Performing Relations

Abstract

This article investigates the role of corporeal habits in the creation and maintenance of a relational layer among humans and robots in a performance context. Drawing on Gregory Bateson's work, this article analyses how habits shape this relational bond in two robotic artworks: Ruairi Glynn's Performative Ecologies and Paula Gaetano Adi's Anima. These case studies illustrate how the performance space provides an ideal ground to experiment with habit formation at the interface of humans and robots. Moreover, this article offers an embodied and relational perspective to human–machinic interaction that embraces ways of relating to highly different nonhuman corporealities.

INTRODUCTION

Communication and interaction are first and foremost about relationships. Beyond what is being said, the roles we perform with one another and how we frame the ways in which we might act, precede and make possible any interpretations of our interaction. This perspective on communication and interaction can contribute to thinking human–robot performances from an embodied and relational standpoint that connects both agents without subsuming their differences.

The performance of corporeal habits is key to the creation and maintenance of such a relational layer. I define habit formation as an ecological cuing of behaviour, which relies on a dynamic interplay of constancy and change within the movement patterns enacted by the participants of an interaction. Additionally, habit co-adaptation involves the materiality of the participants’ bodies, the situational framing of the encounter and the co-creation of relationship dynamics.

To investigate the performance of habit among humans and robots, this article employs two concepts in the work of Gregory Bateson: ‘metacommunicative framing’ and ‘analogue communication’. These concepts serve as lenses to look at the robotic artworks of Ruairi Glynn and Paula Gaetano Adi. Such perspective illustrates how embodied relationality through habit formation can take place among highly different human and non-human corporealities.

Performance practice resonates with Bateson’s concepts and his focus on relationality and the framing of interactions. The performance space is thus an ideal testing ground to experiment with habit formation at the interface of humans and robots. Fundamentally, the robotic artworks examined in this article demonstrate that the coadaptations of corporeal habits, how they reflect and co-create relationship dynamics, as well as the way they are framed, introduce new ways of connecting between humans and non-human bodies.

CORPOREAL AND RELATIONAL COMMUNICATION IN CYBERNETIC SYSTEMS

Bateson was a researcher in many fields, such as anthropology, the social sciences and biology, who, from the beginning of the cybernetic movement, was part of the interdisciplinary exchange in the so-called Macy Conferences, 1943–54 (Pias 2016). These meetings brought together an array of scholars, such as Norbert Wiener, Heinz von Foerster and Margaret Mead, to discuss the principles of organization in goaloriented systems. Cybernetics, a name coined by Wiener in 1948, is a field that considers humans, animals and machines as systems that regulate themselves by their own mechanisms, paying special attention to how they adapt to their environments, as well as how they communicate among themselves and with one another.

In virtue of how some concepts developed and were given importance, scholars have distinguished two waves: first-order and secondorder cybernetics (see Clarke and Hansen 2009). During first-order cybernetics emphasis was placed on homeostasis, and therefore stability in a system, with Claude Shannon’s information theory model serving as an inspiration for many cyberneticians, especially Wiener. In second-order cybernetics, more attention was given to the role of context and the subjectivity of the observer, consequently bringing up more relational ways of developing systemic knowledge.

Bateson was highly influenced by the main concepts of the field, especially second-order cybernetics. Moved by his anthropological background, Bateson tended to focus on the realm of animal communication and psychology, with an interest in pathologies of communication (Ruesch and Bateson 1951; Bateson 1972, 1991); however, his insights on relational communication seem to pervade his research. This makes Bateson’s theories an interesting starting point when considering embodied and relational encounters among humans and non-humans. Bateson’s concepts, furthermore, help me develop a relational and situational understanding of habit, which allows for new connections to emerge amid different human and non-human corporealities. In order to delve deeper into what Bateson was proposing, I focus on two concepts: ‘metacommunicative framing’ and ‘analogue communication’.

METACOMMUNICATIVE FRAMING AND ANALOGUE COMMUNICATION

Bateson was fascinated by what non-verbal behaviours do, rather than what they express. For him, that which behaviours try to communicate is less relevant than their function, which is the creation and maintenance of relationships among communicators. Communication, therefore, is not equated with the message expressed, but with a more primal and affective connection that reflects and cocreates relationships among communicators. An example that Bateson poses is that of a cat and its owner. The meowing of the cat does not refer to a particular object or intention. Rather, it emphasizes a connection among them, namely, one of dependency. What is at stake here is a relationship between two agents: the communication enacted does not say much about the cat on its own, or the owner alone, but rather it says something about the relationship that is created and unites them.

Such insights are collected in what Bateson named ‘metacommunication’: the act of framing an interaction, of establishing the conditions under which we can perform with one another. As Bateson highlights, ‘to act or be one end of a pattern of interaction is to propose the other end. A context is set for a certain class of response’ (1972: 275). I thus understand the concept of metacommunicative framing as the establishing of the conditions under which we can understand and interact with one another. Behaviours create frameworks that allow for other behaviours to be interpreted, and afford possibilities for further interaction, while foreclosing others. Moreover, that which is constantly re-negotiated in communication and interaction is how we relate to one another.

This leads us to one of Bateson’s most relevant distinctions: that between ‘analogue’ and ‘digital’. Even if he acknowledges that the two systems usually go hand in hand, he differentiates digital communication as a verbal, literal system of binary quantities, like human language, from analogue communication as a non-verbal system, whose signals are of a continuous variable quantity. For Bateson, then, the analogue system is better equipped to reflect on and enact relationships among interactants.

The relationships performed through the analogue system of communication are outside of the realm of reflection and intentionality. They are rooted in embodiment, rather than in rational consciousness and representation. As Bateson claims, in the communication between two human beings there is a common anatomy and experience of bodily functioning that gives a basis for how things can be understood and interpreted (Ruesch and Bateson 1951). This type of interaction serves as a foundation for later, more complex, modes of coding, including digital systems.

These initial considerations of an embodied basis for symbolic language were further developed by defenders of embodied and enacted approaches to cognition. One of the founders of that field, Francisco Varela, was himself a cybernetician who, after his work with Humberto Maturana, focused on a critique of representationalist perspectives on cognition, along with Rosch and Thompson (Varela et al. 1992). In The Embodied Mind, Varela, Rosch and Thompson provide instances of that embodied subsymbolic level of understanding, such as ‘kinaesthetic image schemas’: general human cognitive structures that derive from bodily experience and that are the basis for later linguistic abilities. Similarly, Maxine Sheets-Johnstone (2011) also convincingly argues for a pre-symbolic tactile-kinaesthetic basis of interaction, claiming that humans are able to engage in meaningful encounters by means of the physical and affective coordination of their movement dynamics. Examples are kicking, sucking or swallowing: spontaneous movements that create acts of sense-making in our interaction with the environment and others. Only later, and on the basis of these ‘corporeal concepts’ (Sheets-Johnstone 2011: 118), can we develop conceptual and symbolic understandings of the world (see Lakoff and Johnson 1980).

Bateson’s proposals about different modes of communication should thus be seen in light of these later developments. However, he foregrounded an essential aspect of the movement layer of interaction that has not been given the same importance in embodied cognition studies, that is: how such movements necessarily involve a co-creation of relationship dynamics.

HABIT CO-ADAPTATION AS A RELATIONAL LAYER

Such a position, as well as Bateson’s comments on our common biological (and cultural) background, might lead us to believe that successful communication can only take place among living beings that share similar characteristics and conditioning. However, it is important to note that cybernetic theories led Bateson to reconsider human–machinic boundaries, perceiving biological, social and mechanical systems as being ruled by similar parameters. Moreover, Bateson’s paradigmatic example of analogue communication deals with a human–animal encounter (that of a cat and its owner). Therefore, although Bateson did not explicitly deal with human–robot interaction, his theories surpassed such distinctions as they are based on a systemic view of agents – biological or mechanical – and on a shared embodiment in the world.

To better understand such a relational layer of communication that is not based on similar corporealities, I refer to George Spencer-Brown’s (1973) response to Bateson’s work. Spencer-Brown links Bateson’s analogue communication to what he calls ‘communion’, which precedes and creates the basis for communication.

‘Communion’, according to Spencer-Brown, is a form of getting to know one another, even between humans and non-humans. I propose to re-read this modality of ‘communion’ in the light of habits that enable a relational co-adaptation among two parties that allows for a primal understanding.

From this lens, the connection between humans and non-human bodies is facilitated by the embodied co-adaptations of habits, their reflection and co-creation of relationships among interactants, and the way they are framed. As stated before, habits are understood in this article as movement patterns acquired through repetition, and performed without conscious thought, which are relational and situated. Furthermore, the notion of habit highlights Bateson’s interest in the constant re-negotiation of embodied relationships among interactants through habit’s dynamic interplay between constancy and change, as stated by Clare Carlisle (2014).

Even if habit is usually considered within living systems, I suggest that thinking the phenomenon of habit at the interface of humans and robots can aid in opening new ways of connecting to non-human corporealities, foregrounding an embodied relationality in their interaction. To further develop this, I analyse practices of contemporary robotic art because, when viewed from the lens of Bateson’s concepts, their work materializes this embodied relationality between humans and non-humans through the performance of habits.

HEAR ME WITH YOUR BODY: RUAIRI GLYNN’S PERFORMATIVE ECOLOGIES

Ruairi Glynn is an installation artist and director of the Interactive Architecture Lab at the Bartlett School of Architecture in London. His works deal with aliveness in interactive objects and environments, drawing influences from cybernetics and performance art (Glynn 2019). In this section, I will read Glynn’s robotic artwork Performative Ecologies through the lens of Bateson’s concepts in order to explore how an embodied relationality through movement habits can take place among humans and robots.

Performative Ecologies was first shown in the exhibition ‘Emergence, a Show of Artificial Life Art’, curated by Simon Penny and David Familian at the Beall Center for Art and Technology in 2009 (Penny 2011). The artwork is composed of four attention-seeking robot dancers that orientate themselves and search out for audience members (fig. 1). Each of them consists of two elements: a head part on top of a motorized neck, where an infra-red video camera is installed; and a tail part, which turns and lights up. Once the cameras detect that the attention of the audience has been caught, the robots begin a movement pattern with their tails, controlling at the same time how long a person stays during the dance. Each robot performs autonomously thanks to a ‘genetic algorithm’ that creates their movements¹. However, the robots’ autonomy becomes affected by the audience because the movement phrases that are more successful in catching the human’s attention continue to be reproduced, while the less successful ones are forgotten.

Figure 1. Performative Ecologies by Ruairi Glynn (2009). Photo courtesy of the artist

Interestingly, Performative Ecologies does not always need the human input. When not attended by the audience, the robots communicate among themselves, sharing their most successful performances and consequently creating new phrases together. In this case, ‘successful performances’ are understood as those movement phrases that capture the attention of the audience more notably, which is measured by the cameras on top of the robots. This parameter is wired into the genetic algorithm of the robot dancers and informs the development of future performances. Instead of being conditioned by a pre-programmed choreography, the robot dancers thus form a cybernetic system in which the various constituent elements interact to co-create what I am calling ‘movements habits’ that are repeated and transformed in the course of the performance.

Movement habits in this robot performance are understood as ‘an aptitude, a skill or facility’ and not so much as a ‘tendency or inclination’ (Carlisle 2014: 7). They are considered habits because they ‘are acquired through repetition and can be executed without much thought or attention’ (ibid.). In Performative Ecologies, it is also possible to see how the robots’ movement habits, as John-David Dewsbury (2015) points out, consist of three elements: first, a material stimulus (or a cue), in this case the interaction between robots and humans, and among robots, that prompts movement. Second, a routine, as the patterned movement responses that the robot dancers enact. And third, a reward when the performed habit catches the attention of the audience, which is translated into the genetic algorithm in the form of a reward function that allows the habit to continue. The robot’s movement phrases evolve always in tension between constancy and change. Due to the attention of the audience and the collaboration among the robots, their movement habits are modified over time. There is, however, a ‘tendency to stay the same’ (Carlisle 2014: 17), as the proposed modifications by other robots are not accepted and incorporated if they are highly different from their current patterns.

Performative Ecologies proposes a ‘kinetic conversational environment’ as inspired by the cybernetician Gordon Pask’s ideas on conversation theory (Glynn 2008). Pask, also a second-order cybernetician like Bateson, established a difference between communication and conversation, with the latter marking a more creative exchange between different agents that co-create and negotiate their goals in the process of interacting. Glynn, inspired by Pask’s theory, claims that a conversation between humans and machines can help in creating real interactivity in a system, rather than mere reactivity – that is, producing the same output in response to a specific input. Conversation is then ‘a form of social communication that promotes a circularity of inter-actions where participants contribute to a shared discourse negotiating their actions and understandings’ (Glynn 2008). How these relations are established among humans and robots will be further clarified through Bateson’s concepts.

One essential way in which the possibilities of the conversation are established is through a metacommunicative frame, that is, through the act of establishing the conditions under which we can relate to each other. Fundamentally, this is done through the use of the term ‘ecology’. Considering interaction as an ecology allows it to be perceived as a mangle of organisms reacting and adapting to one another and their environment. As Glynn highlights, ‘this notion of ecology touches on the rich web of communicative exchanges that occur in complex interacting systems, [as well as the fact that] the agents within them are adaptive to changing circumstances’ (2019: 82).

Materially, the robots also afford specific interactions: the way in which their ‘heads’ are constructed signal not only a clear front and back but also a desire for establishing a connection, a call for attention (fig. 2). Interestingly, the robots are not positioned in a traditional dynamic in which they are the only entertainers for the humans. As Cleland (2010) shows, during the exhibition people compete for the attention of the robots, calling and gesturing towards them. The visitors consequently adapt their behaviour to be part of that cybernetic system. Gemeinboeck and Saunders similarly observe that ‘the robots in fact elicit the audience to perform with them in order to expand their dance repertoire’ (2015: 166). Here habit’s dual quality is also perceptible, inasmuch as the public is lured by the constancy of the movement patterns performed – which, with time, allow them to associate each robot to specific idiosyncrasies – and the change in the robots’ repertoire that is brought about precisely by this interaction. This re-negotiation then happens in the interplay of embodied relations between humans and robots, in how the visitors adapt their behaviours to attract the robots’ attention and on how their interactions modify the same habits they feel drawn to.

Figure 2. Performative Ecologies by Ruairi Glynn (2009). Photo courtesy of the artist

The relationships established in this system are consequently of two types. The first is competition, as robots and humans compete for the attention of their conversational partners. The second one is collaboration. Not only do the robots share among themselves the habits that have resulted in success, but humans and robots are able to collaborate in creating a new movement phrase, both by providing attention and by kinaesthetically moving their bodies so that new movements are recorded and added to their habit repertoire.

With regards to Bateson’s analogue communication, on the one hand, the cybernetic conversation created between humans and robots is purely corporeal since the language of the robot dancers is uniquely based on the creation, learning and sharing of movement habits. On the other hand, the fact that humans can modify the robots’ bodies in order to teach a movement phrase opens the opportunity for an embodied know-how, a corporeal knowledge to be transmitted among them.

To sum up, Ruairi Glynn creates a space where an embodied conversation among humans and robots through the performance and transferral of movement habits can take place. He emphasizes how relationship dynamics of competition and collaboration are interwoven, by creating an embodied ecology of human and non-human bodies. However, because of the work’s focus on a mostly visual attention, Performative Ecologies does not take the encounter among different corporealities to its ultimate consequences. This is precisely what the following case of Paula Gaetano Adi offers: a more radical approach to the embodied relationality between humans and non-human machinic others.

BREATHING OUR WAY TOGETHER: PAULA GAETANO ADI’S ANIMA

Gaetano Adi is an Argentinian artist and scholar who works on robotics, sculpture, performance, and interactive installations from a decolonial perspective. In her work, she investigates alternative embodiments for autonomous robotic agents, as well as new ways of engaging with them. Anima, a name inspired by the idea of soul, animism and animalism (Gaetano Adi 2010), is an amorph soft robot that swells and begins to inhale and exhale in an exaggerated manner every time that a human approaches it (fig. 3). This robotic breathing occurs through one orifice of its body: a small hole that dilates. Breathing is not like other involuntary bodily reactions because we can learn to control and modify it, creating a habit out of it, as practices of mindfulness show. Anima not only breathes but adapts its patterns of breathing in response to the proximity of humans. This foregrounds the initially involuntary reaction and brings it to the level of a perceptible, though mostly affective, corporeal engagement, making space for not only influencing the robot’s breathing patterns, but also attuning our own to it. Anima’s breathing, then, opens the possibility for modifying and creating new habits by making salient aspects that usually go unnoticed or are deemed as non-significant. And it does so from a purely corporeal perspective, one that is radically different to ours.

Figure 3. Anima by Paula Gaetano Adi (2009). Photo courtesy of the artist

Anima does not move, talk or see the audience; its only way of communicating with humans is by means of its breathing. As Gaetano Adi claims, Anima has a purely corporeal language, which proposes, in opposition to an artificial intelligence, an ‘artificial corporeality’ (2010: 2). This corporeal language, even if it reflects human capacities, does not aim at human-likeness. Anima’s way of breathing emanates from its own particular non-human embodiment. Starting from the design of an amorph shape, Gaetano Adi goes on to ask herself ‘which kind of actions this new embodiment demands?’ (5).

The artist aims to ‘embody some sort of encounter between different kinds of bodies’ (2). This encounter does not take place in the realm of the familiar, nor does it connect two bodies with the same embodied experience (see Rhee 2018). Nonetheless, Gaetano Adi opens a space where a relationship can be established with this alien, artificial being, where corporeal habits can be attuned. How this is achieved can be better analysed through the lens of Bateson’s concepts.

With regards to metacommunicative framing, Gaetano Adi argues that, when creating an artificial body, she does not focus on the making of specific behaviours or gestures, but rather attempts to model ‘potential actions and reactions between the human viewer and that artificial artwork’ (2010: 4). She thus creates ‘an open space for a possible ‘‘dialogue/interaction’’ between them’ (ibid.). Her design process consists of the making of frames where specific relations can take place. This space of possibility again relies on habit’s tension between constancy and change, the known and the unknown, the recognizable and the variability.

In the case of Anima, this is done through the foregrounding of the initially imperceptible coadaptation of breathing patterns by its way of displaying the robot and its material affordances. Gaetano Adi chooses to exhibit the immobile robot on top of a pedestal, in the middle of an art gallery. With no distractions around and lit up, the robot is focalized as the centre of attention in its initially static materiality (fig. 4). This framing allows for a more contemplative gaze on the part of the public, in an encounter where time has slowed down, and that allows the subtle automatic reactions of the robot to be perceived. Breathing, this instinctual and lifeconserving capacity of our bodies, intelligible in its simplicity but usually imperceptible due to its automaticity, is elevated to the focus point of the piece.

Figure 4. Anima by Paula Gaetano Adi (2009). Photo courtesy of the artist

The minimal but enticing behaviour, at once recognizable but also estranged because of its appearance in an alien morphology, inspires curiosity and encourages a corporeal impulse to respond. The soft materials of the robot invite a close, physical approximation from the audience. Through the accelerated breathing as viewers approach and establish physical contact with it, the robot invokes vulnerability and fragility, leading scholars like Kayla Anderson to claim that they express a ‘sort of robotic anxiety’ (2014: 354). The framing of this encounter therefore affords a relationship of care, of vulnerable exposure and of sympathy on the part of the viewer, a connection created on the basis of our most intimate and pre-reflexive bodily reactions.

Concerning analogue communication, Anima reflects and deconstructs what these corporeal habits can be in non-human embodiments, by materializing them in amorphous bodies ‘with no front/back, up/down distinction’ (Gaetano Adi 2010: 5). Devoid of the usual physical distinctions that allow us to create a connection in corporeal similarity, the audience is confronted with the affective impact of ‘a general shape that embodies a certain kind of behaviour’ (ibid.). These behaviours are below the threshold of conscious reflection and are designed to incite a primary form of corporeal attunement between humans and nonhumans. The audience, when in close proximity with Anima, engages in a modification and adaptation of their own behaviours (looking, touching, breathing) in response to the robot’s reaction to their presence, creating ultimately a corporeal dialogue among them, in the form of a feedback loop. This analogue communication, furthermore, does not attempt to be logical or deploy a univocal interpretation. Rather, it invites ambiguity and open-endedness in the encounter of this being – meaning being cocreated in the moment through the performance and the subtle co-adaptation of human and robot corporeal habits.

To sum up, Anima allows for a form of embodied and relational connection to emerge between highly distinct human and non-human embodiments through several strategies: first, the framing of the communicative act as a contemplative, slowed down environment where the perception of subtle corporeal habits, like breathing in response to human contact, is foregrounded; second, in the pre-reflexive attunement that happens between human and robot when co-adapting to each other’s habits; and, finally, in the creation of possibilities for interaction through the design of the embodiment, movement and materiality of the robot that invite an intimate, vulnerable and caring relationship dynamic between human and robot.

CONCLUSION

Before and beyond symbolic communication, humans and non-humans engage in forms of embodied interactions to create and maintain meaningful relationships among them. This mode of embodied and relational communication takes place through the coadaptation of corporeal habits and provides an interesting basis to explore how humans and non-anthropomorphic robots could connect despite their radical otherness. It is through this layer of embodied communication, which mostly occurs outside of conscious control, that we establish how we relate to one another, creating the basis for additional modes of communicating.

The two robotic performances discussed above provide cases of materially instantiating what this embodied relationality could be among highly different corporealities. The manner in which Performative Ecologies and Anima frame human–robot interactions, through its focus on the materiality of the robots as well as on the ecological scene in which they take place, highlights how our encounter with non-human bodies is always based on embodied ways of relating to each other. It is, therefore, through the process of attuning ourselves to radically different bodies, thanks to the pre-reflexive adaptation of our corporeal habits, that we can enlarge our understandings of communication and interaction, reframing who we are in the process.

Notes

1 For a video of the robot performance see Glynn (2007).