Abstract
It is now routine for anthropologists to study those who exercise power and control wealth and status in any number of societies. Implicit in anthropology’s long-standing commitment to apprehending societies in their totality, and explicit in the call to study up, paying attention to power is just one of the routine things that anthropologists do in the course of their fieldwork. That said, many theoretical and ethical norms in the discipline are calibrated to allow researchers to both know about and protect those with relatively little power who made up much of anthropology’s original topical area of interests. By contrast, studying people who exercise power entails special ethical and theoretical consideration. This article enumerates some of those considerations, and suggests that anthropologists need to have coherent theories of social action in addition to theories of social meaning. The article also suggests that some canonical disciplinary ethical norms are inappropriate for the study of the powerful for empirical and practical reasons.
Above all, [they] ask themselves questions: Why have [they] come here? With what hopes or what objectives? What exactly is the nature of anthropological research? Is it a normal occupation like any other profession, the only difference being that the office or laboratory is separated from the practitioner’s home by a distance of several thousand kilometres? Or does it result from a more radical choice, which implies that the anthropologist is calling into question the system in which [they] were born and brought up? (Lévi-Strauss, 1973: 376)
The dilemma
For a long time now, anthropologists have explicitly trained their gaze on those who have and exercise power; who set the conditions for the use of power; who have wealth and or status; or who set the conditions for the acquisition of wealth and or status. Often, this endeavour is referred to as ‘studying up’ (Nader, 1972); and anthropologists (as well as sociologists, e.g. Khan, 2012) have spent a lot of time thinking about how to go about doing that up-studying. Much of this thought has taken the form of methods innovation and reimagining what an appropriate field site should look like for an anthropologist who can’t live, work and recreate all in one place, with one group of people (i.e. do traditional village fieldwork and join the ‘Malinowski Reenactment Society’; see Souleles, 2018: 53).
Generally speaking, anthropologists have come to the conclusion that researchers, when studying up, should imagine multiple field-sites for a single project (Marcus, 1995); use methods other than participant observation (Gusterson, 1997); and think about their informants as existing in and across networks (Souleles, 2018) which materialize in occasional, semi-public ‘interface’ events (Ortner, 2010). Taken together, this tactical reflection and reimagining of what fieldwork can look like has been exceptionally generative for anthropological research, leading to studies on everything from Hollywood film production (Ortner) to nuclear weapons scientists (Gusterson) and private equity financiers (Souleles) among many, many others.
For all this success, I suggest that there is room for reflection on two other features of studying up that have received a bit less attention than research pragmatics: theoretical framing, and ethics (though, still following Nader, 1972: 5, 19–21). Anthropology was classically concerned with people with relatively little power who lived on the periphery of capitalist and imperial systems, or who were oppressed at the centre of such systems. As such, much of the discipline’s theoretical stock in trade and ethical schemes, still today, take disadvantage and lack of power as a default both for dictating ethical conduct and setting the parameters of what could and should be known about the people we study. Moreover, some of the more exciting developments within anthropological theory and practice, developments that have moved the discipline away from its more conventional antecedents, will be helpful in conversations around studying up. What I will suggest is that there are specific forms of non-local theorizing necessary to study up and that those specific theoretical forms, in turn, suggest the limitations of anthropology’s ethical schema, reliant as it is on informed consent and forms of collaboration, as it applies to studying up.
To illustrate the theory necessary and the ethical problems incumbent on studying up, I will draw on some observations from fieldwork I’ve conducted on algorithmic trading on financial markets in the United States.
Anthropology 101
From the fall of 2017 through the fall of 2020 I was a member of the Algofinance project at the Copenhagen Business School. The goal of the project was to understand the changes that have come to financial markets as a result of the automation of trading and the networked interconnection of financial exchanges (e.g. Beunza, 2019; MacKenzie, 2018; Pardo-Guerra, 2019). My role in the project was to conduct interviews and observational work with traders, investors, regulators and technologists across American financial hubs. In total I interviewed 69 investment professionals (of our project’s corpus of 182 interviews), and conducted a month of observational field work with a group of options market makers, all across a number of trips and miscellaneous office and conference visits to Chicago, New York, Los Angeles, Washington DC and San Francisco. My fieldwork, in turn, has generated scholarship on the distribution of ignorance on financial markets (Souleles, 2019a), the shifting demographics and semiotic ideologies of traders (Souleles, 2020a), and the way present-day financial markets shape the distribution of wealth in American society (Souleles, n.d.).
For me, cutting across all of this research has been a more or less basic anthropological impulse to closely document, by necessarily eclectic methods, the everyday life and sensibilities of the people who trade and then somehow to connect that to the larger distribution of wealth and poverty, work and worry in American society (Hart and Ortiz 2014; Ortiz 2020). My anthropological work on finance, in turn, fits into a few of the more recent turns and normative agendas in economic anthropology and the anthropology of finance: specifically, I’m curious about how particular forms of capitalism are generative of certain ways of life and hostile to others (Bear et al., 2015; Rofel and Yanagisako, 2019; Tsing, 2015; Yanagisako, 2003). For me, this agenda has meant trying to answer questions about how and why finance people are able to set the material conditions of other people’s lives in the course of their professional activity, which sees them identify, create and allocate huge amounts of their society’s wealth (Souleles, 2019b, 2020c; Ortiz 2020). Ultimately, my approach to finance, inspired as it is by recent theorizing in economic anthropology, meets back up with the old call to study up and understand how some small amount of people set the material conditions for many other people’s lives.
Despite this up-to-date theoretical grounding and my eclectic use of methods – much of what I do (and what others seem to do) is basically classic cultural anthropology: an individual researcher, with varying amounts of local collaboration in the research enterprise, attempting to explain the ambient shared meaning and assumptions that characterize and animate that group of people through extremely close attention to the mundane round of their daily lives. For me at least, this manner of attention seems to be one legacy of anthropology’s commitment to the study of those on the periphery of imperial and capitalist systems, or those downtrodden and degraded at the centre of those same systems. This close attention is meant to correct stereotypes and prejudice which are, seemingly always, ascendant in other parts of the anthropologist’s world. The idea is that close ethnographic attention illustrates the essential, unique and deserving humanity of the people that we study and can rebut racial essentialism or any number of other forms of chauvinism (as in Keane, 2003).
These insights about the redemptive potential of positive representation have bled into anthropological research ethics, which increasingly suggest as a disciplinary default that people should consent to research and, in turn, have some control over what anthropologists write about them (Bell, 2014). This ethical stance, though, presumes a sort of power dynamic between the researcher and those they study such that the people studied need to be protected from the researcher. It’s a stance which either sees the researcher and the people they work with as equals or sees the researcher as more powerful. More to the point, it’s a stance that, as a default approach to research, does not make much sense in studying up.
Gatzambide-Fernández (2015) has illustrated the limitations of an overly deferential approach to ethnographic description and research ethics in describing some trouble he ran into at an elite boarding school in the north-eastern United States. As part of his research, Gatzambide-Fernández sought permission from the school and afforded them the opportunity to review his work, pre-publication. This all seemed to go well until writing about ‘wienie night’ came up and the school didn’t want him to publish. In Gatzambide-Fernández’s telling, wienie night is a picture-perfect case study in homosocial bonding – it’s a semi-secret, all-male, shirtless competitive hot dog eating contest in which nearly all participants eat, cheer and puke communally under the attention of a raucous crowd. Moreover, Gatzambide-Fernández suggests that wienie night is an example of how elite boarding schools, despite propaganda to the contrary, still serve as sites in which (largely) white male future elites consolidate their class position and status vis-à-vis the larger society via forms of extracurricular homosocial bonding.
Now, this wienie-heavy account of how the school creates elites runs counter to a vision of a boarding school that might see it as a place in which the most deserving, talented people receive an exemplary education which then allows them to meritoriously compete for the levers of society’s power. Naturally, too, as we might expect, the school didn’t want Gatzambide-Fernández to write about wienie night, as it could hypothetically undermine the meritocratic narrative that is presumably part of why the larger society tolerates privilege factories like boarding schools. The schools sees its antiseptic public image as essential to its ongoing role and position in society; and in the school’s mind, Gatzambide-Fernández’s account would undermine this public image. Gatzambide-Fernández ultimately wrote about wienie night against the school’s wishes and suggested that because researchers create narratives about the people they study, such researchers are ‘inexorably implicated in the production of elite status even as [they] seek to expose it’ (2015: 1132). Given this, researchers on elites will on occasion need to adopt an ‘un/ethical’ position in relation to their subjects. The way Gatzambide-Fernández sees it, when empirical findings contradict the publicly facing narratives that certain elites want, such representation isn’t neutral. And in these instances, the researcher is presented with an ethical dilemma – either, on the one hand, censor one’s findings and bolster elite representational propaganda, or break prior ethical commitments one has made with one’s research partners or discipline and publish a narrative which undermines the elite status of the people you’re working with. Gatzambide-Fernández chose the ‘un/ethical’ path and wrote about wienie night.
In what follows, I’ll suggest that these sorts of ethical dilemmas are unavoidable when anthropologists study up. More to the point, these ethical dilemmas point to both specific forms of theory that are necessary to grapple with studying up, and ways in which anthropologists should reconsider their own disciplinary ethics codes. Elites are often interested parties seeking to manipulate their own representation to protect their position in society; anthropological practice needs to anticipate and reflect this. Anthropological theory needs to account for this sort of representational economy. And anthropological ethics needs to allow for situations in which the people anthropologists work with have more power than the anthropologist. In order to show this sort of theory and illustrate this manner of ethics, I’ll deliberately walk through a case from my own research on trading, a case in which some of my findings were also censored.
But first, some background.
John Henry redux
When we talk about finance in the United States, we can’t help but talk about ‘markets’. In turn, what people generally mean when they talk about markets are ‘stock markets’. Stock markets, or equity markets are public exchanges on which people participate in constant rolling auctions to buy and sell ownership stakes (called stocks, shares, or equities) in publicly traded companies. These shares represent fractional ownership in a business. To give an example, at the time of writing, Disney stock were trading at about US $102 per share, and there were around 1.8 billion shares outstanding which put the total market valuation of the Walt Disney Company at around US $180 billion. If you owned one share, you would own one 1.8 billionth of the Disney corporation, and would be entitled to that same fraction of profits if the Disney corporation elected to pay profits as dividends to shareholders. Moreover, if you could somehow amass more than 50%, or somewhere north of 900 million shares of Disney stock, you would control the company and could presumably make as many Star Wars movies as you like and put Mickey Mouse into the public domain should you so please (thought other shareholders might sue you for violating your fiduciary duties to the other, minority shareholders). In any event, this is generally what people talk about when they talk about ‘markets’ – capital, investor markets for shares, ownership and control of businesses. Unfortunately for our purposes, this homology of markets and finance as equity markets masks a whole wider world of money, debt and trading.
Beyond stocks, and their attendant markets, are debt markets where governments’ debt (as in US Treasury bonds, for example) and corporate debt issuances are traded. Debt takes the form of bonds. In exchange for money now, the debt issuer pays a certain percentage of the loaned amount of money over a fixed period of time, at the close of which the issuer then pays back the principle. In addition to debt, there are huge ‘derivative’ markets on all manner of things.
A derivative is some sort of financial product that is based on (or derives its price from) another thing. If we stick with Disney, one simple example of a derivative would be a Disney stock option. Owning a Disney stock option would allow you to buy (‘call options’) or sell (‘put options’) on some amount of Disney stock at a fixed price, over a fixed period of time. For example. If you had a two-year option to buy 100 shares of Disney stock at US $200 per share, this would mean that at any point between now and the expiration date of the stock option in two years, you would be entitled to buy 100 shares of Disney at US $200 per share. If Disney stock climbed to US $300 per share, you would make US $100 per share using your option. If Disney stock never crested US $200, your option would expire, useless (‘out of the money’). Stock options, then, as a specific type of derivative, derive their own price, in part at least, from the price of some underlying asset, in this case some Mickey Mouse stock (for a much deeper discussion of derivatives, see MacKenzie, 2006).
Once you become aware of derivatives, you start to appreciate just how vast and complicated the world of financial markets is. In addition to derivative instruments based on stock, there are derivatives (futures as well as options) on everything from money, to interest rates, to volatility, to bundles of stock, to commodities, to weather, and even to crypto currency. Derivatives can be standard or custom, traded on exchanges or over the counter. Moreover, these derivatives (as well as equities and debt) are often packaged for convenience into bundled financial products, allowing investors to do a sort of one-stop shopping. All these financial instruments, too, exist and often trade at the same time and often have some manner of relationship to one another. The sums of money these markets represent are staggering as well. To give some further sense, the market cap of the S&P 500 stock index, a composite of 500 large, regularly traded companies stood at around US $22 trillion at the time of writing. The global bond market represents about US $100 trillion in debt, again, at the time of writing.
What I hope is apparent at this point is some sense of the scale, complexity and importance of financial markets to capitalist societies. Financial markets are the privileged way that most wealth flows through these societies. Building the dams, carving the canals, and standing to open or close the floodgates of all this wealth are those who work in, on and around finance, the people that I’ve been trying to study (up) in one way or another since 2012. I hope, too, it’s apparent by now the reason that I think financiers are important to know about in the context of a capitalist society. Not only do they allocate wealth, but they are often responsible for the structural conditions under which the allocation of wealth is possible (see further Souleles, 2019b, 2020b). Again, from a disciplinary point of view, this is a classic case of studying up. So, what then has been going on in financial markets that is of ethnographic and by turns societal interest? Well, nothing short of a miniature industrial revolution.
Over the last forty years or so, markets have electrified, computerized, networked and automated (see Beunza, 2019, Preda, 2017 or Pardo-Guerra, 2019 for a detailed account of the history and some of the effects of these shifts). Part of what this means is that the vast majority of trading on financial markets happens faster than humans can follow, algorithmically on servers located in the suburbs of financial centres, and is monitored by traders in their own offices on an ever-proliferating number of screens and a constantly multiplying number of computer systems. It’s an open question, too, how these shifts have affected the ways that people on markets are behaving in their never-ending professionally sanctioned channelling of society’s wealth.
One common approach to understanding how the financial wealth system has shifted has been to pay attention to the rise of ‘quants’, that is, hard scientists, mathematicians and quantitative analysts, and reflect on how they think about markets and financial problems (Derman, 2004; Hansen, 2020a, 2020b; Souleles, 2020a, n.d.; see also Lowrie, 2017). This approach looks at the way in which quants apply the techniques and tools of their own professional life (data modelling, hypothesis testing, algorithmic development and big data analysis) to the world of finance. What this literature demonstrates is how the priorities and career aspirations of science, maths and technology have made the financial world look less informal and networked, and more cold, automatic and rational, and has left little space for old-school, in-person, gut-following pit and phone traders. Also, this literature points to the ways in which the rise of algorithmic trading has led to unpredictable, novel forms of market behaviour, such as the recent onset of ‘flash-crashes’ in which markets can lose or gain up to 10%percent of their value in just a few minutes (Borch, 2016).
The rise of the algorithms and the quants is dramatic and they merit all the attention they have received. Their aggregate effect is inescapable in the world of finance. However, these extensively documented tales of quants and automated finance are not the only narrative about how people are behaving as markets have changed. A parallel story has to do with gambling.
All bets are off
As noted above, the Algofinance project was trying to understand how the rise of automation and computers has changed financial markets. Given this, it was a bit of a surprise to me that I saw gambling feature over and over again in various forms and firms across my fieldwork. For reasons that will soon be apparent, permit me to offer a slightly imprecise composite sketch,
I first heard about gamblers in trading rooms in the course of asking informants about the sorts of people they hire in their firm. One informant noted that in the United States, due to the vagaries in gambling legislation and regulation, some large online poker sites had been pushed out of business. As such, a number of professional poker players found themselves out of work. This was right around the time that algorithms were on the rise. In this informant’s account of things, these former poker players were perfectly suited to be traders in quant-inflected firms. After all, a professional poker player has the attention and stamina to stare at six screens for eight hours a day and monitor trades mostly executed algorithmically. This ability to focus coupled with a gambler’s instincts would allow the poker-trained trader to follow the algorithms and modify their strategies if necessary. Moreover, they could hit a metaphorical stop button if trades got too out of hand, and bring in the quants to diagnose and code a solution to whatever problem the market has cooked up.
I heard about gambling in a firm in a second context as part of an orientation for a group of new hires at a trading firm. In addition to the more typical education about how markets work, what various financial instruments are, and what the firm’s trading strategies and computer systems were, new hires were taught to play poker and about the different strategies they could employ to bet on their hands. The idea here was that building a portfolio and taking trading positions required the ability to evaluate the probability of events and make a sequence of bets that could profit from the likelihood of various market outcomes at the expense of other market participants. The tacit presumption is that betting and odds making in poker is directly translatable to betting and odds making on financial markets (see further Bjerg, 2011).
The final time I heard about gambling was in a number of firms in which sports and political gambling happened as a matter of course throughout the work day. In this case, traders would bet with each other, taking odds bets on sports and politics as matches and elections of interest came up. Again, these traders are weighting the likelihood of a particular outcome of some event, and betting accordingly. One further wrinkle – this sort of betting can match up with the office’s hierarchy, in which one’s direct supervisor would be responsible for keeping an open betting-book (and becoming a ‘bookie’) for subordinates who want to bet on something. In this scenario the subordinate would say, ‘Will you give me a line on, say, a college basketball game’s point spread?’ The supervisor would then quote odds, and the subordinate could haggle over whether those odds were fair. If they came to some sort of agreement, the subordinate would take the bet. The supervisor, in this instance would be generically responsible for keeping an open book for the folks that work closely with them. The idea is that this would be one more expected feature of a trading workplace, and sort of makes a general cultural sense in the context of taking trading positions. The bookie, too, is only able to do their job successfully if they price bets in such a way that, on average, they make money. In a funny way, this minor sports betting mirrors some of the hierarchical features of trading firms specifically and market capitalism more generally: you’re able to play, but on average, the house, the people above you in the hierarchy, the people with more money, the people who run the game and structure its conditions, well, they win.
Now, it’s worth clarifying a few things about this data on gambling. First, none of us in the Algofinance project had any plans to study gambling in the context of trading, nor, honestly did we have any idea that it existed (so, two cheers for fieldwork). Second, I don’t have systematic data on how typical or widespread gambling is in trading firms, either in practice, or in the conception of a trader’s job. Parsimoniously, from my own data I can confidently say that these various forms that gambling takes happen in at least a few American trading firms. Given the networked nature of working in finance, and how often many people change jobs and trading firms, though, it’s likely a bit more widely spread than I can verify. Third, this manner of understanding one’s work and behaving in the workplace is extremely different from how scientific, algorithmic-developer-quants often explain and conduct their work.
One final contextual note: while this sort of odds-counting wagering seems reminiscent of old-school finance, in which people traded in person, in groups on trading floors and in trading pits, and perhaps even has a genealogy that connects it to those sorts of traders, this gambling is absolutely concurrent and coeval with the rise of the algorithms and demonstrates one other way of acting within contemporary markets. After all, those former professional poker players are watching screens; those interns are learning a feel for odds as they get ready to manage computer systems; and betting as a hobby is just one more layer on trading as odds making and position taking. How, then, should an anthropologist interpret these pockets of gambling?
The classic starting places are those social theories that facilitate the thick description of a particular group’s ambient social life in that group’s own terms; that is, anthropologists are trained to start with the local, and build from there. Fortunately, gambling, play and risk-taking are all long-standing topics of inquiry within the social sciences, and making this level of interpretation is straightforward. Erving Goffman (1969) made the point that in certain social situations there is a relationship between the amount someone risks in a game of chance and their potential to win or lose social status. It’s pretty easy to see how this helps make sense of gambling in the trading room. Whether wagering one’s own money or the firm’s trading account, we can imagine a social scene in which improved status clearly accrues to those who are expert in making successful, risky wagers.
Similarly, Clifford Geertz (1973), in his essay on cockfighting in Bali, suggests that the pattern of wagers on cockfights grants a window into the social structure and alliances in Balinese life. More to the point, the event of the cockfight crystallizes, oppositionally, the way in which groups might form against one another; and the amount wagered is a decent guide to how large and evenly matched given groups are. The entire event of the cockfight, too, is immanent with metaphors for Balinese social values. Perhaps not nearly as straightforwardly applied as Goffman’s interactional approach, still, Geertz’s reflection on gambling and deep play has application in parsing gambling in a trading room. Much like the Balinese, traders feel that their wagers afford them the opportunity to act out the way in which they feel the world should be. Wealth should be allocated according to those who understand best the dynamics of markets and capitalism. Each gamble, in turn, can become a microcosmic re-enacting of this profane creed.
Recent studies of gambling have largely bolstered Goffman and Geertz’s claims; Stine Puri’s body of work on wagering in India is largely indicative of this (Puri, 2016, 2020, 2021). Puri (2016) notes that horse race gambling seems to afford context-specific forms of affection and relationships. Puri (2020) also points out that the specific pattern of horse race gambling speaks to how the speculators view society. Puri notes that there is a great deal of momentum betting and crowding in the lead-up to a particular horse race. Puri explains these patterns by talking about how gamblers presume that races are fixed and that the sensible approach for any punter to take is to figure out who is doing the rigging and bet accordingly. In turn, momentum becomes an index of corrupt action. Puri explains that this matches up with a presumption among the people she works with that their society functions via the frequent secret intervention of powerful people, and, cosmologically at least, there is a Hindu perspective that this is about all you might expect for the current fallen state of the world. As Geertz identified family and faction in the cockpit in Bali, so Puri is able to point towards corruption and manipulation at the horse track in India.
For our purposes, what is interesting about my close analysis of gambling in trading is that we are able to see the strengths of normal anthropological analysis and some of its more conventional limitations. In paying attention to what gambling means, we’re able to describe the role it plays in traders’ work lives – as a way they bond and socialize with each other, as well as a way to reinforce proper hierarchy, and as a reflection of how they understand markets as places where gambling logic is helpful. Paying attention to gambling gives us an edge and an angle from which to understand how trading currently works. Risk does relate to status; and wagers do fit into a metaphorically rich cosmology of capitalism. But we don’t, and often can’t know, from this perspective, how specific gambling traders integrate into larger markets, and how those markets integrate into larger society. We also don’t quite understand why writing about gambling could be threatening, as it seems to be an accurate description of what is going on and is foretold in various strands of social analysis and theory. This disjuncture is a key point for studying up and requires a different sort of theory, one that doesn’t always start with the local life-world.
The whole point of studying up is to connect those with power, wealth or status to the rest of the society that they have some sway over, particularly in the contexts in which a society is so large that the effects of the exercise of power are distant from those with power. Thick internal description is obviously a useful starting point, especially when it is a bit outward looking as in the case of gambling, but another layer of analytic attention is necessary – one that is less neutral and descriptive, and that can account for how the powerful are perceived in the larger society, as Gatzambide-Fernández suggests. Anthropology needs to build on closely observed cultural theories of meaning and interpretation, to create larger societal theories of motion, theories with pre-existing categories of analysis that explain how groups interact with each other and build up a larger world.
Anthropology 102
In the past, the version of local anthropological theory that I’ve used to explain deals and transactions in finance (Souleles, 2019b) drew on Marcel Mauss’s (1990 [1950]) idea of a total social fact – that is some event in a society that draws in people, wealth and things, rearranges them according to some value scheme, and then redistributes that set of people, wealth and things in such a way that the society is changed (see further Wendling, 2010). I’ve also used Graeber’s (2001) articulation of value theory to characterize this sort of movement as an instance in which different groups of people’s values, their ideas of how the world should be, come into contact, and get hierarchically rearranged according to dominant power dynamics in a given society.
Seeing financial transactions as total social facts illustrates the larger convergence and dispersal of people, things and wealth that happens when, say, a company like Disney is taken over. Appreciating the fact that some traders see the financial transactions that would lead up to a hostile takeover as being analogous to one’s weekly poker night starts to suggest what values they are bringing to their work, and how they are contributing to the restructuring of society. More to the point, people who work in a taken-over company, likely don’t share the gambler’s sensibility in their day-to-day work, and likely would not choose to have their fate decided by the posturing punters of finance (see, for example, Archer, 2020). For our purposes though, something theoretically sharper still is necessary, since there is also an issue of representational economy and censorship at stake.
In the course of my research and writing on financiers, I’ve also been directly and indirectly censored on a few occasions when I’ve tried to learn and write about things which my informants didn’t appreciate. Indirect censorship happens as a matter of course in studying up. Given inherent access issues, when people refuse to participate in your project, you’re clearly limited in what you’re able to learn about them. This doubly applies to informants who refuse follow up interviews or return field-site visits, or with whom you’re worried about saying the wrong thing and thereby losing access. Direct censorship (in, for example, writing about gambling) is another issue still.
As part of my own training in anthropology, I was taught various decolonizing research methods (e.g. Smith, 1999), and ways to make research egalitarian and participatory (e.g. Lassiter, 2005). The idea with all of this was to ameliorate the invasive nature of much previous anthropology scholarship, and redistribute control over the production of research. Instead of doing extractive science on informants, anthropologists should instead do participatory research with aims in common with consultants, interlocutors, collaborators and friends. Inherent in this reformulation of research was an idea that researchers are able to find common purposes with those they study, and should not do research that those under study do not approve of. This sensibility is further codified in disciplinary ethical norms which enjoin researchers to seek consent and to do no harm. Specifically, the first three injunctions of the American Anthropological Association’s (2012) Principles of Professional Responsibility are, ‘1. Do No Harm,’ ‘2. Be Open and Honest Regarding Your Work’ and ‘3. Obtain Informed Consent and Necessary Permissions’.
In my own research practice, this training and ethical stance has led me to fall over backwards to explain what I’m up to, to seek permissions that allow people to opt out of my projects at various points and, finally, to allow people the opportunity to review my scholarship if their data is used. This last one is the kicker. Ideally this would allow a moment of researcher/researchee concurrence in which the informant, the expert in a given social scene could guide the researcher to a more accurate account of their own life, as well as nudge the researcher to choose a level of anonymity which the researchee feels comfortable with. This, though, presumes some manner of common purpose in the research enterprise. Absent this sort of collaboration or common purpose, this level of openness can become a veto for participants, which my trading informants have made use of around topics they find embarrassing or irrelevant (such as, say, gambling).
It’s worth noting too, that this is another in a long example of anthropology either being subject to research conventions that are an awkward fit for its sort of research, or inventing over-reaching ethics codes of its own. Lederman (2006; see also Katz, 2006) makes the point that US-based ‘Institutional Review Boards’ evaluate research ethics according to a laboratory, experimental or hypothetical/deductive model of science. This in turn casts inferential participant observation field work as inherently unethical. Similarly, Bell (2014) has pointed out that both societally and, crucially, within anthropology’s own codification of ethics, researchers have taken informed consent to be a universal gold standard. Bell suggests, however, that it is generally unclear that people actually understand what they are consenting too in more positivistic research. Bell also points out that even if people understand what they’re consenting to in the context of inductive fieldwork, the nature of anthropological knowledge generation suggests that the substance of the research will likely shift by the end of the project giving the lie to moments or even processes of consent.
This long conversation about anthropological ethics suggests that even in the best case scenario anthropological ethics has done a poor job accounting for the specific nature of the work that anthropologists do and the variety of people they work with. Closer to reality, anthropological ethics sets people up to give an unreasonable amount of latitude to powerful actors who are then able to manipulate the research enterprise or shut it down entirely. Taking consent and mutual agreement over research as an example, it’s worth illustrating why there could be no such common investment in the context of my own research on traders, and why I’ve encountered the blanket veto. Before that though, one further final theoretical detour is necessary.
Unleash the flying hegemonkies
I noted above that value theory allowed me to begin to see how different groups of people engage with one another in society, and how different ideas about how the world should work push groups against each other. I suggested that a trader trading stock, or perhaps an investment banker evaluating a company, likely sees a company and its workers differently than the people working in a company, making a living do. Moreover, much of the tragedy and absurdity of financial capitalism is that people who are increasingly removed from day-to-day economic life are empowered to make decisions about other people’s livelihood. Value theory, as an example of a local descriptive theory that doesn’t connect to the larger society, doesn’t really offer a theory of power and its exercise – it just says that people disagree and compete over value schemas in larger societal life. Nor does it help to explain why the people working in a company controlled by financiers often allow financial gamblers to act the way they do except to note that they’ve lost a value struggle, and that the ideas of the financial gamblers tell us something about why they act the way they do and how they see the rest of society.
To explain the reality of financial power, control and governance with more clarity, we’ll need a theory that doesn’t start as internal and inductive to the research context of a financial firm. Rather we’ll need a theory of society and social interaction that assumes the existence of certain features of the world. To illustrate what that might look like, here I’ll take one example and follow Stuart Hall (1986: 5) by making use of Gramsci’s (1971) ideas of hegemony to ‘sophisticate’ value theory and give it a more critical edge by presuming the existence of class struggle in a capitalist society.
One of Gramsci’s central concerns was the way in which elites and leaders exercise power in a given society. One extremely costly route to the exercise of power is coercion – physically forcing people to do something, as in the North’s victory over the South in the American Civil War that resulted in the end of chattel slavery. This manner of leadership is though, again extremely costly and ultimately difficult to routinize across the whole of society (see for example the failures of Reconstruction after the American Civil War). An alternative to coercion, then, is to legitimate the exercise of power through a sort of general common-sense understanding of how the world should be, towards a ‘hegemony’, and then power matching up and acting according to that common-sense gut notion ‘hegemonically’ (e.g. Kiesling, 2001; Linger, 1993; Weiss, 2011). A priest giving penance to someone coming out of a Catholic confessional generally doesn’t have to whip the person to say their 10 Hail Marys. Presumably the priest’s parishioner already thought the priest’s authority was legitimate since they’re a believer and they elected to show up to confession in the first place. The prayerful penance happens as a common-sense matter of course.
Absolute hegemony in a given society occurs extremely infrequently (only in periods of ‘settlement’) as it requires common understanding across different social classes that often have different histories, leading to different interests (everything from respective location in a given economic mode of production, to varied racial identities, to different locations on the urban/rural continuum). As such, hegemony is best understood as the fragile and ever-shifting product of ideological coalition building, a project that various classes and interests are always trying to bolster or undermine (Brow, 1990). Given that underlying fragility, we’re in a position to understand why I need to offer a non-specific composite account of my trader/gambling findings.
Part of the way that financiers are able to exercise the power that they do in society comes from the public perception of their work, and that perception fitting into certain deserving categories (Souleles, 2017). Finance, as an industry in the US only exists and makes money due to the peculiar combination of hyper-regulation on the one hand, and the privileging, in law, of market-based competition and profit, on the other. A central component of this settlement is that financiers contribute to the common weal and, among other things, maintain orderly public capital markets where the nation’s corporations can seek competitive funding for industry and enterprise. Key to this rosy vision of finance in general, and capital markets in particular, is that traders are acting as prudent, responsible stewards of the nation’s public capital markets (for some of the history of the ethical/legal history of gambling versus speculation versus prudent practical insurance, see Levy, 2012).
On the one hand we might imagine the aura of precision which the rise of algorithms, computers, and networked trading would bring financiers. Sure, tell those stories – they make finance look rigorous, and on the cutting edge. Seamlessly transitioning between pricing futures, and pricing college basketball spreads, though, not so much; and their image starts to fracture. There’s a reason that ‘casino capitalism’ is an epithet. If all investment bankers/hedge fund managers/stock traders do is gamble, why should we allow financiers the position and the power that have? Why should they manage all our money? Surely there are other more reasonable ways to allocate capital, fund worthy enterprise and distribute wealth in society, right? Why not just automate venture capital and legislate postal banking? This representational dilemma is part of the reason I believe some of my work on traders and what happens in the trading room has been directly censored.
Further reflections on studying up
The original aspiration of studying up was to make our picture of a given society whole. Nader’s article introducing the term is particularly interesting because she quotes from her students imagining projects studying entities like the Better Business Bureau, or state insurance regulators. Crystallized in these students’ aspirations is a desire to tie the known structure of their lives to those who exercise power to shape that structure at some remove from their everyday experience. Anthropology does this sort of work, and has developed the tools, techniques, and disciplinary will, to boot, to allow studying up to happen in a routine way. Still, our specific sense of theory, and the ethical prerogatives that come with studying up have not kept up with our practice.
Perhaps most straightforwardly, anthropologists who study up need larger social theory, theories I’ve called ‘theories of motion’ to connect their local thick descriptions to the rest of society. As just one example, I’ve used Gramsci’s notion of hegemony to illustrate the way in which power often operates in accordance with common-sense, everyday notions. In my case, some of those notions were that those who manage money should be and are prudent and careful. Representations that challenge this identity threaten traders’ livelihoods and can lead to stricter regulation and the restructuring of the financial world to the detriment of those who make a lot of money keeping capital markets open. More abstractly, hegemonic processes of common-sense building and the action of power via persuasion connect the world of the traders to other groups of people in American society: to the people who work in the companies whose stock gets traded and to the regulators and legislators who govern markets based, in part, on their perception of them. This is what a theory of motion does – it places a group of people in relation to other groups of people in a larger social field. Studying up, and likely anthropology more generally needs theories of motion in addition to theories of meaning.
The thing about theories of motion, though, is that they clarify power relations and often reveal the operation and action of power in a way that threatens its hegemonic operation. Theories of motion can contradict self-serving justifications for the exercise of power – in part, because they will likely show that hegemonic ideas are contingent, often fragile, and historically specific. Given this, studying up seems to be inherently in trouble with anthropological ethics as currently stated, and would invite further distinction and clarification in what proper ethical anthropological conduct should be. If we are meant to do no harm, it’s challenging to see how studying up is permissible given that, in a way, studying up aims to harm the exercise of specific forms of power even if only by reinterpretation and documentation. More to the point, we may find ourselves in a situation in which a topic is worthy of study, say police brutality or insider trading, but no one will consent to talk to us about it because they get how our study harms their power. Moreover, we might end up in a situation in which a bureaucracy says no (say a police department or Goldman Sachs’s PR desk), but individuals say yes. Should we give up studying the topic because an organization has not consented to participate? Should we give up a project because some individuals consent and others do not? Should we give up a topic if we can find no informants? I think most would say no to some or all of these questions, particularly if they thought the topic was important, but we have developed no ethical norms in anthropology to cope with this kind of situation. Informed consent or doing no harm is likely not possible if the object of study is some exercise of power we object to, or even simply one that we want to scrutinize.
These conundrums multiply when we add in research contexts in which the anthropologist and their research participants are in danger for either doing or participating in research. In police states, anthropologists are often taken to be spies and closely monitored by state intelligence and police agencies (Borneman and Masco, 2015). Not only does this mean that it is difficult to understand who someone is or why they might talk to you, but this sort of research context also creates danger for anthropologist and the people who work with them, people who may at one point or another be taken for collaborators and harmed as a consequence (Driscoll and Schuster, 2018). In relaying the following quote from a key informant explaining why they collaborated with her, Armytage (2020) illustrates these dilemmas in her work on Pakistan’s political and economic elite: first, I am an excellent judge of character. I know the sort of person you are. Second, you cannot hurt me. If you ever used my name and details, I would deny everything and tell everyone you made it up. I have a big family and many friends here, and you are just by yourself … I also had my friends in the intelligence agencies run a background check on you. (2020: 26)
Given these multiplying dilemmas, it seems helpful to me first to distinguish between individual and group harm. It also seems helpful to acknowledge that there are times when the deliberate exercise of representational power we have in our writing should be exercised notwithstanding the censure and critique of our informants, or even the danger it might entail. First to consent and our informants.
In our general practice we already take pains to protect the individual people we talk and interact with. We provide pseudonyms or composite portrayals and check with people to see if there is a level of anonymity that they are comfortable with. We occasionally avoid real names and identifying details in our own notes. Some even destroy some direct records of their research. All told, we seek to spare individuals harm in our description of their larger social world. Moreover, in our ethics reviews, we already distinguish between public figures and issues of general public interest and private people whose lives should remain anonymous. There’s no reason to discontinue these practices. We should still individually seek to protect and mask informants in due consideration of their help in our research enterprise. Groups, on the other hand, merit some further thought.
It’s unclear to me why ethically (as opposed to legally or practically) we should care to seek the consent of places like an investment bank or a police department when we conduct our work. These are institutions – large bureaucracies, powerful financial firms – that act in an outsized way in people’s lives, whether they want these institutions to exist or not. It’s irrelevant to the New York Police Department or Goldman Sachs whether or not I want them to exist or whether or not I want them to control aspects of my life. They precede me, are indifferent to me, yet exert power over me. Given this, I think it follows (again, ethically, not necessarily legally or practically) that we need not afford them the opportunity to approve, reject, or constrain our research.
I think, too, that this calculation changes when there is a threat of direct harm to a researcher or their collaborators. Is your disclosure worth your own life? Your deportation? Your home? How might you seriously answer those questions? Do the people you work with want to risk their lives or their liberty over what you’re writing? Is it possible that by mere association with you, people’s lives are at risk? Are there realistic circumstances in which working with you is a death sentence for your collaborators? Again, how can we seriously answer those questions given the reality of this sort of danger in many field contexts? Insofar as social scientists will be making these sorts of life-and-death deliberations in studying the exercise of power, it is unclear to me what sort of tests and considerations they should weigh – this is so far beyond the fiction of informed consent and the idyll of community participative research. Most researchers would likely allow the acknowledgement that, as a general principle researchers shouldn’t necessarily shy away from important topics just because there is danger and risk. However, there is nothing in formal anthropological ethics as it currently stands that takes seriously the existence of this sort of research and the mitigation of this sort of harm.
Finally, a few notes on writing. One of the leitmotifs in anthropological self-critique since the 1970s is that writing and knowledge, because they shape the representational possibilities for people, are powerful and can do significant harm. For this reason, anthropologists should be exceptionally careful in how they portray people, should normatively presume vulnerability, and write with the ethical commitment to reduce harm and often to proactively help in mind. While I generally admire this commitment, and readily admit that in many conventional anthropological contexts this is important, I think this posture falls apart in several circumstances. Most generally, we have an empirical obligation in the work that we do to accurately reflect the social field. Even when studying the vulnerable there will be those with greater and lesser amounts of power, shaping social life in ways that others don’t like.
Students generally have relatively little power in an academic context. However, this didn’t stop the sociologist Shamus Khan (2011) from writing about sexual aggression and abuse students perpetuate on each other in the context of an elite boarding school he studied, despite push-back from people in the school’s community (personal communication). Moreover, Khan’s writing accurately showed the antecedents present in a school sexual culture that would later allow for an extremely high-profile rape allegation and prosecution (Purdum, 2016). It seems inconceivable to me that we would ignore this sort of an empirically verified feature of a social world, particularly an elite one, because some who are relatively without power would be embarrassed by it or a powerful gatekeeper would seek to shut it down.
In another context, that of studying the process by which fashion models are made and herself becoming a model, sociologist Ashley Mears (2011) had the following to say about her writing: Feminist sociologists value equality and sharing in the knowledge-production process, and many have argued that ethnography represents a methodological ideal because the researcher and the subject can work together and share in the craft of sociological knowledge. While I support this goal in the abstract, once in the field I never felt particularly powerful relative to those I studied. In fact, with the bookers and clients, I felt the opposite: vulnerable. For this reason, I did not feel obliged to share my interpretive authority with my participants. As Stuart Hall has noted, representation is about power. Fashion’s gatekeepers have tremendous power in the modeling market relative to the models they manage. They also have considerable power to shape the values and aspirations of the millions of people who admire (and/or despise) the looks that they produce. (2011: 266)
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement number 725706).
