental systems theory and the scaffolding of language’, describes the mismatch between extended mind and development systems theory in genetic biology. Ch. 7, ‘Dynamical systems theory’, points out that a version of extended mind that takes into account the interaction between organism and environment does not provide support for the extended mind.
Part 3, ‘The embedded and embodied mind’, discusses the embedded view of mind. In Ch. 9, ‘Embedded cognition and computation’, R claims that embedded models are not necessarily defined by explicit encoded rules. A non-computationalist approach is theoretically possible if it is ‘the structure of the environment that causes the developmentally flexible brain to implement some algorithms in the absence of explicitly encoded rules’ (186).
Ch. 10, ‘Embedded cognition and mental representation’, contends that the embedded approach supplements orthodox rule-based computationalism, and moreover that embedded representations are partial, context-dependent and action-oriented, relational and egocentric, and ‘fluidly guide our messy, real-time interaction with the world’ (193). Ch. 11, ‘The embodied view’, evaluates possible interpretations of the embodied mind, making sense of its conceptual relation with orthodox cognition. He points out that the potential disagreement between embodied and orthodox cognition is nonexistent and that embodied models exhibit properties compatible with the orthodox view. Ch. 12, ‘Summary and conclusions’, reiterates R’s rejection of the extended view of cognition and suggests that it is through the active design of experiments and models that the embedded and embodied models can best proceed.