Rethinking the time’s arrow: Beginnings and the sociology of the future

Introduction

Janus, the Roman god of beginnings, is the quintessential representation of the anxiety of ‘knowing’ what cannot be known. Traditionally placed at doorways or entrances, this double-headed god removes us from our normal sense of continuity with the past and confronts us with the challenges the unknown future imposes on us. It is certainly not a coincidence that the first month of our year is named after Janus. Indeed, the anxiety of being in-between past and future, old and new, known and unknown has grown deep roots in our Western culture.

In ‘Orders of discourse’, the text that signals the farewell to his early, archaeological work and the beginning of his genealogical project, Michel Foucault begins by expressing a paradoxical longing to avoid beginnings. ¹ Foucault explains this strange desire to begin without beginning with the dread of standing momentarily outside discourse, a position where the raw power of discourse is fully exposed. Our Western response to this anxiety has involved institutions that, as he explains, ‘solemnise beginnings, surrounding them with a circle of silent attention; in order that they can be distinguished from far off, they impose ritual forms upon them’ (Foucault, 1971: 7). This institutional response, in other words, imposes an order that seeks to domesticate discourse, thus making us less aware of beginnings and thus feel more in control of discourse – and of the future.

Sociology can be seen as part and parcel of this institutional reaction to the ‘respect and terror’ (Foucault, 1971: 10) of beginnings. Indeed, this helps explain both the overwhelming weight the past has in sociological accounts of the present, and, in particular, it seems to account for sociology’s quest for the institutional ‘origin’ that single-handedly accounts for everything that derives from it. Consider Pierre Bourdieu’s ‘genetic structuralism’ and its relative silence about historical discontinuities and the future-oriented character of human action. Indeed, the core conceptual apparatus of practice theory (habitus, field, and capital), including the ancillary notions of struggle and strategy, which involve a degree of intentionality on the part of agents, is fundamentally impervious to the notion that collective and individual agents can start something really new.

The sociological neglect of beginnings is not limited to practice theory, though. In the past 30 years, ‘history matters’ became the rallying cry for historical and sociological institutionalism. To explain current outcomes historical institutionalists study the long-term structural trajectories that shape choices in the present. Sociological institutionalists, though they place a greater emphasis on cognition, likewise see social change as a process of isomorphic adaptation to existing institutional models (Baert and Silva, 2010: 137–138). This focus on issues of social reproduction, not invention or revolution, extends to most other areas of sociological research. Homo sociologicus, whose habitus carries the weight of the past, is opposed to the homo economicus, the rational calculator of future expectations. The weight of structures on the present is determined in terms of the impact of the past, not the future. The future plays a meaningful role in very few sociological approaches. Consider how sociologists today are variously involved in explorations of the materiality of social processes, in studies of the impact of material conditions on social practices, in discussions of different kinds of risk, in cultural analyses of the deep structures of meaning, or intersectional research on race, class, gender, sexuality, and nation. Notwithstanding their epistemological and methodological differences, all these approaches have remarkably little to say about temporality, let alone the future (Urry, 1996). Bringing the future back in means, for instance, that intersectional analyses of one’s biography and body should be as much concerned with renaming historically significant Black women as ‘feminist’ as an act of historical appropriation (Collins, 1993: 13), as with projecting imaginary frames of action that would enable us to begin with a wholly new start. Indeed, identity is as much a question of asserting who we are and where our roots lie, as it is about reinventing ourselves. Institutional trajectories of the past matter, of course, but this article seeks to show the extent to which sociology would benefit from paying more attention to the future. History matters, but the future matters just as much.

Mustafa Emirbayer and Ann Mische have touched upon this issue in an important overview article on agency written some 20 years ago. Their basic claim was simple: prevalent attempts to theorize agency often ignored its temporal character (Emirbayer and Mische, 1998: 963–967), and when they do they tend to restrict the discussion of human agency to its ‘iterational’ (i.e. repetitive, past-oriented) dimension. As a result, while the past (and the present) got the lion’s share of attention, the future-oriented, projective dimension of agency remained on the whole poorly understood (Emirbayer and Mische, 1998: 983 ff.; but see Bergmann, 1992: 86–90). In a series of subsequent writings, Mische (2001, 2007, 2009) published a rich and nuanced body of work where she explores how future projections affect what we do. During the last decade, a number of other authors have followed suit and continued this exploration of the various ways in which the future matters in social-scientific accounts of practice and meaning (Beckert, 2016: 35–98; Flaherty, 2011; Nielsen, 2011). These studies have contributed significantly towards a better sociological understanding of social time by shedding new light on projective/anticipatory action. On the whole, however, sociologists’ understanding of the future imagining remains ‘thin and static’ (Mische, 2009: 702).

By contrast, if sociology were to open itself to the study of beginnings and the disjunctive anxiety that comes with them, sociologists would be in a better position to appreciate how practices/institutions often begin in non-linear, discontinuous ways, and how the disquieting, unknown future shapes the present in no less powerful ways than the safe, known past. Beginnings are among the most critically neglected subjects in sociology. This article attempts to rectify this situation by following the lead of economists, in whose models beginnings and the future feature prominently. Yet, contrary to economists, we should avoid reducing the future to post hoc rationalizations of action. This involves doing two things. On the one hand, sociology needs to pay more attention to the temporal orientations and perceptions of the future that actors nourish; on the other, the capacity to imagine futures ought to play a much larger role in the explanations of sociologists. This does not mean, however, that one should reduce the sociology of the future to the problematic of beginnings. There are futures which do not include any beginnings at all, as there are temporal orientations that actively prevent any beginnings from taking place. ²

That the future should be an integral part of any sociological explanation of social action, agency, or structure is a statement that prima facie most would readily agree. That most sociological approaches still overwhelmingly privilege the past and historical continuities indicates that sociology lacks a concept of beginning upon which to base the projective, future-oriented dimension of human agency and institutional development. The purpose of this article is to advance one such concept.

The article is organized as follows. First, we present the epistemological distinction underlying the existing approaches to beginnings, which opposes the irreversible time of social practices to the reversible time of analytic models. Second, we consider works that have studied beginnings from the perspective of irreversible time of participants. Third, we discuss approaches that reconstruct beginnings from the perspective of the reversible time of the observer. Fourth, we conclude with a sociological conception of beginnings as a heterogeneous duration involving a non-linear succession of temporalities. This promises to equip sociologists with a more robust conception of the future, which, not only for observers but also for participants, is able to fold upon itself. In a sociology that remains overwhelmingly concerned with the logic of reproduction of the past, few conclusions seem more relevant than this.

Claude Lévi-Strauss: Reversible versus irreversible time

Over the years, the starting point for thinking sociologically about time has been Claude Lévi-Strauss’ distinction, originally presented in the first volume of Structural Anthropology (1963), between the irreversible time of social practices and the reversible time of structural models (e.g. Bourdieu, 1990; Giddens, 1979). This is also the starting point of our discussion of beginnings. However, while Giddens’ structuration theory is an attempt to supersede it by moving up to the ontological level and Bourdieu’s genetic structuralism keeps it down at its original methodological level, we construe it as a valuable epistemological vantage point from which to assess existing approaches to the temporal dimension of beginnings. ³

On the one hand, beginnings have been studied from the perspective of the irreversible ‘time’s arrow’. For Lévi-Strauss, this focus on diachrony involves giving analytic priority to the perspective of participants, either of everyday routine or long-term historical patterns. Irreversible time, in other words, represents the linear sequential temporality of real-world social practices. These historical and ecological rhythms of observable reality are a characteristic of what Lévi-Strauss (1963) calls ‘statistical models’ (p.286) and parole. A beginning is here the first dot in the time’s arrow. It has been phenomenology that, in the human and social sciences, has taken the lead in the study of beginnings from the perspective of social agents. Consider Jean-Paul Sartre, for whom beginnings exist only if narrated. True beginnings are the product of fictive meaning-making structures. Real life, by contrast, is mere chronicity: an endless succession of discrete temporal events along time’s arrow. Early in Nausea, Sartre uses Antoine Roquentin’s philosophical reflections to make this point. Here is Roquentin’s apathetic conclusion regarding the possibility of starting something new in life: ‘There are no beginnings. Days are tacked on to days without rhyme or reason, an interminable, monotonous addition’ (Sartre, 1964 [1938]: 57). While Sartre trusts his theory of the novel to rescue beginnings from the torpor of temporal existence, in the sociology of Marcel Mauss (2002 [1950]) one finds the possibility of the development of a ‘phenomenology’ of beginnings that studies how social agents experience and explain beginnings to themselves and others, thus forging bonds that are ‘the same time juridical, economic, religious, and even aesthetic and morphological, etc.’ (p.101): for instance, the act of beginning to attend a university has immediate and irreversible consequences in multiple directions implicating students themselves, their parents, their lecturers, and ‘involve the totality of society and its institutions’ (p.100). Either as merely a narrative effect or as total social phenomenon, beginnings from the perspective of social participants mark the irreversible first step towards a future-yet-to-be, while signalling a rupture with some past state of affairs and giving the present a characteristic flavour or mood of freshness, uncertainty, experimentation, possibility, and hope. ⁴ As a specific kind of perceived temporality, a beginning belongs here to the realm of diachrony, of historical experience and patterns.

On the other hand, beginnings can be studied from the perspective of reversible time. For Lévi-Strauss (1963: 209), reversible time is associated with the structural analysis of langue. It is a quality of ‘mechanical’ models, not empirical reality. The concern here is neither with the nature of things, nor with the subjective perception of reality, but with the objective structures of society and culture as seen from the perspective of the (structuralist) observer. From this perspective, to say that time is reversible is to say that historical transformations can be represented in terms of purely analytic synchronic (or reversible) time in a diagram or model. This mirrors Ferdinand de Saussure’s (1966 [1916]) insistence that language be studied as a system of signs as a complete system at any given point in time, rather than through time. ⁵ From this viewpoint, the analysis of beginnings involves postulating that the ‘primary, fundamental phenomenon’ is the beginning itself, ‘which gets split up into discrete operations in social life’ (Lévi-Strauss, 1987 [1950]: 47). In its most basic sense, the opposite of the beginning is the end. From the perspective of the observer, it is from this fundamental opposition that a beginning acquires its meaning. An imaginary structural model of the beginning would define this meaning synchronically, i.e. as existing outside history and being associated with reversible time. The beginning’s orientation towards the future would not be, therefore, linear and sequential as in the case of irreversible time. Instead, the future could potentially act as the cause of the beginning. It is important to stress that, as already noted, ‘time-reversibility’ is here understood as a quality not of reality, but of the model the (structuralist) observer builds to synchronically represent the discrete operations in social life involved in beginnings. In any case, for the (structuralist) observer, ‘the principle underlying a classification can never be postulated in advance. It can only be discovered a posteriori’ (Lévi-Strauss, 1966: 58). Beginnings, then, can only be known retrospectively, i.e. after their occurrence in the irreversible time of social participants. This is Lévi-Strauss’ greatest contribution and, at the same time, greatest challenge to the study of beginnings. Whereas his distinction between irreversible and reversible time clarifies the terms in which one can think about beginnings, his exploration of Saussurean structuralism seems to rule out the possibility of beginning with a wholly new start.

Beginnings as the first dot of the time’s arrow

This is the starting point of Edward W. Said’s Beginnings: Intention and Method (1975: 29), one of the most detailed and systematic attempts to study beginnings from the perspective of social participants. Despite originating from different disciplinary backgrounds, with different kinds of beginnings as their subject matter, all these attempts share one common aim. This aim is to study beginnings through an exploration of their irreducibly subjective and culturally contingent nature. In the case of Said, who explores beginnings as a ‘writer’ and literary ‘critic’, the polemic with Lévi-Strauss begins where Foucault had left of. With Said, beginnings become the catalyst for moving us beyond the strictures of structuralism towards post-structuralism.

Said (1975) begins by circumscribing beginnings to the act of writing, and defines them as ‘the first step in the intentional production of meaning’ (p.25). Said is here following a long tradition of critical reflection about beginnings as a literary phenomenon, which can be traced back to Aristotle’s (1902, Part 7, par. 2) observation in the Poetics that ‘a beginning is that which does not itself follow anything by causal necessity, but after which something naturally is or comes to be’ (31). Yet Said is intent on showing that establishing the precise point where the narrative begins is much more problematic than what Aristotle’s definition suggests. His first move, despite his emphasis on the intentional agency of the writer, is to adopt a highly impersonal tone and carefully avoiding biographical reflexivity. While some find in this evidence of a ‘phenomenology of beginnings that is not particularly goal-oriented’ (JanMohamed, 1992: 108), i.e. that reiterates rather than questions the Western culture where it is located, this also enables Said to locate his discussion of beginnings in the broader intellectual context of the critical re-examination of the structuralist paradigm that attracted a good deal of attention in the late 1960s, early 1970s. Said endorses the then dominant anti-theological critique of origins (Barthes, 1995 [1967]: 128; Derrida, 1978 [1967]: 292; Foucault, 2005 [1966]: 358–365) as a means to distance himself from Lévi-Strauss’ once almighty and now fading structuralist project – the assumption that we are written by, structured by, the discourses that surround us. ⁶ Whereas an origin is ‘theological’ and ‘centrally dominates what derives from it, the beginning (especially the modern beginning) encourages nonlinear development’ (Said, 1975: 372–373), and comes to represent reversal and discontinuity. With the stroke of a pen, beginnings become emblematic of an intellectual movement that affirms itself as poststructuralist.

This general positioning frames Said’s distinction, reminiscent of Janus’ double-headed figure, between two dimensions or aspects of beginnings. One aspect has been the object of the phenomenological reduction of Edmund Husserl, ‘the perpetual Anfänger (beginner)’ (Said, 1975: 48). Primarily conceptual in nature, Said (1975) labels this aspect of beginning ‘intransitive’ (p.77). Intransitive and conceptual beginnings are ‘very much a creature of the mind, very much a bristling paradox’: a fiction, they refer to the fact that we need to assume a beginning to begin (Said, 1975: 50, 77). Few have captured this paradox better than George Eliot (2014 [1876]), who opens Daniel Deronda with the observation that: ‘Men can do nothing without the make-believe of a beginning’ (p.1). Not only paradoxical, intransitive beginnings are also tautological, perpetually at the beginning, perpetually caught in the ‘circuit of beginnings about to begin’ (Said, 1975: 77). The Greeks, of course, knew this all too well. The myth of Sisyphus, who the gods condemned to ceaselessly roll a rock to the top of a mountain, whence the stone would fall back of its own weight, is not only, as Camus (1989 [1942]) stresses, about the absurdity of futile labour: it is also about the monstrosity of being imprisoned in an eternal beginning. Associated with silence, this aspect glances backward at what is perpetually lost and inwards, with ‘no object but its own clarification’: in short, locked outside language they represent ‘the Word’, as opposed to ‘words in language’ (Said, 1975: 73).

This other, ‘transitive’ aspect of beginnings is associated with the intentional activity of the individual mind and the irreversible time of social participants. Although forward-looking, this aspect of beginnings involves the writer to engage in ‘an activity which ultimately implies return and repetition rather than simple linear accomplishment’ (Said, 1975: xiii). In other words, when one begins to write, one continuously begins again and again. This aspect of beginnings leads to the project being realized, aimed at an end however distant and undetermined it may be. It denotes the constructive quality of language and is ‘suited for work, for polemic, for discovery’ (Said, 1975: 76). Beginning to write, again and again, is fundamentally to produce a series of displacements. In this specific sense, to begin is to make or produce difference. It is about beginning something wholly other than what was. Likewise, literature should be thought of not as a sequential relationship of one work to previous works but as an adjacent relationship, discontinuous with what comes before (Said, 1975: 10–12). This crucial feature is characteristic of the ‘problem-or-project-directed beginning’ (Said, 1975: 50). In short, for Said (1975), to begin is to make or produce difference, difference that is the result of combining ‘the already-familiar with the fertile novelty of human work in language’ (p. xiii).

Behind Said’s work, Husserl’s phenomenology is also the starting point for the sociology of Alfred Schutz. Yet, by combining Husserl’s philosophy with insights from Max Weber and Henri Bergson, Schutz offers a superior account of transitive beginnings to Said’s. Schutz’s phenomenological sociology is not limited to perspective of the ‘writer’ or ‘critic’; rather, it is aimed at reconstructing the perspective of the social participant as such. To this end, Schutz (1967) learns from Weber to treat action as a subjective and sociological category. This motivates him, following Husserl, to build a phenomenology of human consciousness; contrary to Husserl, however, Schutz (1959) sets out to erect this phenomenological project around the notion of ‘intersubjectivity’ so that the social sciences could benefit from it.This move leads Schutz to revisit Henry Bergson’s (1965) treatment of time (see also Bourdieu, 1977 [1972]: 105). More than any other modern thinker before him, Bergson’s temporal-psychic ontology emphasizes time-reversibility as the defining characteristic of the flowing time of introspective experience – the durée (duration). In durational time, as linearity is potentially turned around, the future does not necessarily follow the present. This is because, according to Bergson, the introspective experience of the future involves the convergence of a number of different temporalities within a certain rhythm.This is how Bergson (1965) describes the durée: ‘the flowing of the water, the gliding of the boat, or the flight of a bird, the uninterrupted murmur of our deep life, are for us three different things or a single one, at will’ (1965: 52 – our emphasis). Yet there is a fundamental limitation to Bergson’s concept of duration: it restricts the analysis of durational time to the perspective of participants. If time-reversibility is merely a defining feature of subjective experience, how is sociology supposed to develop an objective account of social time that is not simply linear and sequential?

Schutz’s answer to this (crucial) question is as simple as it is radical. Social time, even in the case of our innermost subjective experience of it, is thoroughly intersubjective (Schutz, 1967). Crucially for the purposes of this paper, Schutz’s account of social time is built around the idea of transitive beginnings. We continuously begin to make sense of the world, and of our place in the world, by leaving the durée and its irreversible time; this interrupts the flow of introspective experience and allows for reflective thinking. Contrary to Bergson, who rules out the possibility of durational time and spatial time intersecting, Schutz invests heavily in showing that out of the fusion of these two temporalities emerges a flux that he labels ‘the lived present’. Schutz’s ‘lived present’ is a continuous flux of experience composed of a series of transitive beginnings, looking backward to the past as one moves forward in spatial time. Meaning resides in the ‘lived present’, Schutz argues, as it is in the present that one can look back into the past or into the future. This is why the ‘undivided total self’ can only exist in the present, for ‘if the self in a reflective attitude turns back to the working acts performed and looks at them modo praeterito this unity goes to pieces. The self which performed the past acts’, Schutz explains, ‘is no longer the undivided total self, but rather a partial self, the performer of this particular act that refers to a system of correlated acts to which it belongs’ (1945: 538). This partial self, whose resemblance with G.H. Mead’s (2011: 9) concept of the ‘Me’ is readily recognized by Schutz (1945: 538), is the self of transitive beginnings. The undivided total self, by contrast, presupposes simultaneity and intersubjectivity. It emerges out of the simultaneity of the flux of consciousness of the other towards the self and the flux of consciousness of the self towards itself, a fusion of fluxes out of which emerges the ‘us’, a thoroughly intersubjective achievement. This undivided total self, again akin to Mead’s ‘social self’, is the self of beginnings as such. From the perspective of the complete social self, a beginning never really is only the first dot of the time’s arrow. A beginning begins at the first dot of the time’s arrow but, as a situation with a certain duration, its end can occur some time later. Where beginnings begin and end depends on the interpretation made by social participants: beginnings depend on how a situation is defined by an individual or a group as the starting point or early stage of a given real-world practice or institution.

Schutz’s phenomenological insights mark the high point of the sociological theorizing on beginnings and social time from the perspective of participants. A case in point is Emirbayer and Mische’s (1998: 977) work on projectivity. This line of inquiry has led to two distinct developments in the contemporary social sciences. One has extended Schutz’s work by means of longitudinal, large-n studies. Transitive beginnings are here analysed not from the perspective of the individual social participant, but from the perspective of the collective social participant. The best example is Flaherty’s (2011) sociology of time, whose analysis of how agents ‘work’ time (e.g. how they ‘steal time’, ‘make time’, and so on) clarifies the extent to which beginnings can be viewed as a kind of ‘social work’ or activity in which time itself is ‘worked’ very much as any other socio-cultural resource. ⁷ The other development has been spearheaded by anthropologists, who have been conducting a growing number of ethnographic studies dealing with how social agents deal with the unknown future, namely the projective or anticipatory dimension of these agents’ action. These studies show how, even from the perspective of participants, time-reversibility is a quality of beginnings.As social participants start a new project, the future emerges in the form of anticipations they make in the present (Hage, 2003; Simone, 2006); how Bergson’s concept of durational time can help illuminate how social agents start new life projects by pre-figuring the end-point as a likely failure, then turning inwards their future anticipations through a series of internal reversals (Nielsen, 2011); how to start something new involves hopes and aspirations regarding the redefinition of one’s individual or collective existence (Appadurai, 2004; Miyazaki, 2004, 2006); how to begin entails activating the imaginative potentials of unknown futures (Crapanzano, 2003, 2004, 2006); how the fictional nature of beginnings is integral to the ways in which political events, such as constituent moments, affect subjective capacities for future orientation (Bourdieu, 2000; Guyer, 2007; Vigh, 2006); and how to begin has a specific materiality, involving the manipulation of physical artefacts around ‘anticipatory infrastructures’ that represent the materialization of desired futures (Miller, 2005; Nielsen, 2008; Thrift, 2005). As important and significant as these studies are, the fact remains that they remain resolutely in the domain of the perspective of the social participant. To see the extent to which intersecting temporalities may unfold beyond the participant’s perspective, one needs to move beyond their scope towards the analysis of beginnings as a philosophical or conceptual problem.

Beginnings from the perspective of the observer

Perhaps the most successful of these philosophical attempts is found in Hannah Arendt, whose frequent reflections on ‘the human capacity to start something new’ granted her the epithet of the preeminent ‘theorist of beginnings’ (Canovan, 1998: vii). Beginnings are here approached from the perspective of an observer who identified herself as a ‘political theorist’ and, perhaps even more accurately, as ‘something of a phenomenologist’ (Walsh, 2008: 361). Yet Arendt’s aim is not to merely reconstruct beginnings as a lived experience, as in Schutz. Rather, Arendt’s goal is to reconstruct beginning as a universal and fundamental category, i.e. as a conceptual, intransitive beginning.

Arendt’s reconstruction of beginnings is the linchpin of her version of Existenzphilosophie of the shared world of human institutions and interactions. Conceptually, it is articulated around the notion of ‘natality’, ‘man’s pre-eminent experience’ since ‘in possessing a principle of beginning, he is able to give meaning to his birth’ (Bowen-Moore, 1989: 23). This is not, therefore, an epistemological project aimed at providing explanations of the causes and consequences of social and political phenomena. On the contrary, Arendt operates in a resolutely ontological level. Her interest resides less in providing yet another classificatory typology of human activity than in developing a critical inquiry of the basic structure of human existence. In The Human Condition, a treatise of phenomenological philosophy first published in 1958, Arendt distinguishes vita activa (‘active life’) from the realm of ideas and thought. Human active life is structured, according to Arendt, around three basic elements: ‘labor’, ‘work’, and ‘action’. Depicted in a hierarchical, temporal relation to each other, where the least dignified instantaneous activity oriented to mere survival is contrasted to the fabrication of a world of lasting entities and, more importantly, to the most elevated and most human activity of all, the immortal capacity to speak in concert with others, each one of the three basic elements of vita activa is equally defined by our capacity or ability to create something new. Yet the hierarchical ordering of these trans-historical ⁸ dimensions reflects a similar valuation of the normative worth of each kind of beginning: an immortal speech, whose message reverberates anew as each new generation is confronted with it, remains for Arendt the ultimate beginning.

The relationship between these universal categories, however, has changed as a result of the shift towards modernity. Arendt’s thesis of the ‘rise of the social’, which points to the gradual colonization by labour of the spheres of work and action, has proven immensely influential, despite the apparent contradiction between the purportedly ‘universal’ character of these categories and her analysis of their historical change (but see Walsh, 2008: 347). In her view, the ‘rise of the social’ represents a major existential threat to humankind. The valorization of ‘labor’ means a saturation of our societies with consumer goods with no lasting value at both the expense of the ‘fabrication’ (work) of goods with permanent value and, even more importantly, at the expense of human freedom: as the ontological orientation towards ‘action’ diminishes and the public sphere decays, maxime in totalitarian regimes, our ability to begin with a wholly new start fades in tandem with them. This is also the reason behind Arendt’s deep suspicion of the social sciences and, in particular, of sociology (Walsh, 2015: 43). Sociology, given its (merely epistemological) foundations, is blind to the true (ontological) nature of the relative valorization of labour. Sociologists, rather than viewing the rise of the animal laborans as an existential threat, tend to reduce it to a mere historical ‘process’ that is also the necessary precondition for its development as a scientific discipline. This also makes social scientists, who adopt the immediate temporality of labour-oriented society as their own, fundamentally ill-equipped to appreciate the eternal as the quintessential social time: their explanations are typically ‘continuist’, focusing on typologies, processes, and causal relations, rather than on origins and discontinuities. This makes them, in other words, unable to appreciate the true nature of human freedom, which involves the ‘breaking with automatic processes, interrupting routines and commencing a train of fresh events’ (Baehr, 2002; Baehr and Walsh, 2017: 17). For Arendt, then, the freedom to start something wholly new gives human beings the extraordinary capacity to reverse the rigid linearity of the time’s arrow.

Arendt’s interest in discontinuities, of course, reflects the much broader tendency towards ‘caesurism’ (Martins, 1974: 280) that characterized social theory in the 1960s and 1970s, a development that accompanied the societal demise of grand narratives (e.g. modernist ideologies) at the time. ⁹ This interest in rupture and concomitant rejection of evolutionist and progressivist approaches to history, epitomized by Foucault’s epistemological break and Kuhn’s paradigm shift, comes, in the specific case of Arendt, associated with an interest in not only beginnings but origins. This is because, in her view, to oppose the reductionism entailed by modernist forms of historicism – linear or cyclical, evolutionist or Marxist – meant the rehabilitation of classical ways of thinking and being. Arendt’s interest in the origins of Western civilization was anything but antiquarian; instead, her critique of modernity is particularly acute in that it mobilizes the immortal wisdom of the classics to castigate present-day social and political arrangements. Consider, for instance, her historical exploration of the origins of totalitarianism (Arendt, 1994 [1951]). More than a genealogy, Arendt offers her readers a study into the essence of totalitarianism, whose historically unprecedented character only comes to light when contrasted with the timeless nature of human freedom. There is a price to pay for Arendt’s ontological exploration of origins, however. Time-reversibility comes here associated with a quasi-mythological quest for the origins of human existence. This raises several problems. As already noted, the tension between the universal and the historical character of the elemental categories of labour, work, and action is never really resolved in Arendt’s philosophy. ¹⁰ Likewise, the conceptual distinction between origins and beginnings is blurred at best (e.g. Kang, 2013: 141–142). Even more problematic, at least for our purposes here, is Arendt’s profession of the sacredness of beginnings.

Beginnings are sacred in the sense that the disclosure of the origin of a phenomenon is believed to reveal its essence. This disclosure is not gradual and process-like. It is a radically contingent event, a rupture in the time–space continuum. Arendt here follows Heidegger in the belief that the task of reinvigorating the Western tradition entails a return to the origins of Greek thought. Likewise, both Arendt and Heidegger refuse to account for these ‘origins’ according to conventional historiographical methodology (Kang, 2013: 148). Consider the case of democracy. ‘Not historically, of course, but speaking metaphorically and theoretically’, Arendt tells her readers in The Human Condition, one can deduce the essential meaning of democracy from the lived experience of a few fifth-century Athenians’ experience of ‘appearing before fellow men’, a reality that ‘needed Homer and “others of his craft” in order to be presented to those who were not there’ (1998 [1958]: 198). But somehow Arendt, who, like her readers, ‘were not there’, is able to see democracy being thrown into the world through a momentous event: Pericles’ funeral oration to the citizens of Athens. Instead of seeing in the iterative nature of this speech, first narrated in Thucydides’ History of the Peloponnesian War and countless times since then, the true intersubjective source of its power as a democratic symbol, Arendt (1998 [1958]: 197) gives it a quasi-sacred status. Not the recounting of Pericles’ words, but the speech itself is seen as an epoch-making event, i.e. the meaning of democracy is not inherently relational, emerging out of its endless historical iterations. Rather, Arendt’s (1998 [1958]) phenomenology of beginning operates with a resolutely substantialist mode of thought according to which, through its ‘performance’, an unprecedented way of doing politics – ‘the organization of the people as it arises out of acting and speaking together’ (p.198) – burst into existence.

No less problematic is the fact that Arendt falls prey to the fallacy of conflating the perspective of social participants with the perspective of the observer. Her philosophical reconstruction of beginning, by suggesting that the origins of democracy can be traced back to a primordial first point located both in the time’s arrow and in geographical space, follows rather than questions the time irreversibility of the perspective of the social participant. Arendt is not the first to incur in this error. A similar error can be found in Lévi-Strauss, for whom the fundamental structures of the human mind could be ultimately shown to function according to the same logic underlying structural models. In The Constitution of Society, one of the most influential social ontologies of the second half of the 20th century, Giddens incurs in the same mistake. ‘The durée of daily life’, Giddens (1984) writes, ‘operates in something akin to what Lévi-Strauss calls “reversible time”’ (p.35; for a development, see Sewell, 2005). A possible reason why such different authors end up ‘providing a mythological solution to a mythological problem’, to use Bourdieu’s (1990: 5) apt phrase, seems to be the ontological assumptions they all share. Bourdieu’s (1990) alternative consists in distinguishing the time of science from the time of action in methodological grounds (Lizardo, 2010: 667), as to avoid incurring in the error of

those who play on the inevitable ambiguity of a learned discourse which borrows from religious experience the words used to describe that experience, in order to produce the appearances of sympathetic participation and enthusiastic proximity and to use the exaltation of primitive mysteries as the pretext for a regressive, irrationalist cult of origins. (5)

As hinted above, Bourdieu’s (methodological) structuralism has remarkably little to say about ruptures, discontinuities, and revolutions. Bourdieu’s ‘actors are in a state of continuous adaptation ( … ) to their external environments’ (Alexander, 1995: 135–136, our emphasis). The ‘increasing historicism’ of Bourdieu’s ‘research and conceptual apparatus’ (Steinmetz, 2018: 605) culminates in his elaboration of the concept of habitus. This concept is virtually impossible to conciliate with the idea of beginning. For instance, it makes little sense to ask whether ‘habitus’ – that system of ‘durable, transposable dispositions, structured structures predisposed to function as structuring structures’, i.e. a principle responsible for ‘the generation and structuring of practices and representations’ (Bourdieu, 1977 [1972]: 72; see also Bourdieu, 1968, 1990: 53) – begins. Even though habitus is introduced as to enable Bourdieu (1985) to ‘break away with the structuralist paradigm without falling back into the old philosophy of the subject or of consciousness’ (p.13), it ends up mirroring Bourdieu’s fundamentally continuist understanding of statistical patterns (‘practice’), which, in turn, is projected back into the practices themselves. Habitus does not exist in the time of science, but in the time of action, as it were. As a ‘metaphor of the world of objects’ (Bourdieu, 1990: 77), it mimics the irreversible time of social structure; it does not provide an independent synchronic representation of it. As a result, despite Bourdieu’s (1990) best efforts in avoiding structuralism’s ontological claims, he ends up offering yet another ontology of the social in the form of an endless and circular account of ‘regulated improvisations’ (p.57) that renders starting something wholly new utterly unthinkable. In a word: there is no beginning to habitus. At most, one could perhaps venture to say that habitus begins as socialization begins, or that to begin again is always a function of changes in habitus’ ‘relation to a field’ (Bourdieu, 1988: 8). But this is to effectively elude the question of beginnings. Ultimately, Bourdieu’s theory of social reproduction leaves no analytical space for the beginning of practice.

And yet practice begins, time and again. New events keep succeeding in the ‘clothesline’ of Cartesian time, to use John Levi Martin’s (2018) expression. Yet even Martin’s (2009) ‘analytic’ structuralism, the latest chapter for the project of a specifically ‘sociological’ structuralism, is insufficient to prevent him from reducing ‘events’ to the ‘observable aspects’ or ‘surfaces’ of structures (see also Martin, 2017, Martin, 2018: 197), however inherently open to change structures are construed. As in most other sociological approaches, beginning is a category absent from Martin’s conceptual apparatus. As a result, he is left with no alternative but to resort to the concept of ‘origin’ to account for the historical emergence of institutions such as the state, the army, and political parties in the United States (Martin, 2009: 189–320). Despite its many qualities, Martin’s (2011) sociological structuralism remains fundamentally an ‘analytical historical sociology’ (Lizardo, 2010: 10) that accounts for present outcomes in terms of the reproduction of the past, with little or no reference to how the future helps define the present and the past. In a crucial sense, then, Martin is putting us today before the exact same difficulty that Lévi-Strauss posed to Said, Foucault, and Barthes half a century ago. Giddens’ and Bourdieu’s subsequent attempts to resolve the antinomies of structuralism by focusing on practice, although positive, did little solve the problem as their theories remain conspicuously silent about the possibility of starting something really new.

Conclusion

This article suggests sociology should adopt Janus, the god of beginnings, as its icon. This would mean embrace Janus’ challenge and incorporate in its intellectual labour both the continuity with the past and the disruption brought about our images of the future. The formulation of a sociological concept of beginning is the first step in this direction. Prior to this, however, let us briefly return to Martin’s original source of inspiration – Lévi-Strauss’ structuralism. The point we wish to raise concerns the nature of his distinction between the reversible time of ‘mechanical’ models and the irreversible time of ‘statistical’ models. For Lévi-Strauss, this is first and foremost a methodological distinction. Yet he was also keenly aware that underlying this methodological distinction laid a deeper epistemological distinction: a naturalist understanding that construes models as mirroring the ‘natural’ divisions of reality. Likewise, Martin’s (2009) analytic structuralism is founded upon a naturalist distinction between model and reality: Martin operates under the assumption that there is a representational fit between the two so that reality (say, the US army) is congruent with the model (a ‘semi-ordered tree’) (p.16, 232 ff.). Crucially for our purposes here, Lévi-Strauss was also aware of the extent to which this epistemological distinction between model and reality was a porous one. In fact, he recognizes not only the co-existence of (conscious or unconscious) cognitive schemes used by participants in their social activities and the structural models of the analyst, but goes on to suggest that there is no reason to suppose that the latter are necessarily superior to the former (Lévi-Strauss, 1963: 282). Yet this passing recognition was as far as he was willing to go in the exploration of the porous relations – indeed, constitutive interplay – between analytic models and observable reality. Lévi-Strauss’ commitment to Saussurean structuralism meant that his efforts remained epistemologically focused on langue, and, methodologically speaking, on the development of ‘mechanical’ models and their characteristic synchronic, reversible time.

If we are to instead explore the ways in which langue and parole help constitute each other we need, first of all, to leave behind naturalist epistemology and the model of representation adopted by the structuralists – the so-called congruence model – in favour of a strongly constructivist one (see e.g. Vieira, 2017). A constructivist model of representation enables us to conceive of beginnings in ways that the more rigid congruence model does not. According to the former, social and political reality has no meaningful ‘natural’ existence apart from the act of representation, which is construed as constitutive of human experience. For instance, political preferences do not exist prior and independently of the process of political representation. Assuming this is true it follows that a beginning from the perspective of the irreversible time of social participants (reality) exists only insofar it is represented in terms of a certain model by either participants themselves or social analysts, some of which involve time-reversibility as a key feature.

A second, subsequent move involves recuperating Walter Benjamin’s (1969) idea that social time is not ‘homogeneous, empty time’ (p.261) but heterogeneous time, i.e. the unsettling possibility that linearity – both for observers and participants – can be potentially turned around. Anthropologists have recently been exploring Benjamin’s insight by reference to objects, whose ‘patine’, for example, has been found to induce non-linear temporal associations in social participants (Dawdy, 2016: 25). However, to restrict the principle of heterogeneous time to the perspective of social participants fails to do justice to the full implications of Benjamin’s insight, for whom it extends to the level of ‘explanations’ as well. As Arendt writes in the introduction to Benjamin’s Illuminations, Benjamin uses the figure of the ‘collector’ to operate a ‘strange inversion’ of the linear sequence of past, present, and future. ‘The figure of the collector, as old-fashioned as that of the flâneur’ (Arendt, 1969: 45) is not a figure of the past. An inhabitant of the present, the collector probes the past in search of original objects: ‘The genuine picture may be old, but the genuine thought is new. It is of the present’, Benjamin observes (cited in Arendt, 1969: 44). With this movement, the collector ‘destroys the context in which his object once was only part of a greater, living entity’ (Arendt, 1969: 45). Hence begins a new phase in the retrieved object’s social life: it is now a genuine and original collector’s item, whose future market value will likely increase with time. Benjamin’s principle of heterogeneous time helps us move beyond linear, continuist sociological understandings of social time. In particular, it enables one to realize that social practice and institutions begin whenever and wherever they, to paraphrase Benjamin, ‘shoot through’ a model.

Our contention is that events and institutions begin when and where social participants start to see, and act upon, the world from the perspective of a certain model. The future, no less than the past, is crucial in the definition of the model. But which model is this, exactly? It is important to insist that ‘models’ are not a ‘social structure’, at least not in the sense of a set of relations or a ‘recurring pattern of social interaction’ (Martin, 2009: 9) that is an isomorphic representation of reality. The models we have in mind here are models or logics of social action: they are the principles that help organize the contents (ideology, discourse, political mobilization, etc.) of social and political action. In this precise sense, they help define or bring about those contents. Model and reality do not exist in neatly separate domains. It is out of the dialectical relationship between contents and logic that social experience unfolds.

As Mead and Schutz have taught us, beginnings are not to be reduced to the first dot of the time’s arrow. But we should not forget that social participants’ beginnings exist insofar they shoot through a model: there is no beginning without the fiction of assuming a beginning to begin. Following Arendt’s castigation of continuist explanations and her concomitant affirmation of the human capacity to start something really new, we suggest we should not limit ourselves by structuralist explanations. These explanations, despite their internal variety, tend to either ignore beginnings’ true disruptive character and impose the same model on all practices or fail to appreciate the extent to which reality comes into being as it shoots through different models, some of which are not based upon a linear succession of past, present, and future.

Yet structuralist explanations are right in assuming that heterogeneous durations that disrupt the time’s arrow are not a merely subjective experience. Observers can study beginnings objectively. This forces sociology to finally come to terms with Janus’s challenge and incorporate the future in our explanations. However, this requires moving beyond the continuism of structuralist explanations, including Bourdieu’s one-model-fits-all solution and realize that practices shoot through a variety of models. In addition, the concept of beginning as a heterogeneous duration provides social theorists with a much-needed tool to avoid reducing events to the first step towards what lies ahead on a linear scale. On the contrary, they must come to terms with the possibility that every time a new event starts a unique duration that swells as it unfolds is created. This momentary swelling of time occurs as different temporalities intersect whose succession is logically discontinuous (Deleuze, 1988 [1966]: 37; Latour, 2005: 179). Rather than the first temporal moment in a sequence that follows each other like pearls on a string, a beginning is a separate moment whose internal relationship with other moments is not linearly sequential. As a result, the progression of past, present, and future may be analytically reversed and (assumed) effects may be revealed to be the actual causes: the beginning of a project is made possible by a certain set of factors and circumstances governed by a determinate logic, even though the latter chronologically follows the former. This does not mean, of course, that chronological linearity dissolves nor that beginnings lose their distinct qualities. We merely wish to emphasize that time folds in multiple ways, and that beginnings can be seen as one possible configuration of several different temporalities whose unfolding is not necessarily linear. At such moments, multiple futures co-exist as people enact and recreate multiple futures.