The theoretical foundations of enaction: Precariousness

Enaction is an increasingly influential approach to cognition that grew out of Maturana and Varela's earlier work on autopoiesis and the biology of cognition. As with any relatively new scientific discipline, the enactive approach would benefit greatly from a careful analysis of its theoretical foundations. Here we initiate such an analysis for one of the core concepts of enaction, precariousness. Specifically, we consider three types of fragility: systemic, processual and thermodynamic. Using a glider in the Game of Life as a toy model, we illustrate each of these fragilities and examine the relationships between them. We also argue that each type of fragility is characterized by which aspects of a system are hardwired into its definition from the outset and which aspects are emergent and hence vulnerable to disintegration without ongoing maintenance.

Effects of guided exploration on reaching measures of auditory peripersonal space

Despite the recognized importance of bodily movements in spatial audition, few studies have integrated action-based protocols with spatial hearing in the peripersonal space. Recent work shows that tactile feedback and active exploration allow participants to improve performance in auditory distance perception tasks. However, the role of the different aspects involved in the learning phase, such as voluntary control of movement, proprioceptive cues, and the possibility of self-correcting errors, is still unclear. We study the effect of guided reaching exploration on perceptual learning of auditory distance in peripersonal space. We implemented a pretest-posttest experimental design in which blindfolded participants must reach for a sound source located in this region. They were divided into three groups that were differentiated by the intermediate training phase: Guided, an experimenter guides the participant’s arm to contact the sound source; Active, the participant freely explores the space until contacting the source; and Control, without tactile feedback. The effects of exploration feedback on auditory distance perception in the peripersonal space are heterogeneous. Both the Guided and Active groups change their performance. However, participants in the Guided group tended to overestimate distances more than those in the Active group. The response error of the Guided group corresponds to a generalized calibration criterion over the entire range of reachable distances. Whereas the Active group made different adjustments for proximal and distal positions. The results suggest that guided exploration can induce changes on the boundary of the auditory reachable space. We postulate that aspects of agency such as initiation, control, and monitoring of movement, assume different degrees of involvement in both guided and active tasks, reinforcing a non-binary approach to the question of activity-passivity in perceptual learning and supporting a complex view of the phenomena involved in action-based learning.

Embodiment in online psychotherapy: A qualitative study

As a result of the COVID-19 pandemic, many therapists and patients have been required to switch to online sessions in order to continue their treatments. Online psychotherapy has become increasingly popular, and although its efficacy seems to be similar to face-to-face encounters, its capacity to support the implicit nonverbal and embodied aspects of the therapeutic relationship has been questioned and remains understudied.

OBJECTIVES

To study how embodied and intersubjective processes are modified in online psychotherapy sessions.

DESIGN

Taking the enactive concept of participatory sense-making as a guiding thread, we designed an interpretative phenomenological analysis to examine the experiences of embodiment in online therapy.

METHODS

We conducted phenomenological semi-structured interviews with patients and therapists who have recently switched from face-to-face encounters to online modality.

RESULTS

Adjustments in verbal and nonverbal behavior, gaze behavior, management of silences, and displacements of non-intentional and pre-reflective patterns onto reflective ones are reported as necessary to compensate for changes introduced in the online modality.

CONCLUSIONS

From an enactive perspective, such adaptations manifest regulatory processes aimed at sustaining interactive dynamics and coordinating the primordial tension between relational and individual norms in social encounters.

PRACTITIONER POINTS

We examine different aspects of embodiment that practitioners should take into account when switching from face-to-face to online encounters with their clients. Online communication systems can alter aspects of the therapeutic relationship, such as its structure, its fragility, and its significance. Video calls afford new forms of intervention such as integrating the experience of patients with their self-image, incorporating information about their habitual environment into the process, and adopting less confrontational therapeutic styles.

Placebo From an Enactive Perspective

Due to their complexity and variability, placebo effects remain controversial. We suggest this is also due to a set of problematic assumptions (dualism, reductionism, individualism, passivity). We critically assess current explanations and empirical evidence and propose an alternative theoretical framework—the enactive approach to life and mind—based on recent developments in embodied cognitive science. We review core enactive concepts such as autonomy, agency, and sense-making. Following these ideas, we propose a move from binary distinctions (e.g., conscious vs. non-conscious) to the more workable categories of reflective and pre-reflective activity. We introduce an ontology of individuation, following the work of Gilbert Simondon, that allow us to see placebo interventions not as originating causal chains, but as modulators and triggers in the regulation of tensions between ongoing embodied and interpersonal processes. We describe these interrelated processes involving looping effects through three intertwined dimensions of embodiment: organic, sensorimotor, and intersubjective. Finally, we defend the need to investigate therapeutic interactions in terms of participatory sense-making, going beyond the identification of individual social traits (e.g., empathy, trust) that contribute to placebo effects. We discuss resonances and differences between the enactive proposal, popular explanations such as expectations and conditioning, and other approaches based on meaning responses and phenomenological/ecological ideas.

Critical integration in neural and cognitive systems: Beyond power-law scaling as the hallmark of soft assembly

Inspired by models of self-organized criticality, a family of measures quantifies long-range correlations in neural and behavioral activity in the form of self-similar (e.g., power-law scaled) patterns across a range of scales. Long-range correlations are often taken as evidence that a system is near a critical transition, suggesting interaction-dominant, softly assembled relations between its parts. Psychologists and neuroscientists frequently use power-law scaling as evidence of critical regimes and soft assembly in neural and cognitive activity. Critics, however, argue that this methodology operates at most at the level of an analogy between cognitive and other natural phenomena. This is because power-laws do not provide information about a particular system's organization or what makes it specifically cognitive. We respond to this criticism using recent work in Integrated Information Theory. We propose a more principled understanding of criticality as a system's susceptibility to changes in its own integration, a property cognitive agents are expected to manifest. We contrast critical integration with power-law measures and find the former more informative about the underlying processes.

Comment: How Your Own Becoming Feels

Mascolo (2020) successfully defends a relational, developmental approach to emotions. I draw parallels between his perspective and the enactive approach, in particular with the concept of participatory sense-making. I suggest that the need to understand emotions developmentally reveals a deeper link between affective life and human unfinishedness, namely, that emotions are collectively constituted ways of regulating human becoming.

Picturing Organisms and Their Environments: Interaction, Transaction, and Constitution Loops

Changing conceptions of the relation between organisms and their environments make up a crucial chapter in the history of psychology. This may be approached by a comparative study of how schematic diagrams portray this relation. Diagrams drive the communication and the teaching of ideas, the sedimentation of epistemic norms and methods of analysis, and in some cases the articulation of novel concepts through pictographic variants. Through a sampling of schematic representations, I offer a concise comparison of how different authors, with different interests and motivations, have portrayed important aspects of the organism–environment relation. I compare example diagrams according to the features they underscore (or omit) and group them into classes that emphasize interaction, transaction, and constitution loops.

Rediscovering Richard Held: Activity and Passivity in Perceptual Learning

Understanding the role of self-generated movements in perceptual learning is central to action-based theories of perception. Pioneering work on sensory adaptation by Richard M. Held during the 1950s and 1960s can still shed light on this question. In a variety of rich experiments Held and his team demonstrated the need for self-generated movements in sensory adaptation and perceptual learning. This body of work received different critical interpretations, was then forgotten for some time, and saw a surge of revived interest within embodied cognitive science. Through a brief review of Held’s work and reactions to it, we seek to contribute to discussions on the role of activity and passivity in perceptual learning. We classify different positions according to whether this role is considered to be contextual (facilitatory, but not necessary), enabling (causally necessary), or constitutive (an inextricable part of the learning process itself). We also offer a critique of the notions of activity and passivity and how they are operationalized in experimental studies. The active-passive distinction is not a binary but involves a series of dimensions and relative degrees that can make it difficult to interpret and replicate experimental results. We introduce three of these dimensions drawing on work on the sense of agency: action initiation, control, and monitoring. These refinements in terms of causal relations and dimensions of activity-passivity should help illuminate open questions concerning the role of activity in perception and perceptual learning and clarify the convergences and differences between enaction and ecological psychology.

Learning to find spatially reversed sounds

Adaptation to systematic visual distortions is well-documented but there is little evidence of similar adaptation to radical changes in audition. We use a pseudophone to transpose the sound streams arriving at the left and right ears, evaluating the perceptual effects it provokes and the possibility of learning to locate sounds in the reversed condition. Blindfolded participants remain seated at the center of a semicircular arrangement of 7 speakers and are asked to orient their head towards a sound source. We postulate that a key factor underlying adaptation is the self-generated activity that allows participants to learn new sensorimotor schemes. We investigate passive listening conditions (very short duration stimulus not permitting active exploration) and dynamic conditions (continuous stimulus allowing participants time to freely move their heads or remain still). We analyze head movement kinematics, localization errors, and qualitative reports. Results show movement-induced perceptual disruptions in the dynamic condition with static sound sources displaying apparent movement. This effect is reduced after a short training period and participants learn to find sounds in a left-right reversed field for all but the extreme lateral positions where motor patterns are more restricted. Strategies become less exploratory and more direct with training. Results support the hypothesis that self-generated movements underlie adaptation to radical sensorimotor distortions.

Deterministic Agent-Based Path Optimization by Mimicking the Spreading of Ripples

Inspirations from nature have contributed fundamentally to the development of evolutionary computation. Learning from the natural ripple-spreading phenomenon, this article proposes a novel ripple-spreading algorithm (RSA) for the path optimization problem (POP). In nature, a ripple spreads at a constant speed in all directions, and the node closest to the source is the first to be reached. This very simple principle forms the foundation of the proposed RSA. In contrast to most deterministic top-down centralized path optimization methods, such as Dijkstra's algorithm, the RSA is a bottom-up decentralized agent-based simulation model. Moreover, it is distinguished from other agent-based algorithms, such as genetic algorithms and ant colony optimization, by being a deterministic method that can always guarantee the global optimal solution with very good scalability. Here, the RSA is specifically applied to four different POPs. The comparative simulation results illustrate the advantages of the RSA in terms of effectiveness and efficiency. Thanks to the agent-based and deterministic features, the RSA opens new opportunities to attack some problems, such as calculating the exact complete Pareto front in multiobjective optimization and determining the kth shortest project time in project management, which are very difficult, if not impossible, for existing methods to resolve. The ripple-spreading optimization principle and the new distinguishing features and capacities of the RSA enrich the theoretical foundations of evolutionary computation.

Sensorimotor strategies for recognizing geometrical shapes: a comparative study with different sensory substitution devices

The sensorimotor approach proposes that perception is constituted by the mastery of lawful sensorimotor regularities or sensorimotor contingencies (SMCs), which depend on specific bodily characteristics and on actions possibilities that the environment enables and constrains. Sensory substitution devices (SSDs) provide the user information about the world typically corresponding to one sensory modality through the stimulation of another modality. We investigate how perception emerges in novice adult participants equipped with vision-to-auditory SSDs while solving a simple geometrical shape recognition task. In particular, we examine the distinction between apparatus-related SMCs (those originating mostly in properties of the perceptual system) and object-related SMCs (those mostly connected with the perceptual task). We study the sensorimotor strategies employed by participants in three experiments with three different SSDs: a minimalist head-mounted SSD, a traditional, also head-mounted SSD (the vOICe) and an enhanced, hand-held echolocation device. Motor activity and fist-person data are registered and analyzed. Results show that participants are able to quickly learn the necessary skills to distinguish geometric shapes. Comparing the sensorimotor strategies utilized with each SSD we identify differential features of the sensorimotor patterns attributable mostly to the device, which account for the emergence of apparatus-based SMCs. These relate to differences in sweeping strategies between SSDs. We identify, also, components related to the emergence of object-related SMCs. These relate mostly to exploratory movements around the border of a shape. The study provides empirical support for SMC theory and discusses considerations about the nature of perception in sensory substitution.

Spinal circuits can accommodate interaction torques during multijoint limb movements

The dynamic interaction of limb segments during movements that involve multiple joints creates torques in one joint due to motion about another. Evidence shows that such interaction torques are taken into account during the planning or control of movement in humans. Two alternative hypotheses could explain the compensation of these dynamic torques. One involves the use of internal models to centrally compute predicted interaction torques and their explicit compensation through anticipatory adjustment of descending motor commands. The alternative, based on the equilibrium-point hypothesis, claims that descending signals can be simple and related to the desired movement kinematics only, while spinal feedback mechanisms are responsible for the appropriate creation and coordination of dynamic muscle forces. Partial supporting evidence exists in each case. However, until now no model has explicitly shown, in the case of the second hypothesis, whether peripheral feedback is really sufficient on its own for coordinating the motion of several joints while at the same time accommodating intersegmental interaction torques. Here we propose a minimal computational model to examine this question. Using a biomechanics simulation of a two-joint arm controlled by spinal neural circuitry, we show for the first time that it is indeed possible for the neuromusculoskeletal system to transform simple descending control signals into muscle activation patterns that accommodate interaction forces depending on their direction and magnitude. This is achieved without the aid of any central predictive signal. Even though the model makes various simplifications and abstractions compared to the complexities involved in the control of human arm movements, the finding lends plausibility to the hypothesis that some multijoint movements can in principle be controlled even in the absence of internal models of intersegmental dynamics or learned compensatory motor signals.

A genealogical map of the concept of habit

The notion of information processing has dominated the study of the mind for over six decades. However, before the advent of cognitivism, one of the most prominent theoretical ideas was that of Habit. This is a concept with a rich and complex history, which is again starting to awaken interest, following recent embodied, enactive critiques of computationalist frameworks. We offer here a very brief history of the concept of habit in the form of a genealogical network-map. This serves to provide an overview of the richness of this notion and as a guide for further re-appraisal. We identify 77 thinkers and their influences, and group them into seven schools of thought. Two major trends can be distinguished. One is the associationist trend, starting with the work of Locke and Hume, developed by Hartley, Bain, and Mill to be later absorbed into behaviorism through pioneering animal psychologists (Morgan and Thorndike). This tradition conceived of habits atomistically and as automatisms (a conception later debunked by cognitivism). Another historical trend we have called organicism inherits the legacy of Aristotle and develops along German idealism, French spiritualism, pragmatism, and phenomenology. It feeds into the work of continental psychologists in the early 20th century, influencing important figures such as Merleau-Ponty, Piaget, and Gibson. But it has not yet been taken up by mainstream cognitive neuroscience and psychology. Habits, in this tradition, are seen as ecological, self-organizing structures that relate to a web of predispositions and plastic dependencies both in the agent and in the environment. In addition, they are not conceptualized in opposition to rational, volitional processes, but as transversing a continuum from reflective to embodied intentionality. These are properties that make habit a particularly attractive idea for embodied, enactive perspectives, which can now re-evaluate it in light of dynamical systems theory and complexity research.

Behavioral metabolution: the adaptive and evolutionary potential of metabolism-based chemotaxis

We use a minimal model of metabolism-based chemotaxis to show how a coupling between metabolism and behavior can affect evolutionary dynamics in a process we refer to as behavioral metabolution. This mutual influence can function as an in-the-moment, intrinsic evaluation of the adaptive value of a novel situation, such as an encounter with a compound that activates new metabolic pathways. Our model demonstrates how changes to metabolic pathways can lead to improvement of behavioral strategies, and conversely, how behavior can contribute to the exploration and fixation of new metabolic pathways. These examples indicate the potentially important role that the interplay between behavior and metabolism could have played in shaping adaptive evolution in early life and protolife. We argue that the processes illustrated by these models can be interpreted as an unorthodox instantiation of the principles of evolution by random variation and selective retention. We then discuss how the interaction between metabolism and behavior can facilitate evolution through (i) increasing exposure to environmental variation, (ii) making more likely the fixation of some beneficial metabolic pathways, (iii) providing a mechanism for in-the-moment adaptation to changes in the environment and to changes in the organization of the organism itself, and (iv) generating conditions that are conducive to speciation.

A minimal model of metabolism-based chemotaxis

Since the pioneering work by Julius Adler in the 1960's, bacterial chemotaxis has been predominantly studied as metabolism-independent. All available simulation models of bacterial chemotaxis endorse this assumption. Recent studies have shown, however, that many metabolism-dependent chemotactic patterns occur in bacteria. We hereby present the simplest artificial protocell model capable of performing metabolism-based chemotaxis. The model serves as a proof of concept to show how even the simplest metabolism can sustain chemotactic patterns of varying sophistication. It also reproduces a set of phenomena that have recently attracted attention on bacterial chemotaxis and provides insights about alternative mechanisms that could instantiate them. We conclude that relaxing the metabolism-independent assumption provides important theoretical advances, forces us to rethink some established pre-conceptions and may help us better understand unexplored and poorly understood aspects of bacterial chemotaxis.

A ripple-spreading genetic algorithm for the aircraft sequencing problem

When genetic algorithms (GAs) are applied to combinatorial problems, permutation representations are usually adopted. As a result, such GAs are often confronted with feasibility and memory-efficiency problems. With the aircraft sequencing problem (ASP) as a study case, this paper reports on a novel binary-representation-based GA scheme for combinatorial problems. Unlike existing GAs for the ASP, which typically use permutation representations based on aircraft landing order, the new GA introduces a novel ripple-spreading model which transforms the original landing-order-based ASP solutions into value-based ones. In the new scheme, arriving aircraft are projected as points into an artificial space. A deterministic method inspired by the natural phenomenon of ripple-spreading on liquid surfaces is developed, which uses a few parameters as input to connect points on this space to form a landing sequence. A traditional GA, free of feasibility and memory-efficiency problems, can then be used to evolve the ripple-spreading related parameters in order to find an optimal sequence. Since the ripple-spreading model is the centerpiece of the new algorithm, it is called the ripple-spreading GA (RSGA). The advantages of the proposed RSGA are illustrated by extensive comparative studies for the case of the ASP.

How (not) to model autonomous behaviour

Autonomous systems are the result of self-sustaining processes of constitution of an identity under precarious circumstances. They may transit through different modes of dynamical engagement with their environment, from committed ongoing coping to open susceptibility to external demands. This paper discusses these two statements and presents examples of models of autonomous behaviour using methods in evolutionary robotics. A model of an agent capable of issuing self-instructions demonstrates the fragility of modelling autonomy as a function rather than as a property of a system's organization. An alternative model of behavioural preference based on homeostatic adaptation avoids this problem by establishing a mutual constraining between lower-level processes (neural dynamics and sensorimotor interaction) and higher-level metadynamics (experience-dependent, homeostatic triggering of local plasticity and re-organization). The results of these models are lessons about how strong autonomy should be approached: neither as a function, nor as a matter of external vs. internal determination.

The contribution of active body movement to visual development in evolutionary robots

Inspired by the pioneering work by Held and Hein (1963) on the development of kitten visuo-motor systems, we explore the role of active body movement in the developmental process of the visual system by using robots. The receptive fields in an evolved mobile robot are developed during active or passive movement with a Hebbian learning rule. In accordance to experimental observations in kittens, we show that the receptive fields and behavior of the robot developed under active condition significantly differ from those developed under passive condition. A possible explanation of this difference is derived by correlating receptive field formation and behavioral performance in the two conditions.

The circular topology of rhythm in asynchronous random Boolean networks

The analysis of previously evolved rhythmic asynchronous random Boolean networks [Biosystems 59 (2001) 185] reveals common topological characteristics indicating that rhythm originates from a circular functional structure. The rhythm generating core of the network has the form of a closed ring which operates as a synchronisation substrate by supporting a travelling wave of state change; the size of the ring corresponds well with the period of oscillation. The remaining nodes in the network are either stationary or follow the activity of the ring without feeding back into it so as to form a coherent whole. Rings are typically formed early on in the evolutionary search process. Alternatively, long chains of nodes are favoured before they close upon themselves to stabilize. Analysis of asynchronous networks with de-correlated (non-rhythmic, non-stationary) attractors reveals no such common topological characteristics. These results have been obtained using statistical analysis and a specifically developed bottom-up pruning algorithm. This algorithm works from local interactions to global configuration by eliminating redundant links. The suitability of the algorithm has been confirmed by both numerical and single lesion analysis. The ring topology solution for the generation of rhythm implies that it will be harder to evolve rhythmic networks for big sizes and small periods and for bigger number of connections per node. These trends are confirmed empirically. Finally, the identified mechanisms are utilised to handcraft rhythmic networks of different periods showing that a low number of connections suffices for a large variety of rhythms. Random asynchronous update forces the evolved solutions to be highly robust maintaining their performance in the presence of intrinsic noise. The biological implications of such robust designs for molecular clocks are discussed.

Evolving spike-timing-dependent plasticity for single-trial learning in robots

Single-trial learning is studied in an evolved robot model of synaptic spike-timing-dependent plasticity (STDP). Robots must perform positive phototaxis but must learn to perform negative phototaxis in the presence of a short-lived aversive sound stimulus. STDP acts at the millisecond range and depends asymmetrically on the relative timing of pre- and post-synaptic spikes. Although it has been involved in learning models of input prediction, these models require the iterated presentation of the input pattern, and it is hard to see how this mechanism could sustain single-trial learning over a time-scale of tens of seconds. An incremental evolutionary approach is used to answer this question. The evolved robots succeed in learning the appropriate behaviour, but learning does not depend on achieving the right synaptic configuration but rather the right pattern of neural activity. Robot performance during positive phototaxis is quite robust to loss of spike-timing information, but in contrast, this loss is catastrophic for learning negative phototaxis where entrained firing is common. Tests show that the final weight configuration carries no information about whether a robot is performing one behaviour or the other. Fixing weights, however, has the effect of degrading performance, thus demonstrating that plasticity is used to sustain the neural activity corresponding both to the normal phototaxis condition and to the learned behaviour. The implications and limitations of this result are discussed.