Daniel Beunza’s (2019) Taking the Floor is not simply a landmark book in the sociology of finance.¹ It is also a contribution to the social sciences more broadly, both methodologically and perhaps even theoretically.

Beunza’s main method in the research reported in the book was participant observation, conducted with great skill. It is worth remembering that in 1999, when Beunza began his work, participant-observation studies of finance were rare. The research that led to other classic studies of this kind (notably Zaloom, 2006; Godechot, 2001, 2007) was already underway, but still largely unpublished, although there was one major completed study by Heath et al. (1995).

The relative rarity of these studies (and even today there are still too few of them) is largely because they involve “studying up” in the sense of Nader (1974): studying those who are more powerful or higher on conventional socioeconomic scales than the researcher. Research access to elites — especially observational access — is notoriously hard to negotiate. It is not surprising, therefore, that Daniel had two false starts before finally achieving access to the trading room that is the focus of Beunza (2019).

Because of the scarcity of fully successful participant-observation studies of finance, Beunza could have written a perfectly satisfactory book simply by packaging together the papers that he wrote — in the main, in collaboration with David Stark — on the basis of this first-rate participant observation (the single most influential of these papers was Beunza & Stark, 2004). Beunza did not take this easy option, and this is what makes his book especially interesting methodologically.

He went back to key informants, and he re-studied the trading room not through observational ethnography but through what was essentially oral history. This led him to realize that what he had taken to be a trading floor in its natural state was actually a very deliberate construction, above all by the man whom Daniel calls “Bob.” Observationally, the trading room looked as if it was “organic” (in terms of the classic distinction between “organic” and “mechanistic” management formulated by Burns & Stalker, 1961), or “heterarchical.” Heterarchy is “a mode of governance that differs from a hierarchy of command and the conceptual hierarchy of cognitive categories” (Stark, 2009, p. 5). Power and authority, though, were more at work than was observationally apparent, and what Beunza had observed was less “natural” and less typical of Wall Street trading rooms than he had imagined.

“Going back” using oral history interviewing to the site he had researched ethnographically thus led Daniel to radically different conclusions. Taking the Floor is therefore of methodological importance, a counterpoint (as José Ossandón has pointed out to me) to the influential critique of interview-based research by Jerolmack & Khan (2014).  

The theoretical significance of Taking the Floor concerns its relations to the actor-network theory of Bruno Latour, Michel Callon, Annemarie Mol and many others. Beunza is among the many scholars in the sociology of finance — including myself — deeply influenced by actor-network theory. For example, a major contribution of Beunza and Stark, 2004 was their emphasis on the role of “market devices” and the physical layout of the trading room. What is most distinctive about the actor-network approach is of course its emphasis on the constitutive role of things, technologies and other non-humans in social life. That emphasis continues in Taking the Floor’s discussion of devices such as trading robots, the spread plot in merger arbitrage, and the like.

There is, however, a persistent strain in Daniel’s work of dissatisfaction with an over-exclusive emphasis on the non-human. It comes out, for example, in work he has been conducting in relation to the pandemic, interviewing senior bankers about working at home versus being in the office and about the loss of technologically unmediated “incidental” interactions. It also comes out in as-yet-unpublished work he has done with Yuval Millo on the automation of trading on the New York Stock Exchange, and in the way in which existing social relations in the NYSE’s trading rooms were quite deliberately preserved in trading algorithms.

The dissatisfaction is also explicit in Taking the Floor (Beunza, 2019):

[M]y conception of Bob’s trading floor had neglected the role of humans and conjectured instead a trading room where objects did the heavy lifting work of coordinating people and enforcing norms. […] I now saw the inaccuracy of such a view. It was frames, norms, shared experiences, personal energy, and other managerial elements that were equally, if not primarily, responsible for the collaboration and restraint. (p. 170, emphasis in original)

There is also an intriguing implicit connection between the book and an article (Strum & Latour, 1987) that is perhaps the best introduction to actor-network theory. Latour’s co-author Shirley Strum is an eminent primatologist, an astonishing researcher who became, in effect, an ethnographer of baboons. Baboon societies, Strum shows, involve complex processes of negotiation and are far from the simple, stable hierarchies that earlier researchers had imagined. They are, however, limited in time and space, essentially to the span of face-to-face interaction, while we humans can build societies that have large-scale histories and geographies, that last through time and spread through space. Why can we humans do that, while baboons largely do not?

Baboons are actually not that much less clever than we are. Famously, the evolutionary psychologist Robin Dunbar has argued that the typical group size among primates correlates with neocortical ratio, in a roughly log-linear relationship (Dunbar, 1992). What has attracted attention to Dunbar’s argument is what happens when you input the typical neocortical ratio of homo sapiens into his equation, which suggests a human group size of around 150 — “Dunbar’s number” (e.g., Dunbar, 2010) — a “finding” that has entered popular culture. It comes up in Taking the Floor as an influence on Bob's desire to limit the number of traders in the trading room (Beunza, 2019, pp. 34–36)

The case specifically for Dunbar’s number has been criticized widely (see, e.g., Lindenfors et al., 2021). But that does not reduce the interest in Strum and Latour’s question. If mental processing capacity is not the full explanation of the durability and spatial extension of human societies, what is? One possible answer is of course language, but the actor-network theory argument is that we humans can live in “big” societies because objects make it possible for us to do so. The actor-network critique of orthodox sociology is that it is in effect baboon sociology: analyzing human society as if it were baboon society.

How completely should we turn our back on baboon sociology? The theoretical virtue of Taking the Floor, and of Beunza’s work more generally, is precisely its caution in this respect. It would go too far to say that Daniel is at heart a baboon sociologist, but his work reminds us of the importance of a nuanced answer to this fundamental question.