The Mechanical Heart:
Speculative human-robot intimacy in Japan
A Thesis Presented by
JUELING HU
Submitted to the Graduate School of Humanities at the University of Amsterdam in partial fulfillment
of the requirements for the degree of RESEARCH MASTER OF ARTS
February 2021
© Copyright by Jueling Hu 2021 All Rights Reserved
The Mechanical Heart:
Speculative human-robot intimacy in Japan
A Thesis Presented by
JUELING HU
Supervisor: Prof. Misha Kavka
Second Reader: Prof. Thomas Poell
Address: Baishi Gu Liancheng F2401, Yuhu Dist., 411100, Xiangtan, Hunan, P.R. China Phone Number: (+86) 177-1044-1994; (+31) 623115492
Private Email Address: jueling.hu@gmail.com Student Number: 12245852
Abstract
Futuristic ideas about robots as social actors are reconfiguring the relationships between humans and technology. Social robots are imagined and designed with not only instrumental functions but also subjective qualities to be human-like existences. These seemingly paradoxical features challenge the rigid dichotomies between human and machine – which was considered to be historically and culturally neutral – and call for new understandings of human-robot relationships in dynamics.
The present thesis examines the emerging relationships between humans and humanoid social robots in a series of social-technical imaginaries (including popular science fiction and speculative designs). Particularly, it explores how the ideal forms of robot subjectivity and human-robot relationships are imagined in Japanese works. Such works not only become globally influential, setting the agenda for the future directions of robotics production global-wide, but also allude to heterogenous future aspirations that reveal alternative potentials to those in the West. In these
imaginaries, the future-oriented technology is integrated into a timeless tradition of Japan’s symbiotic human-robot society, in which robots are comfortably embraced as alternative members.
Drawing on a posthumanist approach, the thesis develops a relational ontology to unfold human-robot relationships, which are found to be necessarily intimate by design. The new ontology of robotics proposes that the qualities of robots are not intrinsic but defined by the dynamic relationships between humans and robots, which lead to the thesis’ particular focus on the moments of human-robot encounters. Through close-reading popular science fictional anime such as Chobits and Plastic Memory as well as speculative designs including those from Sugano and Ishiguro Laboratories, it finds that robots, as alternative social members, are imagined to be parabolically similar to humans: while they are humanlike subjects with “hearts”, they also possess mechanical features to confirm their identities. This tension corresponds to the mutual but human-dominant intimacy: robot
subjectivity enables the attachment to be established by both sides whereas their mechanical features reaffirm human mastery over robotics technologies.
TABLE OF CONTENTS
Abstract iv
Introduction 1
Chapter 1. Robot, An Intimate Machine 7
Designed human-robot intimacy 7
Intimate robot, a dilemma 10
Relational ontology of robotics 14
Chapter 2. Embracing Robot-Symbiotic Society in Japan 18
Robotics as a Japanese tradition 19
Towards a future of human-robot symbiosis 22
Chapter 3. Robot Subjectivity: Parabolically Human 26
Society members with mechanical hearts 26
Mechanical heart as an assemblage of sensors and actuators 30
Being parabolically similar to humans 34
Chapter 4. Human-Robot Intimacy: Mutual but human-dominant 37
Mutual attachments between robots and humans 38
Human-dominated Intimacy 43
Conclusion: Beating with Mechanical Hearts 47
LIST OF FIGURES
Figure 1. Automata, the origin of robots ... 20
Figure 2. Four fields of robotics technologies ... 24
Figure 3. Robotics applied different fields in the society ... 29
Figure 4. TWENDY-ONE ... 31
Figure 5. WAMOEBA-2 ... 32
Introduction
There is far greater subtlety to a humanoid robot than being a mere human replica. This subtlety can be observed when a robot’s metal body wears the old shirt of a family member, moving on their wheels to accompany their “grandmother” for a walk (Willinger, 2019), or when a robot is asked by a human colleague about their favorite cuisine and then answers in a half-joking way, “I can’t eat” (エ リカちゃんねる [Erica Channel], 2018), or when the owner of a robot sees the lights of his home – among all the starry lights of thousands of homes in the city –and feels the warmth of being waited for and relied upon by his robot (Gatebox, 2016).
In these scenes, the metal bodies, the designed programming language, and the occasional glitches recall the robots’ presence as machines, but at the same time the companionship,
communication, and attachment introduced by humanoid robots are also experienced by humans in physical and emotional forms. Rather than objectifying robots as a pile of humanoid metal
components, humans also develop specific tactics – in both cognitive and emotional dimensions – to socialize with humanoid robots (Nass and Moon, 2002). A type of relationship between humans and robots that transcends the strict human-robot divide seems to be manifest in all these interactions: what it means to be a robot is constantly slipping and shifting between a mechanical object and a communicative subject. A blurred and dynamic middle ground opens up in the human-robot relationship, since the robots simultaneously possess technological-material features and occupy communicative roles.
This middle ground becomes particularly clear in a series of imaginary works and designs of human-robot relationships in Japan. Having the reputation as a “nation of robots” (Schodt, 1988), Japan leads the way in the development of robotics technology. Since the mid-twentieth century, the science fiction genre in Japan has offered an alternative set of imaginary schemes that are different from the anthropocentric understanding of human-robot relationships frequently seen in Western media (Zeeberg, 2020). This type of socio-technical imagination does not put humans and robots at two opposite ends: robots are neither imagined as substitutes for humans, replacing humans in the role of intimate partners or subverting humanity to become rulers of the earth, nor are they required to become “real” humans in order to join human society. Rather, in Japanese robot fantasies, robots are both themselves as well as the family members, friends or colleagues of humans. A robot can, as depicted in Astro Boy (Tezuka, 1952), have superpowers to fight for social justice, while at the same time he looks like the young boy next door and calls his creator “father”. The robot can also be the heroine of Nichijou (Arawi, 2011), a young girl who feels unconfident because of the giant clockwork
on her back – unlike humans – but eventually realizes with the help of her human friends that she can still live a "normal" life even though she is a robot.
The blurred dynamics in which robots possess mechanical features and some level of
humanity can also be found as a dominant phenomenon in Japanese speculative designs. In particular, under the vision of constructing a “symbiotic human-robot society” (“人共存ロボティクス社会” in Japanese), the developments and designs in the field of humanoid robotics in Japan display an ambition to create mechanical members of society (Council on Competitiveness-Nippon [COCN], 2019). As one of the solutions to the labor shortage faced by Japan's super-aging society, the design and manufacture of robots, inspired by human capacities, conceive of them as alternative human resources (Siciliano and Khatib, 2008). On the one hand, robots are expected to contribute to human society specific services such as reception, nursing, and medical care. On the other hand, due to the strongly social nature of their services, humanoid robots are designed not only to imitate humans in terms of certain functions like kinematics, senses, behavior, and language, but also to have the ability to understand, express, and transmit emotions.
Why does the intersection between mechanical objectivity and humanity occur so often in the imagination of human-robot interactions? By examining the imaginative moments in Japanese science fiction and speculative design when humans and robots encounter each other, this thesis seeks to understand the conceptual issues of robot ontology and human-robot relationality that underlie this middle ground. In particular, it addresses the following questions: what is the significance of humans interacting with robots using certain social tactics, rather than treating them as mere inert, mechanical assemblages? What kind of robot subjectivity is demonstrated in these imaginations of humanoid social robots? And what kind of human-robot relationships correspond to this type of robot subjectivity?
The relationships between humans and robots have been researched and debated in a variety of disciplines. Technological developments and related ethical concerns, concepts of humanity, and embodied experiences of human-robot interactions have been fundamental to robotics studies, with the field led primarily by roboticists, ethics philosophers, and cultural studies scholars. Sociologists have also provided unique contributions to the understanding of human-robot interactions from a macro perspective, noting the entanglements between robot designs and power relations based on discriminatory patterns such as gender and race. My research benefits from and builds on the contributions of such scholarship. However, amongst these studies, analyses of robots’ abilities to enable humans to gain a sense of authentic connection appear together with concerns that robots prevent humans from building “real” relationships. The dilemma lies in the paradoxical nature of humanoid robots – they are human-like but not humans – and the reason why scholars may draw
opposite conclusions from the same phenomenon is in large part that both presuppose an absolute distinction between humans and robots, humans and non-humans. Robots are either expected to be capable of replacing people, or they are presumed unable to do so. Both sides fail to develop a satisfactory theory that would explain the more dynamic and subtle situations that exist in human-robot interaction.
In response to this gap, the thesis draws on theological philosopher Martin Buber’s theory expounded in I and Thou (1923/1970) as an alternative resource for understanding human-robot relationships. Seeking to understand the relationships between humans, the world, and the divine Thou (deity), Buber develops a two-fold dialogical framework of I-It and I-Thou; the former refers to the relationships between the subject and the instrumentalized other, whereas the latter points to authentic relationships between subjects with their whole beings. Buber’s theory becomes particularly helpful to my research as it emphasizes the formation of subjectivity within relations. For him, “I” am not isolated, but exist in the relationships to others, relationships that happen in moments of
“dialogue” and “encounter”. The contingent feature of dialogues and encounters determines that one’s relations with a particular other can switch between I-Thou and I-It. Rather than reducing human relations to simple either/or, Buber’s two models of our foundational relations to others are interchangeable and always dynamic. His philosophy of dialogue offers a theoretical tool to understand robots in an ontological dimension, based on a notion of robot subjectivity that is non-dichotomous but relational. Following this idea, in order fully understand the connotations of robots and human-robot relationships, I situate robots back in the dynamic networks where humans and robots interact with each other and investigate the moments of human-robot encounters.
In order to examine the encounters between humans and robots in detail, I turn to assemblage theory, which offers a methodological tool to explore the interactions happening inside an encounter. Although Buber’s theory of relationality and assemblage theory have distinct philosophical traditions, the theories can be made to converge because the core of the I-Thou relation is about a “participatory intimacy” (Popova, 2018), about the feeling of “I” being connected. Thus, in Buber’ philosophy of dialogue, the encounter between I and Thou is not only about linguistic and discursive dialogues, but it is also affective and material, which is about the space “in between” the subject and the other (Zank & Braiterman, 2020). These dialogical and affective features resonate with the concept of assemblage. An assemblage is understood as a temporary arrangement between different heterogeneous elements (including practices, materials, affect, discourse, ideas), which emphasizes in-between-ness and is always in process (Hipfl, 2018).
In the thesis, assemblage theory serves as a methodological perspective that brings to light human-robot relationships in dynamic networks. It highlights the forces within an assemblage: the
intertwining with social constructions and shaped by individual sensations and experiences. Applying a Spinozist-Deleuzian approach, I view the moment of human-robot encounter as a result of different forces interacting with each other inside a dynamic assemblage. More specifically, I identify the points of encounter at which relationships between humans and robots are produced, perceived, and interpreted, and trace human-robot interactions within dynamic networks.
I choose to focus on human-robot interactions through popular Japanese science fiction works and speculative designs. With knowledge of all relevant fictional works and material designs1, I
particularly examine science fiction anime Chobits (CLAMP, 2002) and Plastic Memory (Fujiwara, 2015). Both of both of these are pioneer works depicting robots as family members and have gained great popularity2. I also read the milestones of speculative designs, such as the early generation of
humanoid robots WAMOEBA and TWENDY-ONE, developed by Sugano Laboratory of Waseda University, and the currently most advanced humanoid robots invented by Ishiguro Laboratories of Advanced Telecommunications Research Institute International (ATR). The latter has been widely reported on not only in Japan but also in the English world (e.g. Calugarearu, 2017; Mar, 2017; Warburton, 2017).
Science fiction anime and speculative designs are important for my thesis because, firstly, while the field of robotics has witnessed dramatic growth in recent decades, most humanoid robots still exist in the form of blueprints or speculative designs. Secondly, science fiction and speculative designs offer not merely techno-utopian fantasies, but rather “collectively held, institutionally stabilized, and publicly performed visions of desirable futures, [which are] animated by shared understandings of forms of social life and social order attainable through, and supportive of, advances in science and technology” (Jasanoff & Kim, 2015). These imaginaries do not emerge in isolated works, but rather flow through different themes and forms, as they are shaped by robotic innovation policies, financial support from the government and corporations, and constant interactions and negotiations with the audience. Hence, I have chosen science fiction anime and speculative designs not only because they explore possible directions for future large-scale humanoid robot production (Johnson, 2011), but also because they allude to heterogenous future aspirations and practices that reveal normative as well as new, alternative potentials to construct ¨the sense of the world and the relations it holds” (Bendor, 2018, p. 133).
1 Websites such as Ruijianime and Anikore offer nearly complete lists of anime about AI and robotics, whereas websites such as Robotstart provide a list of robots that are currently available in market or laboratories. My research has benefited from them to map the field of robotics-related science fiction anime and speculative designs.
In this thesis I do not distinguish between texts (science fiction) and physical materials (speculative designs), but rather read them across the board as providing imaginative contexts for human-robot interaction. Understanding both science fiction and speculative design as parts of socio-technological imaginaries, I seek to interpret the details of imaginative human-robot interactions and the social contexts in which such interactions take place. I am able to do so because, first, the interactions between humans and existing speculative robots — even those with machine-learning capabilities — are scripted. The “actual” as well as fictional interactions are thus both shaped by pre-prepared imaginative texts. Second, both science fiction and speculative design need to deal with the physical principles and the technological-material features of robots when they conceive the specific scenes for human-robot interactions. In other words, human-robot relationships constructed in science fiction and speculative design involve both discursive and material dimensions.
Based on the above approach, my study of human-robot relations leads me to focus on the globally influential Japanese imagination of robot subjectivity and relationality. To analyze an assemblage of speculative texts and designs, I develop my thesis around the key concept of “the mechanical heart”; while humanoid robots are conceived to possess “hearts” that humans can feel, relate to, and resonate with, the robots’ mechanical features are also highlighted to distinguish them from humans. Building on the fact that humanoid robots are situated on a middle ground as both technological objects and communicative subjects, my central argument is that robot subjectivity is conceived to be parabolically similar to but ultimately different from humanity. On the one hand, humanoid robots are infinitely close to humanity, which is necessarily confirmed through the
physical, cognitive, and emotional closeness – namely, intimacy – generated during the interactions between humans and robots. On the other hand, the technological features of robots, in the social imagination, also function as a testimony that robots are inferior members of family and human society who should be harnessed and controlled by humans. That is to say, the relationships between humans and robots involve both seemingly mutual attachment (intimacy) and a human-dominated power dynamic.
In this thesis, I develop the above argument through four chapters. While the first two chapters aim to bring theoretical and social contexts to my argument, the third and fourth chapters develop the central argument by investigating robot subjectivity and human-robot relationality as imagined, conceived, and designed in science fiction anime and speculative designs. In the first chapter, I enter into dialogue with existing human-robot research, attempting to provide a posthuman theoretical framework for the development of robot ontology that breaks with the dichotomy of humans and robots. I argue that humanoid robots are conceived to possess some level of subjectivity and, in order to do so, they are designed to be necessarily intimate with humans. In Chapter 2, I narrow my view to robotics in Japan and analyze how the imagination of human-robot interactions
unfolds within the specific history and cultural background in Japan. In particular, I focus on the process of “taming” robotics, in which robotics technology is construed as a part of Japanese culture and institutionalized economically and culturally as a way of maintaining the existing social norms. It is because robotics technology is understood in a timeless Japanese tradition and conceived to sustain the existing systems that robots can be embraced as members of family and human society. In Chapters 3 and 4, I discuss the features of robot subjectivity and human-robot relationships, respectively, in both their discursive and affective dimensions. In the third chapter, I trace the discourse of the robotic “heart,” which in a Japanese context refers to subjectivity, and argue that in order to enable humans to “feel” the robotic heart, robots are designed to be intimate with humans – physically, cognitively, and emotionally close to humans. The fourth chapter then argues that the intimacy which is imagined to exist between humans and robots is both mutual and human-dominant.
Chapter 1
Robot, An Intimate Machine
Designed human-robot intimacy
The development of social robots challenges the conception of dichotomous human-robot relationships. According to the Springer Handbook of Robotics, social (or sociable) robots “are designed to interact with people in a natural, interpersonal manner – often to achieve positive outcomes in diverse applications such as education, health, quality of life, entertainment,
communication, and tasks requiring collaborative teamwork” (Breazeal, Dautenhahn, & Kanda, 2016, p. 1935). They are often designed to have zoomorphic and/or anthropomorphic figures, of which3
those with human-like appearances are also called “humanoids” or “androids”. With many different names such as service robots, partner robots, communication robots, or human-symbiotic robots, the ultimate goal of social robots is to take on the roles of members of society who co-exist with humans and participate in their daily lives. This requires them not only to understand human intentions and emotions on a cognitive level, but also to communicate with people in verbal and non-verbal forms, and even to resonate with humans’ feelings. More importantly, in this type of design, robots are no longer seen as mere metal slaves or servants, but as considerate and reliable partners (Dautenhahn, 2007).
Recently, the ambition to push robots to join human society has manifested in an increasing number of social robot designs which aim to offer communication services or task-oriented support in daily life. Social robots are designed to offer care or assistance to humans, taking on roles such as teachers (e.g. robot Nao) and nurses for the elderly and children (e.g. robot Pepper). They are also designed to conduct tasks as alternatives to human labor such as receptionists (e.g. robot Pepper or Junko Chihara), TV announcers (e.g. robot Erica), chefs (e.g. robot Moley), or astronauts (e.g. robot Kirobo). Other robots are designed to provide companionship and even help humans rejoin society. For instance, some of them are supposed to be as “friends” for children and the elderly (e.g. robot PALRO, Telenoid, or RoBoHoN) or therapists to support for children with autism (e.g. CommU). In order to realize the above various robotic functions based on social interactions (e.g., assistance, support, companionship), a plethora of literature has suggested that “human-like” language usage, behaviors, and emotion expressions are crucial features in building social robots. This
understanding links to a vision of social robots that express themselves and respond with a
human-3 Note: in the thesis, when refer to humanoid robots, I use relative pronouns such as who or whom and third personal pronouns such as they, them (when a robot is designed as being gender neutral), or gender-specific
like manner in human-robot communication. This has led to robotics research on social scripts (C. Edwards, A. Edwards, Spence, & Westerman 2016), pre-designed linguistic triggering cues (Xu & Lombard, 2016, as cited in Westerman et al., 2020), vocal fillers in human-robot interactions (e.g., um, uh, ah, hmm, etc.) (Goble & C. Edwards, 2018), vocal features (human voices vs. synthetic voices) (Xu, 2019), and a sense of humor (Rossi, Cimmino, Matarese, & Raiano, 2019). Other studies focus on building non-verbal expressive features for social robots, such as behavior generation (Arkin, Fujita, Takagi, & Hasegawa, 2003), kinesics, haptics, and proxemics-related robotics
development (Brooks & Arkin, 2007), as well as physical body designs linked to sensory connections (Kerruish, 2016). Similarly, robots’ ability to express their “emotions” are considered as another essential features of social robots (Kirby, Forlizzi, & Simmons, 2010).
In these series of studies, roboticists are dedicated to not only creating social robots that talk like humans but also ensuring the robots to act, behave, and express themselves in a humanoid way. Those human-like linguistic cues and expressive features are considered essential for social robots, as they enable the capacities deemed necessary for robots to establish intimate and effective social interactions with humans (Breazeal, 2003; Duffy, 2003; Fong, Nourbakhsh, & Dautenhahn, 2003; Dautenhahn, 2007).
More importantly, with robots capable of social cues, humans are able to perceive and respond to robots as social entities. The interactions between humans and robots are reported to become more natural and understandable from the users’ side (Breazeal, 2002). In other words, social cues enable the users to acknowledge robots’ social presence, and hence to experience “virtual (para-authentic or artificial) social actors […] as actual social actors in either sensory or non-sensory ways” (Lee, 2004, p. 44). Following this idea, apart from equipping social robots with more human-like expressive and responsive features, roboticists also try to develop ways to enable humans to perceive social robots’ social presence so as to establish effective human-robot communication.
The idea of strengthening robots’ social presence derives from the theoretical framework “Computers as Social Actors” (CASA) in the field of human-machine communication. In 1990s, based on a series of empirical studies investigating the interactions between humans and
computers/televisions, Byron Reeves and Clifford Ivar Nass found out that humans tend to project their social rules and expectations into their interactions with machines (1996). Specifically, they reported that during the interactions, humans impose gender stereotypes onto computers and even ethnically identify computer agents; humans can also apply politeness and reciprocity into the communications with computers (Nass & Moon, 2002). Based on their research, Nass and his
colleagues proposed the theoretical framework CASA to understand humans’ social response towards machines on a behavioral and social-cognitive level. This framework has shed light on the
since been applied to analyzing the interactions between humans and communicative AI agents or social robots (e.g., C. Edwards et al., 2016; Kim, Park, & Shyam Sundar, 2013; Spence, A. Edwards, C. Edwards, & Jin, 2019). Roboticists believe that social cues – such as certain bodily reactions, communication patterns, and emotional expressions – increase the chances that humans “attribute motives and personality to the animated” non-human objects and therefore establish connections by anthropomorphizing artificial entities (Epley et al., 2007; Von der Pütten et al., 2010, as cited in Küster et al., 2020).
The above two approaches to building effective human-robot communications – designing social robots with humanoid features and enhancing humans’ perception of robots’ social presence – demonstrate that effective human-robot communications are conceived based on the robots being recognized as social entities. The core of human-robot relationships therefore lies in whether social robots are able to be sensed, recognized, and experienced as subjects who are not only physically similar to humanity but also cognitively and emotionally related to humans. This leads to designs of social robots’ mechanical bodies that resemble physical features of humanity, such as conversational properties, including triggering cues, vocal fillers, or voice features that help improve the robot’s responsiveness on both a cognitive and a material level, and the emotional expressions that stimulate the connections between humans and robots in the emotional dimension. In short, the designs of social robots are dedicated to establishing physical, cognitive, and emotional connections between humans and robots.
It is worth noting how closely these conceptions and designs of social robots resonate with the concept of intimacy. Despite the complexity of definitions of intimacy, this thesis adopts the terminology of human-computer communication, understanding intimacy as closeness that transcends narrow ideas linked to sex but is connected to individual sensation, feeling, and lived experience. More specifically, my research develops the definition proposed by Genevieve Bell et al. in their research (2003, as cited in Borenstein & Arkin, 2019, p. 300) that investigates the role of media technology, including computers and wearable electronic devices, in establishing intimate relationships between humans. For them, technologically enhanced intimacy encompasses the following:
• Cognitive and emotional closeness with technology, where the technology may be aware of and responsive to our intentions, actions, and feelings.
• Physical closeness with technology, either on or within the body.
• The use of technology to express our intentions, emotions, and feelings towards others (p. 300)
In this definition, intimacy takes two forms through media technology. Firstly, it refers to the connections between humans that are enabled by technology, which helps individuals to identify
participates in shaping intimate relationships but is not the subject of intimacy. Secondly, it also refers to the intimacy between humans and technology. As technology becomes ubiquitous, permeating personal spaces and intertwining in everyday practices, it exists in individual’s private space, interacts with our bodies, and thus becomes intimate to people in the physical and material dimension.
Moreover, being as interfaces, it also modulates humans’ sensations, emotions, and knowledge of intimacy, which makes it cognitively close to humans.
Both layers of the definition can be applied to understand human-robot relationships, as robotics technologies here not only serve as technological interfaces that enable the communications between humans but also communicative subjects who directly establish connections with us. It means that the intimacy enhanced by robotics technologies can be presented as physical, cognitive, and emotional closeness between humans and robots. This kind of closeness echoes exactly the visioned human-robot relationship in the field. As I elaborated above, roboticists believe that in order for social robots to fulfill their designed function, they should resemble humanity and enable humans to perceive their social presence in physical, cognitive, and emotional dimensions. That is to say, the
relationship between humans and robots is necessarily intimate by design.
Roboticists believe that, with intimacy, the collaborations between humans and robots can become more reliable (e.g. Bethel, Stevenson, & Scassellati, 2011; Birnbaum et al., 2016; Kahn et al., 2015). When it comes to human-symbiotic robots that offer assistantship in daily life, intimacy is considered essential to enable these social robots to fulfill tasks such as curing depression (e.g. Abdollahi et al., 2017) or providing companionship for elderly people (e.g. Okubo et al., 2018) or children with autism (e.g. Richardson, 2018). Here, intimacy is seen as an effective means to build trust and friendly interaction with certain target groups and thus it becomes a main element in designing social robots. As Jason Borenstein and Ronald Arkin assert, “the design of robots will progress in such a way that the technology will effectively be able to establish intimate relationships with a broad range of human beings” (Borenstein & Arkin, 2016, p. 301).
Intimate robot, a dilemma
In an emerging world where robotics coexists with humans, one of the first questions is how we can deal with the (intimate) relationships with robots. Namely, what are the features of the (intimate) relationships we establish with robots, if they are not merely inert tools for human communication but are designed to provide physical services and emotional companionship as communicative subjects? What is the difference between this type of intimacy and its counterpart intimacy between humans?
The question of intimate relationships between humans and robots was first addressed by scholars in robotics ethics. This field has noted the dualistic, and even seemingly contradictory, properties of human-robot relationships. On the one hand, robots are used by humans as instruments, since the primary function of human-symbiotic robots is to provide services. The relationship between humans and robots in this case is subject-object. On the other hand, because of the social nature of their services, such robots are also designed (or are expected to be designed) to have the capacity to establish emotional connections with humans. In this case, a machine communicates with humans as a subject, and the relationships between them potentially becomes subject-subject. In order to deal with the situation that robots can be both instruments and subjects, two tendencies have emerged in robot ethics studies. Firstly, robots are designed to be anthropomorphized, which poses the risk of building (pseudo-)intimate relationships as a substitute for intimacy between humans. Secondly, the
relationships between humans and robots are essentially instrumental, which could lead to humans objectifying and instrumentalizing their social connections. In both of these arguments, the absolute distinction between robots and humans serves as a basic premise, where robots exist as
(anthropomorphized) objects while humans as subjects.
Building on the dichotomy of human/non-human and subject/instrument, robot ethics scholars claim that the intimacy between humans and robots is inherently unilateral, with humans projecting their own emotions onto robots (e.g. Scheutz, 2012). To achieve this projection, robots are designed to be human-like in the aspects of appearance, embodied features, social behaviors, etc. (e.g. de Graaf, 2016). Robots imitating humans are deemed to enable the human users to blur the boundaries between robotic objects and human subjects. In this way, they can perceive the subjective presence of robots by attributing their experience (e.g. sensory experience including feelings and emotions) and agency (e.g. capability of taking actions such as using language, logic, and planning) to robots (H. M. Gray, K. Gray, & Wegner, 2007). Since there is only one side (human) that possesses “real” subjectivity, the relationship between humans and robots is not mutual, but unidirectional – only humans in this kind of relationship can perceive their own existence. Humans’ singular, disembodied, and isolated subjectivity prevents the possibility to establish authentic intimate relationships between humans and robots, which produces a series of “distorted models of relationality” (Richardson, 2019).
In this framework, the mechanisms of producing human-robot intimacy are understood in two keywords: anthropomorphization and instrumentalization. On the one hand, the
anthropomorphization of robots is considered a premise of human-robot relationship. As shown above, robot ethics scholars’ discussions about the consequences of the “one-way” relationships with robots are based on their assertion that this type of intimacy is established by humans projecting their own experiences and agency. Such “pseudo” relationships contain the risk of replacing the
design level, the arguments about one-way intimacy link to the concerns that robots are designed to “seduce” humans to build emotional connections. For instance, the design of human-robot
relationships may mimic the psychological process of “human-to-human binding phenomena” (Kidd, Taggart, & Turkle, 2006), and thus possess the potential risk for designers to exploit humans’
emotional vulnerability (Nyholm & Frank, 2019), influence consumer decisions by emotional manipulation (Scheutz, 2012), or deceive humans (N. Sharkey & A. Sharkey, 2010). At the social level, the understanding of the “pseudo” relationship is presented in discussions of the risk of people projecting their emotions onto objects (robots). A common argument is that the attachments between humans and robots may prevent humans from socializing with each other (e.g. Turkle, 2011). This type of statement of robots replacing humans has been studied in different demographic segments. For example, scholars argue that when interacting with robots, children can project their emotions onto robots and therefore have the “illusion” that robots care for “me”. Children may therefore become overly dependent on robots and jeopardize the development of their social relationships (e.g., Borenstein & Pearson, 2013). Similarly, using service robots to take care of the elderly may alter in the care practices and even threaten human dignity (e.g. Laitnen, Niemelä, & Pirhonen, 2016).
Despite the above concerns that robots may be too humanoid, the relationships between humans and robots can be understood as essentially instrumental. The risk of this instrumentalized human-robot communication lies in the possibility that it may retroactively shape the way humans interact with each other, resulting in the social connections between humans being objectified and instrumentalized (e.g. Wales, 2020). For example, concerning human-robot intimacy (including love, desire, and sex), praise for the convenience to humans is co-presented with criticism of fetishism. As a scholar who has actively written about robot sex in the recent decade, David Levy states that sex robots will be widely accepted within fifty years. In his book Love and Sex with Robots Levy approaches the intimate/sexual relationships between humans and robots according to the idea of fetishism. He uses historical materials to explain that humans can accept robotic partners because we are capable of establishing attachments with non-human objects. He therefore believes that humans will embrace robots: now that humans are able to establish the attachments with robots, robots can satisfy humans’ needs better (than humans) based on their precisely designed functions. In response to Levy’s statement, Richardson responded that treating robots as “things” and unfolding intimacy merely in a functional level may lead to the objectification of intimacy (2016).
Current robotics ethics studies provide an approach to understanding the robot ontology according to a human/machine, subject/object dual paradigm. They indeed recognize and respond to the phenomenon that human-symbiotic robots are conceived and designed to be intimate with
humans, while at the same time they are supposed to function as tools. However, the dichotomy set up by robot ethics studies fails to unfold the complexity of human-robot relationship. Within the
paradigm of human/machine, subject/object dichotomy, the answers to the question of whether robots can be successfully anthropomorphized correspond to two contradictory orientations: when the answer is “yes”, the concerns focus on robots replacing humans and establishing intimacy with humans, as they are “too humanoid”; when the answer is “no”, concerns arise that, due to robots’ “essential” nature as instruments, they are not only unable to establish intimate relationships with humans, but also present the risk of objectifying intimacy itself.
I juxtapose the two opposed arguments here not to either argue for or against them. Instead, I mean to highlight the (seeming) duality of robotics. In other words, the two types of risks (or
possibilities) possess the potential to exist at the same time in one human-robot relationship. In the above case of robotic partners, the opposing arguments that robots may replace humans as intimate partners (e.g. Levy, 2007) or cause humans to objectify intimacy (Richardson, 2016) demonstrate humans’ abilities to perceive robots’ subjective presence while using them as tools. This echoes with Peter H. Kahn and his colleagues’ empirical study (2006) on preschool children’s interactions with AIBO, a robotic dog produced by Sony. Studying 80 preschool children, the research recorded the subject’s psychological and behavior responses towards the robotic dogs, in order to investigate how the children can perceive and understand these artifacts. The research found that when the children were talking about their interactions with AIBO, they tended to describe the robots as stuffed dogs, but when they were actually interacting with AIBO, the way they responded to the robots was closer to what they would react to living puppies. Similarly, in their observations of children interacting with robots such as My Real Babies, Aibos and Paros (sic), Sherry Turkle et al. found that while children portrayed the robots as close companions, they also used technical terms (e.g., batteries, maintenance) to explain the robots’ present state or behaviors (2006). In these cases, the robots were treated as “relational artifacts” (ibid); in other words, they were viewed as artifacts, which also had the potential to establish attachments with humans.
In fact, the decision to view a robot as an intimate subject or an instrument depends on the interactions between humans and the robot. In other words, robot subjectivity is relational and must be located on a spectrum between intimacy and instrumentality. The location of robot subjectivity depends on a series of slips, dislocations, and transformations within human-robot interactions. This idea of relational subjectivity also resonates with Martin Buber’s philosophy of dialogue under the framework of “I-Thou” and “I-It”. In the following section, the thesis will borrow Buber’s
philosophical framework to develop new ontological understanding of robotics that transcends the dichotomy of human and non-human. In doing so, the thesis aims to examine the process of forming robot subjectivity and human-robot relationships within dynamic interactions.
Relational ontology of robotics
In his short but inspiring work I and Thou (1923/1970), Buber explores our relation to others – including humans, animated things, and the Divine Thou. He believes that relations exist as “encounters” and develops a dialogical framework of I-Thou and I-It. “I-It” refers to the type of relationships when the subject treats the object as an instrumentalized other, applying their empirical experiences to recognize the object and utilizing their knowledge to manipulate it; on the other hand, “I-Thou” points to authentic relationships between subjects, when the subject encounters the other with their whole existence and realizes themselves in the process of relating to others. In particular, when “I” meet “Thou”, I am no longer a subject that experiences and manipulates a certain object; my relation to “Thou” is not driven by any of my needs. In this I-Thou relation, the encounters cannot be captured by any instrumental language. It is the mutual and holistic relation between two authentic subjects that exceeds the Cartesian relation of subject and object, and leaves space for the ineffable intimacy between subjects.
In the I-Thou and I-It framework, the hyphen is crucial. As the basic elements of a relation, the features of “I”, “Thou”, and “It” are not intrinsic, but rendered at the moments of “dialogue” and “encounter”. Namely, a subject is not isolated, but defined through the dynamic interactions with others. In the I-Thou relation, both subjects can only be spoken as a whole: they fully depend on, completely engage with, and mutually acknowledge each other’s existence. As Buber puts it, “who says You does not have something; he has nothing. But he stands in relation.” (p. 55). On the other hand, in the I-It relation, “It” as a thing only exists when it bonds with the other: “it is only by virtue of bordering on others” (ibid). That is to say, “I” am situated in open-end relations which are shaped by certain “encounters”. The contingent feature of encounters determines that one’s relations with a certain other can switch between I-Thou and I-It. Rather than reducing human relations to simply either/or, Buber’s two models, the foundation of our relations to others, are interchangeable and always dynamic.
Buber emphasizes understanding subjectivity and relationality through encounters. The key point is that, for Buber, the I-Thou and I-It relations are not discrete or abstract concepts, but are ontological reality (Zank & Braiterman, 2020). In other words, our relations to the world are not intrinsic, but are constantly happening encounters. This theoretical framework enables him to avoid the dichotomy of subject/object and human/non-human, and to locate subjects on the continuum between human and nature. John Shotter believes that this type of relational ontology is what is lacking in the field of robotics (2019). He argues that the current research in the field only focuses on what robots can or cannot do “at the moment”, according to a Cartesian relationality between
independent, countable entities that relate to each other logically. However, this mechanistic order overshadows the “relational things”. The term refers to those who do not belong to a current certain
state, but are “in the transitions from one state of affairs to another” and point to future potential. He also calls them “hermeneutical-dialogical things”, and notes that they can only exist in the dynamic relations that are constantly (re)established between us, others, and the environment. Therefore, Shotter argues that only talking about the current state and functions of robots will fail to consider our living bodies, “the dialogical nature of […] our living relations to the rest of our surroundings”, and the dynamic relationships shaped by the entanglements between ourselves and others.
In this light, Buber’s philosophy of dialogue provides a more dynamic perspective to understand human-robot relationships, possessing the potential to transcend the rigid dichotomy. It helps to investigate the process of interactions in which humans and robots participate in shaping and confirming each other’s existence, with a dimension of temporality involved. Based on this
framework, Beata Stawarska explores the relationships between humans, artificial intelligence (AI), and robotics. In the case of an interactive robot called Pepper, Stawarska argues that, although Pepper is not viewed as a fellow human because of its technological features (e.g. installed with cameras and a screen), it still possesses the potential of becoming a subject. The emotional affirmations and comforts that flow between humans and robots demonstrate that the I-Thou relation can happen between human and machine (2019). In this case, the author does not focus on whether Pepper possesses properties similar to humans (he also argues that it will cause uncanny effects when robots’ appearances are too humanoid), but rather emphasizes the interactions themselves and their
consequences.
The subjects Buber mentions as potentially involved in I-Thou relations are not only humans, but also animated things (e.g. tree) and reality (e.g. consciousness, god). While it is not possible to know what Buber would have thought of robots, some scholars believe that the relationships between humans and robots can only be “I-It”. As Kathleen Richardson argues, since these artifacts are invented for the purpose of fulfilling instrumental functions, even if they are meant to reach the level of self-sufficiency, this type of relationships is ultimately different from those between humans (2019). The author compares the relationships between humans and robots with the attachment model between feral children and robots and claims that “the absence of interaction with other human beings renders man behaviorally non-human” (p. 78). It means that even if the attachments can be
established between humans and robots, humans in this case cannot be considered as “I”. In other words, for Richardson, human-robot relationships in I-Thou model are not possible. However, that the relationships between humans and robots are different from those between humans does not mean that the human-robot attachment cannot be perceived, sensed, and experienced. Assuming the
inauthenticity of the relationship between humans and robots simply because they are different from the human-human counterpart not only neglects humans’ possible psychological state (by devaluing
the relations as being not “real” or “distorted”), but also loses the chance to understand the potential of human-robot relationships. More importantly, to what extent can the concept of humanity be pure?
While the subject that Buber refers to includes humans, animated objects, and god, the hybrid subject that posthuman theorists focus on helps to expand the connotation of being a subject.
Posthuman theories argue that a human is not a pure, natural existence, but is open to technological modification (Gane, 2006) and may become entangled in a complex of human and machine
(Pepperell, 2003). As Donna Haraway notes in her influential work Simians, Cyborgs, and Women: The Reinvention of Nature (1991), the assertion of human as “self-contained being” is not solid. Instead, the boundaries the assertion is based on, between humans and animals, organisms and machines, and the realm of the physical and non-physical, collapse in the face of the cyborg. The concept of relational subjectivity that posthuman theories insist on focuses on technologically modified bodies, challenging a series of fixed boundaries between bodies, humans, and technologies. For posthuman scholars, a “human” is produced through a series of vital forces on the interface of the body (Åsberg, 2009). In other words, the posthuman body does not end at the skin, but itself becomes an interface which enables and is shaped by the entanglements between diverse, heterogeneous “biological/natural and material/technological actors” (Duff, 2018, cited in Andrews, 2019).
The posthuman discussion of relational subjectivity, from a technological perspective, makes it possible to further interrogate the materiality of subjectivity. As Katherine Hayles (1993) analyses, although information technology and embodiment tend to be divided into two distinct realms of the immaterial and material, the computer-based information system cannot be an abstract concept independent from materiality. Rather, it points to an embodied virtuality, a new form of subjectivity that is born at the interface between bodies and technologies (Gane, 2006, p. 432). This argument not only reminds us that the state of “being human” can be produced when bodies meet machines, but also that it is important to think subjectivity together with materiality. In recent decades, an increasing number of scholars have drawn inspiration from Gilles Deleuze’s works. In particular, the Deleuzian notion of assemblage opens new possibilities for posthuman scholars to understand subjectivity within (material) relationality (Deleuze & Guattari, 1987). For example, using a Spinozist-Deleuzian
approach, Rosi Braidotti argues that subjects“compose a relational community, defined as a
nomadic, transversal ‘assemblage’ that involves non-human actors and technological media” (2019, p. 32). Here, an assemblage is understood as “a relation” (Buchanan, 2017, p. 465), a temporary
arrangement of heterogenous elements (such as discourses, materials, affects, and practices). Inside an assemblage, the elements can be constantly plugged in and out, and the contingent connections between elements form the assemblage itself (Hipfl, 2018). Hence, subjectivity in the posthuman theories with a Deleuzian approach is understood not as a result, but as a process and a “becoming”.
The dichotomy between humans and non-humans is thus further challenged by posthuman theories, as the linear notion of human nature is replaced by a broader “nature-culture” continuum (Haraway, 2003). Although only a few scholars have applied posthuman theories to robotics-related issues (e.g. Wennerscheid, 2018; Chesher & Andreallo, 2020), the approach has proved helpful in other related fields for understanding bodies as the outcomes of vital forces between material, discursive, social, and aesthetic actors. For instance, it reveals how life technologies, such as geological and biological science, have been embedded in both human bodies and daily practices, shaping the way we perceive the world and interact with each other (Grosz, Yusoff, & Clark, 2017). In terms of robotics research, although from a distinct philosophical tradition than Buber’s philosophy of dialogue, the Deleuzian approach of posthumanism helps us to understand the formation of
subjectivity in “encounters” beyond the human, and to read the encounters in a more concrete way. Such encounters exist not only in the dialogues between subjects and objects, but also in the moments when material, discursive, affective, and social actors meet each other.
Based on both of these theoretical approaches, the thesis aims to establish a new relational ontology of robotics in order to understand what it means to be a (intimate) robot and what the connotations are of building (intimate) relationships between humans and robots. More specifically, revolving around “encounters”, I try to understand robot subjectivity and relationality in dynamic networks, so as to overcome the existing dichotomies between humans and non-humans. Rather than focus on what robots can or cannot do at the moment, I pay attention to the future potentials that are generated through the dialogues, interactions, and inter-shaping between humans and robots (Shotter, 2019). I therefore understand human and robot subjectivity not as being fixed or coherent, but as contingent and open-ended entities that are produced in a series of interactions between the actors. The encounters are therefore not only between humans and robots, but also between the elements that constitute them.
Chapter 2
Embracing Robot-Symbiotic Society in Japan
Japan holds a special position in the field of robotics. As one of the world’s largest manufacturers of robots and accounting for more than half of the global supply (The University of Tokyo, 2020), Japan is leading the way in robotics innovation. In particular, Japan has been actively involved in imagining, defining, manufacturing, and marketing social robots, promoting their participation in human social functions such as domestic services, medical care, nursing care, hospitality services, etc. (Statista, 2019). These efforts in Japan are accompanied by its reputation as a “nation of robots” (Schodt, 1988). It is known as a place where people love machines and treat them as “priceless friends” (Hornyak, 2006, as cited in Šabanović, 2014). Compared to the West, Japan is widely considered as having a high acceptance of robots and actively embracing them as members of human society (Zeeberg, 2020). This positive attitude towards robots is situated in the unique socio-cultural history of Japan. For instance, it is argued that national technological traditions, Shinto animism, and popular culture have profoundly influenced attitudes toward robots (e.g., Kaplan, 2004, 2011; Kitano,
1970/2006; Jensen & Blok, 2013).
In this chapter, I aim to trace a history of robotics in Japan. Rather than simply searching for the historical roots of Japanese robot culture, I question how the narrative of a robot-symbiotic society (“人とロボットが共存する社会” in Japanese) is embraced by retelling a robot tradition that
emphasizes a homogeneous, continuous, and linear “Japanese-ness”. In particular, I focus on the way that the future-oriented robotics technology has been tamed into alignment with a linear Japanese history (Kaplan, 2004). I argue that as robotics technology is considered to belong to the cultural tradition, the vision of a futuristic robotics society is therefore comfortably linked to the past. Thus, robotics technology, as a part of Japanese culture, is conceived to preserve the existing social functionality and normativity. Moreover, apart from defining the expected functions of robots, the narrative of robotics as a Japanese tradition also legitimizes the imaginary of robots as society members. By referring to animist culture, robots are justified as animated things that have long been co-living with humans. Here, I attempt not only to explore how the worldwide imaginary of symbiotic robotics was influenced by certain Japanese socio-historical situations, but also to provide the
necessary context for my later analysis of robot subjectivity and the human-robot relationship. I also remind my readers that this imaginary/philosophy of symbiotic robots should not be marked as “Japanese” – even though this is how Japan is trying to brand itself – nor be treated as the Other to the English world. Rather, as this type of Japanese imaginary/philosophy is constantly in dialogue with the world, it should also be positioned in a dynamic global system. Therefore,
investigating the socio-cultural background in the Japanese case helps us to understand, on a more abstract and theoretical level, how the ideas about robots as family/society members are shaped by certain political, economic and cultural synergies.
Robotics as a Japanese tradition
The history of Japanese robot culture is often traced back to the Edo period (1603-1867). According to 2014 White Paper on Robotization of Industry, Business and Our Life, published by the New Energy and Industrial Technology Development Organization (NEDO), the origin of robots can be traced back to automata, a type of wooden puppet driven by spring, clockwork, gears and linkage mechanism. This technology was introduced from the West during the Edo period and then developed independently by local craftsmen (Shirai, 2002). The major outcome of the development of
indigenous mechanical techniques was Karakuri ningyō (“からくり人形” in Japanese), local-designed clockwork wooden dolls. These puppets were mostly used for entertainment, including serving tea, performing dances, shooting arrows, or participating in religious ceremonies. They became a popular form of technological entertainment since the 17th century (Satō, 2012) along with
the rise of the merchant class and consumer culture of early urban society (Sheldon, 1983).
Nowadays, karakuri ningyō has been viewed as an important cultural heritage in Japan. The narrative to trace the robot history back to karakuri ningyō is not only found in the NEDO 2014 white paper, but frequently in both academic literature and cultural reports (e.g. Matsuhira, 2006; Suematsu, 2006; Hotta, 2009; Satō, 2012; Doi, 2015).
Figure 1. Automata, the origin of robots
The three illustrations show, respectively, Western automata, Japanese tea serving dolls, and the cover of Karakuri Zui (“Illustrated Compendium of Clever Machines”, Hosokawa, 1796), the first publication in
Japan to document detailed instructions to construct clocks and automata. The sequence of the three illustrations also suggests a process of localization of the technology (NEDO, 2014, pp. 1-4)
However, while born from cross-cultural technological exchange, the origin of robotics – clockwork technology – is recognized and emphasized as being “Japanese” and “unique”. Most scholars in Japan pay little attention to the precise origin of the clockwork technology, only lightly mentioning that it “was imported from the West during the Edo period”. Instead, the cross-cultural background often functions to prove the “indigenous talent” (Kovacic, 2018, p. 578). Compared to the origin, the national isolation policy implemented soon after the technology importation is highlighted. Historians argue that due to the isolation policy, clockwork technology was developed in Japan independently and integrated into Japanese culture via the joint efforts of local craftsmen (e.g. Japan Clock & Watch Association, 2020). To support this narrative, the “uniqueness” of Japanese
clockwork technology is emphasized. For example, Yasuyuki Shirai (2002) argues that, although Japanese craftsmen had begun to learn the techniques by imitating European works, they gradually moved away from manufacturing Western-style replicas and created their own characteristics such as applying Japanese time-keeping methods. Moreover, the distinct feature of Japanese mechanical technology is also underlined. Yoshikazu Suematsu (2006) believes that the greatest property of Japanese clockwork dolls is that they are designed to be social. Comparing Japanese tea-serving dolls with the European counterpart around the same time, he argues that the Japanese ones are unique in
dolls are considered to have been used on social occasions where the host would serve tea to other guests, their actions are designed based on the human-symbiotic environment; the mechanized dolls function to draw the guests’ interest and facilitate human communication (p. 4).
Research on the technological origin of robotics in Japan does not only seek to emphasize its “Japaneseness” and “uniqueness” on a technological level, asserting that the Japanese clockwork designs are rooted in a social human-symbiotic environment. In addition, on a cultural level, it also seeks to construct a cultural tradition, where animated objects live together with humans and
participate in humans’ religious and social life. Narumi Satō, for example, describes the Japanese tea-serving dolls in the following way on Nippon.com, a well-known, multilingual website for the promotion of Japanese culture:
In contrast to the cold, mechanical impression of an automaton, karakuri ningyō inspire in many people an emotional affinity or even a feeling of love. The dolls seem less like a decoration or toy than something with a heart and life force of its own. (2012). Such narratives attribute the uniqueness of clockwork dolls to their ability to establish
emotional connection with humans: a doll is not merely an object, but a thing that contains vital forces and triggers human attachment. This idea is often legitimized by referring to Japanese religious culture and theatre practices. The Shinto tradition of animism in particular is considered to be fundamental to this idea. Rather than adhere to the boundaries between humans, animals, spirits, and machines, animism believes that humans co-live with their animated counterparts and that the life force can pass through one to another (Jensen & Blok, 2013). In addition, traditional theater is also tied to the uniqueness of Japanese clockwork dolls. This technology is widely used in a ningyō jōruri (“人形浄瑠璃” in Japanese), a traditional Japanese puppet theatre that has flourished since the Edo period (Suematsu, 2006). In this traditional cultural arena, the artisans create puppets as humanoid as possible in stories about love and devotion; the puppets enact life and express emotion, and the audience empathizes with the puppets performing on the stage. Through the practices of making, performing, and watching, the artisans, the puppets, and the audience work together to enhance the projection of emotions and to create the meaning of the puppets as animated objects (Natori, 2006).
Through this historical reinterpretation, robot technology that derives from transnational innovation has been transformed into Japanese cultural tradition, with particular emphasis on the features of being human-symbiotic and animated. It marks this modern technology as a collective effort of Japan’s technological innovation throughout the centuries and thus justifies this technology as a part of the cultural essence (Šabanović, 2014). In other words, the reconstruction of
“Japaneseness” enables robotics to avoid being understood as a technological other but to embraced as a part of material culture.
Linking the new technology with traditional culture and art is meant to establish natural connections between Japanese technology, Japanese people, and Japanese society. Firstly, this practice is expected “to give the public ‘a deeper understanding and awareness of science and technology’ and to improve their adoption and acceptance in society” (Government of Japan, 2006, pp. 61–62, as cited in Šabanović，2014). Discourses such as “karakuri ningyō as the origin of robotics” or “robotics as a Japanese tradition” function as historical evidence: now that robotics has begun to participate in defining and shaping Japanese culture – a culture based on the ideas of animism and human-machine symbiosis – this technology should be accepted as a part of daily social/cultural practices. Secondly, the historical reinterpretation of robot technology has legitimized it as “a part of national industrial plans to innovate Japanese society and invigorate the economy” (Kovacic, 2018, p. 573). With joint efforts among the government, corporations, and research institutes, robotics, especially social robotics, became a Japanese trademark. Robotics technology is not only accepted in local practices, but also presented as a symbol of Japan’s competitiveness in the global technology market.
Frederic Kaplan conceptualizes the process of integrating robotics technology into Japanese tradition as “technology taming” (Kaplan, 2004, p. 2). This transfers robotics technology, which was born from worldwide dialogue, collective imagining, transnational theorizing, and technology exchange into a part of homogeneous and linear Japanese culture. Through investigating a series of myths and popular cultural products, Kaplan argues that this writing of Japanese culture encourages “the artificial reproduction of nature” and incorporates “an aesthetic dimension to the quest for recreating life-like creatures” (p. 2). While it dissolves the boundary between the natural and artificial (p. 4), it also absorbs the new technological innovation without losing its national and cultural
essence. In this way, modern robotics technology is “mastered and harmoniously integrated in society” (p. 4). Taming this new technology brings with it a promise: in Japan, new machines and technological changes will not influence the nature of humanity and society. Rather, since robotics technology, as a tradition, has already existed in Japanese history for a hundred years, it can comfortably be conceived as a technological solution able to fix existing social problems while preserving social normativity and cultural traditions.
Towards a future of human-robot symbiosis
The historical narrative that tames robotics technology into being a part of homogenous Japanese culture helps legitimize certain futuristic conceptions of a robot society. The promise that (new) robotics technology will not affect Japan’s cultural essence makes it easier to use robotics in daily practices. Particularly, the reconstruction of the tradition of “human-symbiotic” and “animated”
other words, it is because robotics technology is placed in a linear history that robots can be imaged as alternative society members co-living with humans, rather than as a Frankensteinian technology.
The development of robotics in Japan is situated in the concern for Japan’s ageing society. The 1st Conference of Artificial Intelligence 1969, under the theme of “can computers be alternatives
of humans?” (“コンピュータはどこまで人間の代行が可能か”in Japanese, my translation), discussed how the field of robotics should respond to the growing problem of the aging society and labor shortage in the postwar 1960s (Kinoshita, 1995, p. 153). As a milestone, the conference set the tone for subsequent research. Even today, the aging society and labor shortage are still the major motivation for robotics development in Japan. The ICT Conference 2013: Visioning Super Aging Society (“ICT 超高齢社会構想会議” in Japanese, my translation) reports that Japan needs to build a robotics-centered smart society to solve the above two issues. In this vision, different service robots are used to guarantee the physical and mental well-being of the elderly and to help them rejoin the society. Robots are expected to stimulate communication, offer nursing care, and assist the elderly to use social and commercial infrastructures (MIC, 2013). Similarly, the NEDO White Paper 2014 also underlines the significance of building a robotics-symbiotic society. It claims that service robots not only make a contribution to filling the shortage of skilled workers in industry and agriculture (p. 2-21) but also enrich people’s family and social life (p. 2-28).
Meanwhile, roboticists believe that the more a robot resembles a human, the more likely they can become alternative labor. For instance, roboticist Hideki Hashimoto and Kazuyuki Morioka argue that human intelligence and physical bodies demonstrate the value of human resources: “humans possess intelligence and perfect design of biological bipedal bodies as a result of thousand years of evolution. […] With our intelligence and physical bodies, we are the most perfect fit for the
environment” (2001, pp. 268-269, my translation). Hence, they assert that “robots, in order to be the alternative labor, should imitate humans” (ibid). These manifesto-like sentences claim that, in order to fill the labor shortage, robots must necessarily be humanoid. Moreover, for robotics, to be
“humanoid” does not merely mean to possess a human-like appearance. Pioneer roboticist Masahiro Mori (1927-) describes the three levels of imitating humans: essence, appearance, and functionality (2014). For Mori, to design robots with human-like bodies and functions is to reach the ultimate goal of creating robot subjectivity. Robots are eventually expected to become social existences that are able to communicate with humans and to process complex information in a human-symbiotic
environment. Roboticist Shigeki Sugano (1997) also highlights the robotic capacity to “understand the human heart” in the creation of humanoid robots. He argues that, only when “heart-to-heart
Based on the challenges posed by Japan’s aging society and the assumption that humanoid robots can most effectively serve as alternative labor, a vision of robotics society has emerged in which humans and robots co-live with each other. Supported by Ministry of Economy, Trade, and Industry (METI) and NEDO, EXPO 2005 AICHI was held under the theme of “We Live in the Robot Age”. It not only presented a blueprint of robotics society, but also concretized Japan’s national plans on robotics technology (EXPO 2005 AICHI, 2021). In the EXPO, archetypes such as cleaning, guiding, security guarding, and nursing care robots showed the future direction of robotics
technologies promoted by the nation. Since EXPO 2005 AICHI, the idea of robotics entering daily life has become increasingly popular, while human-symbiotic robots have become the major targets to be imaged, designed, and developed in the field (Matsuhira & Ogawa, 2009). Nobuto Matsuhira and Hideki Ogawa summarize the fields targeted by robotics in their Toshiba Special Reports (ibid). They draw two axes – targeted markets and types of robotic operation systems – and divide robot service into four zones: domestic (up) or business (down), operational (left) or autonomous (right) robots (Figure 2). Operational robots are designed to be in professional surgical or industrial scenes, executing single tasks such as walking, lifting, or operating other machines. On the other hand, autonomous robots are supposed to live with ordinary people, conducting daily tasks related to communication and domestic help such as doing housework, enabling communication, or nursing care. The report reveals one major trend in the field – applying the technology to the environment of human-robot coexistence.
Figure 2. Four fields of robotics technologies
From Mastushira, N. & Ogawa, H. (2009). 人との共存環境を目指すロボット技術 [Robot Technologies Aiming for Coexistence with Social Environment]. Toshiba Review, 64 (1), p. 4.
