What is ‘affective infrastructure’?

Introduction

Geographers, anthropologists, media scholars, and other theorists of social space have been examining infrastructure systems of all kinds through the momentary feelings and pervasive collective moods that they generate. In doing so, several have described this knot of relations through the concepts of ‘affective infrastructure’ or ‘infrastructures of feeling’. Such a move responds to earlier theorizations of affect which took it to be ‘beyond infrastructural’ (Massumi, 2002: 45). By contrast, affective infrastructures gesture toward emotionality as one component of ‘the complex conditions for other processes, events and relations’ (Anderson, 2014: 105). Nonetheless, the use of 'affective infrastructure' emerges alongside several similar endeavors to specify how affect conditions. Affective economies attend to how affect circulates, exchanges, and binds social subjects in manners reminiscent of political economy (Ahmed, 2004). Affective architectures speak to the design and construction of space, with a focus on interiors (Adey, 2008; Micieli-Voutsinas and Person, 2020). And affective atmospheres, with their nondeterminative or ambiguous causality (Anderson, 2014: 156) and sensuous materiality (Michels, 2015) emerge within immersive or diffuse spheres or airy spaces (Anderson, 2014: 149). Does an affective perspective on infrastructure, or an infrastructural perspective on affect, provide a different vantage point from which to evaluate the way that affects condition politics and everyday life? McCormack (2017) seems to suggest as much by examining infrastructures of affective atmospheres. Yet, as Dawney (2021: 410) puts it, ‘redefining infrastructure as generative field runs the risk of expanding the concept so much that it becomes too broad to be precise’. This article thus seeks to clarify what might be the specificity, distinguishing features, and significance to the infrastructural relation of affect.

In section one, I first describe those who take ‘affective infrastructure’ to indicate the affects generated by concrete infrastructure systems (and indeed, by infrastructural conditions or relations of all kinds). In section two, I attend to how ‘affective infrastructure’ pinpoints how affects compose an element of the infrastructural relations of other formed social systems, as in the ‘affective infrastructure of political organization’. Scholars are interested in affective infrastructure because it seems to indicate the potentialities and the limits that might emerge in collective life, transversally spanning received categories and teleologies that underlaid technological determinism, functionalism, anthropocentrism, economism, and various other idealisms.

Why would ‘affective infrastructure’ have appeared in this moment? In two interludes woven between the two sections outlined above, I speak to potential intellectual histories that have shaped the problem space to which the concept responds. Finally, in section three, I argue that attention to affective infrastructures is reflective of a concern with, on the one hand, a non-essentialist and historically situated thesis of technical alienation that leads us to questions of the coloniality of infrastructure, and on the other hand, political-affective organization, which leads us to questions of the infrastructures of relationality proposed by those engaged with Black and Indigenous anti-colonialism and socialism. Juxtaposing technical alienation and political affective organization might spark geographers to further elaborate how their ratio might be understood and thus composed differently.

The concept of ‘affective infrastructure’ is positioned in an in-between zone. It indicates a renewed attention to the mediation, layering, alienation, and organization of systems and subjects along certain pathways potentially aligned with or in opposition to the conditions of capitalist political economy. The concept's convincing ability to provide explanatory power – to explain both resignation and rebellion – makes it potentially significant rather than too abstract or all-encompassing. As we will see, it traverses psychoanalysis, Black studies, Marxism, feminist and queer theory, Indigenous studies, postcolonialism, and, of course, affect theory.

This analysis is not meant to replace the singularity of anyone else's account of ‘affective infrastructures’, or the constitutive differences among them. Diverse intellectual and political traditions can be divergent. Nor do I want to make any claim to either a fully comprehensive review of the extremely extensive interdisciplinary literature on affect theory or infrastructure studies; this is a selective account. One way of understanding this practice is as an attempt to read different accounts of ‘affective infrastructure’ as neither fully individuated, nor related solely through the internal lineage of theorists they draw upon, but rather as construction of what Spinoza calls a ‘common notion’. These would be concepts that tend to more adequately encapsulate the sympathies among collective bodies, who by using their minds would then be able to mobilize a common notion to diagnose more precisely how they are determined to act and thus potentially act more forcefully together (Casarino, 2008: 32; Ruddick, 2010: 29–30). The ‘test’ of this concept – like any – is whether it gathers an active collective. An effective common notion would increase our powers to more adequately understand the ways we are determined, and thus potentially act in ways that increase or aid the capacities of our collectives. It might not work, or it might work to different degrees for different readers. It is still indispensable to continue our situated empirical and historical analyses of concrete infrastructure systems and the social forms which carry them, spring up around them, administer or govern them, or attempt to repair or stop them from functioning. Nonetheless, a philosophical elaboration through a common notion could clarify the patterns and processes which geographers and other scholars are subject to in diverse situations around the world.

In this article, I largely use the Spinozist understanding of affects as indexes of the transformation of capacities to affect and be affected, and the ideas of those capacities, each of which traverses multiply situated bodies, histories, and scales (Spinoza, 1985: 493 [EIII D3]; Jaquet, 2018). This follows Pile (2010), Anderson (2014: 11–12, 79–93), and Read (2016: 111–12) in pushing back on understanding affect as solely an ontologically distinct realm inaccessible to consciousness. I reject a strong and fast distinction between affect and ‘named’ emotions, while still understanding affect to be more pre-individual and named emotions more individuated. Infrastructures can be understood as a certain kind of conditioning, principally though not exclusively technological, in which a constellation of material relations and flows facilitates some other social and/or material activity or set of activities (Ballestero, 2018; Larkin, 2013). ‘Infrastructure’ refers less to a category of ‘things’ than a particular style of intensive relation or conditioning exemplified by transindividual relations. Such a definition allows the wiggle room to encompass relations among ‘inorganically organized’ concrete infrastructure systems (Ash, 2015: 84) such as roads, energy grids, water pipes, or media networks, as well as the unsettling situations in which people, animals, land, bodies, or ecosystem services function ‘as infrastructures’ for other processes (e.g. Andueza et al., 2021; Barua, 2021; Carse, 2012; Simone, 2004).

The affects of infrastructure

Undoubtedly one of the most exciting scholarly research arcs of recent years has been the examination of the quite-broad range of affects produced by infrastructure systems. In a well-cited intervention both indicative and productive of this groundswell, Larkin (2013: 336) makes the argument that ‘the aesthetic dimension of infrastructure (rather than its technical one)’ is the ‘dominant’ mode through which infrastructure appears. Though a strong claim, it spurred a proliferation of evidence that demonstrates how infrastructural systems of all types are involved in producing a full range of complex affective responses (Harvey and Knox, 2012). ‘Infrastructures excite affects and sentiment’ (Appel et al., 2018: 26) – joyful, sad, angry, and more ambivalent. The production of infrastructure hinges on a kind of ‘promise’ of delivery, whether such futural discourse or the technical system itself is developmental, imaginary, uncertain, or unfulfilled (Carse and Kneas, 2019). Further, the felt affects of infrastructure are shaped by spatial histories, social positionings, and claims ‘about what is properly human’ (Henderson, 2008: 29). The specific temporalities of infrastructure systems – expectation, repetition, decay, repair, maintenance, and everydayness – are generative of specific kinds of affects, such as despair, frustration, or hope (Ramakrishnan et al., 2021a). Might we understand these as not mere bodily affections, but affects that index transformations in power?

The concept of ‘affective infrastructure’ can be understood as a response to calls to better account for ‘how specific emotions are spatialized and (re)produced across time’ (Henderson, 2008: 28). Take, for instance, a (post)colonial public hospital in Papua New Guinea. Though it might be encountered simply as a set of buildings in a contemporaneous present, it is an accumulation of buildings and wings constructed with different social arrangements or purposes in mind. How one feels walking through the wards is a product of historical intentions, yet also yields unintended consequences based on personal knowledge and perception, the interpretations of a guide, and/or one's social position in relation to coloniality. Street (2012: 45) argues that it is precisely this layered aspect of this space which accentuates the ‘ambivalent experiential quality’ of such infrastructures. For workers, the sick, and the scholar, the Mandang Hospital's spatial elements beget both hope in public provisioning and resignation to failure. Rather than argue that hospital spaces were engineered or designed to produce certain affects (among different, segregated populations), Street shows how the layering of different planning objectives and experiences of hospital space generate affective ambivalence. Importantly, ambivalences are further felt differently by workers, businesspeople, expats, and patients. Infrastructures and their histories which are generative of encounters also are generative of emotion, which can bind or separate, and increase or decrease powers.

The Iquitos-Nauta Peruvian road was seemingly designed with the single purpose to connect vehicular transportation between two places, thus changing the speed and rates of connection between them with all the promised benefits. And yet this promise is unfulfilled, as construction proceeds with ‘a series of stops and starts’ (Knox, 2017: 371). The road begins to serve new purposes for different people. Frustrated by the lack of promised progress, locals walk for hours in order to reach and protest regional government officials to finish the construction. Knox reframes the politics of this road by starting from people's messy relations with infrastructures and how they might engender political action. ‘[E]mbodied, affective engagements with infrastructure’, Knox surmises, offer a ‘better starting point’ than theories of the ‘familiar grid of power relationships between the state, the market, the public, and the corporation’ (2017: 368, 368–69). This is a tough normative claim, but the empirical account is convincing: pursuing a project which ethnographically attends to the constellation of bodies, spaces, and materials that form sets of expectations for transformation and failure allows compelling and complicated stories of how political affect emerges in between bodies and capacities, a method sometimes shorthanded by more abstract historical categories.

Though both of these articles use ‘affective infrastructure’ in their titles, these authors do not necessarily mobilize it as a specific concept. And yet, these two articles have productively generated a series of cross-disciplinary conversations that do work to develop ‘affective infrastructure’. Despite differences of theoretical tradition, in these works ‘affective infrastructure’ operates as a hinge or mediator. Knox (2017: 375) highlights the ability of affect to designate a ‘concatenation of forces that ebb and flow and manifest in and between bodies’ and the temporalities of concrete infrastructures which explain ‘how these embodied relationships become related to politics’. It is the capacity for, say, potholes to ‘rupture’ otherwise-smooth expectations or intentions that creates a gap for another form of politics to emerge (see also Dawney, 2013). Street (2012: 54) argues that because a whole range of emotions ‘reside in the landscape itself’, they can become more contestable or uncertain than transitively engineered. The actualization of multiple – sometimes contradictory – feelings demonstrates that a ‘surplus’ of potentiality (Read, 2016: 166) exists in the relationship between concrete infrastructural layerings and different groups of people.

The ‘engineering’ of a specific human emotion is certainly not always a conscious or primary goal of the operator or designer of a given infrastructure system. Concrete infrastructures are often built to facilitate the movement of some other material, which itself may be organic or inorganic. Yet some concrete infrastructures – megaprojects and media systems, for example – are explicitly designed with emotional spectacle in mind. Urban rearrangements are necessary to produce early participatory media like MTV's glass studio in Times Square as well as failed precursors like the quaint Warner-Amex QUBE system, a 1977 interactive television device (Rosati, 2019). Taking these as ‘infrastructures of feeling’ means considering how they are constructed as ‘the spatial and material groundwork’ or the ‘concrete conditions’ of ‘social feelings’ (Rosati, 2019: 252, 253). The political upshot of this insight, Rosati (2019: 256) argues, entails an understanding of the necessity to remake the city if we are to remake ourselves ‘beyond the autocratic powers of private ownership’.

How might we understand the connection between the instrumental purpose and affective effects of infrastructures? Appel et al. (2018: 19) suggest:

Like the building of infrastructure itself, emotional and affective investments, too, are not forged once and for all, but take time to be formed, and can change from positive to negative over a period of time. In other words, the affective relationship between people and infrastructure, while being shaped by notions of futurity, is not always positive and may instead result in deferral, ruination, suspension, abandonment, and repurposing.

As in the examples of hospitals and roads, expectations are differentially shaped by social and historical relationships. But in what way are ‘emotional and affective investments’ like ‘the building of infrastructure itself’? Is it simply that they share the component of temporal instability, such that emotional and affective investments may be characterized by deferral, ruination, suspension, abandonment, and repurposing? Or is there a stronger sense beyond analogy, that because these sorts of affects emerge from a certain historical moment of technical materiality that they reproduce a different regime of affects?

Numerous empirical studies elucidate the ‘dialectical and relational processes between infrastructure and affective experiences, across intersecting axes of social difference’ (Ramakrishnan et al., 2021b: 670; see also Amin, 2014; Wilson, 2016). The proliferation indicates that there is something irrevocably compelling to social analysis that uses infrastructure as an opening for understanding social feelings. Yet there are two problems that this literature poses for the concept of affective infrastructure. First, however infrastructures are defined, they are essentially affective. As a set of relations, a concrete technical system will ‘affect and be affected’, both in its internal technical relations (Ash, 2015) and with regard to those human and more-than-human bodies and landscapes whose powers are transformed. This is especially true if our understanding of affect entails an understanding of a related body's increasing or diminishing powers. What, then, gives the specificity to the concept of ‘affective infrastructure’ that is not already given in a relational concept of affect? Second, is there an historical specificity to concrete technical systems that makes them particularly generative of a specific repertoire of affects? I will return to answer these questions in section three. First, unpacking an understanding of conditioning affects and technical mediation is necessary.

Infrastructures of affect

If the affects of infrastructure have formed an explosive area of interest, many theorists of political affect have likewise been compelled to investigate the way that affects bind or dissolve social formations or performances. This interest is already given in the Spinozist definition of affect as the partial consciousness of a body's ‘power of acting’ as transformed – ‘increased or diminished’ (Spinoza, 1985: 493 [EIIID3]). The question for these scholars is more historical-political than ontological: which sorts of conditions or encounters increase or diminish the power of which specific, collective actors? Do they have patterns, and can they be recalibrated or rearranged by political movements? Is it possible ‘to question how relations are formed and then endure despite conditions of continual change’ (Martin and Secor, 2014: 431)?

Part of the reason such an analytic concept might emerge is given in contemporary politics, which have seen a resurgence around the world of attempts to explain the rise of right-wing populist movements. Anderson and Secor note that populist performances and those who seek to apprehend and critique them rely on ‘collective affects…as causes of or conditions for’ (2022: 2) the emergence of the political right. Affect is supposed to help explain the two-way relationship in right-wing movements between unorganized mass desires and insecurities and their amplification, resonance, and capture by reactionary parties, leaders, media, and organizations. But the problem, Anderson and Secor explain, is that focusing on a given social group's emotional reaction risks drawing attention from that which conditions that reaction. In a prior interview, Anderson further considers the role of social media platforms in precipitating affect. First, he suggests that ‘Every media infrastructure is simultaneously part of an affective infrastructure’ (Kemmer et al., 2019: 32) – so far this would be consistent with the set of literature outlined above. Continuing, however, Anderson clarifies ‘I use “infrastructure” as an open term for the channels and forms that mediation takes as affective life forms and deforms—that is, the material conditions for the circulation and distribution of intensities that then fold back into the infrastructure such that the material and affective become indivisible’. Thus while from the perspective of ‘affects of infrastructure’ the ‘infrastructure’ is to be found in the interstice of the material system, here affective infrastructure names the simultaneous material-affective conditioning activity of the circulation and channeling of certain political events. ¹ As Anderson and Secor argue (2022: 4–5), it is the social media platform's capacity to foster disruptively excessive affective responses – anger, frustration, enjoyment – that heighten its efficacy for the political right.

Importantly, the concept of ‘affective infrastructure’ has also been used in analyses of radical left politics, demonstrating that the scholarly analysis is not drifting into a binary of emotion versus rationality. For Clough (2012), the practices of anarchist organizing build and are built through affinity and solidarity in their political organizing against the Republican National Convention in Minneapolis. ‘Affective structure’ is used to describe a ‘feedback loop’ between ‘affinity as feeling and organization, and the emotional content of direct action as a galvanizing moment that contributes vitality to the movement’ (Clough, 2012: 1674). Thus the event and its affects interpenetrate and (ideally, for these organizers) reverberate with each other, increasing the power to act of the groups in question. In a similar example, Bettocchi examines how the failure of concrete infrastructures in a queer squat in London generates both stress and, as an interlocutor puts it, ‘a kind of solidarity … quite intrinsic to a space’ (quoted in Bettocchi, 2022: 6). Yet in fostering an active affect of solidarity, these modes of organizing – and their spaces – also become an important target of intervention for federal and local policing. Frequent preemptive arrests and raids as well as infiltration of radical organizing spaces allow state actors to attenuate solidarity with paranoia. If ‘structure’ indicates something of political form or organization, infrastructure could be that which animates and sustains it.

One way of clarifying this difference could be to suggest that infrastructure is ‘not identical to system or structure’ if we take infrastructure to be ‘defined by the movement or patterning of social form’ (Berlant, 2016: 393). This is not an argument against form, but rather that infrastructure is the genesis of structure or form rather than the name for it. Structure, by contrast, would be ‘that which organizes transformation’ (Berlant, 2016: 394). An ‘affective infrastructure’ for Berlant has a ‘relative autonomy to structural political imaginaries’ (2016: 414). Berlant's apposite example is ‘the commons’ which, as geographers might agree (e.g. González-Hidalgo, 2020: 15), is not by nature liberatory but involves a certain ambivalent layering of potentially contradictory affects shaped by a historical-political situation. While the question of how this happens is ultimately somewhat underdeveloped in Berlant's argument, their set-up is one which seeks to orient social struggle toward not only the present but the condition for something emerging next, which is what distinguishes commoning as a social infrastructure from privatization. And while Ruez and Cockayne (2021) also highlight the important role of ambivalence (rather than indemnity) to such a project, it is important to also note that space for ambivalence is precisely what is precipitated as the result of an injunction: ‘to build affective infrastructures that admit the work of desire as the work of an aspirational ambivalence’ (Berlant, 2016: 2014). For Berlant, ambivalence is the form of organization of desire that certain affective infrastructures would allow or nurture rather than foreclose by presenting alternative fantasies.

What sorts of organizational form might be sustained by an affective infrastructure? Dean's retrospective on U.S. communist organizing in history and the present provides an analogous attention to the relationship between an affective infrastructure and political organization. Returning to crowd theory and group psychology, Dean's goal is to reconstruct how the party form creates ‘practical optimism’ among its subjects, holding open not just collective endurance but also politics as a project of moving together in a certain direction (2016: 220). Building on this argument, Kamish (2021: 4) further analyzes the way that the Black Panther Party sought to become the vanguard ‘“affective infrastructure” for sustaining revolutionary feeling and political action’. Thus the ‘affective infrastructure’ that the party or the squat layers in and through its spaces augments the collective's power of acting. It is the form which allows people to ‘become bigger, able to do more than they could have imagined’ (Dean, 2016: 209, emphasis added). Furthermore, it is through the organization of this infrastructure that such power coheres and is sustained, across time. The organizational form ‘provides an apparatus for this endurance’ (Dean, 2016: 217). ² Though in a politically divergent way from the analysis of anarchist organizing, Dean's analysis of the party as a form still resonates: the point of an organizational infrastructure is to hold open space in which the energy of formative collective affects can be sustained rather than dissipate (as in crowd phenomenon) or become re-routed by the state.

Drawing attention to such affective infrastructures leads us to consider a fuller range of organizational forms which might sustain desiring subjects than the squat, commune, or party (Nunes, 2021: 230). Among these other forms might be the ‘infrastructures of feeling’ that Gilmore (2017: 237) suggests can be found in the long history of transnational abolitionist movements. For Gilmore, ‘infrastructure of feeling’ defines how ‘ideology becomes material as do the actions that feelings enable or constrain’ (2017: 237). Gilmore is concerned with understanding how abolition geographies and the Black Radical Tradition alike might build sustaining narrative arcs toward liberation, and further embed these arcs in spatial arrangements that reproduce and grow their power. Gilmore turns to infrastructure of feeling to elucidate the ‘consciousness-foundation’ which ‘underlies our capacity to select, to recognize possibility as we select and reselect liberatory lineages’ in a deliberately capacious manner (Gilmore 2017: 237, emphasis added). The practice of abolition geography, as a kind of organized historical materialism for liberation, mobilizes a movement ‘back in time-space’ to select and reselect ancestors as a kind of emotional sustenance, a way of finding ‘alternatives to the despairing sense that so much change, in retrospect, seems only ever to have been displacement and redistribution of human sacrifice’ (228). It is an open question what forces compose that which underlies our capacities to select, but it is clear again that these include ‘feelings [that] enable or constrain’ actions (237).

If Gilmore's brief formulation is provocative, then Choi's expansion of ‘infrastructure of feeling’ demonstrates its utility in different situations. In an examination of ‘sensory and affective infrastructures’ of post-tsunami crisis in Sri Lanka, Choi makes the argument that as ‘everyday scaffolding’ (2021: 86), humor and superstition become components of social survival of ongoing socio-ecological crisis. This argument resonates with contemporary analyses of Indigenous resistance to U.S. settler colonial infrastructure systems, in which scholars posit that against the life appropriating and destroying material infrastructure systems of settler colonialism (symbolized most potently by oil pipelines), Indigenous counter-infrastructures of care, solidarity, and resistance offer alternative ‘forms of politics’ (Barney, 2021; c.f. LaDuke and Cowen, 2020; Spice, 2018). While certain forms of insecurity, paranoia, and fear are crucial supports for right-wing populisms, and the security state and policing they desire (Bosworth and Chua, 2021; Woodward, 2014), it is not the case that ‘negative’ affects cannot be part of social endurance. Choi demonstrates that in contrast to extant concepts of ‘social resilience’, infrastructures of feeling allow for a more capacious understanding of how critical affects such as anger or indignation can still form elements of endurance (see also Henderson, 2008). This perceptive point is sometimes underplayed in ‘affirmative’ readings of affect influenced by Spinoza and subsequently Deleuze (though see Stolze, 2020). The question in any case is how the partial communications or ‘concatenations of affects’ (Balibar, 1994: 27) come to orient a collective such that it begins to act such that ‘the Minds and Bodies of all would compose, as it were, one Mind and one Body’ (Spinoza, 1985: 556 [EIII P18Schol]).

Though some of these authors still include analysis of concrete infrastructure systems, what defines their attention to a second ‘side’ of affective infrastructure is a much more metaphorical, analogical, or even allegorical use of the concept of affective infrastructure to indicate, as it were, the ‘pipes of all sizes’ (Jameson, 2020: 49) that make up transindividual political psychology. Nonetheless, in surveying these some conceptual questions emerge. First, what would differentiate the concept of ‘affective infrastructure’ in these analyses from the generic concept of an ‘affective condition’? There need not be a difference, but I wager the alternative is perhaps more interesting. If affective infrastructure does indicate something more specific than condition or atmosphere, a second question arises: does ‘infrastructure’ denote simply the channeling or circulation of affect (as in one reading of ‘pipes and cables’)? Or is there something in the technological materiality of infrastructure that helps us understand political organization? I return to these questions in section three after addressing first the question of determination.

The two ‘sides’ of ‘affective infrastructures’: Technical alienation and political-affective organization

In the first section, I described how ‘affective infrastructures’ can help us understand how inorganic concrete infrastructure systems yield certain sorts of affects. In the first interlude, I also described how an earlier set of debates raised the issue of whether certain infrastructures are thought to be metaphors, models, or grounds for understanding the psyche. In the second section, I described how the concept of affective infrastructures can help us understand how feelings are one element conditioning political organizations or forms, without determining their appearance or being subsumed by their formal structure or organization in a transitive/transcendent or expressive manner. In the second interlude, I offered that the concept of conditioning or determination at stake resembles, and can potentially benefit from, a recollection of certain Marxist debates. I flagged a series of questions along the way:

What gives the specificity to the concept of ‘affective infrastructure’ that is not already given in a relational concept of affect?

A non-essentialist concept of alienation is ‘the relation of relationlessness’ (Jaeggi, 2014: 1), a deficient appropriation of one's capacities. It has long been noted that a certain concept of ideological and economic alienation bridges Spinoza and Marx (Matheron, 2020: 137; Ruddick, 2010: 24), but it is less often noted that Simondon levels a critique of the ‘compact and resistant, but alienated human reality within the industrial world’ (Simondon, 2017: 119; though see Marcuse, 1991: 23–28; Read, 2016: 104–107). Though Simondon contests these terms (2017: 254), surely one component of such alienation is the separation and subsequent encounter between means of production (including tools and later machines) and the relations of production that organize them (class relations). In becoming organized as components of a machine rather than users of a tool, the knowledge, skill, and capacity of workers ‘vanishes as an infinitesimal quantity’ into but one component of the machine (Marx, 1976: 549). Such alienation results in new structures of feeling, such as the ‘age of cynicism’ and piety characteristic of liberal humanism (Deleuze and Guattari, 1983: 225). Understanding alienation, and the Marxist theme of commodity fetishism, in this Spinozist fashion reveals that the mode of production organizes the world such that it appears the emotions we feel are due to the immediately experienced objects, rather than the socio-economic system in which they and we are arranged (Read, 2016: 121, 148; Ruddick, 2017: 125). The concept of infrastructure is given an immanent understanding, as the metastable solute out of which different sorts of affective politics are individualized or formed, the conditions not of our choosing.

What differentiates this theory of technology from a ‘flat’ ontology is not only that technical systems are taken to carry their own form of inorganic affect (Ash, 2015), but also that it is a historical form of an ontology of technics, ‘bringing history and ontology into conversation in a way that ontology cannot stand outside of history’ (Ruddick, 2020: 510). Consequently, the political dimensions and felt effects as forces of domination are crucial, particularly acute in the situation of colonial and settler colonial infrastructural systems, which carry promises of life but can be imposed through injury and death (Barney, 2021; Curley, 2021; Salamanca, 2016; Spice, 2018). Wakefield (2020) in particular highlights that making a system into an infrastructure – always an achievement – entails the availability of such relations to be apprehended, known, secured, and ultimately governed. Furthermore, foregrounding the transindividual dimension prevents any exit from this thought of alienation into a methodological individualism, which would result in the suggestion that the worker or user is alienated from their individual existence by becoming part of a collective or a machine. Instead, the style of relationship fostered by the modern capitalist form of infrastructure is one of passivity and diminished or constrained collective power, wherein both bodily and intellectual capacities are offloaded from people to the technological material.

The first specific perspective that ‘affective infrastructure’ could have (beyond ‘general conditioning’) would be to not only point us toward the transindividual dimension of infrastructure–human interaction being mediated by all sorts of affective relations (Simone, 2021), but also a diagnostic of the specific ways infrastructure bifurcates and separates people, in an uneven fashion, from their powers of acting – and from their ability to adequately perceive the ultimate causes of their diminished powers of acting. As Jordhus-Lier et al. (2022: 320) argue in an examination of oil politics in Norway and Nigeria, such ‘alienation implicates different groups of people in each other's working lives, consumption patterns and practices of social reproduction in ways that simultaneously mask and highlight interdependencies’. The authors posit this critique toward forms of accounting for oil assemblages that overemphasize instability and thus fail to explain or disclose the systemic maintenance and adaptability of oil extraction economies. Though they describe product, labor-process, human, and social alienation, the concept of technical alienation further closes the gap between Marxism and the (generally non-Marxist) assemblage geographies they critique. So too can such a project attune us to maintenance and repair as potential forms of ‘de-alienation’ as a ‘politics of reconnection’ (Jordhus-Lier et al., 2022: 332–334).

Similarly, Ramakrishnan et al. (2021a: 11) argue that ‘the temporal connects the material, affective and the social aspects of infrastructure, and so attention to temporality gives us the full weight of infrastructures in flux’. While reading this work's careful attention to precisely how the fragilities of an infrastructure system to decay and repair ‘map onto’ affective registers, I was struck by the possibility of mapping the other direction – as if we could take Eve Sedgwick's proposition of reparative reading as a feminist practice (see Ruez and Cockayne, 2021) and understand it through infrastructure. To return to a point made earlier, ‘emotional and affective investments’ might be understood to be like ‘the building of infrastructure itself’ (Appel et al., 2018: 19) because they are engaged in common spatio-temporal processes, like construction, decay, and repair. As Berlant puts it (2016: 393), repair follows from a certain attachment to a broken object/context and reflect on that attachment and how it might generate ‘a form from within brokenness beyond the exigencies of the current crisis’.

In such constrained or broken conditions, the concatenation of affects provides an excess of potential relationships through which another thinking and politics could emerge (Ruddick, 2010: 36). But a problem with quicker considerations of this ‘surplus…which opens the possibility for change’ (Read, 2016: 166) is, quite obviously, that any ‘selection’ of affects or ancestors is ‘extremely hard, extremely difficult’ (Deleuze, 1997: 145). How could those excess possibilities be actualized? The second ‘side’ of affective infrastructure speaks to how such selection could be organized to grow and sustain power (as potentia) to the extent it can be institutionalized (as potestas) without attenuating itself or producing alienation (Nunes, 2021: 129). What makes political organization like infrastructure is its ability to foster, to use a psychoanalytic term, a kind of transindividual transference (Pile et al., 2019: 3–4, Dean, 2016: 184).

For psychoanalysis, transference is the complex process by which unconscious elements actively and socially shift from one or several sets of people or objects to another people, object, or set of objects. Social organization not only shapes the affects of those who engage in it; it also helps them assume the role of social analysts themselves. Thus ‘affective infrastructures’ can be understood as the specific social and spatial settings, contexts, or conditions that generate affective forms (Pile et al., 2019: 5) – the necessity of ‘a mediating object’ (Guattari, 2015: 80) for transference to occur. Political organization can be understood as a practice of collectively apprehending and augmenting these mediating settings to produce social analysis and grow social power. This is a way of conceiving, as Brazilian theorist Nunes (2021: 122) argues, political organization as the concatenation of ‘socio-technical assemblages, operating at whatever scale, that take on the role of coordinating collective effort and orienting it in determinate directions’. The function of the affective infrastructure of the organization is thus to act ‘as the analyst that helps the analysand come into contact with their unconscious and interpret their desire’ at the same time as ‘this infrastructure infuses members with will, enthusiasm and hope’ that grows their power (Nunes, 2021: 228, 229). Political organization repeatedly appears as that which is explained as affective infrastructure because the ‘infrastructure’ referenced is in fact a technology of power. Again, this is not to deny the constraints of the situation still exist – as Guattari puts it in his article on transference, ‘It is, therefore, not a question for us of conceiving the alienating and de-alienating phenomena of the group as things-in-themselves, but rather as the varying sides – that would be differently expressed and developed depending on the context – of a similar institutional object’ (Guattari, 2015: 77, emphasis added).

The ‘institutional object’, settings and situations, or infrastructures, encountered by anti-colonial, socialist organizing are exemplary not simply as a case study, but as the very sources of such analysis. Alienation can be understood as a historically separating and ‘invasive’ infrastructural relation (Spice, 2018), in which infrastructural systems serve as ‘colonial beachheads, establishing the conditions for future dispossession, displacement, and marginalization’ (Curley, 2021: 388). Consequently, as the Indigenous socialist collective the Red Nation argues, ‘The critical infrastructure of Indigenous worlds is, fundamentally, about responsibility and being a good relative’ (2021: 142). Anishinaabeg political activist Winona LaDuke and geographer Deborah Cowen draw a division between the colonial form of infrastructure and a provisional, ‘slow, transformative feminist work of social re/production’ (2020: 264). Though it could appear as a quick comparative move content with focusing on how Indigenous thought has ‘resonances’ with a European thesis of relationality (as warned against by Dorries and Ruddick, 2018: 620), I want to point out that The Red Nation (2019) precisely involutes the socialist tradition of Marxism as supposedly emanating from the West, reminding us instead that the revolutionary project is fully realized in anti-colonial infrastructures of caretaking relations (see also Dunbar-Ortiz, 2016). The second ‘side’ of ‘affective infrastructures’ thus seeks to articulate, and make us conscious of, the need to stave off economic, technical, and ideological alienation by building the forms of organization that through group transference of affects, grow and sustain the organizational infrastructure to produce collective subjects who can analyze and diagnose the social situation.

Not every thinker who has picked up a concept of affective infrastructure will see themselves in either – let alone both! – of these ‘sides’. Furthermore, the heuristic division of the concept of affective infrastructure into two sides is not meant to be a separation; the sides are a ratio. This section provisionally and dialogistically elaborates the concept, pushing it toward what I see as its fullest, most consistent, and most powerful outline as a tool of analysis. But the ‘litmus test’ of any such concept is in its effects (Ruddick, 2010: 40), whether it clarifies well enough that it gathers and grows the powers of those who speak around and through it.

Conclusion

Through a series of readings, this article has sought to explicate how the concept of ‘affective infrastructure’ is currently being used by geographers and political theorists, principally concerning collective political action. I have sought to clarify two related uses of ‘affective infrastructure’ by examining the affects of infrastructure and the infrastructures of affect. I have spoken to potential prehistories of the concept in affective geography and 20th century Marxism, but these are not at all meant to be exhaustive – simply my own contribution to the selection of sources to which it might continue to speak. Along the way, I have clustered a series of other concepts around it: mediation, endurance, technical alienation, temporality and repair, and political organization. Affective infrastructure might be able to help geographers, political theorists, and those remaking political space and organization to become more adequately conscious of the elements determining their affects, thus creating new avenues for remaking them differently. The details of any given situation will matter, though; there are no guarantees. Nonetheless, to me the concept of affective infrastructure is most helpful for allowing for an understanding of determination, including but not limited to the economic sphere, which both avoids transitive ‘economism’ or an ‘anything-goes’ contingency.

By definition, any field of affective infrastructures is a skein, a cacophony, a battleground. The concept must indicate relational mediation or a ratio of at least two ‘sides’, rendering certain collectives to understand themselves as more or less active or passive. The two ‘sides’ of the problem are not meant to be understood in opposition. Indeed, within the broader Spinozist framework I am riffing on, ‘actions and passions are not opposites. The crux of the problem is increasing the active part of the body in the production of affects’ (Jaquet, 2018: 105). Understanding the causes of the ratio of passions and actions could help us modify the the relations that produce it. In describing social life as such, we would begin to see affective infrastructure as offering two different vantage points on one process, thus perhaps giving us the crucial insight into how one might begin to conceive as material the ‘social existence’ that, through a complex set of infra-processes, ‘determines consciousness’ (Marx, 1992: 425).

The success of any such project depends less on finding the concrete ontology that would then ground political struggle wherever and whenever (Povinelli, 2021). Rather, it is a wager that – with the thinkers gathered in this paper – we can be more adequately historically and spatially situated – hereish and nowish – to ask the questions of infrastructure systems, transindividual affects, and political organizations to reflect upon how our powers of acting might be modified through our reflection on their composition.