Anthony Chemero

The question whether cognition ever extends beyond the head is widely considered to be an empirical issue. And yet, all the evidence amassed in recent years has not sufficed to settle the debate. In this paper we suggest that this is because the debate is not really an empirical one, but rather a matter of definition. Traditional cognitive science can be identified as wedded to the ideals of “smallism” and “localism”. We criticize these ideals and articulate a case in favor of extended cognition by highlighting the historical pedigree and conceptual adequacy of related empirical and theoretical work.

Markman and Dietrich¹ recently recommended extending our understanding of representation to incorporate insights from some “alternative” theories of cognition: perceptual symbol systems, situated action, embodied cognition, and dynamical systems. In particular, they suggest that allowances be made for new types of representation which had been previously under-emphasized in cognitive science. The amendments they recommend are based upon the assumption that the alternative positions each agree with the classical view that cognition requires representations, internal mediating states that bear information.² In the case of one of the alternatives, dynamical systems³, this is simply false: many dynamically-oriented cognitive scientists are anti-representationalists.^4,5,6

The complex systems approach to cognitive science invites a new understanding of extended cognitive systems. According to this understanding, extended cognitive systems are heterogenous, composed of brain, body, and niche, non-linearly coupled to one another. This view of cognitive systems, as non-linearly coupled brain–body–niche systems, promises conceptual and methodological advances. In this article we focus on two of these. First, the fundamental interdependence among brain, body, and niche makes it possible to explain extended cognition without invoking representations or computation. Second, cognition and conscious experience can be understood as a single phenomenon, eliminating fruitless philosophical discussion of qualia and the so-called hard problem of consciousness. What we call “extended phenomenological-cognitive systems” are relational and dynamical entities, with interactions among heterogeneous parts at multiple spatial and temporal scales

The ideas of continental philosopher Martin Heidegger have been influential in cognitive science and artificial intelligence, despite the fact that there has been no effort to analyze these ideas empirically. The experiments reported here are designed to lend empirical support to Heidegger’s phenomenology and more specifically his description of the transition between ready-to-hand and unready-to-hand modes in interactions with tools. In experiment 1, we found that a smoothly coping cognitive system exhibits 1/fβ type positively correlated noise and that its correlated character is reduced when the system is perturbed. This indicates that the participant and tool constitute a self-assembled, extended device during smooth coping and this device is disrupted by the perturbation. In experiment 2, we examine the re-organization of awareness that occurs when a smoothly coping, self-assembled, extended cognitive system is perturbed. We found that the disruption is accompanied by a change in attention which interferes with participants’ performance on a simultaneous cognitive task. Together these experiments show that a smoothly coping participant-tool system can be temporarily disrupted and that this disruption causes a change in the participant’s awareness. Since these two events follow as predictions from Heidegger’s work, our study offers evidence for the hypothesized transition from readiness-to-hand to unreadiness-to-hand

Thomas Stoffregen's "Affordances and Events" makes many points that are forgotten all too often--if they are realized at all--by adherents to the ecological perspective in psychology. He is to be applauded for this. But he compiles these points to make a very strong and very sweeping claim about the validity of a broad swath of research that is done by ecological psychologists. In particular, he argues for the following conclusion: "More generally, I have suggested that events may not be perceived; it may be that only affordances are perceived. If so, this would raise serious questions..

Summary. The “New Computationalism” that is the subject of this special issue requires an appropriate notion of representation. The purpose of this essay is to recommend such a notion. In cognitive science generally, there have been two primary candidates for spelling out what it is to be a representation: teleological accounts and accounts based on “decoupling.” I argue that the latter sort of account has two serious problems. First, it is multiply ambiguous; second, it is revisionist and alienating to many of the potential allies of the “New Computationalism”. I also suggest that teleological accounts do not suffer from these problems, making them more appropriate as the foundation of any new computationalism.

Using hypersets as an analytic tool, we compare traditionally Gibsonian (Chemero 2003; Turvey 1992) and representationalist (Sahin et al. this issue) understandings of the notion ‘affordance’. We show that representationalist understandings are incompatible with direct perception and erect barriers between animal and environment. They are, therefore, scarcely recognizable as understandings of ‘affordance’. In contrast, Gibsonian understandings are shown to treat animal-environment systems as unified complex systems and to be compatible with direct perception. We discuss the fruitful connections between Gibsonian affordances and dynamical systems explanation in the behavioral sciences and point to prior fruitful application of Gibsonian affordances in robotics. We conclude that it is unnecessary to re-imagine affordances as representations in order to make them useful for researchers in robotics

We provide a taxonomy of the two most important debates in the philosophy of the cognitive and neural sciences. The first debate is over methodological individualism: is the object of the cognitive and neural sciences the brain, the whole animal, or the animal--environment system? The second is over explanatory style: should explanation in cognitive and neural science be reductionist-mechanistic, inter-level mechanistic, or dynamical? After setting out the debates, we discuss the ways in which they are interconnected. Finally, we make some recommendations that we hope will help philosophers interested in the cognitive and neural sciences to avoid dead ends

The ecological approach to perception-action is unlike the standard approach in several respects. It takes the animal-in-its-environment as the proper scale for the theory and analysis of perception-action, it eschews symbol based accounts of perception-action, it promotes self-organization as the theory-constitutive metaphor for perception-action, and it employs self-referring, non-predicative definitions in explaining perception-action. The present article details the complexity issues confronted by the ecological approach in terms suggested by Rosen and introduces non-well-founded set theory as a potentially useful tool for expressing them. The issues and the tool are brought to focus in the concept of affordance that is the basis for explanation of prospective control of action in the ecological approach

In this talk, we defend extended cognition against several criticisms. We argue that extended cognition does not derive from armchair theorizing and that it neither ignores the results of the neural sciences, nor minimizes the importance of the brain in the production of intelligent behavior. We also argue that explanatory success in the cognitive sciences does not depend on localist or reductionist methodologies; part of our argument for this is a defense of what might be called ‘holistic science’

Several articles have recently appeared arguing that there really are no viable alternatives to mechanistic explanation in the biological sciences (Kaplan and Bechtel; Kaplan and Craver). We argue that mechanistic explanation is defined by localization and decomposition. We argue further that systems neuroscience contains explanations that violate both localization and decomposition. We conclude that the mechanistic model of explanation needs to either stretch to now include explanations wherein localization or decomposition fail or acknowledge that there are counterexamples to mechanistic explanation in the biological sciences

What would it take for an artificial agent to be treated as having moral value? As a first step toward answering this question, we ask what it would take for an artificial agent to be capable of the sort of autonomous, adaptive social behavior that is characteristic of the animals that humans interact with. We propose that this sort of capacity is best measured by what we call the Embodied Turing Test. The Embodied Turing test is a test in which intelligence is operationally defined in terms of autonomous, adaptive interaction with the environment and with other animals. Three versions of the Embodied Turing test were performed with a SONY AIBO robot. Human participants were asked to differentiate between AIBO in a human-controlled mode and AIBO in a software-controlled mode. Our results indicate that the human participants were guessing at how AIBO was controlled. Our data reveals that people do not have enough experience with robots to accurately evaluate its behavior. This indicates that today’s humans do not have enough experience with artificial agents to treat them as morally valuable.

In her essay --?Information, Perception and Action--, Claire Michaels reaches two conclusions that run very much against the grain of ecological psychology. First, she claims that affordances are not perceived, but simply acted upon; second, because of this, perception and action ought to be conceived separately. These conclusions are based upon a misinterpretation of empirical evidence which is, in turn, based upon a conflation of two proper objects of perception: objectively with properties and affordances.

In this essay we respond to some criticisms of the guidance theory of representation offered by Tom Roberts. We argue that although Roberts’ criticisms miss their mark, he raises the important issue of the relationship between affordances and the action-oriented representations proposed by the guidance theory. Affordances play a prominent role in the anti-representationalist accounts offered by theorists of embodied cognition and ecological psychology, and the guidance theory is motivated in part by a desire to respond to the critiques of representationalism offered in such accounts, without giving up entirely on the idea that representations are an important part of the cognitive economy of many animals. Thus, explorations of whether and how such accounts can in fact be related and reconciled potentially offer to shed some light on this ongoing controversy. Although the current essay hardly settles the larger debate, it does suggest that there may be more possibility for agreement than is often supposed

A recent set of experiments of ours supported the notion of a transition in experience from readiness-to-hand to unreadiness-tohand proposed by phenomenological philosopher Martin Heidegger. They were also an experimental demonstration of an extended cognitive system. We generated and then temporarily disrupted an interaction- dominant system that spans a human participant, a computer mouse, and a task performed on the computer screen. Our claim that this system was interaction dominant was based on the detection of 1/f noise at the hand-tool interface. The inference from the presence of 1/f noise to the presence of an interaction-dominant system is occasionally disputed. Increasing evidence suggests that inference from multifractality to interaction dominance is more certain than 1/f- like scaling alone. In this paper, we reanalyze the data using the wavelet transform modulus maxima method, showing that the human-mouse system displays multifractality. This reinforces our claims that the system is interaction dominant.

In their historical overview of cognitive science, Bechtel, Abraham- son and Graham (1999) describe the field as expanding in focus be- ginning in the mid-1980s. The field had spent the previous 25 years on internalist, high-level GOFAI (“good old fashioned artificial intelli- gence” [Haugeland 1985]), and was finally moving “outwards into the environment and downards into the brain” (Bechtel et al, 1999, p.75). One important force behind the downward movement was Patricia Churchland’s Neurophilosophy (1986). This book began a movement bearing its name, one that truly came of age in 1999 when Kath- leen Akins won a million-dollar fellowship to begin the McDonnell Project in Philosophy and the Neurosciences. The McDonnell Project put neurophilosophy at the forefront of philosophy of mind and cogni- tive science, yielding proliferating articles, conferences, special journal issues and books. In two major new books, neurophilosophers Patricia Churchland (2002) and John Bickle (2003) clearly feel this newfound prominence: Churchland mocks those who do not apply findings in neuroscience to philosophical problems as “no-brainers”; Bickle mocks anyone with traditional philosophical concerns, including “naturalistic philosophers of mind” and other neurophilosophers

This comprehensive new book introduces the core history of phenomenology and assesses its relevance to contemporary psychology, philosophy of mind, and cognitive science. From critiques of artificial intelligence research programs to ongoing work on embodiment and enactivism, the authors trace how phenomenology has produced a valuable framework for analyzing cognition and perception, whose impact on contemporary psychological and scientific research, and philosophical debates continues to grow. The first part of _An Introduction to Phenomenology_ is an extended overview of the history and development of phenomenology, looking at its key thinkers, focusing particularly on Husserl, Heidegger and Merleau-Ponty, as well as its cultural and intellectual precursors. In the second half Chemero and Käufer turn their attention to the contemporary interpretations and uses of phenomenology in cognitive science, showing that phenomenology is a living source of inspiration in contemporary interdisciplinary studies of the mind. Käufer and Chemero have written a clear, jargon-free account of phenomenology, providing abundant examples and anecdotes to illustrate and to entertain. This book is an ideal introduction to phenomenology and cognitive science for the uninitiated, as well as for philosophy and psychology students keen to deepen their knowledge

This article is about a sidebar in James Gibson's last book, The ecological approach to visual perception. In this sidebar, Gibson, the founder of the ecological perspective of perception and action, argued that to perceive an affordance is not to classify an object. Although this sidebar has received scant attention, it is of great significance both historically and for recent discussions about specificity, direct perception, and the functions of the dorsal and ventral streams. It is argued that Gibson's acknowledgment of Wittgenstein's ideas of classification suggests a limited scope of his theory of direct perception?it cannot account for the classification of objects. The implications for both the specification debate and theorizing about the brain's dorsal and ventral pathways are explored. Based on a recent ecological conception of information and direct perception, we ultimately argue that both affordance perception and classification are direct

Situated, embodied cognitive science is all the rage these days. Some (including the present author) have argued that situated, embodied cognitive science is incompatible with realism (metaphysical and scientific). In this paper, I argue that this is a mistake: there is no reason one cannot be both a proponent of situated, embodied cognitive science and a realist. To show this, I point to flaws in two previous arguments against realism. I also recommend a slightly modified version of Hacking’s entity realism as an especially good fit for situated, embodied cognitive science.

Several articles have recently appeared arguing that there really are no viable alternatives to mechanistic explanation in the biological sciences. This claim is meant to hold both in principle and in practice. The basic claim is that any explanation of a particular feature of a biological system, including dynamical explanations, must ultimately be grounded in mechanistic explanation. There are several variations on this theme, some stronger and some weaker. In order to avoid equivocation and miscommunication, in section 1 we will argue that mechanistic explanation is defined by localization and decomposition. In section 2 we will argue that systems neuroscience contains explanations that violate both localization and decomposition on any non-trivial construal of these concepts. Therefore, in section 3 we conclude the mechanistic model of explanation either needs to stretch to now include explanations wherein localization or decomposition fail, or acknowledge that there are counter-examples to mechanistic explanation in the biological sciences. We will also consider consequences and possible replies on the part of the mechanist in section 3.

The primary difference between direct and inferential theories of perception concerns the location of perceptual content, the meaning of our perceptions. In inferential theories of perception, these meanings arise inside animals, based upon their interactions with the physical environment. Light, for example, bumps into receptors causing a sensation. The animal (or its brain) performs inferences on the sensation, yielding a meaningful perception. In direct theories of perception, on the other hand, meaning is in the environment, and perception does not depend upon meaning- conferring inferences. Instead the animal simply gathers information from a meaning- laden environment. But if the environment contains meanings, then it cannot be merely physical. This places a heavy theoretical burden on direct theories of perception, a burden so severe that it may outweigh all the advantages to conceiving perception as