Minimal Cognition

The issue of minimal cognition or basic mentality concerns the elementary ingredients of cognitive processes.  Within the larger discourse of enactive and embodied cognition, a current of research has emerged endorsing the radical proposal that basic minds neither represent nor compute, and, moreover, that the cognitive processes peculiar to them tend heavily (if not constitutively) to be both world-involving and to incorporate extra-neural bodily factors.  Such a stance removes certain obstacles to a naturalistic view of basic minds and, at the same time, is more consistent with the idea of there being a deep continuity between life and mind.  Here I suggest that the phenomenological tradition has resources for bolstering the case for radicalism about basic minds.  Although the classical phenomenology of Husserl and Heidegger may be amenable to some form of representationalism, I propose a phenomenological corrective by appealing to the work of Levinas and Merleau-Ponty.  Levinas ardently criticizes the representationalism of early phenomenology and places in its stead a non-representational account of “sensibility.”  Merleau-Ponty, in addition, with his theory of (non-semantic) sense gives us a way of understanding basic minds synergistically, in terms of the perceiving organism’s embodied interactions with its surroundings, in a way supportive of the radical idea that basic minds are self-organizing dynamical systems.

Living systems employ several mechanisms and behaviors to achieve robustness and maintain themselves under changing internal and external conditions. Regulation stands out from them as a specific form of higher-order control, exerted over the basic regime responsible for the production and maintenance of the organism, and provides the system with the capacity to act on its own constitutive dynamics. It consists in the capability to selectively shift between different available regimes of self-production and self-maintenance in response to specific signals and perturbations, due to the action of a dedicated subsystem which is operationally distinct from the regulated ones. The role of regulation, however, is not exhausted by its contribution to maintain a living system’s viability. While enhancing robustness, regulatory mechanisms play a fundamental role in the realization of an autonomous biological organization. Specifically, they are at the basis of the remarkable integration of biological systems, insofar as they coordinate and modulate the activity of distinct functional subsystems. Moreover, by implementing complex and hierarchically organized control architectures, they allow for an increase in structural and organizational complexity while minimizing fragility. Finally, they endow living systems, from their most basic unicellular instances, with the capability to control their own internal dynamics to adaptively respond to specific features of their interaction with the environment, thus providing the basis for the emergence of minimal forms of cognition.

Within the cognitive sciences, cognition tends to be interpreted from an anthropocentric perspective, involving a stringent set of human capabilities. Instead, we suggest that cognition is better explicated as a much more general biological phenomenon, allowing the lower bound of cognition to extend much further down the phylogenetic scale. We argue that elementary forms of cognition can already be witnessed in prokaryotes possessing a functional sensorimotor analogue of the nervous system. Building on a case-study of the Escherichia coli bacterium and its sensorimotor system, the TCSTsystem, we home in on the characteristics of minimal cognition, and distinguish it from more basic forms of ontogenetic adaptation. In our view, minimal cognition requires an embodiment consisting of a sensorimotor coupling mechanism that subsumes an autopoietic organization; this forms the basis of the growing consensus that the core of cognition revolves around sensorimotor coupling. We discuss the relevance of our interpretation of minimal cognition for the study of cognition in general.

Published in BioSystems (2016), 148, 12-21.

Co-authored with Alvaro Moreno

In this paper we address the question of minimal cognition by investigating the origin of some crucial cognitive properties from the very basic organisation of biological systems. More specifically, we propose a theoretical model of how a system can distinguish between specific features of its interaction with the environment, which is a fundamental requirement for the emergence of minimal forms of cognition. We argue that the appearance of this capacity is grounded in the molecular domain, and originates from basic mechanisms of biological regulation. In doing so, our aim is to provide a theoretical account that can also work as a possible conceptual bridge between Synthetic Biology and Artificial Intelligence. In fact, we argue, Synthetic Biology can contribute to the study of minimal cognition (and therefore to a minimal AI), by providing a privileged approach to the study of these mechanisms by means of artificial systems.

Contemporary plant biologists take seriously the notion of plant cognition or intelligence (Gross 2016; Trewavas 2016), yet there is still no consensus on how these abilities are to be described (e.g., Calvo 2016). Most contentious has been the formulation of the intelligent behaviour of plants as ‘neurological’ (ibid; Alpi et al. 2007). I argue that there is no consensus because there is yet to be a principled way to distinguish between the adaptive behaviours plants display (Cvrčková et al. 2009, 395). What are needed to settle this matter are principled distinctions for the kinds of behaviour attributed to plants. This can be achieved by the application of criteria that generally delineate the cognitive, and that discriminate between kinds of cognitive abilities. To this end, I use an evolutionary taxonomy of cognitive control developed by Kim Sterelny (2003) to establish principled grounds upon which the plant cognition debate can be resolved.

This paper aims to elucidate a kind of ignorance that is more fundamental than a momentary lack of information, but also not a kind of ignorance that is built into the subject's cognitive apparatus such that the subject can't do anything about it (e.g. color blindness). The paper sets forth the notion of cognitive confinement, which is a contingent, yet relatively stable state of being structurally or systematically unable to gain information from an environment, determined by patterns of interaction between the subject and the world. In order to unpack the idea of cognitive confinement the paper discusses niche construction theory, and then, in greater detail, the notion of cognitive niche once proposed by John Tooby and Irven DeVore. Cognitive confinement is here imagined as a pathologized form of cognitive niche. This posit is substantiated by referring to a case that has come to the fore in recent years and raised debate around the world: the rise of so-called filter bubbles. They turn out to be instantiations of a more general phenomenon of cognitive confinement.

This paper analyses conceptual and experimental work in synthetic biology on different types of interactions considered as minimal examples or models of communication. It discusses their pertinence and relevance for the wider understanding of this biological and cognitive phenomenon. It critically analyses their limits and it argues that a conceptual framework is needed. As a possible solution, it provides a theoretical account of communication based on the notion of organisation, and characterised in terms of the functional influence exerted by the sender upon the receiver. It shows that this account can be operationalised in synthetic biology, and that it can supply criteria and guidelines for the design and evaluation of synthetic models.

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

According to the sensorimotor approach, perception is a form of embodied know-how, constituted by lawful regularities in the sensorimotor flow or in sensorimotor contingencies (SMCs) in an active and situated agent. Despite the attention that this approach has attracted, there have been few attempts to define its core concepts formally. In this paper, we examine the idea of SMCs and argue that its use involves notions that need to be distinguished. We introduce four distinct kinds of SMCs, which we define operationally.These are the notions of sensorimotor environment (open-loop motor-induced sensory variations), sensorimotor habitat (closed-loop sensorimotor trajectories), sensorimotor coordination (reliable sensorimotor patterns playing a functional role), and sensorimotor strategy (normative organization of sensorimotor coordinations). We make use of a minimal dynamical model of visually guided categorization to test the explanatory value of the different kinds of SMCs. Finally, we discuss the impact of our definitions on the conceptual development and empirical as well as model-based testing of the claims of the sensorimotor approach.

A current challenge in neuroscience and systems biology is to better understand properties that allow organisms to exhibit and sustain appropriate behaviours despite the effects of perturbations (behavioural robustness). There are still significant theoretical difficulties in this endeavour, mainly due to the context-dependent nature of the problem. Biological robustness, in general, is considered in the literature as a property that emerges from the internal structure of organisms, rather than being a dynamical phenomenon involving agent-internal controls, the organism body, and the environment. Our hypothesis is that the capacity for behavioural robustness is rooted in dynamical processes that are distributed between agent ‘brain’, body, and environment, rather than warranted exclusively by organisms’ internal mechanisms. Distribution is operationally defined here based on perturbation analyses. Evolutionary Robotics (ER) techniques are used here to construct four computational models to study behavioural robustness from a systemic perspective. Dynamical systems theory provides the conceptual framework for these investigations. The first model evolves situated agents in a goal-seeking scenario in the presence of neural noise perturbations. Results suggest that evolution implicitly selects neural systems that are noise-resistant during coupling behaviour by concentrating search in regions of the fitness landscape that retain functionality for goal approaching.The second model evolves situated, dynamically limited agents exhibiting minimal cognitive behaviour (categorization task). Results indicate a small but significant tendency toward better performance under most types of perturbations by agents showing further cognitive behavioural dependency on their environments. The third model evolves experience-dependent robust behaviour in embodied, one-legged walking agents. Evidence suggests that robustness is rooted in both internal and external dynamics, but robust motion emerges always from the system-in-coupling. The fourth model implements a historically dependent, mobile-object tracking task under sensorimotor perturbations. Results indicate two different modes of distribution, one in which inner controls necessarily depend on a set of specific environmental factors to exhibit behaviour, then these controls will be more vulnerable to perturbations on that set, and another for which these factors are equally sufficient for behaviours. Vulnerability to perturbations depends on the particular distribution. In contrast to most existing approaches to the study of robustness, this thesis argues that behavioural robustness is better understood in the context of agent-environment dynamical couplings, not in terms of internal mechanisms. Such couplings, however, are not always the full determinants of robustness. Challenges and limitations of our approach are also identified for future studies.

This special issue highlights the work of some recent participants of a series of Minimal Cognition workshops held at the University of Wollongong in 2018. The goal of these workshops has been to showcase the interdisciplinary work being done in what might be called 'minimal cognition research.' Our aim was, and continues to be, to bring biologists, cognitive scientists, philosophers, and computer scientists into dialogue on how we might understand the minimal criteria for cognition and cognitive behavior. This special issue, titled "Approaching Minimal Cognition," is intended to open up further exchange between the different fields converging on this richly interesting area of research, and to spur debate and discussion about the roots of cognition, the indicators of cognitive behavior, collective cognition, and the possibility of life-mind continuity.

This article provides an answer to the question: What is the function of cognition? By answering this question it becomes possible to investigate what are the simplest cognitive systems. It addresses the question by treating cognition as a solution to a design problem. It defines a nested sequence of design problems: (1) How can a system persist? (2) How can a system affect its environment to improve its persistence? (3) How can a system utilize better information from the environment to select better actions? And, (4) How can a system reduce its inherent informational limitations to achieve more successful behavior? This provides a corresponding nested sequence of system classes: (1) autonomous systems, (2) (re)active autonomous systems, (3) informationally controlled autonomous systems (autonomous agents), and (4) cognitive systems.

This article provides the following characterization of cognition: The cognitive system is the set of mechanisms of an autonomous agent that (1) allow increase of the correlation and integration between the environment and the information system of the agent, so that (2) the agent can improve the selection of actions and thereby produce more successful behavior.

Finally, it shows that common cognitive capacities satisfy the characterization: learning, memory, representation, decision making, reasoning, attention, and communication.

Theoretical discussions and computational models of bio-inspired embodied and situated agents are introduced in this article capturing in simplified form the dynamical essence of robust, yet adaptive behavior. This article analyzes the general problem of how the dynamical coupling between internal control (brain), body and environment is used in the generation of specific behaviors. Based on the Evolutionary Robotics (ER) paradigm, four computational models are described to support discussions including descriptions on performance after a series of structural, sensorimotor or mutational perturbations, or are developed in the absence of them. Experimental results suggest that ‘dynamic determinacy’ – i.e. the continuous presence of a unique dynamical attractor that must be chased during functional behaviours – is a common dynamic phenomenon in the analyzed robust and adaptive agents. These agents show dynamical states that are definitely and unequivocally characterized via transient dynamics toward a unique, yet moving attractor at neural level for coherent actions. This determinacy emerges as a control strategy rooted on behavioral couplings and relies on mechanisms that are distributed on brain, body and environment. Different ways to induce further distribution of behavioral mechanisms are also discussed in this paper from a bio-inspired ER perspective.

Boundaries are prominent ingredients of reality, including-most importantly-the boundaries of organisms and the perceived boundaries of things (their surfaces). It is also customary to think of minds as kinds of bounded loci for thoughts, representations and other internal entities, targeting the borderline between the internal domain and the external world as a genuine barrier. Therefore, not surprisingly, boundaries and surfaces have become targets of disciplined formal-ontological investigations. However, to have a boundary (or surface) is not only a purely topographical feature, for boundaries play certain roles. For that reason, this paper unpacks boundaries in general, and surfaces in particular, in terms of interactions between structured entities, most importantly-living creatures capable of (minimal) cognition. It is argued that boundaries should be thought of as products of, and at the same time as critical constraints imposed on structural coupling between complex beings, including the interactions between the (minimally) cognitive subjects and their surroundings.

This paper criticizes the role of intuition-based ascriptions of cognition that are closely related to the ascription of mind. This practice hinders the explication of a clear and stable target domain for the cognitive sciences. To move forward, the proposal is to cut the notion of cognition free from such ascriptions and the intuition-based judgments that drive them. Instead, cognition is reinterpreted and developed as a scientific concept that is tied to a material domain of research. In this reading, cognition becomes a changeable theoretical concept that can and must be adapted to the findings within this target domain. Taking humans as the best-established existing example of the relevant material target domain, this central case is extended to include all living systems. To clarify what it is about living systems that warrants their role as cognitive target domain, the new concept of cobolism is introduced as a complement to metabolism. Cobolism refers to the systematic ways in which each living system encompasses structures, processes and external events that maintain the fundamental metabolic processes that constitute the core of each living system. Cobolism is perfectly general, applies to bacterial and human cases alike, and provides a general format to describe wildly different cognitive organizations. It provides a clear target for the cognitive sciences to work on, turning them into what we can call the cognitive life sciences. Keywords Cognition · Cognitive science · Mark of the cognitive · Cognitive domain · Biocognition · Cognitive biology · Cognitive life sciences B Fred Keijzer f.a.keijzer@rug.nl

A current challenge in neuroscience, systems and theoretical biology is to understand what properties allow organisms to exhibit and sustain behaviours despite perturbations (behavioural robustness). Indeed, there are still significant theoretical difficulties in this endeavour due to the context-dependent nature of the problem. Contrary to the common view of biological robustness as a phenomenon that emerges internally, this article discusses the hypothesis that behavioural robustness is rooted in dynamical processes that distribute between internal controls, the organism body and the environment. This review highlights the varied perspectives and how they have led to the current focus on robustness as a relational phenomenon. A new perspective is proposed in which robustness is better understood in the context of agent-environment dynamical couplings, in which such couplings are not always the full determinants of robustness. The challenges and limitations of the proposed approach are identified.

There is a growing trend in the cognitive sciences to conceive of cognitive behavior as being distributed across brain, body and environment. However, the implications of such distribution for our understanding of biological robustness, which so far has been related to individual-based mechanisms alone, has rarely been discussed in the literature. We used the Evolutionary Robotics technique to examine the relationship between distributed behavioral mechanisms and behavioral robustness. Two kinds of model agents were evolved for a mobile object-tracking task and tested to see whether they can sustain their behavior despite sensorimotor perturbations. The results indicate that a highly distributed realization of behavior can be (i) detrimental, if it is mostly based on factors that are necessary for the behavior, or (ii) beneficial, if it is mostly based on factors that are sufficient for the behavior. Accordingly, we suggest that future discussions of distributed cognition should take into account that there are at least two different possible modes of realizing distributed behavior and that these have a qualitatively different effect on behavioral robustness.

Fungal organisms can perceive the outer world in a way similar to what animals sense. Does that mean that they have full awareness of their environment and themselves? Is a fungus a conscious entity? In laboratory experiments we found that fungi produce patterns of electrical activity, similar to neurons. There are low and high frequency oscillations and convoys of spike trains. The neural-like electrical activity is yet another manifestation of the fungal intelligence. In this paper we discuss fungal cognitive capabilities and intelligence in evolutionary perspective, and question whether fungi are conscious and what does fungal consciousness mean, considering their exhibiting of complex behaviours, a wide spectrum of sensory abilities, learning, memory and decision making. We overview experimental evidences of consciousness found in fungi. Our conclusions allow us to give a positive answer to the important research questions of fungal cognition, intelligence and forms of consciousness.

This article investigates the emergence of robust behaviour in agents with dynamically limited controllers (monostable agents), and compares their performance to less limited ones (bistable agents). ‘Dynamically limited’ here refers to a reduced quantity of steady states that an agent controller exhibits when it does not receive stimulus from the environment. Agents are evolved for categorical perception, a minimal cognitive task, and must correlate approaching or avoiding movements based on (two) different types of objects. Results indicate a significant tendency to better behavioural robustness by monostable in contrast to bistable agents in the presence of sensorimotor, mutational, and structural perturbations. Discussions here focus on a further dependence to coupled dynamics by the former agents to explain such a tendency.