Traces of the Dead,Actions of the Not-Alive: A Prologue to a Theory of Agentification

Abstract

The author describes how sociological and philosophical discussions of agency tend to center questions of how or why people are agentic rather than who or what is agentic. In contrast, the author poses questions about the agency of things, the agency of non-humans, and the agency of dead humans, using three examples of historical traces—Washington’s refusal of a third term, Jenner’s development of the smallpox vaccines, and Smith’s publication of The Wealth of Nations—as historical examples to examine how non-humans and non-living-humans leave traces that can experienced as agentic. The author then analyzes six theories of agency that might provide explanations for these actions (actants, affordances, switchmen, residue, repression, and ghosts) before turning to his earlier work on the concept of “external authorities.”

Introduction

Debates about agency in both sociology and philosophy tend to emphasize its distinction from constraint (Adams, 2011; Archer, 2000, 2003; Meyer and Jepperson, 2000; Reed and Weinman, 2019; Sewell, 1992; Taylor, 1985). In other words, questions about agency are, whether implicitly or explicitly, also questions about the limiting power of larger forces, whether it is Augustine wondering how free anyone could be in a world preordained by God (Weaver, 1996), Kant worrying that Newtonian causation might make us all so many measurable phenomena (Westphal, 2017), or Pierre Bourdieu positing that our thoughts and actions are almost (though not entirely) reducible to field position (Bourdieu & Wacquant, 1992). To ask about agency is usually to ask about freedom.

Yet there are other, more orthogonal approaches to agency, conversations that move away from disambiguating when people are actually free, or the structured nature of their freedom, whether via “structured structures predisposed to function as structuring structures” (Bourdieu, 1977, p. 72) or anything else. In these conversations, the question is not about how people are agentic but rather how things are agentic (Latour, 2005; Verbeek, 2021), often with a parallel focus on how that agency is experienced by people (Guhin, 2020). How and when do we feel as though something else is an agent, and why? And are we correct?

This distinction can go many ways, and it can again return to the old debates within the North Atlantic intellectual tradition. For example, we might examine the degree to which we think a certain criminal is morally culpable for their crimes. One side might argue that Jean Valjean did steal a loaf of bread and should therefore be held responsible for it (Hugo, 1863). He was free to steal or not to steal, but he stole. Another side might argue that it was Valjean’s poverty that compelled him to do what he did; as a result, Valjean was not entirely free. Again, these are ancient debates that have moved from predestination to Newtonian causation, from an awareness of the environment to an awareness of the genetic basis of behavior and then back to what the social world does to genes and how we think about them (Landecker & Panofsky, 2013).

Yet, I want to shift from asking about an individual person’s freedom—Jean Valjean in this case—and move instead to a question about what I had earlier said about what someone might argue about Valjean—that he was compelled by his own poverty. What does it mean to talk about poverty compelling an action? For example, in his book, The Price of Poverty, anthropologist Daniel Dohan (2003) describes certain “institutions of poverty,” consistently making them the subjects of his sentences. They “shaped how residents. . .struggled to advance by shaping how residents engaged in and combined jobs” (p. 12); “facilitate income generation even as they impede morality” (p. 224); and “spring from and sustain the social, political, and cultural environment of the low-income urban neighborhood (p. 4). Perhaps this is simply metaphor or synecdoche. After all, when we write in an academic paper that The Wealth of Nations says something, we are probably not referencing a talking book but rather the words in the book put there by an author. Or so it would seem. Yet, even if these are words, these words are not the author themselves. And it is still interesting that we can write about what books say and that the turn of phrase makes such ready sense to so many. What do books do? And how might these actions of books be like the actions of an institution like poverty?

Social scientists and philosophers have tended to argue that this kind of agent-talk is more than a metaphor: social institutions (Abrutyn 2013; Friedland & Alford, 1991; Hallett, 2010; Meyer, 2010; Thornton et al., 2012; Turner 1997) do appear to act in the world as “social forces” (Schmaus 2004). These social forces are themselves social constructions (Berger & Luckmann, 1967), only existent because of human consciousness. Yet, this does not mean such forces are reducible to so many individual actors; indeed, the emergence of a social phenomenon as greater than the sum of its parts is a key insight in both social science and social philosophy (Sawyer, 2005; Searle, 1995).

These insights are to some degree as old as sociology itself. Even Weber—that famous methodological individualist—recognized the way that institutional and organizational forms coerced, constrained, and made coherent particular ways of imagining and then living a life (Weber, 1978). Yet, I am not simply making an argument that institutions or any other “social forces” have real effects. Instead, I’m asking about what it means to experience institutions or other non-humans as agentic. As I describe in an earlier study of the “external authorities” of scripture, prayer, and science, I’m flipping the script from previous studies of agency: “while Meyer and his coauthors have examined agency as an institution, they have not looked, at least not explicitly and interactionally, at how institutions are narrated and experienced as agentic” (2020, p. 7–8). Building on this critique, I show how not only institutions but also books, practices, and technologies come to gain an agentic capacity.

This shifts the discussion of agency away from whether living people have agency to whether things that are not people, or at least not living people have something we might call agency. In this case, it is not even clear what poverty is—ontologically, sociologically, politically, etc. Sociologists might call it an institution, but that is not entirely obvious, nor is this quite capacious term itself agreed upon. Still, here is the question: what does it mean to say poverty compelled someone? Or to say science shows something? Or to say that scripture speaks? What are the traces these things leave behind, and how can we recognize agency through examining them?

Relatedly, what does it mean to say, for example, that Durkheim argues something? Durkheim, after all is, dead. Do we mean that he did argue something in his text and the argument has present-day relevance and so we bring it into the present? Or it something spookier? Yet, it is nonetheless sociologically interesting how often people refer to the dead or the not-human as doing things we traditionally associate with people with wills: doing, saying, commanding, forbidding, preventing, etc. The dead leave traces, of course, but what do those traces do?

In this paper, I will review some of the ways that philosophers, literary scholars, and sociologists have thought about non-human agency, as well as the agency of the dead. What does it mean to describe or discuss the dead in present tense, especially those who have written important texts? What does agency look like when mediated through non-humans and through history? After providing some examples that might be helpful to think with, I will review some previous work on questions of agency. Then I will review some of the literature, looking at this spooky action at a temporal distance: path dependence, residue, repression, and ghosts. I will end by thinking about what all of this might reveal about the need for a theory of agentification.

Three Traces

To help me think about these questions, I want to examine three different events, all of which happened within 20 years of each other. Let us go back around 250 years to the English-speaking North Atlantic and meet our interlocutors. Each of these have left vital traces across history, and I’m interested in what those traces reveal about the agency of the non-living and non-human.

The first event happened in 1796. President George Washington, the first president of the United States, chose not to run for re-election for a third term, which he would surely have won (Beermann, 2010; Peabody, 2001). This decision, Washington hoped, would help to strengthen the office of the president while also providing a template for the peaceful transition of power (especially between parties), a tradition which would be tested 4 years later in the transition between America’s second president, John Adams, and its third president, Thomas Jefferson. Of course, in an era marked by capital insurrection and “fake elections,” it remains to be seen whether this tradition can be traced forwards as well as backwards, but Washington, school children across the United States learn, gave it a good start.

The second event happened in the exact same year. Edward Jenner (is said to have) invented the smallpox vaccine, with obviously vital implications as we ease out of our own global plague (Bennett, 2020; Gross & Sepkowitz, 1998; Hammarsten et al, 1979). Smallpox was a devastating disease for thousands of years, and it is the first (and only) disease to have been totally eradicated through vaccination. While something like vaccination had existed before Jenner’s work—it had actually been imported from Turkey earlier—Jenner was the first to vaccinate as we now understand the term, and his invention is considered one of the most important in the history of medicine, with traces obvious across virtually every sphere of modern life.

The third event happened 20 years earlier. In 1776, Adam Smith, after years of study and work, published The Wealth of Nations, a stunning masterpiece of economic, political, and moral theory whose importance continues to reverberate today (Harrison, 2011; Ross, 2010). While economics had been studied for centuries before Smith, and while Smith’s theory of value had more in common with Marx’s than with the “marginal revolution” (Blaug, 1972) that would come around a century later, it is nonetheless a commonplace assumption that Adam Smith, and especially The Wealth of Nations, set the stage for modern economics, with traces clear all over the world (Norman, 2018).

I want to make a few quick points about the three examples I just provided. First, note that all of them can be understood as events. In her recent book, What is an Event? (2017). Wagner Pacifici (2017) shows how events can be at once discrete and continuous. In this sense, the invention of the smallpox vaccine is over and done with, even as the ongoing importance of the vaccine remains, a residue of that originating event we can continually trace back to its origin. That residue endures because of the reshaping of social life that events provide. As Sewell (1996, p. 843) describes them, “Historical events tend to transform social relations in ways that could not be fully predicted from the gradual changes that may have made them possible. . .[events] reshape history, imparting an unforeseen direction to social development and altering the nature of the causal nexus in which social interactions take place”. Note here how Sewell describes events themselves as the subjects of these sentences. It is neither Smith, nor Jenner, nor Washington who radically changes the social structures of their worlds but rather the events in which they have participated. In these cases, it is events—rather than people—that leave traces.

That events are more than the sum of their individual actors is a sociological insight at least as old as Ecclesiastes. Yet even as the sociologically minded strive to decenter individuals, it is striking how often these stories are traced back to individual actions, which is the second point I will make. It is Washington’s choice that still seems to matter in American norms, just as it is Jenner’s vaccine that seems to keep us safe. However, there is an interesting difference here in that for technology, the author seems more easily lost than in politics. The technology seems to take on an agency of its own—what a vaccine can do—rather than always Jenner acting within the vaccine. And this is even more the case when comparing technology to the written word. Because of course we talk about texts as agents all the time—what The Wealth of Nations says about trust-busting, or what The Iliad can still teach us about the horrors of war.

Yet, what is important to me here is the question of tense, which raises my third point. Why is it that Adam Smith still speaks to us in the present tense, at least in the way we narrate him, while Washington and Jenner have done all of their work years ago? I think the fact of present tense reveals something interesting and puzzling about tracing agency, specifically the agency of authors revealed in long-lasing texts: how actions from centuries ago are felt through time as ongoing, an electric current that is, in fact, current. Adam Smith is dead, yet he keeps on speaking. To any academic, that the dead keep on speaking is as obvious as a library’s walls, yet the puzzle remains: the dead, in fact, do not speak. They do not do anything else either. So how are we to conceptualize Smith’s speech, especially if it has some causal purchase on contemporary actions in the world?

To be clear, none of the events I described above are quite as simple as I described them. There is ongoing scholarly debate about Washington’s motives and relevance (Peabody, 2001), Jenner’s centrality (Gross & Sepkowitz, 1998), and Smith’s status as the ur-economist (Brown, 1996). Yet, whether or not the stories I described above are too simple to be true, they are certainly not too simple to be believed. I learned some version of each of these stories before I turned 18, and so do many other Americans. This comes to my fourth point: these stories all contain the construction of certain autonomous “moments” that might well be different from what historians say really happened. Yet as scholarship on Washington reveals, the point of stories about Washington is not simply to reconstruct historical events but to create an iconic figure and reshape a national culture (Schwartz, 1991, 1983). While few scholars would claim that the events described above do not matter at all, many would point to how these events are cultural phenomena that remain malleable.

In other words, alongside the relatively discrete set of historical questions about what Jenner’s vaccine did and continues to do, there are the concurrent political and cultural questions about what Jenner’s vaccine meant and how that meaning has changed (Hobson-West, 2007). The felt agency of these events is a combination of these two sets of concerns, both a historical narrative about what the events produced (which might or might not be accurate) and an ongoing sense of what the events meant and continue to mean. These obviously influence each other as well. For example, the fact that Washington owned people as slaves is increasingly affecting his reputation as “the father” of the United States. This change in Washington’s meaning can then have an effect on the broader willingness of Americans to accept or privilege a narrative account of Washington’s centrality in the establishing of the peaceful transition of power. The meaning of an event can either empower or discourage the popular agentification of the event as a historical actor, regardless of what historians might claim. To trace an event’s power, we must also trace its meaning.

Agent Detection

Before I get to these other theories, it might be helpful to provide a bit more context to how people seem to think about agency.

Some psychologists describe an “agent detection mechanism,” in which humans tend to detect the presence of other agents, even when there are none to be found (Beier, 2008; Beier & Carey, 2014; Bertolotti & Magnani, 2010). Imagine sleeping in a cave. You hear a twig snap outside. Of course, it could be a random twig snapping: falling off a branch, the wind, a host of reasons. But it could also be a tiger or some other predator, or a human who means you harm. Erring on the side of assuming agency has the evolutionary advantage of keeping you alive, while erring on the side of not worrying about it runs the small but real risk of getting you killed. There is some evidence via child psychology that children tend to experience many thing and beings as agentic (a wagon is able to be said hello to, able to be hugged, able to be hit when having caused pain). Indeed, that sort of “human” connection to non-human things extends well into adulthood, and not just for those cast away on an island who develop deep friendships with volleyballs, as Tom Hanks’s character does in Castaway. Animism—the sense that everything around us has a will—has never entirely gone away, as the recent resurgence of studies of animism reveal (Hornborg, 2006; Merewether, 2019; Okanda et al., 2021; Wilkinson, 2017) and as anyone who has yelled at their television can recognize.

This “agent detection mechanism” hypothesis is used most often to account for why religion makes intuitive sense to people: it seems more plausible to many, following a tendency toward believing in agents, that some being causes lightning and seasons and love and death, rather than them simply happening. And we might call those beings gods (Atran, 2002). Yet, the mechanism can perhaps explain a bit more than religion. Just as the seasons can feel too complex and broad to be narrated and understood as the diffuse processes they actually are, so too can social processes, organizations, and institutions. France is much more complicated than any one set of people or processes, yet we nonetheless feel comfortable talking about a thing called “France” as acting in the world, an agent with requests and reasons, hatreds and hopes. The same can be said for how we talk about corporations, as the sociologist Carly Knight (2018) has shown in her work. “Apple” is a vastly complicated organization, but it is described and narrated by humans as an individual. Indeed, according to United States law, Apple is itself an individual person with the rights thereof.

Yet, this does not have to only be a story about organizations, corporations, and states. Think about social processes and institutions as well. What “family” demands of us, for example, is not quite the same as what individual members of an individual family demand or even what one specific family group demands. Similarly, science and religion are quite often narrated and experienced as individual entities with coherent lessons, demands, and intentions. Those within either science or religion—like those working inside Apple or France—might be skeptical that any of these categories are coherent and homogenous enough to be thought of as unitary, let alone as agentic, but nonetheless, such unitary agency is exactly how Apple, France, Religion, and Science are often narrated and described. Science proves, Religion moralizes, France wants, and Apple accomplishes. Recognizing how humans “agentify” institutions and organizations helps us to trace agent-work in action.

Some Previous Theories

In this section, I will argue that many theories get at some of what I’m describing here, but none of them exactly cover the questions I’m asking. I will review the concepts of (1) actants, (2) affordances, (3) switchmen, (4) residue, (5) repression, and (6) ghosts, looking at some of the ways other scholars have engaged the question of non-human agency. I will then turn to my own argument.

Between Actants and Affordances

The sociologist Bruno Latour (2005, 1987) has famously coined the term “actants” as a key part of a “flat ontology” in which everything can act upon anything else (2005, 1987). He really does means everything, from plain old humans down to bacteria and up to a nation or a corporation. What matters is the network of actors, or actants as Latour might prefer, thereby emphasizing how many of those doing the acting are non-humans. Actor Network Theory, as Latour and his colleagues call it, might seem to solve our problems; after all, here is a specific discussion of non-human and material actors. Yet for my purposes, Latour’s interests here presume what I am seeking to describe. I do not take for granted that the vaccine is an actor, for example. I’m just as interested in the cultural and semiotic processes through which the vaccine is an actor as in the role of the vaccine as an agent in history. In other words, for Latour, actants are an etic category; while useful and fascinating, that presumes the question I’m asking, which is precisely how agents become an emic category; how and when do people come to experience the non-human and non-living as agents, and how can we use the traces of that agency to identify the processes involved?

So if actants offer us too much, then affordances (Bloomfield et al., 2010; Ostertag, 2021) might not offer us enough, including Tavory and Jerolmack’s (2014) clever study of affordances in non-human agents. And some of the examples I provided above do have affordances, that is, certain characteristics that “afford” certain behaviors easier than others, like a door to opening, a chair to sitting, or, in Tavory’s (2010) work, a yarmulke to identifying as Jewish. Yet to argue that non-human agents have affordances is not quite the same thing as to argue they are agents, and that is my question here. This is not a study of non-human interaction but non-human agents. So I know turn to four other ways scholars have examined non-human agency and the traces it leaves behind: switchmen, residue, repression, and ghosts.

Switchmen of History

The quote is famous enough, but I will repeat it here: “Not ideas, but material and ideal interests, directly govern men [and women’s] conduct. Yet very frequently the ‘world images’ that have been created by ‘ideas’ have, like switchmen, determined the tracks along which action has been pushed by the dynamics of interest. . .” (1946, p. 280). Now what makes a world image (Tenbruck, 1980)? The concept is a bit muddy for Weber, especially as the seeming universality of world images (not too different from a parallel contemporaneous German concept of worldview) finds an uneasy coexistence with his more famous concept of value spheres, a tension between comprehensive identity and institutional specificity that is repeated in a similar way in Pierre Bourdieu’s theory of fields (itself deeply indebted to Weber). Yet, we can safely bracket the “worldness” of these world images here and instead focus simply on the historical part of this account, which has been especially important in Weber’s ongoing explanations of religious identity (Radkau, 2013).

For example, the personalities of Jesus of Nazareth or the prophet Mohammad matter, argues Weber, because these charismatic figures have a way of relating to and viewing the world that, institutionally contingent or not, set to sail a certain institutional ship, leaving traces for all of history back to these original individuals. Now that ship might, like the ship of Theseus, become entirely replaced over time. Yet, there is still something in its form and its intended purpose—what Aristotle would call its formal cause and final cause—that lives well beyond what Aristotle would call its material cause (the stuff it’s made of) and its efficient cause (the person or thing that made it). Now that does not mean religion is immutable: it is entirely possible, even likely, that Jesus would find the Vatican entirely unrecognizable. Yet, the point is simply that there is probably something that remains of Jesus in the Vatican, just as there is something that remains of bacteria in human DNA. And what we can trace back to Jesus, Weber might argue, is something of the world image.

So what does this have to do with our examples of agency? Let us go back to Jenner inventing the vaccine. There is not necessarily a world image that sticks around in the vaccine itself, but we could imagine various STS critiques—functionally similar to economic accounts of path dependence—arguing that certain forms of technology are sticky products of their time, with cars bringing forward metaphors and relations to horses, for example, when horses are no longer how we organize transportation. Drawing from SCOT (the social construction of technology), Kline and Pinch (1996, p. 765) show how “‘relevant social groups’ who play a role in the development of a technological artifact are defined as those groups who share a meaning of the artifact. This meaning can then be used to explain particular development paths”. My questions are motivated by the same desire to trace a technology’s origin story. Does Jenner act through time through inventing his vaccine? Is his event complete?

I could ask the same question about Washington’s world image, and this gets even more sociologically complicated, as the discrete choice Washington made not to seek re-election may have had a powerful effect, but what Aristotle would call the efficient cause of that choice was not necessarily Washington, as philosophers and sociologists of decision theory might make clear. As Gabriel Abend (2018a, 2018b) has described in some of his recent work, how we figure out what is a choice is not nearly as straightforward as it might appear. But this gets us away from my question of agentification and toward the trickier question of agency versus structure, and that is not where I want to be. The question then is not who or what really made the choice we attributed to Washington or how free was Washington really when he made that choice but rather how we narrate and understand a dead Washington—or a discrete event in Washington’s life—as still having this agentic capacity. And, to continue the question of world images, the question remains, as I described before, a mixture of historical inquiry and ongoing cultural salience. How Washington helped to create a world-image does matter, but just as relevant is how contemporary understandings of America’s world-image (to the extent such a thing exists) narrate Washington as either a founding patriarch or frankly problematic. What are the traces Washington has left behind? Self-government or slavery or both?

But it is The Wealth of Nations that I am most interested in here as this is where world images become most complicated. What is the world image of The Wealth of Nations? Is it the Scottish Enlightenment from which it comes? Or the Austrian and neo-Austrian Chicago schools of economics that have come to claim the book as an ur-text? What is this 1776 book’s ongoing status as an event or as an agent, let alone that of its author? If a world-image can be a switchman of history, then certain texts seem able, at the same time, to keep trains on pre-established tracks and, with enough cleverness, to change the nature of things enough to set new tracks any time. Which is to say texts can be events or agents. And so, perhaps, can their authors.

So what say switchmen about the agency of the dead? The dead do appear to act, switchmen might show, and we can follow the traces they leave behind within the Weberian tradition of sociological inquiry. Yet, how the dead act remains undertheorized within this tradition.

Residue and Repression

Next I want to look at the concepts of repression in Freud and residue in Raymond Williams, especially pulling two important pieces here that could be found in Weber but are better articulated in Williams and Freud. The first, from Williams (2020, p. 47), is that the residual hanging-on from previous societies “still seem to have some significance because they represent areas of human experience, aspiration and achievement, which the dominant culture under-values or opposes, or even cannot recognize. . .”. What is important about Williams’ account of the “remnants” and residue of past societies is the way they function relatively autonomously both from each other and from the society as a whole. The traces of these part eras, in other words, are also things that act. To be clear, Williams does not always refer to these remnants in an agentic way, and the remnants’ agentic capacity is not an explicit part of his argument. Yet, not only does the past lives on even when unattached to the current means of production and concomitant social organization but the past can be experienced as living on through a series of non-human agents. Things like “Washington’s decision not to have a third term,” “the smallpox vaccine,” The Wealth of Nations, and a dead human named Adam Smith. As with Weber, Williams does not quite provide the solutions to the questions I’ve been raising, but he does point to similar phenomena.

The same could be said for Freud’s concept of repression (Billig, 1999, 2008; Freud, 1936, 1957; Ricoeur, 2008), which has obviously been extremely influential. According to Freudian social analysis, cultures and communities can repress certain beliefs or commitments yet not repress them entirely, leaving the seemingly repressed idea with an unrecognized power. One important idea from psychoanalysis is denegation, the French (and now English) translation of Freud’s concept of Verneinung that describes the negation of a repression that is neither entirely accepted nor entirely eliminated: “The negation is neither an acceptance nor an entire rejection of the repressed material, but rather. . .a ‘sublimation’ that is characteristic of a Hegelian synthesis in which the negation of the antithesis is at once negated as a separate proposition and absorbed into a higher synthetic view” (Foshay, 1997, p. 596). Yet, what is useful to me is not the repression itself or even the denegation but the ongoing engagement and experience with a certain concept, experience, or event that is felt to have an agentic power throughout time, even when that event appears negated, repressed, or forgotten. Tracing this repression and denegation through individual or sociological analysis is obviously a key piece of the psychoanalytic tradition, yet I am less interested in unpacking the traces here and more committed to understanding how traces of repressed ideas can reveal the agentic capacity of the non-living and dead.

Other scholars, especially Nobert Elias (Déchaux, 1993; Elias, 1998; Lahire, 2013; Redner, 2015) and Pierre Bourdieu (Bourdieu & Wacquant, 1992; Bourdieu, 1991; Déchaux, 1993; Steinmetz, 2014), have been excellent at showing how psychoanalytic concepts can be useful for sociological problems: Freud might have theorized repression as a psychological process, but that does not preclude repression from being societal and sociological as well. Yet even if the psychoanalytic tradition gets us part of the way there, it remains for others to theorize how that power is understood, experienced, felt, and narrated as an agent.

Ghosts

While Derrida (De La Campa, 1999; Derrida, 2012) has famously studied haunting and ghosts in cultural theory, I want to turn to a more sociological account, that of Avery Gordon’s famous study of “haunting and the sociological imagination” For Gordon (2008),

the ghost imports a charged strangeness into the place or sphere it is haunting, thus unsettling the propriety and property lines that delimit a zone of activity or knowledge. . .the ghost is primarily a symptom of what is missing. It gives notice not only to itself but also to what it represents. What is represented is usually a loss, sometimes of life, sometimes of a path not taken. From a certain vantage point the ghost also simultaneously represents a future possibility, a hope. . ..the ghost is alive, so to speak. We are in relation to it and it has designs on us such that we must reckon with it graciously, attempting to offer it a hospitable memory out of a concern for justice. Out of a concern for justice would be the only reason one would bother (2008, p. 63–64).

In Gordon’s illuminating reading of history, politics, and fiction, she shows how certain people can “see that the not there is a seething presence, [can] comprehend the living effects, seething and lingering of what seems over and done with, the endings that are not over” (p. 195). Drawing as much from Freud and psychoanalysis as from Raymond Williams and other cultural materialists, Gordon argues that “when it appears to you, the ghost will inaugurate the necessity of doing something about it” (p. 206). Gordon emphasizes how elements of the past carry forward in time in ways that can also be agentic, even if they are not the agents we are used to. In this case, it is neither Washington nor his decision to retire, neither Jenner’s vaccine nor Smith’s famous book. These ghosts are those people Washington enslaved perhaps, or those deemed unworthy of Jenner’s vaccines. Perhaps the ghosts are the billions still unlifted by the grand promises many read into The Wealth of Nations. And these are only the obvious ghosts, the ones already lingering right outside of our field of vision. There are many more agentic ghosts we might be too morally obtuse to sense, even elliptically. This study of ghosts parallels a study of repression with implications for the endurance of world-images as well: the pacifism of Jesus is at once present and ignored by Christian warmongers. The ghosts of American Christians’ slave-owning past are repressed and, ironically, strengthened in their repression, similar to Freud’s concept of denegation.

Gordon’s study of ghosts gets the closest yet to what I’m describing of what the dead and the non-human can do, yet its normative and political centering, while vital, can miss more generic sociological possibilities. In other words, Gordon’s study of ghosts focuses on the moral weight of counterfactual history, leading the living towards the experience of an actual relationship with the traces of people and paths left behind. That gets some of the way toward what I’m describing, but not quite all the way there. What about a felt relationship to something as vague and generic as Science? Or France? Or Vaccines? Or Capitalism? Therese broad institutions have a kind of ghostly agency in that they are not living, breathing humans, yet they are not quite ghosts as Gordon describes them.

Agency as More Than Metaphor

In this section of the paper, I turn to my recent book, Agents of God, to argue that my theory of “external authorities” can, with some revisions, answer many of the questions I raised in the earlier section of this paper. In my ethnographic account of two Muslim and two Evangelical high schools (2020), I show how students and teachers at religious high schools in the New York City area describe scripture, prayer, and science as “doing” and “commanding” certain things. I demonstrate that, within these school communities, scripture, prayer, and science are experienced as “external authorities” that solve certain “paradoxes of power” in educational contexts: teachers want students to do what they are expected to do while also wanting those students to want to do it. In other words, while schools are inevitably coercive organizations, sometimes that coercion is not recognized as coercion, especially by teachers, but by students as well. One way this problem is solved is by shifting the agent in question: instead of the teacher experiencing herself as giving a command, she is instead simply an agent of someone—or something—else. I draw on other scholars’ studies of agency (Adams, 2011; Reed & Weinman, 2019; Reed, 2017, 2013) and institutions (Hallett, 2010, 2007; Meyer & Jepperson, 2000; Meyer, 2010) to show how certain kinds of institutions are experienced as relatively autonomous agents within demarcated social locations.

For example, in an interview with a teacher at one of the Evangelical schools I studied, a senior Bible teacher described why he had to oppose gay marriage, even if he was sympathetic to the problems gay people experienced. “I wish I could say it was okay but I have to follow the Bible, and so I can say that I have to say homosexual practice is a sin. . . So I mean, it’s—boy, I wish it was okay but the Bible tells me it’s not” (2020, p. 141). Note how it is not God who is making these claims upon the teacher but rather the Bible itself, making explicit demands and “telling” the teacher to do and believe certain things. I describe a similar process of “agentification” in my chapters on science and prayer.

My theory of external authorities is primarily about certain institutions that are experienced and narrated as agentic, yet it could easily be expanded to any other non-humans. Yet, one key difference from my work and some of the questions I’m asking about here is the question of power. In Agents of God, the defining characteristics of an “external authority” is contained within its name: something like the Bible externalizes authority, giving a teacher or other authority figure in the school the legitimacy of putting the authority for a claim outside of themselves, “the same way that a lever gains strength when its fulcrum is further away from the effort applied” (2020, p. 7). Yet, this focus on authority does not need to be as interactional as I describe external authorities and, in a different way, as Gordon describes ghosts. Instead, the authority contained with my definition of external authorities and the summoning of Gordon’s ghosts can be understood as both the historical causation and ongoing meaning I described in this article in reference to Washington, Jenner, and Smith. And now we have an answer to how and when and why the dead and non-living act. And we know how we can trace those actions. We simply follow what people say they do. In that sense, something like Weber’s world-image is helpful for us not in the etic sense of tracing what remains of Jesus in the Vatican. The world-image argument itself moves from being a theory to a kind of data, an illustration of how people—including Weber—trace the power of long-dead individuals across history and space.

However important the Thomas theorem (Thomas, 1923) may be, I recognize this pragmatism might appear both overly convenient and more than a little glib. Yet, my argument that things are agents when our respondents say they are agents is not intended as either a metaphysical solution or an end to the conversation. Instead, what I am suggesting is an empirical redirection of a philosophical problem, looking at the ways that people do non-human agency through careful study of their interactions with non-human others. This is, in many ways, simply the research agenda of Bruno Latour and his colleagues, though with a key distinction: I am interested here in the explicit agentification of the dead, of organizations, and of institutions. In the research agenda I am suggesting here, such explicit agentification must be at the emic rather than at the etic level, something which is not always the case for Latour’s work. The same could be said for Weber, Williams, and Freud, alongside their interpreters. As with Weber, Latour, and his colleagues’ interpretation of agency becomes at once theory and data, ways to orient how we conceptualize agency and ways to illustrate how those conceptualizations themselves maintain this agent-work.

This focus on agentification’s pragmatic payoff might be effective enough for my earlier discussion of what Washington means, for example, or how people experience the ongoing power of a vaccine, or Jenner’s vaccine specifically, or Jenner himself. Yet, a focus on pragmatic payoff seems to avoid the more complicated philosophical and historical questions I have been asking throughout this paper. It is all well and good to look at how things and institutions and the dead are agentified by people through tracing these processes empirically. But how does such an interactional focus answer the broadly metaphysical questions I have been posing about causation and identity?

I would suggest that the interactional questions and the metaphysical questions are not as far apart as they might seem. A phenomenological and pragmatic focus on what people do and what those actions mean can go a long way to explaining something like historical change and the vagaries of identity: note, for example, how the centrality of events for Sewell hinge not only on what events do but also on how people experience and interpret the actions of those events in their lives and in their social worlds. Of course, my pragmatism here is also a deep constructivism, and to the extent that someone wants a realist account of whether, how, and why certain things and institutions are or are not agents, then, depending on your definition of realism, my account here is inevitably disappointing. Yet, I am not convinced that sociologists, at least as sociologists, are capable of making ontological claims outside of what people do and say and believe. A social ontology is still an ontology (Durkheim, 1995; Searle, 1995) . And to the extent that social ontologies are what we have to work with, then even the purely metaphysical questions I have posed are ultimately empirical ones. If you want to find out if a book can truly speak, then just ask people the book has spoken to. In this sense, to say the dead do things is not fully metaphysical in either the philosophical or mystical sense of the word; it is simply empirical.

Conclusion and Implications for Future Work

So where does that leave the tracing of the events I described above? As I described in this paper’s title, this paper is less a summation than it is a prologue, so I am content mostly to have cleared ground here, showing why previous theories cannot adequately address the problems of agentification I identify here? The remaining questions, as I argue, are empirical ones. What are the conditions under which Jenner’s vaccine maintains its agentic power, and what traces make such agentification more likely? Why does someone like Washington leave so many more traces than the advisors or other contexts than probably had as much casual power as he died? These discussion are more than simply a meta-analysis of the “great man theory of history”: it is surely true that complex social processes are too often simplified into stories of a few powerful individual making discrete choices. The sociological work of tracing is often an explicit counter-argument to this tendency, showing how any one of us is less free than we might believe.

Yet, this again returns to the kind of agency debate I do not want to rehash here. Rather than ferret out the difference between agency and structure, I would rather focus on how agency is experienced by and through humans, and how we trace the experience of the world we have today the actions—past, present, and ongoing—of the non-human and the dead. In the sense, the great man theory of history matters not because it is a bad sociological argument but rather because it is a common folk argument, revealing an empirical tendency toward the ongoing agentic power of unique individuals from the past.

And such agentic power does not end with the dead. We humans regularly turn non-humans, especially organizations or institutions, into singular agents felt to have a certain kind of authority. We can trace that agency-work back to evolutionary origins or the emergence of social forces, but whatever the cause, such agent-work leaves traces in the world—a mandate for vaccinations, a process for political transition, a system of economic order—each of which can be traced back to dead actors or discrete events that seem to continue acting. The author, with apologies to Roland Barthes (Barthes, 1977), seems very much alive, or at least spectrally present. These events seem to maintain a world-image, bringing affordances to future interactions.

Yet, this agency is more than the sum of its affordances, and the empirical question of whether an institution or dead person is agentic, for my purposes, must begin as an emic, on-the-ground process rather than an etic decision for analytic purposes. We can trace back all sorts of reasons why economics is the way it is, with some paths leading back to Adam Smith and some not. Yet, what is interesting to me here is not what we say about Adam Smith but rather how we say things about Adam Smith, that, is the degree to which he—or the world-image of his era—is agentified in the present, or, even more importantly, whether his book does so. In this sense, sociological arguments are as much data about agentification as they are explanations of the process.

Of course, all of this is subject to empirical tests. This is less an ontological philosophical claim that it is a theoretical grounding of empirical sociological work. As both Carly Knight and I have noted in our work, people describe institutions, organizations, and other non-human things as agents, and these descriptions have real-world stakes. It remains the job of other agents to get a better grasp of what those stakes entail.