Ecole 2006 : du 29 mai au 3 juin 2006 - Ile d'Oléron (France)
There are two basic requirements for any paradigm in cognitive science: it must provide a genuine resolution of the Mind-Body problem ; and it must provide for a genuine core articulation between a multiplicity of disciplines, at the very least between psychology, linguistics and neuroscience. The paradigm of enaction provides a solution to these problems by deploying the basic equation "Cognition = Life = Autopoiesis". This research programme is grounded in sensory-motor dynamics; but it is in no way confined to "lower-level" cognition, since the history of life on Earth includes hominisation and continues up to the present day.
First generation distributed cognition noted that the language of classical cognitive science's computational metaphor maps well onto the operation of certain classes of socio-technical systems (see "Cognitive Consequences of Information flow"), but that it fails as a model of individual cognition (see Cognition in the Wild, Chapter 9). Over the past two decades, cognitive science has been shifting from a concept of cognition as a logical process to cognition as a biological phenomenon. Second generation distributed cognition follows this shift and focuses on the nature of embodied activity situated in social and material context (see "Material anchors for conceptual blends"). This re-examination of human organism-environment relations brings to light new classes of adaptive mechanisms and new understandings of the dynamics underlying the operation of distributed cognitive systems (see "The cognitive life of things").
In cognitive science, we currently witness a "pragmatic turn" away from the traditional representation-centered framework towards a paradigm based on the notions of ?situatedness" and ?embodiment", which focusses on understanding the relevance of cognition for action, and the real-world interactions of the brain. Such an "action-oriented" paradigm has earliest and most explicitly been developed in robotics, and has only recently begun to gain impact on cognitive psychology and neurobiology. The basic concept is that cognition should not be understood as a capacitiy of deriving world-models, which then might provide a "database" for thinking, planning and problem-solving. Rather, it is emphasized that cognitive systems are always engaged in contexts of action that require fast selection of relevant information and constant sensorimotor exchange. In the context of such an action-oriented conceptual framework, investigation of the intrinsic dynamics of neural circuits becomes increasingly important. There is ample evidence that the processing of stimuli is controlled by top-down influences that strongly shape the dynamics of thalamocortical networks and constantly create predictions about forthcoming sensory events. Therefore, perceptual processing is increasingly considered as being active and highly selective in nature. I want to discuss recent neurobiological evidence supporting this ?pragmatic turn" and the implications of this view for future research strategies in cognitive neuroscience. Moreover, I will discuss recent evidence suggesting that the temporal structure of neural activity patterns may be particularly relevant to understanding selection and top-down processing in the context of action.
My contribution will bring together two programmes of research. Originally developed independently of one another but with time have come to cohere with one broader theoretical framework. The first is that presented in my book "The Representational and the Presentational" of 1993, the second involves a monograph on the phenomenology of human consciousness that I am writing now. The earlier book is devoted to a comprehensive, and radical, critique of the representational-computational paradigm in cognitive science. According to it, cognition is conceived by means of computational processes applied upon underlying mental representations. My critique is based, on the one hand, on comprehensive empirical inspection of practically all areas of human cognitive behavior, and on the other hand, on conceptual theoretical analysis. The critique highlights a series of factors (or parameters) that I think are both central and basic for human cognition, which the representational paradigm either ignores or regards as merely secondary. When the primary import of these factors is appreciated, one reaches the conclusion that rather than being the basis for cognition, representations are the products of cognitive activity, and that the basic capability of mind is not information processing and symbol manipulation but rather being and acting in the world. With this, the focus of psychological science shifts from the domain of the unconscious to that of the conscious, and instead of process modeling, cognitive theory should adopt a perspective of phenomenology of understanding.
The foregoing conclusions led to my study of consciousness. I have been engaged with the study of the phenomenology of human consciousness for about 25 years now, long before the topic was popular or even considered legitimate for serious investigation. I shall present several lines of such inquiry, all of which come together in the book I am writing at present. These lines concern thought sequences, the systematic typology of experience, and a novel approach for the study of non-ordinary states of consciousness. Together, these have brought me to the conceptualization of what I refer to as a general theory of consciousness, one that encompasses both ordinary and non-ordinary states of mind. A sketch of his theory will be outlined.
The two projects noted entail radical changes in the conceptualization of what psychology is. These changes have to do with alternative types of theory, alternative topics for research, and alternative methods for both empirical research and theory construction.
What is the enactive approach to cognition? Over the last 15 years this banner has grown to become a respectable alternative to traditional frameworks in cognitive science. It is at the same time a label with different interpretations and upon which different doubts have been cast. In this talk, I will elaborate on the core ideas of the enactive approach and their implications. Five inter-related ideas define enactivism: autonomy, sense-making, emergence, embodiment, and experience. I will show how these are coherent, radical and very powerful concepts that establish clear conceptual and methodological guidelines for research. At the same time, the framework is not a closed one and I will look at the problems that arise from taking these ideas seriously. I will show that the enactive approach has plenty of room for elaboration in many different areas and many challenges to respond to. In particular, the conceptual problems surrounding many theories on value-appraisal and value-generation will be explored. I will argue that the enactive view takes the problems of meaning and value very seriously and elaborates a proper scientific alternative to reductionist attempts to tackle these issues by functional localization. Another area where the enactive framework can make a significant contribution is social interaction and social undertanding. The legacy of computationalism and methodological individualism is very strong in this field. Enactivism allows us to see embodied social interaction and coordination at many different levels in an integrated manner, from the emergence of autonomous temporal structures that regulate interaction to the generation of socially mediated meaning.
These discussions will be supported and illustrated with examples from work in evolutionary robotics. The need for synthetic minimal models and their scientific role will be a running theme of this talk. This will lead to interesting methodological implications.
Finally, I will also present some brief speculations about how enactivism could be the right tool to help us bridge knowledge of concrete embodied and situated practices and higher level human cognition. The language offered by this perspective already proves very useful in formulating the problem in a tractable manner. These speculations will centre around the role of play as an activity that allows the development of meaning-manipulation skills as well as a further level of autonomous cognitive self, one that is characteristic of human beings. For the enactive view play is seen as re-creation whereas for computationalism fun is a mystery.
Traditional cognitive science fails to take into account the dimensions of appearing to and being for of the mind. In the light of recent work, it seems that the time has come to move from the "naturalisation of the mind" to a more ambitious and more rigorous programme of research : naturalizing phenomenology. What is at issue is not simply to naturalize intentionality - usually interpreted in anglo-saxon philosophy of mind as a referential, denotational relation of the mind to something other than itself ( aboutness ) - but to naturalize phenomenology itself as a method of access to lived experience. This move implies and signals a switch from previous scientific research on the mind towards a science of consciousness as such.
In view of the interest but also the difficulties of this new programme, and in particular its tendency to neglect phenomenological reduction as a principled method and to ignore the distinction between static phenomenology and genetic phenomenology, this paper proposes a new reading of the major analyses of Husserlian phenomenology and to show how the latter recategorizes, that is to say takes up and displaces the question of the ontological constitution of cognition and the epistemological constitution of cognitive science. On this basis, I propose a new configuration for the possible relation between phenomenology and cognitive science, quite different from the naturalization of the first by the second. This proposition is based on two leading threads :