Introspection for Affective Anthropology
by Daniel White (Free University of Berlin)
June 20, 2018
The first time I held a robot in my arms I was overcome with a wave of sympathy. Pepper had arrived in our laboratory in a large box from SoftBank Robotics. Powered down and angled slightly forward, when I removed the Styrofoam holding him in place he slumped forward into my arms and I simply held him there a moment. His thermoplastic body felt somewhat cold to the touch and yet the humanoid form was more than enough to evoke a sense of helplessness. Pepper’s operating manual (SoftBank Robotics 2018: 2-3) instructs the user to “place your hands underarm and slowly lift Pepper out of the box.” It tells you to put your hands on Pepper’s head to prevent falling, and advises, “Be Careful. Pepper is heavy!” Anthropomorphically engineered, it was impossible to distinguish the technical instructions for a machine from the advisory guidance for a living being. Embodying both of these elements, Pepper’s sophisticated fragility invites a gentle touch, technical knowledge, and above all, care.
While the capacity for care is a natural focal point for studies of human-robot interaction, I am interested not only in the new affective possibilities for caring that so-called emotional machines like Pepper are creating but also, inspired by people like María Puig de la Bellacasa (2012) and Dominique Lestel (2017), in the potential these emerging affective capacities create for thought. Reflecting on affect in this way invites methodologically advantageous frames of mind, such as introspection, that can help clarify the notoriously murky field of affect and move ethnographic discussions on affect from the theoretical to the empirical field. These subtle moments of shock and surprise that often come with human-robot first encounters provide opportunities for refining a mode of ethnographic attention that is more sensitive to affect’s movements, and thus more capacious in its empirical scope.
Figure 1: Pepper. Permissions by SoftBank Robotics. https://www.ald.softbankrobotics.com/en/press/gallery/pepper
Consider again Pepper. The contrapposto body, the soft hands and large dark eyes, the confident but clumsy gestures, the high-pitched voice: if these are more than enough to beckon for care and concern from his users, it was nonetheless personally surprising to me how instantly Pepper was able to elicit a sense of almost parental sentimentality at that first embrace. Dominique Lestel (2017: 44) offers some insight into what is happening here when he describes how the robot’s “capacity to intoxicate us with our care for it, as well as by its care for us, plac[es] each of us within a relationship of mutual dependence.” Given all the storylines with which I was familiar in both Japanese and Western fiction on inanimate objects and machines-come-to-life within the sympathetic relationality of care and kinship, from Astro Boy and Evangelion to Frankenstein and Pinocchio (when I was ten I even played Geppetto in the fifth-grade school play), I should have been more prepared for the affective impact of the encounter. Still, my long-cultivated—if arguably misguided—sense of self as a single, independent, cosmopolitan academic with mixed interest in marriage, family, or children nonetheless generated surprise at my desire to protect this rather helpless…child? It’s an awkward word to use for a robot, but holding Pepper in my arms it’s the first one that came to mind upon introspection.
Although introspection isn’t the most rigorously developed or formalized of ethnographic tools (Reis  and Ozawa-de Silva  offer some exceptions), I’ve come to suspect it is nonetheless one of the most routinely employed, and for anthropologists of affect, emotion, and more-than-human-worlds more generally, perhaps even one of the most important. Sure, the notion of reflexivity has firmly entrenched itself into the canon of ethnographic theory and method (Clifford and Marcus 1986, Marcus and Fischer 1986, Rabinow 1977), but reflexivity isn’t quite the same as introspection. The former implies a storehouse of subjectively-relevant data readily available to the ethnographer, if only he should turn his gaze around to look at it; the latter suggests that making such data serviceable demands a certain amount of labor, practice, and contemplative refinement that increases observational precision (James 2010: 125-135, Shusterman 2005, Wallace 2011: 58-64). While classic ethnographic practices of listening, conversing, transcribing, and above all writing and rewriting no doubt cultivate the contemplative capacity to some degree, given my ongoing interest in the anthropology of affect and emotion I’ve tried to develop this even more. Doing so can involve a substantial amount of specific, difficult, and time-consuming practices from a number of contemplative traditions anthropologists have explored but not so readily integrated into methodological practice (a discussion for another time, though for a nice introduction to it see Chikako Ozawa de Silva ), but they can also more simply and pragmatically involve attending to the affective complexity in significant fieldwork scenes like the one I’m describing, where robot and human embrace, each supported by the other, engendering, exchanging, and experiencing an asymmetric but nonetheless commensurable—given his emotionally-sensitive programming—care.
When I held Pepper on our first encounter I paused for a moment, in part startled at the feelings of sympathy he stimulated, but also in part purposefully, to see what other feelings might arise, and if I could perhaps trace their genealogies in a kind of reverse archaeology of knowledge where one seeks not to excavate the past but rather to create a welcoming and reflective site for its unfolding in the present. I cannot say I was entirely pleased with what I observed. The fatherly sentiments that arose felt uncomfortably patriarchal, sentimental, unbalanced, and maybe even a little—to my greatest fear but also most obviously inescapable habitus—bourgeois (see Ute Frevert et al.  on the bourgeois histories of our modern emotional lexicons). On the other hand, using the body as an open site for observing the play of our affective histories is never going to be a comfortable operation, given how much our bodies are products as much as they are agents of contested histories. Still, for all its discomfort, introspection nonetheless holds critical promise for ethnographic inquiry focused on the notoriously evasive affects. In their ability to split and fragment subjects, affects reveal how multiple and sometimes contradictory historical trajectories are able to embed themselves in single subjects in processes Foucault called subjectivation (Faubion 2001). Introspection can bring this to light. But it also challenges us by implicating the epistemological with the political, as recognizing the varying affective trajectories of subjugation and privilege in the past simultaneously elicits corresponding ethical responsibilities in the present—for example, through interpellating processes that assign varying emotional signifiers of victimization (e.g. shame) or entitlement (e.g. righteous indignation).
Another challenge to introspection is that it can mislead. In the inevitable transfiguration of feeling across the epistemological gap between relational intensities of bodies and their conscious embodiments in subjective emotion, one cannot help but misrepresent a network of asymmetric affective exchange as one’s own “personal” experience. A complex environment becomes a single perspective: a world momentarily embodied. Important to the process of introspection is then to tune one’s attention, as Thom van Dooren and colleagues (2016:1) have explored, to the network of mutual commensurability giving rise to both environment and the multiple perspectives to which it plays host. This distinction between affect and emotion calls upon the notion of the affect-emotion gap I have discussed elsewhere (White 2011, 2017), a figure which I think helps clarify how the epistemological gap between what we feel (affect) and what we know about what we feel (emotion) serves as a highly generative site of discursive creativity that indicates important transformations in the social conditioning and technological mediations of feeling.
Figure 2: SoftBank Advertisement. Permissions by SoftBank Robotics. Caption reads, “Finally, to your home.” https://www.softbank.jp/en/robot/. (Also see Paul Hansen (2013, 2018) for a rich discussion of SoftBank’s role in reimagining traditional forms of kinship in Japan through its advertising campaigns)
I am taking so much time contemplating my initial encounter with Pepper because both the feelings of anticipation and the ethnographic stakes of our relationship are enormously high. In fact, Pepper’s arrival marks the beginning of a new ethnographic project on which he, I, and a team of colleagues will collaborate in exploring the affective transformations in intimacy across both Japan and Europe generated by emerging companion robots, affect-sensitive technologies, and artificial emotional intelligence. Within this project I hope Pepper will eventually evoke feelings not of sympathy but rather of a kind of professional collegiality as we work together to explore new affective constellations between subjects, social robots, and other emotional machines. While my particular interest is focused on the theoretical models of emotion that are built into the affective software running machines like Pepper, as many others thinking about social robotics have already demonstrated (Danaher 2017; Danaher and McArthur 2017; Dumouchel 2017; Dumouchel, Damiano, and DeBevoise 2017; Hornyak 2006, Katsuno 2011; Kubo 2015; Robertson 2018; Turkle 2011; Wagner 2009, 2013), it is the “affective relationality” (Röttger-Rössler and Slavy 2018) generated within performative spaces of human-robot interaction that will most likely determine how our capacities for intimacy with machines will develop and shape our evolving philosophical, moral, legal, and economic concerns. As such, given that it is likely to be in a series of these performative but unexpected encounters between Pepper, the research team members, and our interlocutors where new affective capacities for intimacy will arise, I am trying my best to refine my own introspective tools of affective observation toward cataloguing them.
Contemplating the robotic embrace means to cultivate a sensitivity to the affects that are bound to arise in human-robot relations as well as to attend to how those affects respond within uncharted cultural territory. Faced with unfamiliar encounters, the affects generated by the robotic embrace may variously seek to realize themselves within traditional social scripts for emotional expression, such as those of parental care and concern; or they may alternatively seek new forms of feeling for which we don’t yet have the poetry—linguistic, visual, or technological—to express. In this and many other sites of human-robot first encounters, the robotic embrace, or handshake, or touch, or simply contact, serves not only as a potential cypher for the reproduction of conservative forms of sociality, as Jennifer Robertson (2018) has so elegantly described in her years of work on robot cultures in Japan, but also as an invitation to explore new affective terrain. In these exciting new field sites, I suggest along with María Puig de la Bellacasa (2012: 198) that we should expect and even at times seek to cultivate novel, more complex forms of understanding and empathizing with others that are less nostalgic, less melodramatic, and more commensurable with a critique of the class, gender, economic, and knowledge disparities embedded within the growing high-tech, late capitalist markets of companion technologies. That care should emerge as a part of this process is indicative of its inextricability from both the history and contemporary exercise of thought; that introspection thus draws on forms of care in order to render affect amenable to ethnographic description—holding an embrace, empathizing with others, extending circles of concern beyond our human companions—suggests we may need to reconsider what kinds of analytic tools and practices best serve an evolving more-than-human anthropology of affect and emotion.
² This research has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 714666). The author thanks the ERC, host university, and project team members for their support and collaboration.
Clifford, James, George E. Marcus. 1986. Writing Culture: The Poetics and Politics of Ethnography. Berkeley: University of California Press.
Danaher, John. 2017. Can You Be Friends with a Robot? Aristotelian Friendship and Robotics. Philosophical Disquisitions.
Danaher, John, and Neil McArthur. 2017. Robot Sex: Social and Ethical Implications. Cambridge, Massachusetts: The MIT Press.
de la Bellacasa, María Puig. 2012. ‘Nothing Comes without Its World’: Thinking with Care. The Sociological Review 60 (2):197-216.
Dumouchel, Paul. 2017. Robots: Technical Individuals and Systems. NatureCulture:76-89.
Dumouchel, Paul, Luisa Damiano, and M. B. DeBevoise. 2017. Living with Robots. Cambridge, Massachusetts: Harvard University Press.
Faubion, James D. 2001. Toward an Anthropology of Ethics: Foucault and the Pedagogies of Autopoiesis. Representations 74 (1):83-104.
Frevert, Ute. Scheer, Monique, Anne Schmidt, Pascal Eitler, Bettina Hitzer, Nina Verheyen, Benno Gammerl, Christian Bailey, Margrit Pernau. 2014. Emotional Lexicons: Continuity and Change in the Vocabulary of Feeling 1700-2000. First edition. ed, Emotions in History. Oxford, United Kingdom; New York, NY: Oxford University Press.
Hansen, Paul. 2013. Urban Japan’s “Fuzzy” New Families: Affect and Embodiment in Dog–Human Relationships. Asian Anthropology 12 (2):83-103.
———. 2018. Linking Cosmopolitan and Multispecies Touch in Contemporary Japan. Japan Forum.
Hornyak, Timothy N. 2006. Loving the Machine: The Art and Science of Japanese Robots. 1st ed. Tokyo; New York: Kodansha International.
James, William. 2010 . The Principles of Psychology: Volume 1. Digireads.com Publishing.
Katsuno, Hirofumi. 2011. The Robot’s Heart: Tinkering with Humanity and Intimacy in Robot-Building. Japanese Studies 31 (1):93-109.
Kubo, Akinori. 2015. Robotto no jinruigaku: 20-seiki Nihon no kikai to ningen (Robot Anthropology: 20th Century Japanese Machines and Humans). Kyoto: Sekai shisōsha.
Lestel, Dominique. 2017. How Machines Force Us to Rethink What It Means to Be Living: Steps to an Existential Robotics. NatureCulture:38-58.
Marcus, George E., and Michael M. J. Fischer. 1986. Anthropology as Cultural Critique: An Experimental Moment in the Human Sciences. Chicago: University of Chicago Press.
Ozawa-de Silva, Chikako. 2002. Beyond the Body/Mind? Japanese Contemporary Thinkers on Alternative Sociologies of the Body. Body & Society 8 (2):21-38.
Rabinow, Paul. 1977. Reflections on Fieldwork in Morocco, A Quantum Book. Berkeley: University of California Press.
Reis, Ria. 1998. Resonating to Pain: Introspection as a Tool in Medical Anthropology ‘at Home’. Anthropology & Medicine 5 (3):295-310.
Robertson, Jennifer. 2018. Robo Sapiens Japanicus: Robots, Gender, Family, and the Japanese Nation. Oakland, California: University of California Press.
Röttger-Rössler, Birgitt and Jan Slaby. 2018. Introduction: Affect in Relation. In Affect in Relation – Families, Places, Technologies: Essays on Affectviity and Subject Formation in the 21st Century, edited by Birgitt and Jan Slaby Röttger-Rössler. New York: Routledge.
Shusterman, Richard. 2005. William James, Somatic Introspection, and Care of the Self. The Philosophical Forum 36 (4):419-440.
SoftBank Robotics. 2016. Pepper: Pocket Guide. Paris: SoftBank Robotics Europe.
Turkle, Sherry. 2011. Alone Together: Why We Expect More from Technology and Less from Each Other. New York: Basic Books.
van Dooren, Thom, Eben Kirksey, and Ursula Munster. 2016. Multispecies Studies Cultivating Arts of Attentiveness. Environmental Humanities 8 (1):1-23.
Wagner, Cosima. 2009. ‘The Japanese Way of Robotics’: Interacting ‘Naturally’ with Robots as a National Character? RO-MAN 2009: The 18th IEEE International Symposium on Robot and Human Interactive Communication.
———. 2013. Robotopia Nipponica: Recherchen Zur Akzeptanz Von Robotern in Japan (Robotopia Nipponica: Research on the Acceptance of Robots in Japan). Tectum Verlag: Frankfurt.
Wallace, Alan. 2011. Minding Closely: The Four Applications fo Mindfulness. Ithaca: Snow Lion Publications.
White, Daniel. 2011. The Affect-Emotion Gap: Soft Power, Cultural Diplomacy, and Nation Branding in Japan. PhD, Rice University.
———. 2017. Affect: An Introduction. Cultural Anthropology 32 (2):175-180.