In 1996, William Sewell published one of the most important works in historical sociology, “Three Temporalities: Toward an Eventful Sociology” (working paper from 1990 available here!). The essay argued that classic works of sociology too often followed the wrong approaches to temporality. Iconic works of historical sociology often explicitly or implicitly invoked teleological or experimental modes of temporality. In contrast, Sewell argued that historical sociologists should instead approach temporality as eventful. Since then, historical sociologists have largely strived to do just that.
While thinking about Sewell’s schema for a paper I’m writing on Polanyi, I realized that the typology was useful for making sense of a very different context: roleplaying games (RPGs). In this post, I’ll briefly explain Sewell’s argument by showing how the three temporalities map onto three ways of playing D&D (or, really, any similar tabletop RPG). Unlike with historical sociology, I will not be making an argument that one way of playing is “better”, though I do think we can learn something from Sewell’s typology about why the modes of play analogous to teleological and experimental temporality are frequently unsatisfying, and about the ingredients of a good eventful campaign.
Experimental Temporality & The Sandbox
Statisticians and quantitative social sciences often refer to a “data generating process.” The idea is that any given observation in a dataset can be understood as being produced by some underlying, probabilistic dynamic. The goal of all the fancy math is to try to infer features about that underlying process. Experimental temporality involves treating the past the way that quants treat survey data, as outcomes of a (mostly) unchanging random process whose distribution we are determined to uncover.
This metaphor is not perfect; even historical sociology in the experimental temporality mode tends to be more interested in processes and mechanisms than (some) quant work. But the deep logic is similar, as Sewell shows in his analysis of the Theda Skocpol’s States and Social Revolutions. Sewell argues that Skocpol’s attempt to hew closely to Mills’ method for comparing cases leads her to ignore, among other things, how revolutions interact over time, how the fact of the Russian Revolution changes the discourses and possibilities for the Chinese Revolution, and so on. For Skocpol, at least in her most programmatic mode in the book, each case is separate, and helps to illustrate the necessary and sufficient features of social revolution, which is itself imagined as a stable sort of process that sometimes happens when the right features happen to line up.
Experimental temporality corresponds to a mode of play known as the sandbox. The campaign (like history for Skocpol) is imagined as a series of basically unrelated, sometimes even probabilistically generated, challenges to overcome. The characters are effectively playing in a giant sandbox. Clearing out one dungeon changes nothing fundamental about the nature of the world, about the “structures” (in Sewell’s terms) that define it. Each adventure is serially independent of the last. There’s no consequential greater story arc, there’s just the thin threads that characters chose to follow that lead the characters from one adventure to the next.
Sandbox campaigns can be a blast! But like experimental temporality, they can often be a let down. In particular, players (read: historical actors) do not experience any agency. Their actions cannot have lasting consequences because the structures in place are fixed; the next dungeon will just be a harder version of the previous one. In a sandbox-style campaign (experimental temporality), players (historical actors) cannot fundamentally alter the world they inhabit.
Teleological Temporality & Railroading
In some ways, the polar opposite of experimental temporality is teleological temporality. While experimental temporality assumes that the structures underlying the world (the data generating processes) are fixed, teleological temporality assumes not only that these structures change, but they are progressing (or occasionally regressing) towards some fixed endpoint. Change is not only possible, but inevitable.
Classically, Marx (at least at times) is the most obvious teleological thinker in the historical sociology canon. Capitalism follows Feudalism, and works according to a very different logic. In turn, Socialism will follow Capitalism. The Revolution is coming, and once it occurs, the entire system will work differently. The structures of capitalism will fall by the wayside, replaced by a new set of structures.
Mid-20th century historical sociology was rarely quite so deterministic as Marx himself, but Sewell shows how teleological temporality enters into projects as distinct as Immanuel Wallerstein and Charles Tilly’s. The whole notion of “modernity” tends to lead us to think teleologically. Some aspects of the world are throwbacks, holdovers, or fetters (in Marx’s language); others are progressive, forward-looking, on the path towards the next thing (for better or for worse).
Teleological temporality corresponds nicely to what gamers call “railroading.” In a game characterized by railroading, the Dungeon Master (DM) running the game has a plan. Playing in a railroaded game feels a bit like acting out someone else’s script, or reading through someone else’s novel. The characters are involved in epic, world-shattering events, but the conclusions to those events are pre-determined. Kingdoms rise and fall, gods are born and destroyed, the world after looks different from the world before, but…
…the players still have no agency. While in a sandbox, players have the illusion of choice between adventures (none of which can alter the structures of the world), in a railroad, the players get to experience transformations of the structure of the world, but no choice in how those transformations play out. Similarly, in teleological temporality, historical actors show up and play their appointed roles (“revolutionary vanguard”, etc.), but their choices cannot fundamentally alter the direction of history.
Eventful Temporality & Eventful DMing
The final mode that Sewell elaborates is eventful temporality. Sewell’s definition is worth quoting:
Eventful temporality recognizes the power of events in history. Social life may be conceptualized as being composed of countless happenings or encounters in which persons and groups of persons engage in social action. Their actions are constrained and enabled by the constitutive structures of their societies. Most happenings reproduce social and cultural structures without significant changes (Giddens 1984; Sewell 1992). Events may be defined as that relatively rare subclass of happenings that significantly transform structures. An eventful conception of temporality, therefore, is one that takes into account the transformation of structures by events. (Sewell 1996: 262)
Sewell contrasts eventful temporality with both experimental and teleological temporality. While experimental temporality emphasizes uniformity of causal laws and independence of events, eventful temporality focuses on the heterogeneity of causal laws (things work differently at different points in time) and interdependence (exactly what happens during one event may change what track history runs on, and thus events at time two are partly a function of what happens at time one). Similarly, while teleological temporality denies the importance of contingency and assumes a fixed endpoint, eventful temporality allows for radical contingency.
In a teleological mode, any individual decision cannot alter the course of history; in eventful temporality, it very much can, if the decision happens to be made during an event. Sewell’s own analysis of the storming of the Bastille in the French Revolution is probably the most famous “eventful” argument. In this account, while structural forces may have partially determined that a revolution of some sort would occur, the particular, contingent choices made during important events shaped exactly what the revolution would mean, and how it would transform the structures of French society.
Unlike the sandbox and railroading, there’s no single term that directly captures this flavor of temporality as translated into the context of D&D. Advice to DMs is mostly given in terms of “don’ts” – don’t just provide a sandbox, don’t ride your players too hard onto a railroad, etc. But I think it’s worth thinking about, for lack of a better term, eventful DMing as its own mode, rather than simply a middle ground, or an avoiding of pitfalls from the other two.
For Sewell, eventful temporality is satisfying for historical explanation because it acknowledges both the importance of enduring structures and the possibility for agentic action that can transform those structures in unpredictable ways. A good RPG has exactly the same features! Players in an eventful D&D game need some structures. The world should feel resilient – there should be pressures to do things a certain way, act a certain way, adventure in a certain way. And yet, players should also be able to make choices that can radically upset those rules. But those events should be “relatively rare.” Every encounter need not be an event, in Sewell’s terms. The rules of the world should stay relatively fixed for stretches of time, to let the players understand and experience them, to become immersed. And then, when events occur (perhaps from the player’s own actions, following according to the structures in place), the players’ own choices determine in broad strokes where the campaign (and the world) go next.
In concrete terms, imagine a campaign where the players are servants of a monarch who sends them on various missions. Fetch the MacGuffin. Slay the Trogdor terrorizing the peasants. And so on. The world has structures in place: a monarchy, territorial boundaries, good and bad creatures and so on. But the players’ actions may lead to a crisis. Perhaps fetching the MacGuffin opens a rift with the neighboring kingdom and leads to a war. Perhaps the players slay the Trogdor, only to realize that it was created by the evil actions of the monarch, and the players were sent to cover up the experiment gone awry. The players then have meaningful choices to make: which kingdom do they back? Do they expose or confront the monarch? And from there, the basic structures of the world may change, in more or less radical ways.
An eventful campaign is one with enough structure to make it possible for transformational events to occur, and thus for the actions of the players to meaningfully alter the world. An eventful campaign is one in which the players experience agency. And, for many players, that’s the most fun part.
In historical sociology, the stakes are a bit different. The goal is not fun, but rather better historical understanding and explanation. But the route to get there is delightfully similar: an understanding of historical change that explores both enduring structures and contingent, agentic, transformational events.