Beyond micro and macro to contingent, heterogeneous networks

Discourse analysis is typically seen as “microanalytic,” focused on small pieces of social life. This type of work is often contrasted with “macroanalytic” approaches that focus on more spatially and temporally extensive processes and organization. Most forms of discourse analysis presuppose a “macro-micro dialectic,” in which purportedly homogeneous “macro” processes constrain events and actions but are also simultaneously constituted by “micro” events and actions. Once we move beyond the speech event to study pathways across events, however, it becomes clear that a macro-micro account does not suffice. Under scrutiny, neither “macro” nor “micro” represents a coherent level of explanation, and combining them does not solve the problem (Wortham, 2012). Instead of imagining two allegedly homogeneous scales, we must ask which of many potentially relevant processes and resources—from interactional through local and global, from instantaneous through those emergent over months, years, or centuries— facilitate and constrain focal actions and how these heterogeneous processes interrelate in a given case. Wortham, Nichols, Clonan-Roy, and Rhodes (2020) develop this argument further.

By criticizing a “macro-micro” account, we do not mean to dismiss the insights contained in such an approach. Those who study “macro” processes have shown that constraint plays an important role in meaningful human experience. As we have argued earlier, following many others (e.g., Bourdieu, 1972; Silverstein, 1992), any analysis of signs within a discrete event presupposes standards of value and cultural models from beyond the “micro” speech event. Participants in discursive interaction themselves presuppose more widely circulating models as they construe discourse, and analysts must do the same. An analysis limited to the micro cannot even make sense of micro-level events themselves. Individual actions, interactional patterns, and the social processes accomplished across pathways of events are all constrained by processes at wider temporal and spatial scales.

Despite this important point, however, analysts interested in the “macro” too often explain their core insight about constraint with reference to allegedly homogeneous “structure.” As the term is typically used, structure in fact includes very heterogeneous processes: practices of capitalist exchange that have emerged over millennia, practices of European colonization that occurred over several centuries in the past millennium, the global movement of people and ideas that has accelerated over recent decades, the emergence of new styles that emerge and disappear in a few years, and the adoption of a new approach over months or years in an organization or a family hoping to change. Each of these constrains, but these heterogeneous practices and models are not one kind of “macro” thing that constrains in the same way. Any process that takes place at a wider scale can constrain processes at narrower scales (Lemke, 2000). A novel presupposition established over a few seconds of interactional time can constrain the action of subsequent speakers even though transient interactional accomplishments do not normally qualify as “structure.” Constraints can also emerge from local, more evanescent patterns that mediate and sometimes undermine more enduring regularities.

“Macro” and “structure,” then, are often misleading terms. Constraint is crucial to explaining discursive and other human processes, but it is not a homogeneous entity grounded in one “macro” level of explanation. Many types of material objects, ideas, and practices constrain in various ways. Constraints can support or undermine each other, and various constraints always operate in any given case, something Blommaert (2018) calls “polycentricity.” Sometimes, a constraint is institutionalized in enduring practices, but in other cases institutions play no important role in effective constraint. Sometimes, constraints are established within a small group and remain irrelevant to others, but they can nonetheless powerfully limit action within the group. Sometimes, constraints are ephemeral, salient for a while but then disappearing quickly. Instead of assuming that an analyst’s favorite type of “structure” plays a crucial role in constraining thought and action, we must investigate the heterogeneous constraints actually influencing our object of study in specific cases.

“Micro” has similar strengths and weaknesses as an analytic category. Discourse analysts’ emphasis on contingency, on the unexpected patterns and actions that can emerge despite sociocultural expectations, captures a crucial aspect of social life. Habits change, and new ways of understanding emerge, sometimes in far-reaching ways. To explain change, we must account for the emergence of unexpected practices, models, and behaviors. As long as this insight about contingency and emergence does not imply that more enduring patterns do not constrain, it is crucial. But if we construe emergence as necessarily arising from the “agency” of individual actors or from “micro” interactional improvisation, we make a mistake. Contingent patterns emerge at various scales, sometimes at scales more extensive than individual actions or discrete events.

The concept of “agency” carries misleading assumptions about autonomous rational minds and about the individual as the relevant level of analysis for meaningful phenomena. Using this sort of “agency” to explain unexpected, emergent patterns is incomplete and often unproductive. Individuals do sometimes develop novel plans and execute unexpected actions, and these sometimes influence larger social patterns, although this is always mediated through collective artifacts. Contingent emergence, however, is more often accomplished by groups. Sometimes, this happens interactionally, when participants create an unexpected pattern through improvisation (Sawyer, 2003; Schegloff, 2007). It can also happen over longer scales, such as when a group engages in what Shotter (1993) calls “joint action” that cannot be reduced to the contributions of discrete individuals. A couple can create a new way of relating, across months or years, and its emergence is not typically reducible either to the actions of an individual or to a pivotal event. A system that includes both humans and tools can develop new practices, such that we must describe the whole system of individual dispositions and actions, along with nonhuman contributions, in order to explain change (Latour, 2005; Michel & Wortham, 2009).

Most emergent patterns cannot be reduced to individual actions in discrete events. Sometimes, seminal actions by individuals do make a crucial difference, but most change does not happen in this way. More often, change emerges across events, through co-constructed and sometimes unintended contributions. In explaining contingent emergence in a given case, we must determine the levels of analysis relevant to the focal phenomenon. If we assume that either individual agency or interactional creativity is the only way to explain emergence, we will misinterpret cases that centrally involve resources and processes from other scales.

Discourse analyses that rely solely on “macro” or “micro,” “structure” or “agency” will thus be inadequate. A simple combination of the two will not suffice either. Many have proposed a “macro-micro dialectic,” moving back and forth between the constraining force of the macro and the creative potential of the micro. On such an account, enduring practices and models constrain events, while action in particular events reproduces or helps to transform those sociohistorical patterns. Giddens (1976) articulated an early version of this argument, calling it “structuration,” but many others adopt a similar position. We agree that both constraint and emergence are crucial, but we disagree that “macro” and “micro” are the right categories for analyzing them. Constraint and emergence occur at various scales. “Macro-micro” accounts misconstrue heterogeneous resources from various scales, forcing them into two artificial scales instead of exploring how diverse resources become relevant in any given case. Instead of one monolithic type of structure, various types of constraining processes occur at various scales. And innovation occurs not only through individual intentions actualized in discrete events but also from collective processes at various scales (Wortham, 2012).

Latour (2005) provides a useful theory of how people mobilize resources from various scales. He argues that the social world is constructed out of heterogeneous “assemblages.” For any focal phenomenon, a “network” has been constructed, and this network is heterogeneous in both scale and type. Ideas, objects, and dispositions from different scales come together to do the relevant work (Lemke, 2000). In Maurice’s case, for example, the process of social identification depended on resources from various scales: the decades-long evolution of Americans’ expectations that girls will be more successful in school; the local intensification of this gender stereotype in Mrs. Bailey’s classroom; the centu-ries-long exclusion of African American males from mainstream American institutions; whatever ontogenetic trajectory led Maurice to desire both academic success and high status as a male; and so on. A network is also heterogeneous in type. Material objects matter, like the spatial organization of the classroom, with certain students sitting nearer the teachers. Ideas matter, like the models of personhood circulated in the media that represent African American males in certain ways. And embodied dispositions matter, like the tension between Maurice and Mr. Smith that was palpable in the classroom.

Latour helps us see how analyses of typical events or simple macro-level regularities obscure the heterogeneous cross-event processes that actually constitute social life. Instead of relying on sociological essentializations, we must study the complex emergence of identities, groups, registers, habits, and models across events at various scales. Different research questions and different focal phenomena will require attention to different configurations of resources. In order to explain complex processes like learning, social identification, and enregisterment, analysts must attend to a configuration of heterogeneous, interconnected resources across several scales instead of attending to one or two scales selected a priori. Participants’ sense of the unproblematic, “natural” establishment of social facts in any given case depends on a contingent configuration of resources from various scales.

Discourse analysis beyond the speech event is a method that helps analysts overcome the tendency to collapse analyses into “macro” and “micro.” Our approach helps analysts look beyond one or two scales and uncover intermediatescale regularities that take shape across pathways of linked events. In this book, we have shown how discourse analysis beyond the speech event allows us to capture the heterogeneity of relevant resources and study the contingent emergence of social actions that draw on resources from various scales. In practice, participants mobilize heterogeneous resources and establish links across events in ways that accomplish complex social actions and outcomes. Our approach to discourse analysis beyond the speech event provides one important methodological window into these processes.