I assess the relevance of John Dewey’s well-known article, “The reflex arc concept in psychology”, for a historical revision of the emergence of the recent embodied approach in the cognitive sciences. In particular, I try to identify its specific contribution in the shift from Dewey’s conceptual analysis to the way in which, during recent years, certain research programs have developed their methodological profile and put it to work in experimental and modeling practices. My hypothesis is that, under a certain interpretation, Dewey’s article plays the role of the main intellectual precursor in the development of embodied cognitive science and, specifically, the related dynamical approach. Relevance: The article assesses the historical roots of the recent group of research programs in “embodied cognitive science,” with particular attention to the underlying shift in methodological profile. Among the conglomerate of programs that can be included within the general denomination of embodied cognitive science, the article focuses on the dynamical approach, which stands out, on the one hand, for its radical character vis-à-vis classical approaches and, on the other hand, for its characteristic brand of cognitive-scientific approach.

In the enactive paradigm of cognitive science, development plays a crucial role in the realization of cognition. This position runs counter to the computational functionalism upon which cognitivist and classical artificial intelligence systems are founded, especially the assumption that cognition can be achieved by embedding pre-formed knowledge. The enactive stance involves a progressive phased transition from cognitive capacity to cognitive capability, highlighting the role of development in extending the timescale of a cognitive agent’s prospective abilities and in expanding its repertoire of effective action. We review briefly some necessary conditions for cognitive development, drawing on examples from developmental psychology, illustrating the ideas by looking at the ontogenesis of instrumental helping and collaboration in infants, and identifying some of the essential elements of a developmental cognitive architecture. We then focus on the fact that enactive systems are operationally-closed, autonomous, and self-maintaining. Consequently, there are organizational constitutive processes at play as well as behavioural ones. Reconciling these complementary processes poses a significant challenge for the creation of complete model of development that must show how constitutive autonomy is compatible with and may even give rise to behavioural autonomy. We conclude by drawing attention to recent research which could provide a way of addressing this challenge.

Key words: behavioural autonomy cognitive architecture constitutive autonomy development enaction ontogeny phylogeny value systems

The reciprocal coupling of perception and action in cognitive agents has been firmly established: perceptions guide action but so too do actions influence what is perceived. While much has been said on the implications of this for the agent’s external behavior, less attention has been paid to what it means for the internal bodily mechanisms which underpin cognitive behavior. In this article, we wish to redress this by reasserting that the relationship between cognition, perception, and action involves a constitutive element as well as a behavioral element, emphasizing that the reciprocal link between perception and action in cognition merits a renewed focus on the system dynamics inherent in constitutive biological autonomy. Our argument centers on the idea that cognition, perception, and action are all dependent on processes focussed primarily on the maintenance of the agent’s autonomy. These processes have an inherently circular nature – self-organizing, self-producing, and self-maintaining – and our goal is to explore these processes and suggest how they can explain the reciprocity of perception and action. Specifically, we argue that the reciprocal coupling is founded primarily on their endogenous roles in the constitutive autonomy of the agent and an associated circular causality of global and local processes of self-regulation, rather than being a mutual sensory-motor contingency that derives from exogenous behavior. Furthermore, the coupling occurs first and foremost via the internal milieu realized by the agent’s organismic embodiment. Finally, we consider how homeostasis and the related concept of allostasis contribute to this circular self-regulation.

The strong version of the life-mind continuity thesis claims that mind can be understood as an enriched version of _the same_ functional and organizational properties of life. Contrary to this view, in this paper I argue that mental phenomena offer distinctive properties, such as intentionality or representational content, that have no counterpart in the phenomenon of life, and that must be explained by appealing to a different level of functional and organizational principles. As a strategy, and following Maturana’s autopoietic theory of cognition, I introduce a conceptual distinction between mind and cognition. I argue that cognition corresponds to the natural behaviour that every living being exhibits in the realization of its existence, and that, viewed in that way, cognition is a dynamic process of structural coupling that, unlike mental phenomena, involves no representational contents. On the basis of this distinction, I try to show that while life suffices for cognition, it does not suffice for mind. That is, that the strong continuity is not between life and mind but between life and cognition.

Key words: Life-mind continuity thesis, autopoiesis, mind, cognition

This thesis is about the ontology of living beings as natural systems, their behavior, and the way in which said behavior, under special conditions of social coupling, may give rise to mental phenomena. The guiding questions of the thesis are: 1) What kinds of systems are living beings such that they behave the way they do? 2) How, through what kinds of mechanisms and processes, do living beings generate their behavior? 3) How do mental phenomena appear in the life of certain living beings? 4) What are the natural conditions under which certain living beings exhibit mental phenomena? To answer these questions the thesis first assumes, then justifies and defends, a Strict Naturalistic (SN) stance with respect to living beings. SN is a metaphysical and epistemological framework that, recognizing the organizational, dynamic and structural complexity and peculiarity of living beings, views and treats them as metaphysically ordinary natural systems; that is, as systems that, from the metaphysical point of view, are not different in kind from rivers or stars. SN holds that if in natural sciences rivers and stars are not conceived as semantic, intentional, teleological, agential or normative systems, then living beings should not be so conceived either. Having assumed SN, and building mainly on the second-order cybernetic theories of Ross Ashby and Humberto Maturana, the thesis answers question 1) by saying that living beings are (i) adaptive dynamic systems, (ii) deterministic machines of closed transitions, (iii) multistable dissipative systems, and (iv) organizationally closed systems with respect to their sensorimotor and autopoietic dynamics. Based on this ontological characterization, the thesis answers question 2) by showing that living beings’ behavior corresponds to the combined product of (i), (ii), (iii) and (iv). Points (i) and (ii) support the idea that living beings are strictly deterministic systems, and that, consequently, notions such as information, control, agency or teleology—usually invoked to explain living beings’ behavior—do not have operational reality but are rather descriptive projections introduced by the observer. Point (iii) helps to understand why, despite their deterministic nature, living beings behave in ways that, to the observer, appear to be teleological, agential or “intelligent”. Point (iv) suggests that living beings’ sensorimotor dynamics are closed circuits without inputs or outputs, where the distinction between external and internal medium is, again, an ascription of the observer rather than a functional property of the system itself. Having addressed the basic principles of living beings’ behavior, the thesis explores the possible origin of (truly) mental phenomena in the particular domain of social behavior. Complementing Maturana’s recursive theory of language with Vygotsky’s dialectic approach the thesis advances, though in a still quite exploratory way, a sociolinguistic hypothesis of mind. This hypothesis answers questions 3) and 4) by claiming that the essential properties of mental phenomena (intentionality, representational content) appear with language, and that mind, as a private experiential domain, emerges as a dialectic transformation of language.

Traditionally, computational theory (CT) and dynamical systems theory (DST) have presented themselves as opposed and incompatible paradigms in cognitive science. There have been some efforts to reconcile these paradigms, mainly, by assimilating DST to CT at the expenses of its anti-representationalist commitments. In this paper, building on Piccinini’s mechanistic account of computation and the notion of functional closure, we explore an alternative conciliatory strategy. We try to assimilate CT to DST by dropping its representationalist commitments, and by inviting CT to recognize the functionally closed nature of some computational systems.

Key words: computational theory, dynamical systems theory, cognitive science, representation, functional closure.

In cognitive science, computationalism is the thesis that natural cognitive systems are computing systems. Traditionally, computationalism has understood computing and cognitive systems as functionally open systems, i.e., as systems that have functional entries through which they receive inputs, and exits through which they emit outputs. In opposition to this view, enactive theory claims that natural cognitive systems, unlike computing systems, are autonomous systems whose functional organization does not have inputs and outputs. Computationalism and enactivism seem to share an assumption that computing systems are input-output functional systems. In this paper, such an assumption will be critically reviewed by appealing to the cybernetic notion of functional closure. The notion of functional closure, as elaborated in Maturanas cybernetic neurophysiology, refers to a closed functional network in which, due to the circularity of the dynamics, we cannot distinguish inputs and outputs as intrinsic functional properties of the system. On the basis of this conceptualization, it will be argued that some paradigmatic cases of computing systems (notably a physically realized Turing machine) are actually functionally closed systems, and therefore computing systems without inputs and outputs. If this analysis is right, then the incompatibility that enactivists see between computing systems and organizationally closed functional systems would no longer hold, as it would not be true that computing systems must necessarily be understood as input-output systems.

Key words: computationalism enactivism functional closure turing machine input-output system

Recently, Michael Wheeler (2017) has argued that despite its sometimes revolutionary rhetoric, the so called 4E (embodied-embedded-enacted-extended) cognitive movement, even in the guise of ‘radical’ enactivism (REC), cannot achieve a full revolution in cognitive science. A full revolution would require the rejection of two essential tenets of traditional cognitive science, namely internalism and representationalism. Whilst REC might secure antirepresentationalism, it cannot do the same, so Wheeler argues, with externalism. In this paper, expanding on Wheeler’s analysis (2017), we argue that what compromises REC’s externalism is the persistence of cognitively relevant asymmetries between its purported cognitive systems and the environment. Complementarily, we argue that an antirepresentationalist ancestor of enactivism, the autopoietic theory of cognition, is able to deliver and secure externalism, thus offering the explosive combination (i.e., antirepresentationalism plus externalism) that Wheeler claims us needed for a revolution in cognitive science.

Maturana’s cognitive perspective on the living state, Dretske’s insight on how information theory constrains cognition, the Atlan/Cohen cognitive paradigm, and models of intelligence without representation, permit construction of a spectrum of dynamic necessary conditions statistical models of signal transduction, regulation, and metabolism at and across the many scales and levels of organisation of an organism and its context. Nonequilibrium critical phenomena analogous to physical phase transitions, driven by crosstalk, will be ubiquitous, representing not only signal switching, but the recruitment of underlying cognitive modules into tunable dynamic coalitions that address changing patterns of need and opportunity at all scales and levels of organisation. The models proposed here, while certainly providing much conceptual insight, should be most useful in the analysis of empirical data, much as are fitted regression equations.

Key words: cognition, living state, mutual information, phase transition, rate distortion

Radical enactivism (REC) and similar embodied and enactive approaches to the mind deny that cognition is fundamentally representational, skull-bound and mechanistic in its organisation. In this article, I argue that modellers may still adopt a mechanistic strategy to produce explanations that are compatible with REC. This argument is scaffolded by a multi-agent model of the true slime mould Physarum polycephalum.

Key words: minimal cognition, radical enactivism, mechanistic explanation