Extended cognition and the priority of cognitive systems Action editor: Leslie Marsh Robert D. Rupert Department of Philosophy, University of Colorado at Boulder, Boulder, CO, USA Received 2 March 2010; accepted 2 April 2010 Available online 23 July 2010 Abstract This essay begins by addressing the role of the so-called Parity Principle in arguments for extended cognition. It is concluded that the Parity Principle does not, by itself, demarcate cognition and that another mark of the cognitive must be sought. The second section of the paper advances two arguments against the extended view of cognition, one of which – the conservatism-or-simplicity argument – appeals to principles of theory selection, and the other of which – the argument from demarcation – draws on a systems-based theory of cog- nition. The final section contests the claim, made by Andy Clark, that empirical work done by Wayne Gray and colleagues supports the extended view. Ó 2010 Elsevier B.V. All rights reserved. Keywords: Extended cognition; Extended mind; Distributed cognition; Situated cognition; Embedded cognition; Parity Principle 1. Introduction According to the hypothesis of extended cognition (HEC), human cognitive processing extends beyond the boundary of the human organism. 1 As I understand HEC, it is a claim in the philosophy of cognitive science; the extended view is meant to capture a central fact about human beings’ core cognitive faculties and, in doing so, reshape the study of human cognition. So understood, HEC cannot be vindicated by oddball cases, science-fiction thought experiments, or the development of assistive tech- nologies. Such results—even if they constitute individual cases of extended cognition or entail the conceptual possi- bility of it—fall far short of the revolution promised by HEC’s proponents. 2 These results are germane only if they help to reveal what actually are (or should be) the proper- ties of central causal-explanatory interest in cognitive sci- ence, that is, the properties appearing in cognitive science’s best account of everyday, paradigmatically cogni- tive capacities—language-acquisition, inference, categori- zation, planning, theory-construction, perception, and reading—and the associated behavior. Many criticisms of HEC have made their way into print (Adams & Aizawa, 2001, 2007; Block, 2005; Butler, 1998; Grush, 2003; Rupert, 2004, 2008, 2009, 2010; Segal, 1997; Ste- relny, 2004; Weiskopf, 2008; Wilson, 2002). The present essay adds to this critical literature but does so in a construc- tive way: by presenting an independently motivated account of cognition. Given the empirical facts about human beings, this conception of cognition likely countenances very little of 1389-0417/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.cogsys.2010.04.002 E-mail address: robert.rupert@colorado.edu 1 An enormous amount has been written in support of HEC, from a philosophical standpoint: see Clark (1997, 2001, 2003, 2007, 2008), Clark and Chalmers (1998), Dennett (1996), Haugeland (1995), Hurley (1998), Millikan (1993, essays 7–9), Rowlands (1999), van Gelder (1995), Wheeler (2005), and Wilson (2004). For a sampling of the empirical work that has inspired HEC, see Brooks (1986), Dawkins (1982), Donald (1991), Hutchins (1995), and Thelen and Smith (1994). Also see the references provided in note 2. 2 The language of paradigm shift, revolution, and the reorientation of cognitive science can be found in Brooks (1999), Gibbs (2006), Haugeland (1995), Lakoff and Johnson (1999), Spivey, Richardson, and Zednik (2010), Thelen and Smith (1994), van Gelder (1991, 1995), Varela, Thompson, and Rosch (1991), and Wheeler (2005). www.elsevier.com/locate/cogsys Available online at www.sciencedirect.com Cognitive Systems Research 11 (2010) 343–356