INTRODUCTION

Empathy has quickly become one of the most familiar buzz words in the world of business, whether in product design or human resources, customer research or in cross-functional team building. To invoke empathy in the contemporary professional setting is to signal “human-centered” and cutting edge. In recent years, since it was launched into the design and innovation professional vocabulary and then expanded to gain a foothold in both management and entrepreneurship milieus, empathy appears to have become not only a celebrated and desired tool into the innovation toolbox—almost a dispositif in a Foucauldian sense—but also so prominent that its invocation and application seems to have become mandatory. What was once a fresh reminder to business people, designers, and engineers that feelings, perspectives, and emotions, and not only numbers, have an immediate value to their operations and work, today appears to be fast becoming an empty signifier.

Our community is certainly not oblivious to this. As empathy´s celebratory potential began taking on formulaic and mandatory overtones, recent debates around it have offered an increasing array of counterarguments against the use of the term. In this catalyst paper, I do not aim to provide a single concluding vote to either of these camps, but rather to explore to what extent and with what effects could we, as a community of scholars-practitioners, rework the role of empathy in ethnographic research in applied settings. I synthesize what has been said so far of empathy and its role in design and innovation, and I take stock of what these contributions point to. More importantly, in so doing I am looking for clues as to whether there are paradoxes or unresolved tension in the ways empathy has been conceptualized and deployed in our practice which might provide a fresh analytical ground for asking new questions about empathy and how it is used in applied research.

What follows is a preliminary rewriting of the question of empathy. in the first, conceptual part of this paper, Synthesis, I ask: “How can we re-scale empathy in innovation?” As a first step in this endeavor, I suggest that to empathize is a scalar activity—a point whose implications and potentials are largely lost in both practice and writings on the topic.

My position here is animated by my own grappling with the topic, some theoretical, some stemming out of my 24-plus months of fieldwork with a radical innovation moonshot venture pursuing the development of humanoid robots, where I first started out researching questions of identity and team culture, and then became increasingly involved, through participant observation, into development and outreach questions of how to create empathy for robots at societal level. In Part II, Exegesis, I illustrate the limits of empathy in innovation and the study thereof as an example to such grappling.

However, such an understanding of the power and prominence of the affective dimensions of empathy requires that we understand empathy not on a flat scale, as a temporary adoption of a worldview perspective from a point-to-point individual-to-individual (as in a “the researcher” empathizing with “the user”). It requires, rather, a more granular understanding of empathy on a nested scale, one implicating historical, cultural, and social aspects in active interplay with each other, and empathy´s reconceptualization as an inhabiting of affective states and in terms of intermediary experience of the multiplicity of its constitutive affective variants (such as, among others, hope, anger, pain, passion, fear, exhaustion, bravery, weirdness, friction). Ultimately, it allows us to better capture, conceptualize, manipulate and responsibly account for questions of scaling feelings and perspectives in our work. In Part III, Catalysis, I suggest a non-exhaustive list of provocations that might help us reframe the question of empathy.

PART I: SYNTHESIS

Empathy´s meteoric rise to prominence in and dominance of the vocabulary and mindset of the world of design and innovation is part and parcel of the changes design thinking brought in the 1990s (e.g. Leonard and Rayport 1997). As one of the first and most distinct steps in design thinking—back then a novel approach on how to identify and solve problems—the rise of empathy as a concept and as a fundamental step in the innovation process in the last 20 years can easily be pointed to as one of the true success stories of a long-standing and continuously ongoing push for peopling engineering practice and management thinking. As a result, recent decades have seen a substantial number of professionals adopting it as their occupational identity and becoming empathy coaches, empathic strategists or empathy gurus, and entire dedicated “empathy labs” exist both as independent businesses and within large corporations such as Google and Facebook (Stinson 2020). In taking stock of the merits of empathy as part of design and innovation, as well as the challenges and dangers posed by its increasingly near-automatic and formulaic application lately, we must tack back and forth between not only what the term means and what it does, but also place it within a larger understanding and increasing critique of design thinking as the leading framework for innovation.

Originally starting as merely a new product development framework, but then steadily expanding into questions of customer centricity and organizational culture, design thinking practitioners lay claim to have a clear map of “applying the principles of design to how people work” (Kolko 2015; see also Kolko 2014; Brown 2009). The claim was that it created better outcomes, more finely attuned to user needs and “pain points” (Platzer 2018) than hitherto delivered by a remote bird´s-eye view of quantitative approaches. Part of its revolution has been to bring decision-making and product testing outside of the confines of labs and into the real world, placing the designer not only as a creator of specifications and aesthetics deduced from their own imaginary about the world in which their creations will be embedded and will circulate, but also as a validator and generator of real-life insights on how such a potential product would be experienced, and—crucially—understanding and placing the perspective of the user and the user´s reality above one´s own assumptions.

A key differentiator that design thinking claimed for itself in its approach was a kind of empathic perspective-taking that other approaches lacked, catapulting empathy—meaning “in feeling” from the Greek pathos via the German Einfühlung—as the go-to method of tapping into other people´s realities via sharing their inner experiences. What “being in feeling” meant produced a number of definitions, sometimes full of contradictions, which I am about to suggest, points to the weaknesses of adopting empathy as an approach – weaknesses which we should be either collectively moving away from, in favor of more ethnographic thinking, or working to eliminate and make stronger.

Thus, Battarbee et al. have defined empathy as “the ability to be aware of, understanding of, and sensitive to another person´s feelings and thoughts without having had the same experiences” (2014, 2 my emphasis), while a little later in the same text, they suggest and affirm practical approaches to achieving empathy precisely through experience-near techniques, such as, for example, to “participate in grueling endurance events to share athletes´ exhilaration and pain” (2014, 4), recalling to mind Lois Wacquant´s call for embodied methods, an “incarnate study of incarnation by practical example” (2014, 4).

Renowned product designer Jon Kolko describes it thus:

“empathy is about acquiring feelings. The goal is to feel what it´s like to be another person. That goal is kind of strange, because it´s unachievable. To feel what someone else feels, you would needs to actually become that person. You can approximate her feelings, so product research intended to built empathy is really trying to feel what other people feel. Assuming you aren´t actually an eighty-five year old woman, consider for a second what it feels like to be an eighty-five year old woman. This consideration is still analytical, it´s about understanding. You need to get closer to experiencing the same emotions that an eighty-five year old woman experiences, so you need to put yourself into the types of situations she encounters [to] approximate her feelings, leaving your own perspective in order to temporarily take on hers” (Kolko 2014,5)

Michael Ventura similarly notes, “empathy is about understanding. Empathy lets us see the world from other points of view and helps us form insights that can lead to new and better ways of thinking, being, and doing” (2018).

In sum, if one were to approach the concept of empathy as championed by design thinking (e.g. Brown 2009) and applied in marketing and leadership contexts (e.g. Ventura 2018) and product development (e.g. Kolko 2014), the promises that approaching the lived realities of those for whom we design, share moments of (cross) “cultural intimacy” (Herzfeld 2005), and for whom and with whom we ultimately create value are so many, that surely we should subscribe to deploying empathy without second thought. As described in these widely admired and popular approaches, empathy promises straightforward and surefire ways into other people´s realities and offers the ability to quickly and amenably tap into exactly the tranches that we need to understand for the purposes of delivering insight. Those applying empathy are implicitly portrayed as swiftly deploying it as a tool—although, unlike in ethnography, we never see this—a tool which works magically to translate the immanent and immaterial (feelings and lifewords) into the profitable, the material, and the immediate—objects, structures, services. As Jennifer Wong quips cheerfully (or ironically?) in an online article on creating empathic design systems, “…to help solve the UX process problem, inject a bit of empathy” (2019).

Whether it is presented as a tool from the designer´s toolbox (Kolko), as a mindset and a way of being (Ventura), or even perhaps as a medium to be administered (Wong), the one aspect which all proponents and interlocutors of empathy and design thinking agree on is depth: the prize is to understand “deeply” (e.g. Stinton 2020; Kolko 2014) and to achieve “perspective”, often seen as the product of “stepping in other people´s shoes.”

For anthropology, on the other hand, the question of accessing, understanding, and representing, in a formulaic shorthand, “how they feel in their shoes,” has never been a simple affair. The discipline has dealt with the question of fellow feeling as a vehicle to knowledge and as a subject of inquiry in a characteristically discerning manner. It has examined the question of “fellow feeling” (Solomon 1995), and has recognized a difference between empathy, emotion, and affect as three distinct domains, all of which require various levels of engagement with context, focus on embodied experience, and in which narrative and language mediate what is essentially an intersubjective experience that is both slipping, and yet firmly enmeshed, within social and political imperatives and structures (see, for example Lutz and White 1986; Besnier 1990; Beatty 2013; Beatty 2014 for comprehensive reviews; and on affect, Skoggard and Waterson 2015; Stodulka et al 2018; Newel et.al 2018). What is more, the ambiguities and limits of knowing “other people´s minds” has been shown to be always linguistically mediated (e.g. Keane 2008), but also necessarily embodied.

Thus both Daniel White (2017) and Danylin Rutherford (2016) have suggested that affect is largely unspoken and involves an embodied intensity of feeling which in turn gives rise to emotion within the subject. White succinctly captures the historical shift in the field between emotion and affect: “if anthropologists of emotion throughout the 1970s and 1980s had shown how feelings variously fix and stick through different compositions of language and discourse, anthropologists of affect shortly thereafter sought to show how some feelings slip, evade, and overflow capture” (2017, 175). In other words, if empathy is the ability to bridge inter-personal varieties of existence in the search for capturing meaning, it requires a reorienting of cognitive, affective, and bodily states.

Clifford Geertz´s famous skepticism as to whether adopting “the” native´s point of view is analytically valuable comes to mind here, as he argues instead for a “hopping back and forth between the whole perceived through the parts” (1983, 69). This is a subtly scalar proposition of engaging phenomena on a nested scale, and not a singular point-to-point one. Numerous other scholars have further unpacked the density of the concept. Famously, Renato Rosaldo´s poignant essay “Grief and a Headhunter´s Rage” (1993), on understanding murderous grief after the loss of a loved one only after the tragic death of his wife during fieldwork, suggests that there are domains of human experiences which are viscerally comprehensible only to those who have gone through them. More recently, and in a different vein, Douglas Hollan (2008)has argued that empathizing is an intersubjective act not only of feeling but also of imagination—and, crucially, is not the work only of the one empathizing but also requires a reciprocity of emotion and imagination on the part of the one being empathized with. This last point suggests that empathy is a perspective-taking exercise based not only on a singular agent, but is rather the product of two agents taking perspective with respect to each other—meeting on a mutually re-scaled perspectival plain. Finally, C. Jason Throop suggests that “empathy…must always be understood in the context of particular cultural meanings, beliefs, practices, and values…it is significant to explore how empathy is both recognized and enacted by individuals in its marked and unmarked forms but also to examine the specific contexts, times, and situations in which empathy is possible and valued and those in which it is not” (Throop 2010, 772; also Hollan and Throop 2008).

Yet “standing in their shoes” and “seeing like they are seeing” has been deemed increasingly deceptively formulaic.EPIC community members have already put forth a range of thoughtful objections to the preeminence of empathy discourse. Rachel Robinson and Penny Allen (2018), for example, have argued compellingly that empathy is not to be conflated with evidence, and have discussed the many traps in which they perceive empathy can introduce unwelcome and unhelpful bias. Tamura and colleagues (2015) have demonstrated that a “sense of ownership” is much more effective in the innovation and entrepreneurship context than empathy in that it creates more powerful research. John Payne (2016) has commented on Paul Bloom’s (2017) recent arguments against ‘empathy’ as a decision-making rationale. Payne carefully examines the limitations of empathy, noting: “Many of these methods have been repurposed from the social sciences to the needs of design practice. However, when removed from their theoretical foundations and optimized toward identification of user needs, they don’t account for the social implications of the work we do. This needs to change” (2016). Romain, Johnson, and Griffin (2014) have been similarly preoccupied with the ways in which empathy obscures the potentially meaningful to consider tensions between stakeholders in business. Finally, in an even more provocative vein, Thomas Wendt (2017) has argued that empathy is too human-centric, reductive in its Western anthropocentrism, thus essentially rendering the political aspects and questions of power in design essentially invisible, to the detriment of all.

In sum, for professional ethnographers, the way empathy is approached in most design thinking is problematic, stemming from an increasing tension between design thinking and ethnography. As Jay Hasbrouck has elegantly pointed out, design thinking has become “symbiotic in practice, but […] at odds empirically” (2018, 3) with ethnographic approaches, creating an unwelcome conflation between the kinds of questions that design thinking can ask and answer, and those that ethnographic thinking can, in addition to inaccurately framing all human-centric approaches as reductive.

But lest we consider that it is the anthropologists who are particularly critical of the concept of empathy, skepticism of it and its application has also been mounting in parallel in design circles. Without mincing words, Natascha Jen has spoken against it, calling design thinking as a whole “B.S.” for being too prescriptive, and signaling out empathy as specifically problematic: “the word empathy is prevalent in design discourse; people have become experts on design empathy. Back in the day we called it research…” (2018). And, in what is perhaps most damning condemnation because it comes from one of the most authoritative voices in design, Don Norman (2019) has admitted to not believing in empathic design for several reasons. One is because of the inherent inability to design for “many” by immersing yourself in the individual experiences of the one or the very few; another, because very often people´s own understandings of their own experiences and feelings are not immediately accessible to themselves. Finally, in his view, the ways empathy and human-centric design operate at present, they simply cannot solve for the truly wicked problems, such as climate change and hunger, for example—something that Natasha Iskander critiquesd in the pages of the Harvard Business Review as the inherent tendency in design thinking to protect the status quo and to reinforce the position of the designer, but not designed for—even if empathy is employed, because ‘solving for’ is the remit of the powerful (2018).

What emerges as a pattern, then, is that although empathy remains a fruitful, popular, and profitable approach to obtain perspectives and mine them for understandings of experience, its shortcomings are increasingly being exposed. Key among them are that it does not address its political potential and is regularly ahistorical; it does not lend itself readily to understanding contexts defined by uncertainty and complexity; it ignores key questions of the positioning of subjects—including in relation to each other; it can get lost in translation between research encounter and the production of an object. It does not make a critical distinction between reported experience and shared experience; and fails to explain how it deals with the limits of verbal explanation. Further, it does not differentiate critically between cognitive and affective empathy in a systematic manner, or explain when to use which variant. Nor does it address well how empathy operates from within the fraught entanglements of objective and subjective phenomena. Pain is one such phenomena, ironically enough. David Platzer (2018) has given the concept of “pain point” an excellent treatment. However, when the question of what the pain point means is refracted through a careful consideration of the role of empathy in it, it becomes necessary to situate both at multiple scales and levels of analysis: one objective (there is in many cases such a thing as real experience of physical pain, discomfort or unease which innovation addresses) and the subjective, more elusive forms of experiencing them—something which C. Jason Throop, not incidentally also thinking about pain, has termed “intermediary forms of experience”—“much of what we deem to be experience is characterized by … transitions, margins, fringes, by the barely graspable and yet still palpable transitive parts of the stream of consciousness that serve as the connective tissue between more clearly” (2009, 536).

In sum, although an inherently relational phenomenon, both in its reliance on accessing other people´s experiences and in translating them into different metaphysical forms (be they objects and services that circulate often locally and globally), current approaches to empathy fail to factor in something which anthropologists have long understood, examined, and theorized: emotions and affect are as social as they are cultural, and they are socially constructed, always enmeshed at the nested scales of individual and society, always rife with political potential, and always refracted through questions of meaning and power, always contextual, fleeting, incomplete, and elusive.

A good way forward, I believe, is to take a cue from anthropology’s insistence on unpacking what perspective is, and to think of perspectives precisely as scalar phenomena, and, in turn, scales as being a question of perspective and positioning. This is a running theme in both branches of approaches to empathy: the anthropological one and the design thinking one. In many ways, then, both anthropology and design rely first and foremost on taking perspective, which is an inherently scalar phenomenon, as a recent edited volume on the topic has proposed. Drawing on Marilyn Strathern´s definition of scale as “the organization of perspectives on objects of knowledge and enquiry” (2004, xiv in Summerson Carr and Lempert 2016, 5), E. Summerson Carr and Michael Lempert argue in the introduction to their edited volume on the pragmatics of scale that “scaling involves vantage points and the positioning of actors with respect to such vantage points means that there are no ideologically neutral scales [and that] scaling is process before it is product” (2016, 3-4).

Noting various examples of scale from a range of cognate disciplines, from distinctions between “private,” “personal” and “political” to “macro” and “micro,” and even conceptualizations such as “bench-to-bedside” throughout their introduction and the volume, a key motivation is to show “the inherently perspectival nature of scale, asking of our material “whose scale is it,” “what does this scale allow one to see and know” and “what does it achieve and for whom” (2016, 15, original emphasis). In a subsequent chapter, Susan Gal further highlights the comparative logic inherent in both scale and perspective: “scaling implies positioning and, hence, point of view: a perspective from which scales (modes of comparison) are constructed and from which aspects of the world are evaluated with respect to them” (Gal 2016, 91).

Yet in borrowing from anthropology, and re-scaling the process of perspective taking for business contexts in making it faster, less granular and less concerned with language, context, and embodiment as constitutive of empathy, design thinking has lost the kind of granularity that is exactly what makes empathizing a very rare kind of empirical tool for understanding other people´s realities.

Coupled, however, with the proliferation of the discourse of empathy in the business milieu, it would appear that two camps are forming. One is calling for more empathy—scaling it qualitatively and championing a more granular and extended research at the empathy step in the innovation process—and the other is signaling that the concept has ceased to be analytically useful. Where does that leave our field?

I propose that instead of seeking to substitute one´s own perspective for that of the user in attempting to gain perspective through a “like” state, a more ethnographically informed approach to gaining perspective would be to pursue a “with” state. Instead of “seeing like them,” “seeing with them” allows practitioners to position the otherwise wicked problem of capturing and representing others´ experiences in a granular manner by situating the empathizing endeavor at multiple scales at ones. In the next section, I offer two illustrations on the challenges for so doing, and in the final, catalyst section, I briefly touch upon what the opportunities might be if the field takes a turn in this direction.

PART II: EXEGESIS

Admittedly, the grapples that inform my provocations on the need to re-work the concept of empathy into more granular variants are borne out of work different to the commercial projects that are often presented here at EPIC, which focus on issues such as UX, product development, and organizational culture. Rather, the context of my (originally purely academic) work is extreme (cf. Hallgren, Rouleau, and de Rond 2018) in that it is unique and cannot be said to represent most commercial settings in which applied ethnography operates. Specifically, I work with a moonshot startup which dwells uneasily in the space between the commercial demands and expectations of the venturing scene and the scientific requirements and realities of research: an academic venture occupying the outer extreme edge of an already extreme category of innovation, which Sarasvathy (2008, 93) has termed “the suicide quadrant”—where a new product is introduced into a new market. In the case of humanoid robots, it is largely the case that the innovation is so radical, that there is no product, no urgent demand, and no immediate market. Traditionally, this has meant that either only large companies such as Google can afford to have in-house units (such as X, the Moonshot Factory) dealing with such kind of innovation, or that the government gets involved (cf. Mazzucatto 2011). In the case of my collaborators, neither of these were not the case—thus making product development and keeping the startup financially afloat a Herculean task. It faces the kind of slow diffusion and challenging scaling based not so much and exclusively on kind of innovation which does not rely on the quick diffusion cycles of lean driven product development but requiring the slow but steady interpretative understanding of how to disrupt meaning as a necessary early ingredient (Haines 2016).

Yet it is precisely this far-off vantage point that gives me a different vista on questions of how we approach empathy, affect, emotion, and experience more broadly in the search for useful understanding of others. This approach draws on the strengths and contributions of academic anthropology´s unpacking of these questions to which I referred in the previous section. But it also transforms questions from being meaningful and relevant into also being applicable and interventional. Applied ethnography makes such a pivot in its daily operations, which nonetheless do not preclude the ability to draw on and contribute to theory equally well. This point is worth insisting on given that we are still collectively working to end the “theory-practice apartheid” (Baba 2005).

I never intended to study questions of radical innovation, let alone musculoskeletal humanoid robots. Rather, as a scholar I was interested in how startup teams form in the academic context, and how their identities and practices inform team culture. But as I was studying questions of culture, identity, and practice within the setting of a moonshot startup in the academic setting, as is often the case with prolonged fieldwork, I became more and more incorporated into the team, slowly and over the course of many months, through our shared understanding that a sociocultural anthropologist has a legitimate role in a startup developing humanoid robots, especially where sociocultural outreach is concerned.

To be sure, no single paper could capture the multiplicity of angles through which the topic of empathy as a scalar and perspectival project, rather than as simply a method to gain perspective, is refracted in every milieu conceivable in innovation and entrepreneurship. In what follows, I offer two vignettes from my own ongoing work, which serve here to illustrate why I am compelled to question empathy in design thinking. The first instance revolves around questions of the robot´s features and appearance, and questions of gender and race in particular. The second vignette draws on how an unexpected failure of empathy resulted in developing one of the most popular pre-programmed function the robot has: hugging. In both instances, I chart the dilemmas that empathy, as a scalar, perspectival, embodied, and linguistic phenomenon, presents to our current thinking on the topic.