On the 99th page of my dissertation, “The Objective Function: Science and Society in the Age of Machine Intelligence”, the first full paragraph describes how machine learning researchers go about deciding which projects they should work on. Given that the dissertation, as a whole, is concerned with understanding how machine intelligence—data-driven algorithmic techniques like machine learning, data science, and artificial intelligence—produces knowledge and constructs the authority it holds as it moves across a wide range of domains, this page seems a particularly apt representation of the whole. This page also gestures at the idiosyncrasies of fieldwork that allowed me to address these concerns, as the research lab I had found myself working in was particularly reflexive about the work they did. Without the reflexivity of my interlocutors, many of the social processes I was interested in would have likely remained hidden from my view, or at least been much more difficult to access ethnographically. The paragraph in question reads:
“In the sense that most work in machine intelligence involves applying a well-suited technique to a well-posed problem, the exercise these applied machine learning researchers undertook for their research report was merely a highly reflexive version of how machine intelligence is applied generally. Inspired by the horizontal organizational structures common in the technology industry (Scott 1975; Pfeffer and Leblebici 1977), researchers at OTH shared responsibilities and steered many business decisions through collaboration and consensus. The topic-selection process at the lab was emblematic of this. It was run by the researcher whose “turn” it was to write the next report and involved a series of whiteboard exercises in which all lab members would nominate candidates, research their potential as a topic, and then present their favored candidates to the entire team, who would then collectively winnow down the candidates to a few finalists. The report author would ultimately select the topic. The reflexivity of the process these applied machine learning researchers engaged in offers a valuable window into questions that go un-asked and assumptions that go un-examined in more mundane, less research-oriented applications of machine intelligence. The research report I participated in addressed a technique called “multi-task learning” that was first described in the late 1990s (Caruana 1997), but the technique rose to the top of the topic-selection process because it had recently found applications in industry (see McCann et al. 2018), and was more easily integrated with newer machine learning programming packages.”
What this paragraph reveals, for me at least, is that the production of knowledge through machine learning is a collaborative, practice-based process. To fully understand the role it plays in reshaping social worlds requires seeing the many ways it is shaped by organizational prerogatives and industry-specific ways of structuring work practices. But machine intelligence is also shaped by the questions researchers choose to ask, and—by extension—the questions they do not. These situated knowledges (Haraway 1988, Rouse 2002, Katell et al. 2020), I eventually conclude, contribute to the overall conservatism of machine intelligence. Contrary to the trappings they carry in popular imaginaries of technological advancement and futuristic automation, machine intelligence conserves the power of already-existing institutions and it reinscribes social relations of the past into the present and future. Machine intelligence may, on the surface, threaten the authority of powerful actors like judges, doctors, or loan officers by way of automation. But, in reality, machine intelligence preserves the power of the institutions — the courts, hospitals, and banks — that confer authority upon those actors. This is because, in selecting problems, and in addressing them, the work of machine intelligence is to ask questions that extend power, and not to challenge it.
Haraway, Donna J. 1988. “Situated Knowledges: The Science Question in Feminism and the Privilege of Partial Perspective.” Feminist Studies 14 (3): 575.https://doi.org/10.2307/3178066.
Michael Katell, Meg Young, Dharma Dailey, Bernease Herman, Vivian Guetler, Aaron Tam, Corinne Binz, Daniella Raz, and P. M. Krafft. 2020. Toward Situated Interventions for Algorithmic Equity: Lessons from the Field. In Conference on Fairness, Accountability, and Transparency (FAT* ’20), January 27–30, 2020, Barcelona, Spain. ACM, New York, NY, USA. https://doi.org/10.1145/3351095.3372874.
McCann, Bryan, Nitish Shirish Keskar, Caiming Xiong, and Richard Socher. 2018. “The Natural Language Decathlon: Multitask Learning as Question Answering.” ArXiv:1806.08730 [Cs, Stat], June. http://arxiv.org/abs/1806.08730.
Pfeffer, Jeffrey, and Husayin Leblebici. 1977. “Information Technology and Organizational Structure.” The Pacific Sociological Review 20 (2): 241–61.
Rouse, Joseph. 2002. How Scientific Practices Matter: Reclaiming Philosophical Naturalism. Chicago: University of Chicago Press.
Scott, W. Richard. 1975. “Organizational Structure.” Annual Review of Sociology 1 (1): 1–20.
Diego Valdivieso: For those who have not had the opportunity to read your book and have a glimpse of the vicissitudes of Buenos Aires middle class, its taxi industry and how the arrival of Uber was signified and experienced by these actors, could you please summarise the main topics you cover in your work?
Juan Manuel del Nido: In late 2015 I arrived in my hometown of Buenos Aires, Argentina to study the taxi industry. Everyone – taxi drivers, union leaders, the city’s middle class – was convinced that, due to the partisan and political make up of the city and the country, Uber would never arrive. But in April 2016 it did! I had by then amassed mountains of ethnographic material on taxis: their laws, governance, economy, infrastructure, sociocultural practices. As the conflict between the company and the taxi industry escalated, I noticed that a certain segment of the city’s middle class – superficially ecumenic, techno-idealist, managerially-minded, anxiously globalist, performatively cosmopolitan – set the tone that defined public debate around the conflict. In that tone, they increasingly naturalised their own reasons and logics and disavowed the reasons and logics of others – specifically, the taxi industry whose workings I had come to know quite well. They were not just political adversaries, but they and their reasons came to be pathologised, written off and explained away; in a way, they stopped counting, even if they were still there and still protesting. I decided then that this would not be a book about taxis or Uber in themselves, or about precaritisation, platform capitalism, the gig economy, but one examining what I call “post-political reasoning”: the logics, rhetoric, and affects through which people imagine, legitimize, and argue for an experience where it is hard, or impossible, to disagree in certain ways.
Diego Valdivieso: This book may be the outcome of the only ethnographic project carried out while Uber was (crash) landing in a territory. What advantages did this unexpected event bring to your research approach? How did you realise that the unstoppable arrival of Uber was a rich area of research?
Juan Manuel del Nido: It was a truly exceptional opportunity to see abstract logics come to life in a very literal and raw way. Until late March 2016 Uber was viewed as impossible; by mid-April it was allegedly processing tens of thousands of ride requests per day. Before my eyes taxi drivers, judges, federal attorneys and city dwellers began triangulating reasons and arguments that added up to fascinating interpretations of “competition”, “monopoly”, “freedom of choice” and “market forces”, for example, or of the divide between economic and political domains. Drawing from common sense truths, cultural anxieties, urban practices, political conjunctures and other affects and logics, these triangulations would have been entirely speculative or even unpredictable just a few weeks prior. Notions like “competition”, for example, are hugely abstract and ethnographically elusive. I had just spent half a year researching the very relations such notions were reframing – those of the taxi industry, of Peronism, of city and national politics “before Uber”. This knowledge offered me a chance to make anthropological sense of how such abstract notions take concrete forms, something an ethnographer arriving to a place where Uber, or any disruption, has already happened, would struggle to do.
The book was born out of my attempt to turn that luck of being in the right place at the right time into an ethnographic opportunity that I figured others, or even myself in future endeavours, would be unlikely to have: what is it that someone who does not get to see this pivotal break happening is unlikely to capture? On a different level, I was fascinated and alarmed by the fact that, as you say, Uber’s arrival seemed unstoppable. Already anthropologists were critiquing the platform’s precarisation of working conditions, denaturalisation of labour and more, but I do not think then, or now, scholars really asked: how did this come to appear as unstoppable to a certain group of people who are familiar, and may even agree, with our long-rehearsed arguments about labour, precariousness and such? Rather than focusing on denunciation and moralisation, I think we need to produce better ethnographies of how, increasingly, certain ways of thinking – managerial, technocratic, moralised and moralising – are becoming harder to actually challenge, that is, to disagree with beyond the protected confines of our academic seminars.
Diego Valdivieso: Throughout the book you are able to tackle a particular kind of reasoning by focussing on what people see as relevant and the affects, emotions, motivations and aspirations involved in this process. Although you give us some hints, could you share your reflections on how to address ethnographically something like reasoning through logics, rhetoric and affect?
Juan Manuel del Nido: I think it’s about fostering a different ethico-narrative disposition than the prevailing one, aiming to be alert to how things, affects, and rhetoric combine, catalyse or stunt each other, shaping social relations in often quite complex or even counterintuitive ways. To be clear: this is not yet another plea for seeing the world as emergent, contingent, becoming, fragmented, multiple, multifarious and so on, but rather an exhortation to ethnographers to reflect on the difference between pedantic, neurotic and often moralising literalism, on the one hand, and actual critical engagement with our interlocutors and ethnographic encounters, on the other.
For example: when the courts struggled, in technical terms, to block Uber’s services from the territory of the city of Buenos Aires, the middle class saw in that technical difficulty evidence that “economic forces” trumped “the political”. One way of addressing this ethnographically, common in anthropology today, is through a “well, actually…” disposition – a literalist deconstruction of an argument, however spurious. Such pedantic literalism increasingly passes for a kind of rigour, and even for a righteous concern with what we understand to be the truth. What I propose in Taxis vs Uber is to take these ethnographic encounters seriously, instead of literally, to produce an actual critique of how they organise knowledge and how they manage to persuade. In the case of this example, a historical distrust in the courts, the government and the union; an unreflexive, bourgeois sense of one’s own agency and what it demands of the world; and a habit of casually crossing jurisdictional borders, all fed into each other in the reasoning of a certain segment of the middle class and give us far richer insights into how it was even possible to imagine that Uber, or its relations, trumped what scholars see as the political – regardless of the technical, moral, or truth value of such a claim. However spurious, this is the reasoning that made it so hard for the taxi industry to even be heard. The capacity of a kind of reasoning to persuade does not depend on a soundness one can cross examine through a literalist sequencing, as it were, but on how affects, practices, materialities and other ethnographic features, truths in a subtler, and in the actually interesting sense of the term, give flesh to that reasoning as they buttress and propel each other. We miss these configurations if we do ethnography with militant (and increasingly, moralising) pedantry.
Diego Valdivieso: Your book gives fundamental clues to understand the scope of non-expert reasoning and the socio-political consequences that this way of knowing can generate. Beyond the particularities that your analysis of Buenos Aires middle class and the arrival of Uber suggest, what phenomena do you think could be addressed from an approach centred on the distribution of the sensible?
Juan Manuel del Nido: I think it is an immensely generative and underexplored possibility! As Ranciere formulated it, the distribution of the sensible suggests we think of the political as the distribution of parts, roles, and voices in a society or even an epoch, a pattern of differences and proportions whose language we broadly share and where we roughly know our, and others’, part. This distribution is always unequal: parts count in different ways and some parts are there but somehow do not count. Parts, here, is not necessarily byword for class, caste, or any sort of inexorable partitioning of society, but rather particular convergences of bodies, interests, and things in the face of a social question.
Ranciere was thinking about a very big picture, political-system-kind-of-magnitude, but I kept thinking in his terms to understand how different parts of society – unelected experts, gangs, a globalist middle class, a particular industry, an elite producing the dominant aesthetics – make up a social problem at ethnographic level: a conflict over an app, homelessness, building a house, whatever. An ethnography approached through the distribution of the sensible is one that focuses on what emerges ethnographically as being at stake, a site of contention, as disagreement in the broadest sense of the term, by looking at how voices, people, things participate in the distribution of the senses those stakes can have. What are the material or rhetorical cleavages that catalyse, hinder or shape the stakes? How do different parts consolidate themselves, seek to reframe the terms in which other can participate in the stakes, or to change what counts as a valid argument, or presence, and what counts as noise? Which voices give their tone and grammar to the whole, which ones can share a lexicon with which others and which ones are disallowed?
Political anthropologists might recognise here Jonathan Spencer’s invitation to follow disagreement in their ethnography. In that vein, in Taxis vs Uber I found the distribution of the sensible is a particularly powerful approach to examine the increasing moralisation, pathologisation and disavowal of genuine disagreement in our societies. It is increasingly difficult to disagree and remain a political equal: to disagree is to be an idiot, recalcitrant, morally toxic and be written off, or in Ranciere’s terms, to become those whose voices count as noise, “the part that has no part”. In Taxis vs Uber this was the part of the taxi industry and of whoever disagreed with Uber, but increasingly, all around us, this is the part for whoever tries to claim that a particular question, from public health to migration, from an app to climate change, could be asked in different terms, or answered otherwise.
Ariana Gunderson: You write that “cooking involves a code and its instantiations,” (Sutton 2021, 15). Do you consider the code of cooking to be analogous to linguistic codes? If so, how?
David Sutton: This question is really at the heart of what I was trying to do in this book. Because when I started studying cooking, I was very far from a structuralist perspective, and was much more drawn to approaches to cooking as embodied sociomaterial practice. Much of my work on cooking that was based on my video ethnography, especially in Secrets from the Greek Kitchen, focused on skill, tool use, the kitchen as environment, and other concepts that I adopted from people like Tim Ingold and Jean Lave. But what kept nagging at me was that cooking clearly wasn’t just emergent. We don’t just start out with a random set of ingredients and see what bubbles up; we set out to make something. So the whole dialectic between structure and practice that was so much of my graduate training seemed relevant again, and especially in the form that Sahlins writes about, since his approach is all about understanding the riskiness of all practice. And of course he was drawing from and modifying the linguistic-derived approach Lévi-Strauss. And then of course there was Mary Douglas’s work on food categories. So I think that at first I believed that these new approaches were what I needed to understand cooking, but the book is really about reconciling a dynamic structuralism with a more embodied phenomenology.
Ariana Gunderson: Might we consider recipe-writing a process of entextualization? Is the moment of recipe inscription a risky one?
David Sutton: On the one hand I have long felt that the moment of entextualization of recipes has tended to be problematic, a claiming of authorship that has often privileged male chefs over “anonymous” female cooks, a point made by Luce Giard, among others. And this appropriation of power often occurs in the process of inscription, whereas oral transmission is still controlled by ordinary women. So it’s risky from the point of view of who gets credit and who gets forgotten. It’s also risky in the sense that a recipe is always a “moment in time,” as Jacques Pepin puts it, the freezing of a process, which is the opposite of an approach attuned to contingency. So inscription is also translation, a translation of an assemblage of experiences; it is doubly risky. Perhaps triply so because in many culinary memoirs the moment of writing down the recipe from an older relative almost always presages impending death. At the same time, I think that the written recipe has a function, at least as a memory jog. Although the more I think about it, I realize that on Kalymnos this function is served by other people, mostly women that share the matrilocal kitchen space, and who constantly remind each other of the ingredients, proportions, and tricks that are involved in each dish.
Ariana Gunderson: Your research has been rooted in Kalymnos for decades, enabling you to examine long-term change and continuity in this new book. What do you see as the connection between an extended period of study and paying attention to small scale change?
David Sutton: I’ve always admired long-term fieldwork and the insights that come from it; I think it provides insight into continuity and change, or “changing continuities” as my mother, Constance Sutton, described it based on her long-term engagement with Barbados. Given that my initial fieldwork on Kalymnos was about historical consciousness, it’s also been interesting to see how ideas about the past change over time, and especially how small-scale change can lead to bigger changes. But small-scale change is important in other ways, in that you can see it happening ethnographically much more clearly than you can see a change, let’s say, from so-called traditional to modern world views. So I’m suggesting that focusing on something like cooking allows us to see the process of change (and continuity) in action, rather than comparing how things were at two points in time and making assumptions about what happened in between.
I’ve noticed how many social theorists use the metaphor of recipes to talk about various social processes, though as with my comments above, I think the idea of the recipe can be problematic. On the other hand, I like to think about how much of the activity of cooking is similar to anthropology: attention to detail, participant sensing, focus on parts and wholes. Making cooking more explicit as part of our research can illuminate a lot of the social processes that we are interested in.
Ariana Gunderson: In an autoethnographic interlude, you describe recreating your late father’s spinach casserole in search of his voice. This calls up Annie Hauck-Lawson’s use of the concept of food voice to assert agency and the real-world impact of non-verbal, edible communication. Can you speak to how you find the concept of food voice useful in your ethnographic work?
David Sutton: I’ve always liked Hauck-Lawson’s concept of food voice because it can both extend and stand in for other ways that people express themselves. In Greece food voice is expressed at least in part through smell as neighbors pay attention to, and comment upon, the smell of what’s cooking next door. But I think I was most directly influenced by Carole Counihan’s slightly modified use of food voice, or what she calls “food centered life histories.” Especially in her book A Tortilla is Like Life, she uses the concept to get at the very distinct personalities, and distinct life trajectories, of the Mexicana women in southern Colorado that she was studying. I tried to do a bit of that in my previous book Secrets from the Greek Kitchen. In a way I feel like food voice does some of the work for me at the micro-level that gustemology does at the collective level. Both are about exploring peoples’—individual and collective—food-centered world views. I feel that the best ethnography moves between these two levels.
Ariana Gunderson: Did writing Bigger Fish to Fry change how you cook? Do you hope it will change the way readers cook?
David Sutton: I think it did and I hope it does. One of my targets in the book is the idea of culinary perfectionism, what John Finn calls “culinary fascism.” It’s the idea that there is one right way to do some kitchen task, or one best recipe for any dish. I think there are a lot of lingering problematic assumptions in this approach to cooking, which can lead to things like molecular gastronomists claiming to separate old wives’ tales from scientific truths about cooking. My focus on risk and contingency, I hope, challenges the idea of perfection: in other words, I suggest that, like the Kalymnians, we should imagine good cooking as managing contingencies (material, sensory and social), rather than achieving perfection. Also, I think that the idea that I develop from Sahlins that every reproduction is also a transformation suggests a greater willingness to accept and enjoy the differences and similarities when we cook a familiar dish. I think that if we think about what makes cooking cooking in terms not of a product but of a process of confronting all the contingencies that arise both in and out of the space of the kitchen, and developing our own tricks to deal with these contingencies, to improvise, we might develop a healthier, more equanimous attitude, rather than the more dichotomous one of success versus failure, which can lead to stress and frustration.
Ilana Gershon: What is so fascinating about A Village Goes Mobile is how effectively you use your long historical relationship with a West Bengali village to reflect upon how the introduction and decade-long use of mobile phones affects social relationships. Which aspects of your previous research shaped your research foci and your questions?
Sirpa Tenhenun: My previous research in Janta focused on gender, kinship and politics, so I was inclined to continue to observe these aspects of culture and society when I started working on the appropriation of mobile phones. Without my long-term relationship with the community I might not have been able to recognize changes, and how mobile phone use, on the one hand, contributed to these changes and, on the other hand, to how mobile phone use was intertwined with changes. I, for instance, could observe how mobile phone use helped transform kinship relationships. I also witnessed how most people could accomplish more in a shorter time by being able to coordinate their activities with the help of mobile phones. However, the most significant economic change in the village since the turn of the century was not due to the use of mobile phones, but to the agricultural policies. Since farming small plots of land has become increasingly unprofitable, young men from small farms use their phones to find paid employment outside the village.
Ilana Gershon: You discuss the fact that most West Bengali villagers did not have landlines, that their only experience of telephony were the mobile phones. This reminded me of Terry Turner’s discussion of Kayapo videos, that their makers had experienced radio and then video cameras without experiencing all the different media in-between that Euroamericans have. How do you think this affected people’s experiences of telephony – to never have had an experience with landlines first?
Sirpa Tenhunen: Villagers who never had an access to a landline phone experienced the ability to use mobile phones as more spectacular than those of us who have routinely been using landline phones before mobiles became available. This does not mean that the idea of a landline phone did not at all affect how mobile phones were used: initially, when the phone density was low, mobile phones were mostly kept at home and shared by the household members. However, villagers were able to use mobile phones more innovatively since they had not developed routines of landline phone use. For instance, they frequently used the phone’s speaker to share the phone conversations with whoever was present.
Ilana Gershon: How did the introduction of smartphones affect social organization in Janta? Did it change gender relationships?
Sirpa Tenhunen: Smartphones were used to strengthen pre-existing identities but they also offered opportunities and technical affordances which challenged hierarchies. Contrary to their hierarchical position, educated young wives and children could become the phone use experts in their families. Low caste people were able to make identity statements simply by possessing smartphones. Digital sphere constructed through mobile telephony was first associated with male-dominated public sphere but it proved much more malleable than the public sphere outside the home from where women are still largely excluded—the great majority of women continue to be housewives despite women’s growing aspirations to be able to work outside the home. The young men pioneered the use of smartphones—if their older models were still working, they were kept in the house and used by the rest of the family, mainly by women, who usually stay at home more than men do. At the same time, few women who moved outside the home for work or study started to acquire their personal phones. Young women now also preferred smartphones, readily discussing the multiple functions of these phones and demonstrating their ability to use them. Most people stored music and films on the smartphone’s memory card from downloading shops instead of browsing the internet independently. I never saw women listening to music on their phone in public places like men do —listening to music on smartphones through the phone’s loudspeaker is used to ascertain the meaning of the public sphere as a masculine space where men can spend their leisure time. The few men and women in the village who have used their personal phones to browse the internet all had a college education. The ability to browse the internet with one’s phone was, therefore, related to one’s education and wealth rather than merely gender.
Ilana Gershon: What kinds of social relationships or circulation of knowledge changed because phones made it common to have a one-to-one or dyadic conversation, which might have been difficult to achieve before the advent of mobile phones?
Sirpa Tenhunen: Within the sphere of politics, mobile phones provided a channel to contact opposition political leaders discreetly. The use of mobile technology amplified multiplicity by strengthening clandestine political activities and alternative discourses. Women’s increasing access to a mobile phone influenced their relationships with men, but—more crucially— it influenced the kinship code of conduct and kinship hierarchies within families and between kin groups. I observed how phones offered women a channel to express unconventional ideas and exert their will through networking. For instance, a mother could advise her daughter over the phone to not to obey the mother-in-law whose demands were excessive. Thanks to phones, young wives were able to stay in constant touch with their natal families which was unheard of in the past.
Ilana Gershon: One of your findings that I found startling was that mobile phones had contributed to a marked decline in village-level leaders’ power. Could you explain why this is the case?
Sirpa Tenhunen: Mobile telephony was a crucial factor in the rise of the opposition in West Bengal, where the Communist Party had been in power largely through its network of local village leaders from 1977 until 2011. In 2010 opposition activists related to me how mobile phones help them secretly mobilize against the ruling party. Political activists used phones more to organize party meetings and offer political patronage than to organize spontaneous demonstrations and support. The parties’ power used to be largely derived from their role as arbitrators of disputes: any person who feels that he or she has suffered an injustice can call a village meeting, led by local political leaders, during which a solution will be negotiated between the disputing parties. Thanks to mobile phones, patronage could now increasingly be sought from opposition leaders and from outside the village. When I visited the village in 2013, I found that after the end of CPI(M) rule in the village in 2011, not a single general village meeting had been held to solve local problems and disputes. Mobile phone use had helped amplify translocal political networks thereby reducing the power of local village leaders. Phone use for political purposes built on earlier political patterns and meanings, but it made politics faster, more heterogeneous, and translocal.