Within and Beyond the “Fourth Generation” of Revolutionary Theory

Processual Ontology

For Abbott (1988:170), “general linear reality” assumes that “the social world consists of fixed entities (the units of analysis) that have attributes (the variables).” In this understanding, the interaction of attributes leads to stable patterns—patterns that persist regardless of context. Although they claim to be rejecting such a wager, fourth-generation theorists of revolution are often wedded to this notion of revolutions as “collections of properties.” Indeed, debate within current scholarship tends to center around which properties are essential or contingent to particular revolutions or clusters of revolutions. Goldstone (2009), for example, highlights 12 components of “color revolutions,” which he traces from the revolution of the United Provinces against Spanish rule in the 16th century to present-day instances in Ukraine and elsewhere. In his most recent work, Goldstone (2014:16–19) lists the “necessary and sufficient” conditions that induce an “unstable equilibrium” that in turn foster revolutionary situations: fiscal strain, elite alienation, popular anger, “shared narratives of resistance,” and “favorable international relations” such as the withdrawal of support for a client regime by its patron. These conditions are generated by a range of causes, from shifting demographic patterns to new patterns of exclusion, which foster social instability and thus act as the “fundamental causes of revolutions” (Goldstone 2014:21–25). Such analysis, like other fourth-generation approaches, contains the ontological assumption that revolutions consist of certain attributes that can be taxonomized (as in Goldstone’s “suite” of factors that comprise “color revolutions”) or combined (as in Foran’s Boolean analysis of Third World revolutions), albeit with due regard, at least in theory, to the complexity of the social world and to variation within revolutionary experiences.

In contrast to this identification of core revolutionary attributes, this article sees revolutions as historically specific processes. In a strict sense, the diversity of revolutionary instances (as noted by Goldstone in the quote at the beginning of this section) dictates that all explanations are “case-specific”—revolutions are particular assemblages that combine in historically discrete ways. Because the specific processes within which these assemblages cohere is singular, and therefore historically unrepeatable, the timing of revolutionary events is crucial. For example, reforms by a state within a revolutionary situation may succeed or fail depending on when they take place. If reforms take place sufficiently early, they may decompress revolutionary mobilization (as in Morocco and other monarchies in North Africa and the Middle East in 2011); too late, and they are likely to fail (as in Tunisia and Egypt the same year). ⁴ No single attribute can thus be associated, measured, or coded in relation to reform attempts by a state during a revolutionary situation. In a similar vein, when a contentious movement appears is just as important as how it is organized. For example, there may be few differences between the organizational capacity of the Syrian opposition that has fought Bashar al-Assad since the 2011 uprising and the movement that toppled Zine Ben Ali in Tunisia in January 2011. If anything, the former has shown a greater capacity to mobilize and sustain its struggle. The latter was successful not because of a set of fixed, timeless attributes but because it was the first such struggle in the region. Revolutionary waves often become less successful the further away they travel from their original point of instigation (Beissinger 2007; Patel, Bunce, and Wolchik 2011). This is the case for three reasons: First, revolutionaries in states outside the original onset of the crisis often overstate the possibility of success, placing too much weight on dramatic news from elsewhere and drawing firm conclusions from relatively sparse information (Weyland 2012:); second, revolutionaries enact their protests in increasingly inhospitable settings—regimes learn quickly, including how to demobilize their challengers (Della Porta and Tarrow 2012); and third, authoritarian state-society relations do not disappear overnight (Lawson 2005; Way 2011). Such studies indicate that revolutionary scholarship should be concerned less with the fact of emergence than with the timing of emergence.

Fourth-generation approaches to revolution often claim to recognize that revolutions are not static containers composed of fixed attributes. But they do not always sustain this wager in their empirical analysis. The difficulty, as Motyl (1999:23) points out, is that revolutions are not “tangible” objects that can be “touched”:

Revolutions do not exist as materially tangible 3-D objects, in the sense that we say that rocks and trees and airplanes exist as physical things. We can throw, touch or board the latter, we can use all or some of our senses to comprehend their physical reality, but we cannot do the same for revolutions. We cannot, like homicide investigators, draw a chalk line around a revolution, nor can we place it in an infinitely expandable bag. We cannot touch it, taste it, or for that matter even see it. Naturally many eyewitnesses to revolution claim to have seen it, but in reality what they saw were events and processes and people and things that, together, are called revolutions.

Such an understanding reconceptualizes revolutions as “webs of interactions” whose effects change according to when and where they are instantiated (McAdam, Tarrow, and Tilly 2001:13; Tilly 2004:9). Revolutions take place because of particular constellations of events, “not because of a few fundamental effects acting independently” (Abbott 1992:68). To this extent, it makes little sense to ask: “Is x a revolution?” Such an exercise entails the comparison of a processual configuration against an inert checklist of characteristics. But revolutions have no ascribed properties. Nor do they contain fixed attributes. To the contrary, revolutions are sequences of events that attain their significance as they are threaded together in and through time. To put this in Abbott’s terms (1988:179), revolutions are “closely related bundles” whose meaning arises from the order and sequence within which their events are knitted together. ⁵

If revolutions are assemblages that can be understood only through retrieving their temporally specific configurations, perhaps they are best examined on a case-by-case basis? Such research is certainly valuable in terms of its sensitivity to the multitude of interactions that constitute revolutionary processes. However, this mode of analysis is also guilty of generating what Little (1995:52) calls “combinatorial explosion.” Because there are always contingencies and interactions that go unobserved, there can be no “total explanation” of revolutionary processes, however micro level the analysis. If all historical events are overdetermined, in that there are more causes than outcomes (Adams 2005), then all analysis of revolutions underdetermines the “true causal story” by necessity (Little 1995:53; also see Flyvbjerg 2001). Indeed, all theoretical work is an act of foregrounding suppression that simplifies history and constructs the social world into wagers about “why this and not that.” Theories denote what is significant and insignificant about a cluster of historical events. Attributional accounts carry out this task by testing the weight of causal factors that are taken to be significant. Yet such a wager cannot eliminate the effects of the causal factors that lie outside the scope of a particular theory—it simply represses them. In this sense, there can never be theoretical “closure,” particularly given that attributional accounts are especially unsuited to examining the interdependence of “significant” and “insignificant” causal processes (Adams 2005:11–12).

The implication of seeing revolutions as temporally specific assemblages requires a form of analysis in which the researcher amplifies the clusters of events that form revolutions, providing a “rational reconstruction” of how revolutions begin, endure, and end (Jackson 2011:38–39). This task is helped by the fact that even during periods of radical uncertainty like revolutions, social action is not random. Rather, social action is embedded within fields of action that constrain behavior and give meaning to these actions (Fligstein and McAdam 2012; Flyvbjerg 2001). In “normal times,” social orders are relatively stable—they are constituted by fields of action that are patterned in relatively sticky, predictable ways. At the same time, many fields of action—such as gender and class relations—are so deeply embedded that they are resilient to attempts at radical transformation. However, this does not make such fields of action static—there is always a processual dimension to the ways they are produced and reproduced. In “abnormal times,” such as revolutions, these processual dynamics have the capacity to reshape fields of action and categories of meaning quite dramatically (Goldstone and Useem 2012). In this way, revolutions can be understood as attempts to break existing fields of action and embed new webs of interaction. This twin process of displacement-replacement occurs in several fields of action—economic, political, and symbolic—simultaneously.

As the earlier quote from Motyl (1999:23) makes clear, it is not possible to “draw a chalk line around a revolution,” but it is possible to speak of revolutions as “events and processes and people and things that, together, are called revolutions.” Such an understanding means arresting the desire to map revolutions in their entirety in favor of discerning the logical shapes within which revolutions cohere. In this sense, it is helpful to see revolutions as traffic jams rather than solar eclipses (Tilly 1993). Whereas the latter are the result of regular celestial motion that follows a precise schedule under stable conditions, the former vary in form and severity and develop for a number of reasons. This does not mean there are no regularities to traffic jams: They are linked to rush hours, bad weather, roadworks, traffic light sequencing, breakdowns, accidents, and so on (Tilly 1993). One cannot predict traffic jams in the same manner as predicting solar eclipses (e.g., bad weather may or may not lead to a traffic jam), but the combination in which these factors arise yields recurrent patterns. Like traffic jams, revolutions are, at least in part, stable accumulations of interactions. They contain situational logics that emerge as events and experiences cohere to form meaningful fields of action. These fields of action are exposed through the construction of “analytical narratives” that filter revolutionary events into idealized causal pathways. Analytical narratives are “structured stories about coherent sequences of motivated actions” (Aminzade 1992:457–58). They are interpretative to the extent that they identify connections that are taken to be meaningful (Reed 2011). They are also tools of simplification in that they emphasize certain sequences of events and downplay others. But analytical narratives are also systematically constructed and logically coherent, providing a means of differentiation between significant and accidental causal configurations and producing useful insights into concrete instances of revolutionary change (Jackson 2010). Such analysis realizes the hitherto untapped processual impulse of fourth-generation revolutionary theory.

The realization of the processual ontology favored but not actualized by fourth-generation revolutionary theory leads in turn to a configurational account of causation. ⁶ As particular bundles of events, both the sequence within which revolutions take place and the context within which they occur are significant—causal regularities emerge contextually, constituting configurations that are robust but situational. Although causal configurations are contextually located, they constitute relatively stable sites for examining the emergence, durability, and outcomes of revolutions. As Sewell (1996) notes, all revolutionary events are part of broader chains of events. These chains of events have cascading effects in that they both break and reproduce existing social formations—they are “sequences of occurrences that result in the transformation of structures” (Sewell 2005:227; see also Mahoney and Schensul 2006). ⁷ Because they transform fields of action, events are theorizable categories. Sewell uses the example of the fall of the Bastille to illustrate this point. The storming of the Bastille on July 14, 1789, was imbued with significance “beyond itself”: The event was recognized within broader political and symbolic fields, breaking existing configurations and reconstructing categories of meaning, including notions of “people” and “revolution.” This specific event had cascading effects: It challenged existing symbolic repertoires and helped reformulate categories of meaning—it was a rupture that assumed “authoritative sanction” (Sewell 2005:257). Processes like the fall of the Bastille in particular, and revolutions in general, illuminate how moments of temporal heterogeneity morph into common fields of “ruptural unity” as social facts are disrupted and transformed (Steinmetz 2011:59; see also Tarrow 2012). Configurational causal accounts permit researchers to assess the ways in which historical events enable social formations to break down and reemerge.

Events like the fall of the Bastille, therefore, have outcomes that can be traced through the ways in which sequences of “happenings” are casually conjoined. This process of “eventing” sees historical events as assuming relatively stable shape through the interactions between “happenings” and the fields of action within which they are nested (Jackson 2006). Such a move is never complete—alternative readings are always available. But if all theoretical work requires the simplification of historical “mess” into plausible causal stories (Tilly 2006), then analytical narratives of revolution are no exception to this: They are tools by which to assemble historical clutter into significant “plots.” These plots assess the transformation of patterned social relations in and through time. The result is a sense of “followability” to dynamics of revolution: a “narrative intelligibility” in which events are connected to accounts of sequence and order (Gallie 1964).

Such an understanding of revolutions begins to fulfill the promise of fourth-generation approaches to revolution. Current scholarship tends to be caught in a bind: It accepts the multiplicity of revolutionary episodes but retains an “attributional” ontology that requires revolutions to fulfill certain elemental conditions. However, identification of the context-specific interactions that constitute revolutionary processes generates a dual benefit: intimate knowledge of concrete revolutionary episodes and understanding of how revolutions are sedimented within wider fields of action (Fligstein and McAdam 2012; Flyvbjerg 2001). In this understanding of revolution, the tasks of the researcher are fourfold: first, to examine the particular sequences through which revolutions are “evented”; second, to assemble these sequences into plausible analytical narratives that are logically coherent and supported by the available evidence; third, to abstract the causal configurations through which revolutions displace-replace fields of action; and fourth, to assess how these causal configurations explain revolutions in diverse settings (McAdam et al. 2001; Tilly 2004). This mode of research tacks between empirical and theoretical registers while being sensitive to the temporally singular character of revolutions and the possibility of generating insights beyond specific revolutionary episodes. The first step in fulfilling the promise of fourth-generation revolutionary theory lies in the development of a processual ontology that in turn leads to a commitment to configurational causation.

Relational Social Action

The second step in fulfilling the promise of fourth-generation approaches is to move beyond the “analytical bifurcation” (Go 2013) that is often drawn, explicitly or implicitly, between structural preconditions and strategic action. Although fourth-generation approaches usually claim to be doing just this, there remains a sense in which culture, ideology, and leadership are grafted onto structural preconditions to generate a “complete explanation” (McAdam et al. 2001:196; Sharman 2002:1). For example, Parsa’s (2000:12, 25, 279) “synthetic” account of the Iranian revolution focuses on the structural vulnerability of the Shah’s regime and how a “hyperactive state” politicized market interactions. At the same time, the dependency of the Iranian state on foreign backers, most notably the United States, along with elite fracture as the Shah’s patronage system weakened, set the stage within which various groups, from clerics to bazaaris, acted (Parsa 2000:7, 21). In Parsa’s account, state vulnerability provided the structural precondition for the emergence of a revolutionary situation. Once this precondition was established, additional variables, ranging from the formation of opposition coalitions to the mobilization of collective sentiment, explained the timing of the revolution (Parsa 2000:7, 25). Such a dichotomy—empirically present, if theoretically disavowed—is a regular feature of fourth-generation accounts.

This tendency toward analytical bifurcation is problematic in that it reinforces two, equally unsatisfactory, myths: Agent-centric theory builds on the myth of the person as a preexisting entity, and structural accounts build on the myth of society as a preexisting entity. To put this another way, whereas a focus on structure tends to reify relatively fixed patterns of social relations as “things with essences,” an emphasis on agency imagines a preexisting, asocial individual whose motivations, interests, and preferences come prepackaged without recourse to broader fields of action. Both positions are unsatisfactory. Indeed, both rest on an assumption that their basic units of analysis are static, whether this assumes the form of an interstate system, a class, or a volitional subject. Revolutions, however, are not static containers that contain essential traits. As a result, analysis of revolutions cannot assume the stability of a set of universal factors that are easily transplanted to diverse settings. Rather, the ways in which relations between social sites constitute revolutionary dynamics must be given analytic priority. All social structures are relatively fixed configurations of social action, just as all social action takes place within relatively fixed configurations of social ties. There is no nonstructured action free from broader ties, connections, patterns, and interrelations (Emirbayer and Goodwin 1996; Fligstein and McAdam 2012). Social life takes place in “structured contexts of action” (Emirbayer and Goodwin 1996:372)—that is, fields of practice formed by configurations of events and experiences. Extending fourth-generation accounts requires moving beyond binaries of structure and agency and toward a relational approach that conceptualizes social action as taking place within these broader configurations. Revolutions are formed through the constitutive interaction of “entities-in-motion.”

This point is made clearer by differentiating between entities and entities-in-motion (Go and Lawson forthcoming). Entities are the subject of “substantialist” approaches, which see the basic units of enquiry as fixed substances, whether these substances are things (e.g., revolutions), people (as in expected utility models), or systems (as in world systems analysis, which parses a single global structure into core, peripheral, and semi-peripheral polities, each of which is defined through a set of essential attributes). In substantialist thought, entities contain a finite set of core attributes, as in Skocpol’s (1979:4) understanding of revolutions as the “rapid, basic transformations of a society’s state and class structures, accompanied and in part carried through by class based revolts from below.” Skocpol’s definition posits a set of properties that revolutions must contain. Empirical study takes place within this definitional ambit, examining whether events conform with or challenge these predetermined characteristics. Supplementary work in this idiom either refines Skocpol’s definition (e.g., Goldstone’s [2001:403] differentiation between “political” and “social” revolutions) or generates causal claims that flow from the definitional starting point (e.g., Foran’s [2005] analysis of the five essential requirements of Third World revolutions). The study of revolutions tends to see its object of analysis as durable entities possessing essential properties.

The most important problem associated with substantialist thinking is positing revolutions as entities that assume a static, unchanging form rather than as entities-in-motion that are made in and through time. In this way, substantialist thinking elides the eventfulness of revolutions, fostering a fixed idea of revolutions that weakens analysts’ capacity to capture their changing, configurational quality. The hold of Skocpol’s definition, for example, continues to funnel the study of revolution toward a particular view of revolutions associated with the notion of “total change.” Yet the universe of cases that conforms to such an understanding is, at most, 10: England, Haiti, France, Mexico, Russia, China, Cuba, Vietnam, Nicaragua, and Iran. Even in these cases, it is questionable whether Skocpol’s definitional edict can be sustained. If revolutions must be “rapid,” it is difficult to see how China’s three decade–long struggle conforms. If revolutions must transform “state structures,” then France does not qualify—republicanism was a relatively short-lived experiment, eclipsed first by Napoleonic empire and then by the restoration of the monarchy. If revolutions must be “class based revolts from below” and transform “class structures,” then few if any revolutions meet this standard. Revolutions are cross-class coalitions bound up in complex dynamics of continuity and change (Dix 1984; Goodwin 2001).

The salient point is not that Skocpol’s definition is particularly difficult to square with diverse revolutionary experiences. All theories are tools of simplification—their utility arises not from their capacity to explain everything but from their capacity to generate useful insights into particular domains of social life. The problem is that any definition rooted in the attempt to ascribe revolutions with a set of essential characteristics must by necessity freeze history. Such an exercise not only occludes the empirical subject matter it claims to explain, it also fails to capture the sense in which revolutions are entities-in-motion. The result is the flattening of the social world into a static sphere of preexisting social formations. Requiring that revolutions fulfill a set of inalienable characteristics distorts understanding of how revolutions change according to historically produced circumstances. For example, many post–Cold War revolutions have distinct trajectories in terms of their rejection of armed confrontation, embrace of nonviolent repertories, and fostering of despotically weak states (Lawson 2004, 2005, 2015a). Studies that stay within a substantialist framework cannot easily capture such a shift. Rather, a substantialist baseline will see only conforming or nonconforming parts of a preexisting script. More fruitful, I argue, is adoption of a “relational” stance that examines the contextually bound, historically situated configurations of events and experiences (Abbott 1988, Emirbayer 1997; Go and Lawson forthcoming).

In contrast to substantialism, a relational approach holds that entities “are not assumed as independent existences anterior to any relation, but . . . in and with the relations which are predicated of them” (Cassirer 1953:36; see also Emirbayer 1997:287). Rather than presuming that there is an abstract essence to revolutions or an essential set of properties that revolutions must contain, a relational approach gives analytic priority to the historically located events and sequences through which causal sequences within revolutions emerge. Unlike substantialist accounts, a relational approach does not ascribe necessary conditions under which revolutions will arise (e.g., Goldstone 2014). ⁸ Nor does it seek to generate covering laws within which specific episodes of revolutions can be tested and coded (e.g., Foran 2005). Rather, a relational approach seeks to dissolve the binaries that limit effective analysis of the changing form that revolutions assume over time and place. The difference is akin to taking a photograph or shooting a film. Substantialist approaches attempt the former, holding certain conditions constant by taking a snapshot of a particular moment in time, then testing the generalizability of this snapshot to other instances of the phenomena in question. Relational approaches favor the latter, seeing social reality as a moving spectacle that requires analytics to be adjusted to changing conditions. In this latter understanding, the character of social objects cannot be assumed as if the subject of inquiry lay elsewhere. Rather, the particular forms that entities-in-motion assume is the subject of inquiry.

Relational approaches examine the ways in which historical events are generative of how social formations emerge, reproduce, transform, and potentially, break down. As noted in the previous section, in revolutions, as in other domains of social life, social action is connected through “webs of interactions.” These webs of interaction produce relatively fixed patterns of enduring interactions. Although such patterns are open to contestation, they constitute stable sites for the development of empirical enquiry. Just as sequences of revolutionary events can be logically connected through a processual ontology, so too can social action be usefully examined as it accumulates in particular assemblages (Go 2013; Latour 2005). By focusing on instances of change—what Mann (1986) calls “neo-episodic moments”—it is possible to assess how patterns of social relations are disrupted and transformed. Sometimes, this transformation is overtly coercive: in France, more than one million people died in the revolution and the wars that followed; in Cambodia, nearly a third of the population died in violence following the seizure of power by the Khmer Rouge (Goldstone 2014). At other times, this transformation deepens the state’s infrastructural power. The French revolutionary regime transformed provinces into webs of départements, districts, cantons, and communes. It also used symbols, images, and rituals such as festivals to socialize populations into the revolutionary ideal (Ozouf 1991). Such socialization extends to spheres as apparently humdrum as holidays. In Cuba, for instance, the figure of Don Feliciano came to replace the Christmas tree and Santa Claus (Paige 2003).

The contrast between attributional and relational approaches is stark. The former sees the purpose of theorizing about revolutions as (1) the identification of attributes that are necessary and/or sufficient to revolutions and (2) a comparison of these attributes to a range of apparently distinct cases. In other words, attributional approaches study the cross-case variation between a number of apparently independent causal factors. Relational approaches, in contrast, examine the bundles of patterns, sequences, and assemblages that constitute revolutionary episodes. The focus is on interdependent rather than independent causal dynamics, such as those that connect the Haitian and French revolutions with insurgencies in Asia, Africa, and the Arabian Gulf (Anderson 2013). Although the social objects created by such processes are necessarily entities-in-motion, and despite the diversity of revolutionary episodes, comparable mechanisms can be observed in discrete historical cases: for example, the polarization of adversaries into opposing factions; the role played by brokers in unifying disparate opposition groups; and the decertification of the regime by key elites. ⁹ In identifying these mechanisms, the question is not whether a certain condition enables a particular effect but how an effect comes to be possible through a particular assemblage of events and experiences (Hedström and Bearman 2009). Revolutions may not have uniform structures, but they do have shared forms (Thomassen 2012).

An Intersocietal Approach

The third way in which the promise of fourth-generation approaches remains unrealized is in their failure to generate a fully fleshed out intersocietal approach. ¹⁰ The term intersocietal is not intended to mean that objects of analysis must be “societies.” Rather, it is concerned with examining the relationship between “external” and “internal” dynamics wherever these are found: for instance, in ideas that cross borders, among networks of revolutionary actors, or in asymmetrical market interactions. An intersocietal approach is concerned with the ways differentially located but interactively engaged social sites affect the development of revolutions without containing a prior presumption of what these social sites are.

Third- and fourth-generation theorists often claim to have sufficiently incorporated the international aspects of revolutions into their analyses. In response to the relative neglect of international factors by first- and second-generation work, beginning in the 1970s, third-generation theorists (e.g., Goldfrank 1979; Goldstone 1991; Katz 1997; Skocpol 1979; Tilly 1990) included a range of international factors in their accounts. Goldfrank (1979:143, 148–51) argues that the roots of revolutions lay in the “world capitalist system” and its “intensive international flows of commodities, investments, and laborers,” “great power configurations” (e.g., a shift in the balance of power), a “favorable world situation” (e.g., changing client-patron relations), and a “general world context” (e.g., a world war, which preoccupies great powers). Skocpol (1979:14) famously argues that “social revolutions cannot be explained without systematic reference to international structures and world historical development.” She highlights the formative role played by two international factors in the onset of revolutions: the uneven spread of capitalism and interstate (particularly military) competition (Skocpol 1973, 1979). Both of these factors are embedded within “world historical time,” by which Skocpol (1979:23) means the overarching context within which inte-state competition and capitalist development take place. Tilly (1990:186) also highlights the importance of interstate competition, arguing that “[a]ll of Europe’s great revolutions, and many of its lesser ones, began with the strains imposed by war.” ¹¹ Goldstone (1991) widens this focus by noting how rising populations across a range of territories foster state fiscal crises (by increasing prices and decreasing tax revenues), heighten elite fracture (as competition between patronage networks sharpens), and prompt popular uprisings (as wages decline in real terms). Finally, Katz (1997:13, 29) notes how “central revolutions” (e.g., France in 1789) foster “waves” of “affiliated revolutions” (see also Beck 2011; Markoff 1996; Sohrabi 2002).

The retrieval of the international by third-generation revolutionary theorists has been extended by a number of fourth-generation theorists (e.g., Beck 2011, 2014; Foran 2005; Goldstone 2001, 2009, 2014, 2015; Kurzman, 2008; Ritter 2015). Goldstone (2014:19, 21–22) highlights a variety of ways through which “favorable international relations” serve as the conditions for societal instability, and he lists a range of factors, from demographic changes (e.g., rising populations) to shifting interstate relations (e.g., the withdrawal of external support for a client regime), by which international processes help cause revolutions. ¹² As discussed in the following, of Foran’s (2005) five “indispensable conditions” that have enabled revolutions in the Third World to take place, two—dependent development and world-systemic opening—are overtly international. Kurzman (2008:8) notes how a global wave of democratic revolutions in the early part of the 20th century spread over widely dispersed territories, from Mexico to China, and he argues that this wave acted as a “dress rehearsal” for later events, most notably the 1989 revolutions in Central and Eastern Europe. Beck (2011:193) sees such waves as likely to increase “as the level of world culture more rapidly expands,” an argument that finds support in Beissinger’s (2014) database of revolutionary episodes, which shows a marked increase in the depth and breadth of revolutionary waves over the past century. Ritter (2015:5) emphasizes how an international context characterized by the “iron cage of liberalism” traps authoritarian states into accepting at least the rudiments of democratic practices. If authoritarian regimes are to maintain the benefits of ties with Western states, from arms to aid, then they must open up a space for nonviolent opposition to emerge—the structural context of international liberalism provides an opening within which domestic nonviolent opposition can mobilize. ¹³

Given this proliferation of interest in the international components of revolutions, it could be argued that contemporary revolutionary scholarship has solved the “problem” of the international. Many contemporary works are replete with references to transnational empirical connections (e.g., revolutionary repertoires that cross borders), and international factors are often seen as the precipitant cause of revolutions (through relations of dependent development) and the direct outcomes of revolutions (through inducing interstate conflict). ¹⁴ These accounts have gone a considerable way toward opening up a productive exchange between revolutionary theory and “the international”; this article builds on the insights of Goldfrank, Skocpol, Goldstone, Foran, Kurzman, Beck, Ritter, and other pioneers. However, I also seek to extend the insights of this scholarship by demonstrating how “the international” has not yet been theorized “all the way down.” Three motivations lie behind this claim. First, despite increasing attention to the multiple connections between revolutions and the international, this relationship remains unevenly examined, being highly visible in some work (e.g., Beck 2014; Foran 2005; Goldstone 2014; Kurzman 2008; Ritter 2015) yet all but invisible in others (e.g., Goodwin 2001; Parsa 2000; Slater 2010; Thompson 2004). ¹⁵ Clearly, there is much still to do in terms of “mainstreaming” international factors into the analysis of revolutions. Second, use of the international is often reduced to a handful of factors. In Skocpol’s analysis, for example, interstate competition is a surrogate for military interactions, particularly defeat in war. Hence, “wars . . . are the midwives of revolutionary crises” (Skocpol 1979:286). Such a view neglects how a cornucopia of international processes, from transnational cultural repertoires to interstate alliance structures, affect the onset of revolutions. Third, much revolutionary scholarship grafts international factors onto existing theoretical scaffolding rather than integrates such factors within a single framework. This point is worth examining in more depth.

In Foran’s (2005) influential work, revolutions in the Third World are seen as emerging from the interaction of five “indispensable conditions”: dependent development; exclusionary, personalistic regimes; political cultures of opposition; economic downturns; and a world-systemic opening. Two of Foran’s five causal conditions are overtly international: dependent development and world-systemic opening. Yet these factors contain little by way of causal force. The first, dependent development, is a virtually universal condition of core-periphery relations—to paraphrase Skocpol’s (1979:34) comment on the ubiquity of “relative deprivation”: What “peripheral” society lacks widespread dependence of one sort or another on a metropole? Even given Foran’s (2005:19) specific rendering of dependent development as, following Cardoso and Faletto (1979), Evans (1979), and Roxborough (1989), a particular process of accumulation (“growth within limits”), the concept is wide enough to be applicable to every Third World state. This is borne out by Foran’s (2005) own analysis, in which dependent development appears as a near constant of both successful and unsuccessful revolutions. ¹⁶ In other words, the causal weight attributed to dependent development is nil: It serves as the background condition within which revolutions may or may not take place. In this sense, to posit relations between polities as dependent is less to assert a causal relationship than it is to describe the condition of every peripheral state around the world. Without further specificity as to the quality and quantity of dependent development, the term becomes little more than a backdrop.

At first glance, Foran’s (2005) second “international” category—world-systemic opening, by which he means a let-up of existing international conditions through interstate wars, depressions, and other such crises—appears to be more promising. Yet here too, the causal agency of the international is significantly curtailed: World-systemic opening is seen merely as the final moment through which the “revolutionary window opens and closes” (Foran 2005:252). In other words, the structural preconditions that lie behind revolutions are situated elsewhere—in domestic regime type, cultures of opposition, and socioeconomic conditions. World-systemic opening is the final curtain call on a play that has largely taken place elsewhere.

The two international components of Foran’s analysis are thus limited to walk-on roles: Dependent development is the background from which revolutions may or may not occur; world-systemic opening is the final spark of a crisis that has been kindled elsewhere. The sequence through which Foran’s multicausal analysis works is highly significant: international (dependent development), domestic (exclusionary, repressive regimes), domestic (cultures of opposition), domestic (economic downturns), international (world-systemic opening). The fact that Foran’s sequence differentiates international and domestic in this way reproduces the analytical bifurcation that his analysis—and fourth-generation theorists more generally—hoped to overcome. Such a bifurcation occludes the myriad ways in which Foran’s ostensibly domestic factors are deeply permeated by the international: Exclusionary regimes are part of broader clusters of ideologically affiliated states, alliance structures, and client-patron relations; cultures of opposition are local-transnational hybrids of repertoires and meaning systems; and socioeconomic conditions are heavily dependent on market forces that transcend state borders. Rather than integrate the international throughout his causal sequence, Foran maintains an empirical and theoretical bifurcation between domestic and international. And he loads the causal dice in favor of the former.

Foran’s deployment of the international is emblematic of fourth-generation revolutionary scholarship. For instance, Goldstone (2001:146), although clear that international factors contribute in multifaceted ways to both the causes and outcomes of revolutions, is equally clear about the division of labor that exists between these two registers:

Although the international environment can affect the risks of revolution in manifold ways, the precise impact of these effects, as well as the overall likelihood of revolution, is determined primarily [italics added] by the internal relationships among state authorities, various elites, and various popular groups.

In a similar vein, Goldstone’s (2014) recent work makes much of the ways international factors serve as important conditions for and causes of revolutions. Yet, with the exception of noting the propensity of revolutions to stoke interstate war, international factors largely drop out of Goldstone’s account of revolutionary processes and outcomes. In this way, even fourth-generation scholarship that claims to fully incorporate international factors into its analysis can be seen as containing two shortcomings: first, the maintenance of an analytical bifurcation between international and domestic registers; and second, retaining a residual role for the international. As a result, attempts to integrate international factors into the study of revolutions tend to fall into a condition of “add international and stir.” Grafting the international onto existing theoretical scaffolding retains—and sometimes strengthens, albeit unintentionally—the bifurcation between international and domestic. And this bifurcation contains an (often implicit) assumption that the former serves as the secondary dimensions of the latter’s primary causal agency. How might an approach that sought to more thoroughly integrate the international into the study of revolutions proceed?

An intersocietal approach to revolutions starts from a simple premise: Events that take place in one location are both affected by and affect events elsewhere. A number of transnational histories point to the ways revolutionary events contain an international dimension that supersedes the nation-state frame (e.g., Adelman 2008; Hunt 2010; Stone 2002). To take one example, the onset of the French Revolution cannot be understood without attention to the expansionist policies of the French state during the 17th and 18th centuries—between 1650 and 1780, France was at war in two out of every three years. This bellicosity, a product of pressures caused by developments in rival states as well as domestic factors, brought increased demands for taxation that, over time, engendered factionalism in the ancién regime (Stone 2002:259–60) and led to chronic state debt (Hazan 2014). The interactive dimensions of international relations also affected events during the revolutionary period. For example, in 1792, as the Jacobins were losing influence to the Girondins, leading Girondins pressed the state into international conflict. ¹⁷ As France’s foreign campaigns went increasingly badly, the Committee of Public Safety, a leading site of Jacobin authority, blamed the Girondins for betraying the revolution and committed France to a process of domestic radicalization: “the Terror” (see Hazan [2014:299–303] for a critique of the “ideological” use of this term). In this way, domestic political friction induced international conflict that in turn opened up space for heightened domestic polarization. The Jacobins identified the Girondins as “unrevolutionary” traitors, speculators, and hoarders and identified themselves as the guardians of the revolution, a process of “certification” that prompted a wave of popular militancy, most notably the levée en masse (Crépin 2013; Hazan 2014; McAdam et al. 2001:323–27; Stone 2002:194–208).

In addition to the dynamic roles played by intersocietal relations in fostering the revolutionary situation and revolutionary trajectories in France, intersocietal relations also played a fundamental role in the outcomes of the revolution. First, the revolutionary regime annexed Rhineland and Belgium and helped ferment republican revolution in several neighboring countries, including Holland, Switzerland, and Italy. Second, the revolution prompted unrest throughout Europe, including Ireland, where a rebellion against English rule led to a violent conflict and in 1800, the Acts of Union between the United Kingdom of Great Britain and Ireland. Third, the threat from France was met by extensive counterrevolution in neighboring states. In England, for example, habeas corpus was suspended in 1794, and legislation ranging from the Seditious Meetings Act to the Combination Acts was introduced to contain the spread of republicanism. Although the French did not generate an international revolutionary party, many states acted as if they had done just this, instituting domestic crackdowns to guard against Jacques-Pierre Brissot’s claim that “we [the French revolutionary regime] cannot be at peace until all Europe is in flames” (cited in Palmer 1954:11).

An intersocietal approach builds from this understanding of the generative role of flows between and across borders. Empirically, an intersocietal approach charts the ways relations between people, networks, and states drive revolutionary dynamics. The Haitian Revolution, for example, contained multifaceted intersocietal dimensions: its embedding within circuits of capitalist accumulation, slavery, and colonialism; its embroilment in interstate wars; and its impact on the development of uprisings in Latin America and beyond (Geggus 2010; Klooster 2009; Shilliam 2008). Highlighting these empirical connections, whether direct or indirect, realizes the descriptive advantages of an intersocietal approach. To date, development of a descriptive intersocietal approach has been most evident in transnational, global, and economic history (e.g., Armitage and Subrahmanyan 2010). However, the richness of this scholarship has not been matched by work that adequately explores the analytical advantages of an intersocietal approach. Analytically, an intersocietal approach is concerned with how the social logics of differentially located but interactively engaged social sites affect the causal pathways of revolutions. Such interrelations take many forms: the withdrawal of support from a patron, the pressures that emerge from the fusion of “advanced” technologies in “backward” sectors of the economy, the transmission of revolutionary ideas, the diffusion of contentious performances, the desire to emulate revolution and counterrevolution, and so on. In both descriptive and analytical forms, intersocietal interactions are less the product of revolutions than their drivers.

The promise of an intersocietal approach rests on its capacity to theorize what otherwise appears as empirical surplus: the social logics contained within the intersocietal dynamics that constitute revolutionary processes. The concatenations of events through which revolutions emerge are dynamically related to the ways in which social relations within territories interact with those beyond their borders. Intersocietal relations form an interactive crucible for each and every case of revolution, from the desire to “catch up” with more “advanced” states to the role of ideas in fermenting unrest across state borders. The “external whip” of international pressures, added to the uneven histories within which social orders develop, produce an intersocietal logic that has not as yet been effectively theorized in the study of revolutions. It is the task of an intersocietal approach to identify these dynamics and demonstrate their generative role in the formation of revolutionary processes. Although it can be difficult both analytically and descriptively to avoid using nation-state frames, there is no sociological rationale for maintaining the bifurcation between international and domestic. Revolutions are complex amalgams of transnational and local fields of action—they are “intersocietal” all the way down.