The Social Grammar(s) of Future-Making

Abstract

This article outlines a theoretical approach for studying the “grammatical” structure of processes of future-making. Building off theoretical insights about the indexicality and relationality of meaning, I show that the prevalent semantic view of future-making as a movement from imagination to action is unable to theorize the dialogical relationship between the projection of futures and the construction of corresponding trajectories. I propose an alternative that draws on the concepts of enacted sensemaking and situated action to conceptualize projection as a process of organizing indexical practices into relationally defined trajectories into the future. I then reappropriate Leifer’s analysis of how the relations between local actions constitute shared projective horizons in chess to show how this iterational fabric of interconnections defines a relational grammar that coordinates future-orientations. Finally, I use Hutchins’s analysis of distributed cognition to draw these insights together into an account of how the continuous dialogical interaction between projectivity and iteration enacts social grammars, through which images and practical trajectories of future-making are co-constructed.

The question of how “projected” (Mische 2009, 2014) or “imagined” (Beckert 2016) futures feed back into, direct, and coordinate social action has occupied an increasingly prominent place in the sociological imagination in recent years (Beckert and Suckert 2021). Perceptions of, attitudes to, or aspirations for the future have long been a topic of research in sociology and beyond (Beckert and Suckert 2021; Mische 2014:441; Suckert 2022), but this more recent sociological work on futures is distinctive for its focus on the “constitutive” role of explicit projections of the future in shaping social agency in the present (Mische 2014:440). Informed by pragmatist efforts to develop conceptions of agency that make conceptual room for social actors’ “projective capacity to imagine alternative possibilities” within “temporally embedded processes of social engagement” (Emirbayer and Mische 1998:963), this research has sought to overcome the “problematic downplaying of the deliberative component of future projections” (Mische 2014:441), which has left rationalist and practice-centered analyses of futurity in social life unable to account for how projected futures can become “a dynamic force undergirding social change” (Mische 2009:694). As a corrective, this work focuses on how projections of the future are articulated through a “process of imaginative experimentation with projected courses of action” (Mische 2009:696) in which actors visualize the causal relations and processes they imagine will determine the future and how their actions can influence outcomes and then “organize their activities based on this . . . representation” (Beckert 2016:9).

This analytic narrative resonates with prevalent modernist and Western intuitions about intentionality (Duranti 2015:11), the idea of the future as “open” and contingent on human imagination and action (Koselleck 2005; Luhmann 1976), and the growing interest in how (cultural) meaning shapes social processes and formations (Emirbayer and Goodwin 1994; Mohr 2020; White 2008). However, whereas the avowed goal of this body of work has been to provide a more balanced view of the respective weight of future projection and more routine or habitual (or iterational) temporal orientations in shaping processes of future-making, in practice, this often translates into a gradual, first rhetorical but increasingly also analytic and conceptual, decoupling of projection not only from the iterational dimension of agency/future-making but also from the process of future-making itself. For instance, Tavory and Eliasoph (2013:909) argue that projection needs to be understood as a “mode of future-coordination” that is not only analytically distinct but also operates autonomously from temporal orientations grounded in the iterational dimension of agency, which they refer to as “protentional” orientations. Mische (2009:696) draws a similar distinction, arguing that future-making involves a translation of the results of “imaginative experimentation with projected courses of action” into a reconfiguration of “received categories of thought and action.” Tavory and Eliasoph (2013:922) argue that future-making requires an alignment between projection and protention, but they also (at least, rhetorically) assign performative priority to projects for the future that need to be implemented or “‘done’ at the level of protention.”

In other words, scholars increasingly envision the articulation of imagined futures as temporally or logically separated from their (subsequent) translation into corresponding actions or trajectories. In this semantic view of future-making, the “cognitive contours” (Mische 2009:697) or semantic features of projected futures define what constitutes appropriate actions and how to construct trajectories that correspond to these projections. Mische (2009:702), in this sense, speaks of a “movement from cognition to action,” and both Mische and Beckert point out how projections of the future provide actors with (more or less elaborate) “stories” that function as “maps for action” (Mische 2014:444), laying out specific trajectories for “how the present will be transformed into the depicted future through causally linked steps” (Beckert 2016:69). Even if most theoretical analyses of the social process of future-making carefully embed this semantics-centered account within a more sophisticated processual and relational terminology, recent empirical work displays a pronounced tendency to focus on the articulation of imagined futures and their semantic features (Beckert and Suckert 2021) rather than the more messy process of constructing practical congruence between such projections and the “performative work” (Mische 2014:440) of aligning multiple actors’ actions into a practically viable trajectory.

I argue that this “pervasive tendency to problematize futurity in epistemological terms” (Tellmann 2020:346) is due to a lack of a sufficiently fine-grained analytic framework and conceptual tools for adequately theorizing, describing, and analyzing how the temporal orientations to (longer-term) projects and the protentional orientation to immediately unfolding sequences of action are simultaneously co-constructed within the process of future-making. I take inspiration from Emirbayer and Maynard’s (2011) argument that pragmatist theoretical vocabularies frequently remain at a level of theoretical generality that makes it difficult to fold them into concrete analysis at the level of social practice, leading many scholars into a problematic “eclecticism” that mixes together relational and substantialist assumptions and concepts (Emirbayer 1997). Much as they suggest that pragmatist sociology could benefit from integrating concepts and analytic tools from other sociological traditions, in particular, ethnomethodology, to help concretize its theoretical arguments, I argue that work on future-making (itself heavily inspired by pragmatist ideas) could benefit from borrowing insights from theoretical conversations on similar “performative” social processes to stop it from sliding into substantialist understandings of projection and projected futures as “shared” semantic frames of interpretive conventions.

I develop this argument through three steps. First, I problematize how projection is increasingly framed as a discrete social process or activity that is logically distinct from the “temporally embedded process of social engagement” (Emirbayer and Mische 1998:963) in which practical trajectories for bringing about such projected futures are constructed. I build on Garfinkel’s (1967, 2002) analysis of indexicality and Suchman’s (2007) analysis of how plans (or “projects”) are embedded within “situated action” to show that projection needs to be understood as co-constructed with protentional orientations within “collectively organized contexts of action” through which future-making, like agency more abstractly, operates (Emirbayer and Mische 1998:974; Jasanoff and Kim [2009:122] similarly insist that imagination needs to be understood as a process that is endogenous to “an organized field of social practice”). I draw on Weick’s (1988) concept of enacted sensemaking to show how projections of future trajectories are not prior to but are co-constructed with the actions that “implement” them and that projection is more fruitfully understood as a second-order distributed process that constructs a grammatically ordered, sequentially elaborated “network of meanings” than as a “shared” semantic construct that precedes its expression through corresponding actions.

I then extend Weick’s (Weick et al. 2005) analysis of distributed sensemaking by drawing on Leifer’s (1983, 1988) brilliant analysis of how each “local action” in a game of chess (e.g., Suchman’s [2007] “situated action”) contributes to a de-centered or distributed process of projection by virtue of how the syntactical interconnections between pieces (and the horizon of their possible moves) enact a relational “grammar” that endogenously structures the unfolding trajectory of moves. This analysis allows me to show how the projective mode of coordinating future orientations is not separate from but an endogenous result of the interconnections between local actions and the relational order they jointly enact. The analogy between projection in a game of chess and in social processes of future-making allows me to show that although they can be analytically distinguished and even become “disjointed” (Tavory and Eliasoph 2013), projective and protentional temporal orientations are constitutively interdependent because their “dynamic interplay” (Emirbayer and Mische 1998:963) defines and enacts the very setting in which both local or situated meanings and “global” projects become meaningful in relation to one another.

Finally, I draw on Hutchins’s (1995) pathbreaking work on socially “distributed cognition” to draw these two arguments together and show how a relational order of meanings that connect situated actions (a “netdom,” that is, “mixtures of social networks and semiotic domains” [Godart and White 2010:567]) functions as a social grammar of future-making. I build on White’s fundamental insight that “network relations and discursive processes are dual and co-constitutive” (Mische and White 1998:695; see also Godart and White 2010) to show that projection should not be understood in terms of some shared semantic substance but as “the grammatical processes used to articulate the representation within a relevant language game” (Goodwin 2018:283), which constrain and condition what futures can be practically envisioned and operationalized by social actors who are embedded within a process of temporal engagement within a netdom. I show that projected futures are more consistently understood and theorized as a “synthetically construct[ed] representation . . . that emerges out of the interconnections” between indexical processes of sensemaking and organizing (Taylor and Van Every 2000:207).

Futures From Practice: The Indexicality Of Projection

As Mische (2014:447) observes, “projected futures are often hard to see amidst the routine practice of day-to-day life.” As a result, despite their growing popularity as an object of sociological analysis, it has proven challenging to fit the idea of projected futures into existing sociological theories and ontologies of the social (Suckert 2022:418). Building on the intuition that interindividual processes of future-making require actors “to share an image of a future together, even if implicitly—at least enough to coordinate their actions” (Tavory and Eliasoph 2013:909), sociological work on future-making has sought a way to analytically stabilize projected futures as an object of analysis and in particular to isolate their relevant features from the “delicate choreography [required] to maintain actors’ shared orientation towards the future while accommodating motion, ambiguities, and missteps” (Tavory and Eliasoph 2013:909).

Analytically extricating projected futures from the “temporally embedded processes of social engagement” (Emirbayer and Mische 1998:963) they instruct involves (at least) three (interrelated) theoretical problems. First, because “the future is a process that unfolds in a non-linear way as actors investigate it,” images of the future are not static but frequently “revised based on contingent and changing interpretations of an emerging situation” (Beckert 2016:14). Second, because projected futures are continuously (re)negotiated in processes of “communicative interaction within groups, organizations, and institutional settings” (Mische 2014:441), actors share images that are “structurally similar” but whose relevant properties are not “contained fully in the representation of any one [site] nor are they finalized at any moment in time,” similar to processes of “distributed sense-making” in organizational processes (Weick and Roberts 1993:365). Finally, to meaningfully speak of projection as a “mode of future-coordination” that is distinct from “protention” (situational orientations to the immediate future), ¹ it is necessary to analytically differentiate the construction of “projects” and the “trajectories” they inform from mere “protentional sequences”—but also to avoid a “theory of complicity in which protentions and projects . . . work seamlessly in tandem” (Tavory and Eliasoph 2013:919).

Seeking ways to make future projections more analytically tractable, most recent research emulates Mische’s (2009, 2014:447) methodological strategy of focusing on (discursive) “externalizations” of future projections during reflective “future-oriented deliberations” because they render the semantic properties of these projections explicit and thus more easily accessible. Mische (2014:440) carefully outlines how such a methodological focus on “sites of hyperprojectivity,” that is, “communicative settings, somewhat removed from the flow of day-to-day activity” and in which the reflexive deliberation of the future brings “projective grammars” to the discursive surface, is complementary with rather than a substitute for careful analysis of the “performative work” through which futures are coordinated within actual social processes. Subsequent work, however, has been less careful in distinguishing discursive processes in which projective grammars are made explicit by social actors themselves, and thus become more easily accessible for the researcher, from projection as a conceptual dimension of future-making that is analytically constructed by the observer but ontologically endogenous to social process. Instead, this work tends to skip the problem of how to theoretically describe and empirically analyze the “dynamic interplay between projectivity and iteration” (Emirbayer and Mische 1998:963) and focus on the articulation of projected futures and their semantic properties. The result has been a gradual but consequential process of nominalization and objectification in which projection (singular), as an endogenous, reflexive, or deliberative mode of activity that social actors engage in as part of social processes of future-making, gradually becomes confounded with projections or projected/imagined futures (plural), shared images, or interpretive conventions whose semantic properties explain how futures are being performed and with what outcomes.

Such cases of substantialist conflation of “observable processes-in-relations” (White 1997:60, cited in Emirbayer 1997:298) with the concepts introduced to describe them are a common problem in sociological discourse (Bourdieu 2008:31; Emirbayer 1997; Garfinkel 1967:1–11; Smith 2005:55). As Smith (2005: 54) put it in her critique of mainstream sociological theorizing, “concepts are substituted for or displace the actual in which, by whatever indirect means, they originate: the actual becomes selectively represented as it conforms to the conceptual; the conceptual becomes the dominant mode of interpreting the resulting selection” (a move that she provocatively refers to as “blob-ontology”). The result is a reification of conceptual entities whose assumed “actuality” is ascribed to them based on “their imputed effects on people and their doings, but how they exist is not seriously problematized” (Smith 2005:55; Bourdieu [2008:31] and Garfinkel [1967:1–11] formulated similar critiques).

In the case of recent sociological work on futures and future-making, this substitution of a “substantialist” and semantic for a relational, pragmatic, and processual understanding of projection and future-making has been obscured by the “eclectic mixing” of pragmatist rhetorics and terminologies with more substantialist ontological intuitions (despite Emirbayer’s [1997:282] spirited warning against just such a conceptual blending). Specifically, analyses of future-making have slipped (back) into the deeply engrained intuitions and idiom of what Duranti (2015:105) described as the “logical tradition” of thinking about agency and social order. This logical tradition understands and explains social action and coordination in terms of a correspondence of means (sequences of actions) and ends as defined by (shared) “logical orders of intelligibility” (Duranti 2015:105). As exemplified by Searle’s (1985) synthesis, the logical tradition understands action as an expression of a prior intentional state or intentionality that specifies a trajectory or sequential assemblage of individual acts that are “necessary in order to perform the whole task as defined by ‘prior intentions’” (Duranti 2015:16). Thus, in the logical tradition, social action and its coordination across multiple actors (or “shared” intentionality [Bratman 1993], in the sense of defining a “collective” [Searle 1990]) involves a causation that runs from intentional states such as plans or projections to their execution as action.

Whereas relational sociology has attempted to rethink agency (or future-making) as path-dependent and situationally embedded and therefore “as inseparable from the unfolding dynamics of situations” (Emirbayer 1997:294), much recent research of future-making falls back on terminologies and analytic narratives that “isolate . . . intentional states from the larger contexts and activities within which acts of intentionality are realized” (Duranti 2015:109; for a similar analysis, see Bourdieu 2008:31; Smith 2005:75) by describing processes of future-making as “movement ‘from text to action’” (Mische 2009:702), or as involving the articulation of “projects” that are “‘done’ on the level of protention” (Tavory and Eliasoph 2013:922). Although the theoretical reasoning is often more complex than these formulations may suggest, they are symptomatic of a process of “eclectic” conceptual blending that is headed in the direction of what Butler (2010) aptly describes (and critiques) as “a ‘hermeneutic’ version of cultural construction, in which (shared) intentional states or orders of intelligibility are theoretically reified into a metaphysical substance that precedes its expression” through corresponding social action(s) or performative effects.

Recent work on futures and future-making in sociology exhibits many of the symptoms Butler (2010) describes. Although it is unclear precisely how “projections” of the future exist (where they are located, how they are shared, how they “act” on social practice), various performative effects in constituting and coordinating future orientations and trajectories are imputed to their existence as shared semantic conventions (mirroring Smith’s [2005:55] analysis of how agency is attributed to conceptual artifacts). Mische (2014:438), for instance, oscillates between describing future imaginaries as cognitive schemata that “exist in our heads” and as enunciations that exist across de-centered discursive processes because they “do not just happen inside people’s heads, but rather develop via communicative interaction within groups, organizations, and institutional settings” (p. 441) and are “subject to the rules and contingencies of sequentially evolving conversational interaction” (p. 446). Beckert (1996, 2016) likewise strongly nominalizes projection in his account of “imagined futures.” In good pragmatist fashion, Beckert (2016:14) points out that “the future is a process; it unfolds in a nonlinear way as actors investigate it and make decisions,” but at the same time, these decisions are anchored in quite semantically discrete and elaborate “imaginaries of what the future will look like.” Actors first form images of how the future will unfold by “visualiz[ing] causal relations, and the ways they perceive their actions influencing outcomes” and then “organize their activities based on this mental representation” (Beckert 2016:9). These images are not simply sets of loosely related meanings woven into processes of future-making, but “underlying any imaginary of a specific future state is a story of how the present will be transformed into the depicted future through causally linked steps” (Beckert 2016:69), and social actors “use these expectations to coordinate their decisions” (Beckert 2016:11).

Projection as Indexical, “Situated” Planning

One of the most perceptive and trenchant critiques of the idea that the “ordered properties” of social practice and relations are the result of common frames, interpretations, or shared understandings that precede and produce such order was formulated by Garfinkel (1967:35). Garfinkel (1967:28) demonstrated that the mutual intelligibility of social actions is not the result of their correspondence to a shared logical order but an ongoing, practical accomplishment that is produced endogenously “by the methodical practices through which members accomplish sense and accountability” of their activities in relation to one another. ² Strict interpretations of Garfinkel’s work tend to zoom in on the local manifestations of methodical practices and their ordering effects, which analytically “disconnect them from past and future as well as from the ongoing social processes in which the originals were necessarily embedded” (Smith 2005:67). However, as Emirbayer and Maynard (2011) suggest, some of Garfinkel’s theoretical insights and conceptual tools can be deliberately “misappropriated” (as Smith [2005:99] puts it) to render pragmatist arguments about the relationality and processuality of the social more analytically concrete.

In particular, Garfinkel’s (2002:197–218) brilliant analysis of the indexicality of what he calls “instructed action” is useful for understanding what is problematic about the tendency to conceive of the projection of futures as the articulation of shared understandings or interpretive frames rather than as a process or operation that emerges from the “ongoing dialogue or conversation between the iterational modalities of agency and [its] temporal relational contexts” constitutively entangled with the “iterational modalities” of agency (Emirbayer and Mische 1998:981). Tavory and Eliasoph (2013:919) convincingly argue that future-making cannot analytically be reduced to either dimension, warning against a “reduction of . . . trajectories to protentional sequences” and a “theory of complicity, in which protentions [and] projects . . . work seamlessly in tandem.” Clearly, the relative weight of these two dimensions within processes of future-making can vary, with more fully elaborated projections allowing social actors to imagine trajectories that escape the habitual homology between practice and social structure (Mische 2009:696). However, their argument that projects and protention can become “disjointed” and that effective future-making requires coordination between these two dimensions or levels (Tavory and Eliasoph 2013:922–25) relies on terminology and rhetorical frames that suggest projective intentionality and habitual doing are ontologically autonomous and can operate separately from one another, pulling social actors toward miscoordinated and potentially incompatible futures. A closer reading reveals more nuance, but certain formulations also encourage a view of the coordination of future-making in terms of semantic causation, requiring that “actors . . . share an image of a future together, even if implicitly, at least enough to coordinate action” (Tavory and Eliasoph 2013:909, 922) because “any project, as well as any cultural narrative, must be ‘done’ on the level of protention.”

In contrast, Garfinkel’s (2002:197–218) account of “instructed action” conceptualizes what he calls “prospection,” the reflexive projection of the sequential ordering of actions required to achieve an intended outcome, as integrally bound up with the routine methodical practices by which action is made rationally accountable to an unfolding context. As Garfinkel (1967:97) put it, there is a “necessary distinction between a ‘possible future state of affairs’ and a ‘how-to-bring-it-about-future-from-a-present-state-of-affairs-as-an-actual-point-of-departure’” or “operational future.” Even where possible futures are semantically elaborated and “known in a definite way,” they leave the “alternative paths to actualize the future state as a set of stepwise operations upon some beginning present state . . . characteristically sketchy, incoherent, and unelaborated.” Actors construct projections of the future to instruct their actions in the present, but this process is intrinsically indexical, which, as Suchman (2007:80) succinctly summarizes Garfinkel’s analysis, “means that an instruction’s significance with respect to action does not inhere in the instruction but must be found by the instruction follower with reference to the situation of its use.” Prospections and the instructions (specification of sequential ordering of actions) derived from them “have a developing coherence as a part of a course of action that they do not have as instructions at the outset” but which they acquire through their sequential, dialogical interaction with (relevant) context (Garfinkel 2002:41). From this perspective, projection should be seen as a dimension of a dialogical process in which a sequence of (routinized) acts is constructed that is prospectively sequenced to be coherent with a context whose relevant features are unfolded and concretized through the process. As the initial instruction and course of action or trajectory become respecified when particular acts become (partially) “blocked, thwarted, or rendered ineffectual” within a context that includes both material features and others’ witnessable practices as functional components (Emirbayer and Maynard 2011:239), the instructed course of action can (ex post) be described as being coherently instructed by (and rationally accountable to) this context.

Suchman (2007), whose work is deeply rooted in Garfinkel’s reflections (pp. 69–84), has developed a fine-grained analysis of how the projective elaboration of plans, understood as the specification of a “sequence of actions [to] accomplish some preconceived end” (Suchman 2007:52), and the production of situated action (in the sense of Garfinkelian indexicality) go hand in hand—projection is bound up with situated modes of orientation and ordering as much as it helps construct and select protentional orientations within local situations. Suchman (2007:13) shows that planning “is itself a form of situated activity” because “the planned character of our actions is not . . . inherent but is demonstrably achieved.” Actors do not monitor the rational or planned character of their courses of action in terms of logical correspondence with some (shared) semantic projections of the future but, rather, by making them witnessable and accountable to the prospectively constructed features of the situations they themselves unfold and create through their action and in terms of which they can evaluate whether these actions are “close enough for all practical purposes to what we had intended” (Suchman 2007:13; Thévenot [1995] proposes a similar “externalist” and distributed view of planned action). Projection thus needs to be understood not in terms of shared understandings or semantic frames/orders but as the prospective anticipation of pragmatic “settings,” that is, “a relation between agents and the arenas in relation with which they act” (Lave 1988:150), which is endogenously constructed and modified in social process. Projections and plans are “elaborate[d] . . . just to the extent that elaboration is useful; they are vague with respect to the details of action precisely at the level at which it makes sense to forego abstract representation” in favor of protentional orientation within a setting; planning and projection do not “serve as specifications for the local interactions but rather orient or position us in a way that will allow us, through local interactions to exploit some contingencies of our environment and to avoid others” (Suchman 2007:185).

Projection as a Distributed Process of “Enacted Sensemaking”

Garfinkel’s (1967, 2002) and Suchman’s (2007) analyses offer a careful critique of and antidote to substantialist understandings of projection and future imaginaries in terms of semantic orders of intelligibility that “transcend indexicality, that is, the actualities from which they are extracted” (Smith 2005:52; see also Emirbayer and Maynard 2011:239). Instead, they show how the projection of “rational properties of . . . indexical actions is an ongoing achievement of the organized activities” through which social actors construct, maintain, and “share” social settings and their ordered properties (Garfinkel 1967:34). However, ethnomethodological research and analytic terminology is predominantly focused on the problem of understanding local practices and processes of ordering (Hilbert 1990), making it difficult (and somewhat controversial) to extend their insights to describing and theorizing the interconnections between the interaction order and the relational order within which it is set (Tavory and Fine 2020:366). To skirt this controversy, I draw on Weick’s (1988, 1995a, 2010) concept of distributed and enacted sensemaking as a useful resource for scaling up Garfinkel’s insights into the indexical relationship between the projective elaboration of and protentional orientation within settings to processes of future-making in which projection and the setting it instructs are relationally and temporally distributed across “extended relations that coordinate multiple settings translocally” (Smith 2005:49). In other words, projection can be understood as a process of enacted sensemaking that produces a relationally and sequentially ordered set of situated plans and actions that are made mutually accountable due to their mediation by shared cultural form(s) (Godart and White [2010:582] argue for such a view of culture and structure as co-constituted).

Weick (1995a:31) explicitly theorizes sensemaking as an iterative process of constructing an ordered sequence of situated actions in which “people act, and in doing so create the materials that become the constraints and opportunities they face.” By constructing actions to be accountable (“have sense”) in relation to one another as a function of the setting they enact, each action generates further cues about the relevant environment that become input for subsequent sensemaking and elaboration of the setting they enact. One anecdote Weick (1995a:54) repeatedly uses (and incrementally embellishes) to concretize his concept of enacted sensemaking is about a Hungarian military unit that got caught up in a snowstorm during a reconnaissance exercise in the Swiss Alps. After being lost for two days, the unit discovered a map in one team member’s gear, which it used to tentatively identify landmarks, determine their position in relation to them, and incrementally elaborate a trajectory to their destination by iterating this process. However, upon arriving, they discovered the map depicted an entirely different area, located not in the Alps but the Pyrenees.

What makes this anecdote so interesting is what it reveals about the nature of projection, planning, and shared understanding. It seems natural to understand the map as a shared (abstract) “semantic” order the team used to construct a trajectory of “logically corresponding” actions (or “shared” intentionality [Bratman 1993] and thus a “collective” [Searle 1990]). However, the fact that the projected trajectory succeeded despite the map being “objectively” false shows that such a semantic and logical view of projection is misleading. Rather, it makes more sense to think of the map as a (materialized) cultural resource (in the sense of Swidler [1986]), a sort of “projective grammar” (Mische 2014), or more precisely, as providing a system of grammatical categories that help “focus ‘rays of attention’” (as Mische [2009:696] expresses it, following Schütz) on types of topographic features that are crucial to constructing an effective course of action, even if the precise grammatical relations depicted do not map onto the actual environment.

Smith (1999:119–30) developed a wonderful analogy between the process of coordinating situated actions into convergent trajectories and the process of “reading a map.” Maps encode and make available a system of cultural forms that help actors identify relevant (combinations of) typified features to which they can then orient and make their situated actions accountable (Smith 1999:125). A map depicts an environment through a system of multiple (interrelated) grammatical dimensions that focus attention on the relations between particular semantic categories (and the sets of features belonging to these categories). ³ Even where the map does not correspond to the terrain, it still functions as a visualization of grammatical structure(s) that helps coordinate and make multiple situated actions and plans (logically) mutually accountable by ensuring they encode features in ways that are semantically and grammatically invariant across the sensemaking process. The grammatical system thus coordinates situated acts of projection by enabling a reflexive dialogue between “the features iconically represented in the map” and “locally recognizable differentiations” (Smith 1999:126).

What makes this process dialogical is that the map cannot itself “performatively” relate situated interpretations into a “logical order” but depends on what Thévenot (1984) calls an “investment in forms,” a (protentional) “cultural toolbox” (Swidler 1986) of “iterable forms” (Cooren 2000:37; drawing on Derrida 1982) such as interpretive procedures, typifications, and scripts, which, not unlike Garfinkel’s (2002) ethnomethods, allow actors to identify, select, and integrate features into a syntactically correct(ly constructed) setting. As Smith (1999:126) puts it,

[T]he map reader must know how to find the local particularities of the terrain as the objects indexed by the map; they are not contained in cartographic discourse. The relation of referring is brought into being in an actual course of action in which a reader . . . picks out the objects indexed by the map from a field of experience which is always more and other than can be recognized by a reader under the map’s instructions.

Actors thus require a “protentional syntax” and shared semantic typifications to identify grammatical structure(s) in congruent ways across particular acts of sensemaking or projection (Weick 1995a:13) to ensure the setting is “written in the same language” (Latour 1987:254) and thus remains semantically and grammatically invariant across each transformation it undergoes through an act of situated sensemaking. This preserves shared reference to the environment across situated actions by ensuring they remain grammatically consistent, in the sense that the trajectory constructed from a toolbox of (domain-specific) typified acts (“taking the subway,” “walking,” “driving in a car”) are sequentially and functionally ordered in a way that is grammatically consistent with the topographical setting that is being enacted. Projection and planning, then, are not prior to or separate from but are constitutively entangled with the practices and iterable forms of protentional orientation. Rather than a disembedded articulation of a shared semantic frame, they may be better understood as the production and transformation of a relational, pragmatic, and processual setting or “enacted environment” inside of which projection/action continuously bring “events and structures into existence and set them in motion” (Weick 1988:306). Projection, from this perspective, is endogenous to a process of sensemaking in which “portions of the field of experience are bracketed and singled out for closer attention” and can be cumulatively incorporated into the enacted environment, becoming “the source of expectations for future actions” (Weick 1988:306–307). Future-making thus should not be understood as a logical “movement from text to action” or the “doing” of projects at the level of protention but as a process in which both are coproduced in an iterative process of future-making in which “the text is constructed as well as . . . read” (Weick 1995a:7). Here, “text” is a shared (representational) account of how a trajectory of actions is functionally appropriate to achieving a goal (i.e., “planning”) within a specific environment.

Network Structure, Projection, And The Relational Grammar Of Future-Making

Despite the differences in their conceptual terminologies, pragmatist and ethnomethodological ideas and Weick’s (1995a) conception of (enacted) sensemaking offer a number of complementary theoretical insights for understanding future projections not as shared, discrete semantic or logical orders but as a relational order of meanings defining a praxeological setting that is enacted through a continuous dialogue between situated actions and the social processes in which they are engaged (following Smith 1999:110). Weick’s (1995a) analysis of enacted sensemaking, specifically, opens up a path toward theorizing such projections as a pragmatically constituted “consensual domain” (Maturana 1981:16–22) of grammatically structured, sequentially ordered networks of indexical meanings.

In this section, I want to extend this “grammatical” view of projection via the observation that the cultural processes in which meanings are co-constructed should not be conceptualized as autonomous from but, rather, as endogenously interwoven with the relational structure of networks (along the lines of Godart and White’s [2010:567] analysis). I argue that the grammatical structuring of situated meanings into projections of the future can but does not have to rely on externalized or materialized cultural constructs (e.g., a “map”). It can also be accomplished distributedly, by the mediation of situated actions through the system of “interconnections between interaction order[s] and the relational order within which [they are] set” (to borrow Tavory and Fine’s [2020:366] phrasing). Following Garfinkel’s (2019) Wittgensteinian suggestion that games and their rules can provide a useful analogy for understanding how social actors construct shared understandings as they make their respective lines of action accountable to one another, I use the case of playing chess as a prism for understanding how relational order becomes visible to situated action indirectly, by means of shared sets of iterable cultural forms (e.g., scripts, typifications) that produce a de-centered grammatical ordering of situated actions across the process of future-making. I draw on Leifer’s (1983, 1991) brilliant analyses of chess as a means of understanding the relationship between what he refers to as “local” and “global” action to concretize Goodwin’s (2018:283) argument that the (functional) relationship between situated action and a “global” project or plan is not one of logical correspondence but is defined through the “grammatical processes used to articulate [a] representation within a relevant language game,” which, in the case of chess (as in social activity systems), are defined by the relational order of the network domain in which a process takes place.

Projection through Indexical Relations: the Case of Chess

Chess is commonly described as an ideal-typical example of a game based on detailed calculation and elaborately planned trajectories (Besbris and Fine 2023:1135–38). Although its opening stage is heavily scripted by large and elaborate repertoires of routinized sequences, this is not true of the middle game and endgame. The highly stylized terrain of the chessboard provides a topographic “framework [that] defines a ‘global’ level . . . where outcomes have a shared interpretation” (Leifer 1988:500), but in practice, any detailed “ex ante framework” or game plan “will be useless, as evaluations and strategies are in continuous flux” (Leifer 1991:26). The middle game, in particular, is “more like figuring out a route on a map that keeps changing before your eyes” (Garry Kasparov, cited by Besbris and Fine 2023:1137). The idea of constructing a detailed “global” plan or project specifying a detailed trajectory “remains only a theoretical possibility until the number of alternatives is vastly reduced through prior play. Only near the end of the game do players confront the global game, where alternatives and evaluations can be fully shared” (Leifer 1988:500–501). Instead, “the crucial property of local actions is flexibility. Local action cannot be tailored for specific outcomes or cast the player into a clear role, but must be able to serve multiple, unforeseen outcomes” (Leifer 1988:503).

Padgett (1981) describes this practical logic as one of maintaining “ecological control,” which relies on the protentional orientation provided by the functional interconnections between different possible moves to project how specific situated actions open up, close, or shape future trajectories as a function of the chessboard’s topography and the relational order formed by the pieces situated within it. Even highly skilled players rely on the interconnections between local actions within this relationally organized system of possible moves as a “practical regime of engagement” (Thévenot 1995, 2001) that enables them to evaluate individual moves’ rationality against a setting of functional relations that hold these moves accountable against the logic of the game. A game of chess can thus be understood as an “activity system” whose constituent situated actions (moves) are directed toward a shared outcome or object(ive) that provides the “enduring, constantly-reproduced purpose . . . that motivates and defines the horizon of possible goals and actions” for the activities in the system (Engeström 1999:170). In contrast to Weick’s (1995a) mountaineers, however, the individual situated actions in chess are not “co-oriented” (Taylor 2005) toward the joint object(ive) by means of an externalized cultural (or projective) grammar, such as a map, but by a generative grammatical system that is built or woven into the relational structure of the game or activity system itself.

The relational positioning of pieces within the fully determinate topography of the chessboard defines a set of external constraints that shape the semantic meaning of possible moves within the context and process of the activity system. In contrast to Weick’s (1995a) mountaineers, grammatical relations between moves (as a function of the combination of syntactic and semantic constraints) are not determined through a reflexive dialogue between an external grammatical system/set of rules and the indexical construction of situated meanings. Rather, the construction of situated actions is relationally indexed to the joint setting or “consensual semantic domain” (Maturana 1981:16–22) defined by the board, the pieces’ different forms of agency, and their positional relations. This relational order as defined and reproduced in the game functions as an endogenous grammatical structure that ensures the continuous praxeological consistency of the moves. The concrete trajectories that emerge from such local action may be more or less strategically effective, ⁴ but to the extent that situated actions and the calculation of their functional (or grammatical) embeddedness in the relational order of the game consistently rely on a common (institutionalized) cultural toolbox of typifications through which moves are made accountable to one another (much as Schutz [2011:238] describes the condition for mutual intelligibility; Thévenot [1990, 1995] makes a similar argument), the sensemaking process generates a “global competence” without “detour through shared representations” (Thévenot 1995:419; my translation from French original). The relational order of the game thus functions as a grammatical structure that continuously coordinates a set of syntactically linked “trajectories of co-relevant actions” (Enfield 2013:31) through an iterative process in which “actions that elicit actions supply more observations and a narrower range of alternatives for decision . . . until a global outcome is apparent” (Leifer 1988:504).

The Grammatical Structuring of Agency within Relational Network-Domains

Although the moves in a game of chess are ultimately executed by (normally human) global players, chess provides a good analogy for how the relational ordering of interconnections between situated actions can function as a “socially organizing grammar” (Smith 2005:110) that endogenously coordinates situated actions into a (more or less) consistent process of future-making. From the outside, this process appears as a shared projection of a future that explains a set of corresponding actions. Based on my argument here, it could more aptly be described as a distributed process of enacted sensemaking that endogenously organizes a set of relationally connected situated actions into a more or less consistent system of “distributed” (Enfield and Kockelman 2017) or “shared agency” (Bratman 2014) that is directed toward a shared future via a network of interdependent meanings that, from a bird’s-eye view abstracting structural similarity across processual and relational variations, looks like the externalization of a shared semantic image of the future. Building on Weick’s (1995a) analysis of enacted sensemaking, we can also describe it as a process of future-making that is mediated by an “enunciative network” (Dreyfus and Rabinow 1983:52) whose grammatical structure shapes the futures that can be articulated through it. From a temporal perspective, such a “network of meanings” (Godart and White 2010:572) can be described as a system of “semiotic processes [that] enable and constrain, and thus . . . ‘select’ and ‘sieve’ one another” (Kockelman 2013:100) as they “extend into the imagined future” (Mische 2009:698). Through their grammatical interrelations, these processes define and enact “a matrix of possibilities for its future organization. These possibilities are shaped retrospectively by what has gone before, while prospectively each new element projects a horizon of changing trajectories for what might follow” (Goodwin 2018:48). Borrowing Geertz’s (1973:5) vivid image, we might say the situated actions that occur within such a distributed process of future-making are “suspended” within the “webs of significance” they themselves weave, and their grammatical structure is defined by how multiple actors’ respective trajectories of situated actions and sense-making are contextualized by their semiotic linkages. ⁵

Although, as one reviewer pointed out, Goodwin’s (2018) account of “co-operative action” was developed for the analysis of locally witnessable interaction processes, it can usefully be extended to, following White (Godart and White 2010:568; White 2008:7–13), what might be called “netdoms” or “network-domains” of future-making in which relational networks and semiotic domains are constitutively fused together (although they can be distinguished analytically) within social processes. By analogy with Goodwin’s (2018:3–4) analysis, such a netdom can be understood as a relational “sign complex . . . that gives organization to the elements that compose it” and whose (grammatical) structure “is the focus of transformative operations being used by another actor to create a next action.” Within such a network or sign-complex, each “new action is built by decomposing, and reusing with transformation the resources made available by the earlier actions of others” (Goodwin 2018:1). As long as social actors “build new action cooperatively by including in each next action materials placed in a public environment by earlier actions, which leads systematically to the accumulation of structure being organized as resources for the construction of relevant action” (Goodwin 2018:31), the netdom’s “constitution . . . as a web of relationships makes possible systematic, incremental modification of the elements that compose it . . . [and] can be transformed . . . by retaining the patterned relationship . . . while modifying elements within the pattern” (Goodwin 2018:4). From this perspective, a process of future-making can be understood as a “chain of transformative operations [that] is consistent with the way in which a Peircean interpretant incorporates an earlier sign complex as its object, while then itself becoming the object to be incorporated in the organization of subsequent signs” (Goodwin 2018:32). Borrowing from Weick and Robert’s (1993:365) account of systems of distributed sensemaking, we can thus describe the future as the object of such a relationally constituted activity system, which is

being constituted as these activities become more or less interrelated . . . the emergent properties of this object are not contained fully in the representation of any one person nor are they finalized at any moment in time. A single emergent property may appear in more than one representation, but seldom in all. And different properties are shared in common by different subsets of . . . structurally similar representations . . . although these activities are done by individuals, their referent is a socially structured field. Individual activities are shaped by this envisioned field and are meaningless apart from it.

How Disruption Reveals The Narrative Grammar(S) Of Future-Making

Playing chess provides a helpful analogy for understanding how the relational system of interconnections between situated actions can function as a grammatical structure within processes of future-making. Rather than conceptualizing processes of future-making as instructed by semantic convention (shared projections or imaginaries of the future or trajectories for its realization), their organization can thus be understood as accomplished endogenously through social grammars that co-orient situated actions within a relationally structured network of distributed sensemaking that enacts the joint future. However, whereas in a game such as chess situated actions are tightly coupled and thus fully enclosed within a rigid grammatical structure imposed by a fixed set of uniform rules, social processes of future-making are much more “loosely coupled systems” (Orton and Weick 1990). Social actors do not behave as cultural-relational “dopes” (Lynch 2016) whose sensemaking and decisions are seamlessly instructed by a social grammar or script (whether predominantly cultural or relational). Instead, actors navigate and “switch” between multiple partially overlapping but heterogeneous contexts and netdoms (Fontdevila, Opazo, and White 2011; Godart and White 2010; Mische and White 1998), thus introducing a distinctly sequential dimension that may interfere with the routine “co-construction of meaning in the back and forth of interaction” (Tavory 2018:121). The result is that social processes of future-making oscillate between phases in which protentional orientations are effectively aligned into a consistent “grammatical function” and “moments of disjuncture,” when future-oriented sensemaking becomes more reflexive and recursive and social actors articulate more explicit accounts of how their situated actions are directed at specific “anticipated futures that they must continuously calibrate for concerted action to be possible” (Tavory and Eliasoph 2013:911).

Although this means that it is not uncommon for projection and protention to become “disjointed” (Tavory and Eliasoph 2013:922–25), especially across organizational boundaries (see Meyer and Rowan 1977), this reasoning also suggests it is more theoretically accurate and informative not to understand this disjointedness as a relationship between two distinct and autonomous modes of future-making but as a series of “relational ruptures . . . of the ongoing relationships within which . . . interaction occurs” (Tavory and Fine 2020:374). A theoretical focus on sites of projectivity and conceptualization of future imaginaries as externalizations of shared semantic (whether cultural or cognitive) constructs invites a dualist reading of such disjointedness but is hard to reconcile with a more dialogical understanding of projection and protention as functionally entangled dimensions within processes of future-making. Following this line of reasoning, disjointedness between projection and protention is better understood as disruptions within the “inter-situational texture of meaning” (for an elaboration of this concept, see Tavory 2018), the setting that is enacted in the process of future-making. When discrepancies or inconsistencies emerge within the network of meanings and situated actions whose connections define the setting through which the future is enacted, social actors articulate (more) explicit accounts of the functional (and sequential) interconnections between situated actions and meanings to realign them into a consistent grammatical structure of future-making.

In this last section, I therefore build on Tavory and Fine’s (2020) argument that by looking at moments of disruption of the systems of ongoing relationships, we can gain the clearest insights into the interconnection between situated (inter)actions and relational orders. I show that projection becomes particularly visible and appears autonomous during such moments when the grammatical structure of future-making becomes problematic, shifting the sensemaking process into a mode of reflexive explication of how the indexical orientation of situated action to the future is mediated by its syntactical embedding within the grammatical organization of the distributed process of future-making in which they participate.

Distributed Cognition and the Narrative Structure of Future-Making

To develop this argument, I draw on Edwin Hutchins’s seminal work on naval navigation as a form of distributed cognition (Hutchins 1991, 1995; Seifert and Hutchins 1992). Naval navigation offers a particularly interesting case for my argument because it involves a social process of future-making that is implemented through a distributed system of calculative tasks that are rigidly enclosed within a highly formalized grammatical structure that implements a global “normative computational structure” (Hutchins 1995:343). I focus on Hutchins’s (1991, 1995:7–8) detailed analysis of a situation in which the routine process of sensemaking and the distributedly enacted setting through which a ship’s position in its environment is represented and in which the ship’s trajectory and the actions required to implement this trajectory are projected and planned were profoundly disrupted and nearly collapsed. This allows me to zoom in on “the relationship between a moment of disruption and of alignment” (Tavory and Fine 2020:366) and clarify how social actors develop reflexive descriptions of the relationship between their situated actions and the grammatical structure of the networks of meanings (setting) through which they “coordinate lines of actions within and across situations” (Tavory and Fine 2020:366) to enact a sequence of actions that will lead to an intended, shared trajectory.

Simplifying somewhat, naval navigation involves an iterative process, the “fix cycle,” which is a distributed calculative process that enacts a joint setting by continuously plotting a ship’s movements and position in relation to selected features of its environment and through this setting, projects the sequence of actions required to put the ship on a trajectory to an intended future destination (Hutchins 1995:117). The “fix cycle has two major parts: determining the present position of the ship and projecting its future position. [It] gathers various bits of information about the ship’s location in the world and brings them together in a representation of the ship’s position” (Hutchins 1995:26), based on which appropriate moves for constructing a trajectory to the intended destination are calculated (Hutchins 1995:39). Building on White’s theorization of netdoms as a “compound of interrelated stories” (Godart and White 2010; White 2008:20–62), one can describe the fix cycle as a process of future-making that follows a narrative logic, that is, “a logic of causality . . . mediated through context” (Lane and Maxfield 2005:13), as it constructs a sequence of actions that will cause a vessel to reach its destination by combining available types of actions as a function of the constraints imposed by the environment uncovered through the sensemaking process.

Hutchins (1995:343) describes the fix cycle as an activity system consisting of multiple calculative tasks and types of practical actions that are relationally and sequentially interlocked and organized in a way that implements a “normative computational structure,” or from a temporal perspective, a narrative grammar that adjusts the trajectory as a function of the constraints and affordances of the environment in relation to the grammatically possible acts that can link together into a storyline that is causally consistent as a function of contextual features that mediated its course. Hutchins uses more cognitivist terminology that nonetheless captures analogous insights, describing the fix cycle as a “cognitive ecology in which the various representational technologies constitute one another’s functional environments” (Hutchins 1995:168) so that “the behavior of each of the participants provide[s] the necessary elements of the information environment of the other” (Hutchins 1991:34). Each calculative operation constructs meanings that are syntactically formatted in a way that makes them “deictically transposable” (White 2002:307) into subsequent operations (Hutchins 1995:341), meaning their transposition into different situated calculations preserves the overall, structural-semantic invariance of the setting—the functional analogue to the consistency of mapping grammatical categories onto observations in the case of Weick’s (1995a) mountaineers.

Under normal circumstances, the future-making process in naval navigation is highly routinized: The formal scripting of computational tasks secures their protentional co-orientation within the sequential and relational grammatical structure of the fix cycle. This stability ensures the grammatical consistency of the meanings produced across the fix cycle’s operations. As Hutchins (1995:117) puts it, the network of computational task-nodes forms a system of distinct “representational media” whose functional and sequential ordering makes possible projection as a process in which the meanings produced by each node become inputs transformed by the interpretive procedures of the next, enacting a global “representational state” that is “propagated from one medium to another by bringing the states of the media into coordination with one another.” ⁶ This representational state, however, is “an operation rather than a common intersection of overlapping sets” (Garfinkel 1967:28)—it is a processually enacted, “synthetically constructed representation” (Taylor and Van Every 2000:207) of the ship’s movement in relation to its (relevant) environment that does not exist as a unified symbolic representation. ⁷

Despite Hutchins’s (1995) more cognitivist terminology, the future-making process he analyzes can be described in narrative terms once one recognizes that in formal narrative theory, “stories are relational rather than temporal . . . [they] organize meanings and relate them into patterns” (Godart and White 2010:572). Hutchins’s (1995) computational modules, which organize calculative operations carried out by different team members into a shared task by creating syntactic and semantic linkages between them, can thus be analyzed as stories (despite the absence of a temporal dimension in Hutchins’s synchronic description of the fix-cycle’s structure) because each “concatenates meanings in a relational structure with a horizon of possibilities” (Godart and White 2010:567) that is defined by the larger grammatical structure. In narrative analysis, the synchronic dimension of stories is referred to as their respective plots, which “provide templates for how positions [or ‘nodes’ within a netdom or fix cycle] can be related in a meaningful way” by ensuring the elements of the opening scene of the story (its inputs) are transposed or reconfigured in a way that respects their syntactic and functional interrelations in terms of their joint (dramaturgical) setting (Godart and White 2010:576). The normative computational structure of the fix cycle can thus be understood as a narrative grammar that links multiple stories (or from a synchronic perspective, routinized plots) together so that each plot’s structure is conditioned or even “determined by their respective positions within the framework” (Greimas 1987:111).

Each story contributes to the joint enactment of the setting by calculating semantic features so they are “constituted in the light of the overall narrative” or “made functions” of this emerging narrative (Bruner 1991:13). Relevant context “is relayed, in the course of a trajectory, by narrative structures and it is these that produce meaningful discourse articulated in utterances” (Greimas 1987:64). The operation of the fix cycle thus corresponds to White’s (1980:8) definition of narrativity as “the instrument by which the conflicting claims of the imaginary and the real are mediated” because context is made relevant to the specific courses of actions it is possible to implement through the action-scripts available in the narrative via the distributed calculative (sensemaking) process. From a synchronic perspective, the result closely resembles Weick and Robert’s (1993:365) analyses of the “collective mind” in organizational contexts, whose

emergent properties . . . are not contained fully in the representation [or projection] of any one [actor or node] nor are they finalized at any moment in time. A single emergent property may appear in more than one representation, but seldom in all. And different properties are shared in common by different subsets [of local calculations].

Whereas in chess, the grammatical consistency of the process of (enacted) sensemaking is built into the relational structure, in processes of social future-making, the grammatical consistency is implemented through social organization, which must be stabilized endogenously against possible disruptions. In Hutchins’s (1995:167) naval fix cycle, this stabilizing mechanism is provided by “mediating technologies,” that is, material and symbolic devices that “remove from local computations any aspects that are invariant across the spatial and temporal extent of the computation,” meaning features of the environment that are not grammatically relevant to the narrative construction of the trajectory. Together, these devices constitute a “mediating structure” (Hutchins 1995:295–316) that ensures the process of future-making operates without “violat[ing] the allowable transformations of the embedding narrative structure” (Lane and Maxfield 2005:14). In Hutchins’s (1995:154) words, it “constrains the representational states that can be produced to ones that are syntactically correct” and thus stabilizes the grammatical structure and ensures the semantic “invariance” of the setting in terms of which future trajectories are projected and implemented.

Projection and the Reflexive Enactment of Grammatical Structure

What happens when the grammatical structure of this process and thus the semantic continuity of the synthetically constructed setting and representation of its trajectory into the future are disrupted? I omit the precise (technical) details that caused the disruption and near breakdown of the navigational process, which resulted from an electrical failure that disabled a key mediating device, the gyrocompass, and the ship’s rudder (see Hutchins 1995:119–26), and instead focus on the “repair work” (Schegloff, Jefferson, and Sacks 1977) by the navigational team. The gyrocompass is a crucial navigational device that mediates and aligns several calculative tasks simultaneously, notably by ensuring the automated conversion of several distinct navigational observations into cartographic measures formatted for specific calculative tasks. The failure of the gyrocompass disrupted the routine “deictic transposability” of meaning across the sequence of computational tasks in the fix cycle: Because observational data of key environmental features were not translated into the semantic formats required for the routine calculative scripts, the fix cycle could not compute a position or plot a trajectory that would be grammatically consistent with the projective setting. When team members noticed the mismatches between the semantic inputs and the requirements for the calculative syntax of their tasks, incongruencies built up and quickly propagated downstream to other calculative tasks and team members (Hutchins 1995:168). These disruptions were particularly severe because the ship was navigating through a relatively narrow harbor, so the future-making process was “event-driven in the sense that the navigation team must keep pace with the movement of the ship . . . when something goes wrong, it is not an option to quit the task, to set it aside momentarily, or to start over from scratch” (Hutchins 1995:21). Because of the inertia of naval vessels, decisions and actions taken previously will continue to shape the operative possibilities in the present and for the future (Hutchins 1995:39–41), meaning that any future trajectory constructed (and sequence of actions to implement it) needs to be built onto and be made congruent with the previously engaged one.

The navigation crew was able to avoid a full-fledged “collapse of sense-making” (Weick 1993) due to the grammatical “modularity” (Hutchins 1995:320–45) of the future-making/sensemaking process. Because each distinct task in naval navigation is formatted by the mediating technologies that link it to adjacent (upstream and downstream) tasks within the sequentially ordered computational structure, in their training, team members learn how specific mediating technologies function and how their use configures both their own task environment and calculative script and adjacent computational tasks (Hutchins 1995:167). This shared knowledge of particular grammatical modules “becomes a shared resource” allowing adjacent team members to realign their “joint performance of the modular procedure” (Hutchins 1995:24), providing them with “guidance as to the composition of the functional system” and helping them understand how their respective task-module is embedded within its larger grammatical structure (Hutchins 1995:154). We can think of computational modules, in this context, as akin to what Mische (2014:445) calls “genres” of social practice, that is, “recognizable, schematized templates” of functional interdependencies between actions or “configurations of meaning which have developed in the context and bear the imprint of the characteristic usages associated with the activities of a group” (Smith 1999:120).

The chain of mediating devices that grammatically “discipline” local calculative modules by imposing syntactic and semantic formats consistent within a jointly enacted setting (preserving its “invariance” and thus the grammatical consistency of the course of action it informs) thus provides instructions as to what syntactic (re)combinations are possible and likely to fit into the larger grammatical/computational structure. This chain imposes a number of pragmatic “constraints on the correction of errors . . . help[ing] identities cope with the irregularities triggered by uncertainty” (Tilly 1996:589; cited in Tavory and Fine 2020:370) about how their anticipated trajectory or the next story to be constructed fits narratively into the larger narrative process and structure (Godart and White 2010:575). These affordances help coordinate social actors’ efforts to rebuild consistent translations between their future-making activities—they facilitate and instruct a process similar to what Weick describes as “heedful interrelating” (e.g., Weick and Roberts 1993). In the absence of affordances that render the functional embedding of local tasks in a wider relational order legible for local inspection, social actors lock into an “unheedful” focus on their isolated task, which leads to “less comprehension of the implications of unfolding events, slower correction of errors, and more opportunities for small errors to combine and amplify” at the level of the sensemaking process (Weick and Roberts 1993:366). In contrast, a shared understanding of how mediating devices (or social genres) require and impose certain syntactic and semantic transpositions and structural couplings between their respective calculations provides actors with instructions for realigning their calculations in meaningful ways even if they cannot return to their original calculative scripts and restore the original computational structure.

Hutchins’s (1995) description of how such local repairs and reconfigurations were enacted resembles the mechanisms that speakers use in constructing meaningful dialogues or conversations in natural language. Social actors know how their respective tasks have been syntactically “arrayed in parallel structures . . . inviting a mapping from one to another,” and people use this knowledge to project different modes of syntactical coupling so that “the coupled components re-contextualize each other, generating new affordances for meaning” (Du Bois 2014:360). Individuals’ understandings of these lateral “structural parallels invite functional inferences, influencing the situated interpretation of both utterances” (Du Bois 2014:366). In other words, these components provide actors with guidance as to how their particular calculative practice has been functionally embedded within the grammatical structure by foregrounding the semantic and syntactic parallels with adjacent calculations that allow them to produce outputs that are “meaningful” as part of the jointly enacted setting.

Thus, with routine grammatical alignment between their situated calculative practices disrupted, these parallelisms provide actors with a scaffolding for developing explicit projections of how their calculative scripts functionally embed into the larger process of future-making. Returning to White’s theorization of netdoms as compounds of stories (Godart and White 2010; White 2008), we can say that actors use these scaffoldings as the basis for telling stories about how their particular calculative module contributes to the overall construction of a setting and the enactment of a future trajectory within that setting. In other words, actors construct explicit “lexical descriptions” (Hanks 1990:49) of how their calculative practices can be embedded into a situated plot or script that is commensurate with the overall grammatical categories and structure of the global “normative computational structure” (in this sense but using different terminology, see Hutchins 1991:24). These lexically explicit stories project the grammatical interrelations between adjacent plots/stories, allowing actors to reconstruct how their respective scripts are linked within their calculative modules into plots that produce mutually intelligible intermediate outputs that are also “culturally meaningful” in terms of the global grammar of the sensemaking or future- making process (Hutchins 1991:31). The availability of a (although disrupted) mediating structure helps actors reflexively (re)“construct their actions (contribute) while envisaging a social system of joint actions (represent), and interrelate that constructed action with the system that is envisaged (subordinate) . . . [and through] contributing, representing, and subordinating . . . create a joint situation of interrelations among activities” (Weick and Roberts 1993:363). By restoring a “relative symmetry of indexical contexts” (Hanks 1990:47) along the calculative sequence of the fix cycle, the mediating structures help rebuild “the fit between computational and social organization” (Hutchins 1991:35), that is, they implement the original grammatical structure (or something quite close and ontologically continuous) through a modified computational architecture. ⁸

Discussion And Conclusions

The reanalysis of Hutchins’s (1995) account of naval navigation helps us draw together the insights gained across the previous sections into a more systematic account of how the endogenous interplay of projection and protention defines and enacts the grammatical structure of processes of future-making. Of course, projective or deliberative discourses in which a process of future-making “as a whole” becomes the object of an explicit reflection or even reconstruction can provide important insights into its grammatical structure, as Mische (2009, 2014) has argued. Often, such discourses occur in sites that are somewhat (socially, spatially, or temporally) removed from the actual activity systems and processes in which futures are enacted or “made.” However, it is important to draw a clear analytic distinction between such metadiscursive deliberation and the more pragmatically oriented modes of projection that are endogenous to and (directly) constitutive of the process of future-making. Borrowing a concept that linguists use to refer to discourses that reflexively describe (and problematize) how linguistic form conditions the construction of meaning, we might refer to the latter as “meta-pragmatic” (Silverstein 1993) projection or deliberation and to the former as pragmatically embedded or constitutive projection.

The key problem of the semantization of future-making I have diagnosed and critiqued is related to the analytic conflation or amalgamation of these two distinct modes of deliberation and projection (although both can and do occur together, and metapragmatic projections can also occur endogenously to processes of future-making). If our analysis focuses on metapragmatic projective discourses, as in Mische’s (2014) “sites of hyperprojectivity” that are somewhat insulated from the more event-driven flow of activity, we will find reflexive (if necessarily partial and hence polythetic) accounts of the endogenous but relationally enacted “projective grammars” at work in the process of future-making. As a result, a semantic reading of future-making as driven by such shared cultural understandings or narratives appears only natural. However, although metapragmatic accounts do feed back into the “indexical order” of the processes they describe (Silverstein 2003) and may reconfigure their grammatical structure, they must do so from within (or indexically). As I hope my argument has shown, although analytic distinctions such as between projection and protention are crucial to understanding how social actors coordinate and enact futures, it is equally crucial not to slip into substantialist understandings of such concepts but to approach them as “observable-processes-in-relations” and develop a careful relational and processual theorization of their dialogical relationship in constituting processes of future-making. Furthermore, this theorization must be sufficiently fine-grained to avoid convenient but analytically deceptive conceptual aggregations, such as treating discursive utterances as evidence of underlying (shared) projections without clarifying their referents within social practice or processes.

What I outline as an alternative is not a full-fledged theoretical framework of future-making (neither in content nor in expositional form) but, rather, a conceptual platform I hope might support a process of “theorizing” (Weick 1995b) future-making in more genuinely relational and processual terms. It deliberately begins not with images of the future and asks how they are put into practice but, rather, with “the grammatical processes used to articulate this representation” (Goodwin 2018:283) and asks how such relational social grammars are defined and enacted, problematized, and (sometimes) reconfigured. Approaching future-making as a form of (or in analogy with) enacted sensemaking (Weick 2010), although it admittedly risks downplaying the degrees of freedom introduced by collective imagination and deliberation, makes it easier to understand the entanglement and “dynamic interplay between projectivity and iteration” (Emirbayer and Mische 1998:963), or projection and protention, as a relational process rather than a semantic operation. This framework can help us understand projection as a “second-order process that must be accounted for from the underlying dynamics within netdoms” (Godart and White 2010:568) rather than as (the production of) a semantic frame or set of “cognitive contours” that (logically, temporally, or both) precedes the “relational dynamics” of future-making (Mische 2009:697). Instead, we can understand projection as a relational communicative process that endogenously grammatically organizes a relational order of situated meanings within netdoms (Hernes and Maitlis 2010; along the lines of work in organization studies that understands processes of sensemaking and organizing as co-constitutive; see Taylor and Van Every 2000; Weick, Sutcliffe, and Obstfeld 2005).

If ethnomethodology can help concretize some of the core theoretical insights of the pragmatist tradition for concrete sociological analysis (Emirbayer and Maynard 2011), work on sensemaking and organizing-through-communication provides a (severely underutilized) resource for studying future-making that is (more) consistent with its relational and processual pragmatist inspirations and underpinnings. Rather than think about the relationship between projection and protention as one between two autonomous modes of future-coordination, the relationship could, for instance, be more fruitfully understood along the lines of Taylor and Van Every’s (2000) account of organizing as a communicative process. Taylor and Van Every (2000) translate Frege’s (1948) crucial insight that “sense determines reference” (see also Kripke 2008) into a sophisticated account of how situated practices become co-oriented and organized through the interaction between “sub-symbolic” processes (akin to the relational interconnections through protentional forms) and processes of “symbolic” discursivization. Much as in Hutchins’s (1995) account of how mediating structures provide a fabric of protentional orientation on which the more projective co-orientation through shared lexical (or symbolic) descriptions depends, such projections are inseparable from the relational order of meaning, the text, that they help write.