The first two E’s
The autopoietic roots of embedded and embodied cognition
According to the 4E approach, cognition cannot be explained as merely realised by the neural activity of the brain, but since cognitive states and processes are interdependent on the rest of the body and the environment outside the body, cognition should be regarded as embedded and embodied — while acknowledging the organism’s unity and subjectivity, cognition emerges through the brain-body-environment system.
For the moment, we’re going to overlook the problem of defining the nature of relations that would make brain, body and environment all parts of the cognitive system, possibly making this latter extended. This topic will be addressed more in details later on in a dedicated article. At his stage, we will just describe more in practice what would make cognition embedded and embodied (see previous article), and we will do so by drawing on the principles of autopoietic theory, which — in broad terms — can be considered partially coherent with the 4E point of view and is indeed often referred as original root of enactive approaches.
The autopoietic theory provides a definition of cognition which stems from the conception of ‘living being’ as autopoietic system, then conventionally attributing the status of cognitive entity to any system that exhibits autopoietic proprieties.
Maturana and Varela, in what is considered the first formulation of the theory [5], define living beings as autopoietic systems since they are physical systems that through self-contained proprieties and mechanisms of self-production, are able to maintain their delimited structure and organization, and regenerate themselves. According to the authors, the emergence of cognition would be a direct and necessary result of these autopoietic proprieties of living beings (and not exclusively to them) and on this basis, they outline a notion of cognition based on a naturalistic viewpoint, besides pointing out its consequent conventional facet.
In a nutshell, in naturalistic terms cognition essentially consists of two elements: structural coupling and adaptation — that is to say that cognition occurs when in the constant process of self-preservation, an organism exhibits a structural-operational congruence between its processes-organization and the environment in which it lives. As a consequence of this viewpoint, the authors associate the concept of cognition to an additional attribute, in this case of a conventional nature and essentially dependent on the observer [6]. Namely, we would detect a form of cognition every time we observe an organism and we consider its behaviour adequate to the current circumstances, attributing to the entity the capacity of sensing and knowing the environment.
Some 4e approaches, seems to have exasperated the autopoietic viewpoint interpreting the processes of coupling and adaptation — then cognition — as intrinsic proprieties of the autopoietic systems, attributing to these latter an inherent natural subjectivity in the sense-making process. However, postponing the important and controversial discussion about the (ontological) subtext behind this position, we can say that two of the key attributes (E’s) of the 4E approach lies on this autopoietic definition of cognition. Namely, any cognitive phenomenon or intelligent behaviour, as well as physical-chemical metabolic processes, has to be included in the broad set of self-regulative activities of the body interacting with the environment [4]. As Friston clearly summarises — autopoietic systems are “structurally coupled” with their medium, embedded in a dynamic of changes that can be recalled as sensory-motor coupling- [1]
From a 4E perspective, this means that any phenomenon ascribable to the agent, cognitive abilities and intelligent behaviours included, has to be seen as inherent to the agent’s interaction and interdependency with the environment — thus embedded — and that this interaction is conducted inevitably through the body and its structural-functional features — thus embodied –
In other words — it follows that cognition, even in its broad sense of knowing-understanding, is the continuous process of sense-making intrinsic to the constant bodily interaction with the environment, and it cannot reflect any form of self-reliant faculty, secluded from this triadic relationship. Rather, this ability to understand-know the environment has to be seen as always situated and contextual, which means that is a skill determined by the organizational and structural proprieties of the agent’s body, therefore embodied and embedded. [2]
A typical example used to illustrate this definition of cognition at a simple level is bacteria swimming up through a gradient of glucose[3]. The ability of bacteria to understand the environment, considering sugar as “food” (chemotaxis), arises from their adaptive autonomous dynamics which involves an operational relationship with this specific chemical substance for self-maintenance needs. In other words, the bacteria ability to sense and making-sense-of the environment depends on the limited range of operations allowed by its physical-chemical proprieties (body) in relation to that specific environment.
This viewpoint necessarily leads to the rejection or reassessment of any form of internalism, as primarily proposed by traditional cognitive science, therefore, of representational models of cognition (as mentioned in previous articles). However, the fact that the embodied-embedded position can be defended as self-evident observation and pragmatic conclusion, it is not enough to justify a scientific theory of mind based on this view. Indeed, in order to formalise such a notion of cognition it is necessary to agree about the nature of embedment-embodiment, especially about the relationship between cognitive system and environment. For instance, as the issue is often addressed, we need to clarify whether the relation of interdependence that has been outlined above should be acknowledged in constitutive terms or as being of a causal nature, whether the problem should be addressed in radical-naturalistic terms, conventional-pragmatic terms or maybe only in statistical terms. In the forthcoming articles, we will try to provide an overview of the most interesting contributions and proposals on this subject.
[1] Allen M, Friston KJ. From cognitivism to autopoiesis: towards a computational framework for the embodied mind. Synthese. 2018 Jun 1;195(6):2459–82.
[2] Di Paolo E, Thompson E. The enactive approach. The Routledge handbook of embodied cognition 2014 Apr 29 (pp. 86–96). Routledge.
[3] Evan T. Mind in life: Biology, Phenomenology, and the Sciences of Mind. 2007.
[4] Maturana HR. Autopoiesis, structural coupling and cognition: a history of these and other notions in the biology of cognition. Cybernetics & human knowing. 2002 Mar 1;9(3–4):5–34.
[5] Maturana HR, Varela FJ. Autopoiesis: The organization of the living. Autopoiesis and cognition: The realization of the living. 1980;42:59–138.
[6] Villalobos M, Palacios S. Autopoietic theory, enactivism, and their incommensurable marks of the cognitive. Synthese. 2019 Sep 3:1–7.
