So-called “individualists” argue that cognition is located within the physical bounds of agents. “Externalists” argue that cognition sometimes extends beyond agents to include portions of the environment. The externalist position about cognitive processes is sometimes referred to as Active Externalism, the Extended Mind Hypothesis or the Hypothesis of Extended Cognition. The first decade of the twenty-first century has seen vigorous defense of this position on a variety of grounds.

Lately individualists have been mounting a renewed response to externalism. Robert Rupert’s Cognitive Systems and the Extended Mind (hereafter CSEM) is one of the most important pieces of this response. Rupert argues that the recent arguments presented in favor of the more radical aspects of externalism are not convincing. The features of twenty-first century externalism that Rupert finds attractive can be accommodated with in the orthodox individualistic view of the location of cognitive processes.

CSEM is divided into four parts. In (the introductory chapter, Rupert helpfully disambiguates distinct positions that fall under the banner of “situated cognition”. Advocates of situated cognition argue that, to understand cognition, we must not merely focus on agents and brains. Instead we must include the broader contexts which are their natural home. Defenders of the extended mind hypothesis hold that features of these contexts are literally part of at least some cognitive processes. Advocates of “embedded” views of cognition emphasize the importance of the environment but locate cognition within the physical bounds of agents. Proponents of “embodied” views of cognition argue that distinctively bodily processes have a special role in cognitive processes. Rupert argues against the extended mind hypothesis and for versions of the embodied and embedded views that locate cognition within the physical bounds of agents.

Rupert’s arguments against externalism come in the second and third parts of the book (by name, Parts I and II). Rupert’s primary concern in the second part is to clarify just what it is for a system or process to count as cognitive. Such a principle of demarcation is needed for debate over the location of cognition to be substantial. Rupert examines and rejects various principles, eventually settling on a systems-based rule for demarcation. He defends an individualistic understanding of this principle by charting explanatory successes of cognitive science which, he argues, are good evidence that the systems in question are to be located within agents. In the third part Rupert scrutinizes more particular arguments for extended cognition and finds them wanting, often on the ground that individualistic explanations of particular phenomena are preferable to extended explanations.

Rupert turns to embedded and embodied views of cognition in the fourth part (by name Part III) of CSEM. Rupert argues that the important lessons of these remaining sub-varieties of the situated cognition family pose no challenge to the ordinary view of cognition as bodily bounded.

The strengths of CSEM are found both early and late. The introductory taxonomy of various situated positions is helpful. So are the arguments against radical interpretations of the embedded and embodied approaches. The discussion of demarcation principles and the related criticism of the extended view are less persuasive. The attention given to what it is for a process or system to count as cognitive is misplaced for two reasons. First, it threatens to render unclear the link between the individualism/externalism debate and psychological practice, a link which is crucial to CSEM. Cognitive scientists have a range of phenomena to explain, probably with a rough sense of what makes this range a family. The real issue is the location of the processes/systems marshaled in explanations. Given this, the appropriate issue is not what it is for something to be cognitive, but what it is for something to be a system/process..

A second reason for being suspicious of the attempt to pin down a principle demarking the cognitive is that doing so erodes the possibility of clarifying the notion of the cognitive via particular explanations. Perhaps particular explanations which include environmental features offer new information about the nature of cognition. If so, then the extended view offers the possibility of radical change to our understanding of the cognitive. Rupert seems to have no room for such a conceptual possibility, but it seems to me to be one that we should leave open if we are going to follow the best explanations that cognitive science offers with regard to the location of cognition.

This issue has implications for Rupert’s overall argument against the extended view. The most important plank of his critical platform is that the best explanations of cognitive phenomena are individualistic. Suppose that extant explanations rely on a notion of cognition such as Rupert’s. If this notion is called into question by other explanations), then we have important choices to make. We might revise our understanding of what it is for a process or system to be cognitive. If we do this then we should give up our confidence, to some degree, in the particular explanations that used the old notion of the cognitive. Alternatively we might decide that “cognitive” is shorthand for a variety of natural kinds. While some “cognitive” processes might be bodily bound, others, with a different sense of “cognitive”, might not be. In either case Rupert’s route from particular explanations to a defense of bodily bound cognition is at least made more complex and perhaps even undermined.

A review of this length cannot accurately convey the fine detail with which Rupert examines this territory. This is an important book for individualists and externalists because of the care with which Rupert clarifies and presents the positions and issues. Although it cannot be taken as a conclusive case against extended cognition, CSEM presents an important challenge and provides some tools for subsequent discussion. Strongly recommended.

© 2011 Andrew Sneddon

Andrew Sneddon is a professor in the department of philosophy at the University of Ottawa.