Negative urbanism: unknowability, illegibility and ambivalence in the platform city

Abstract

On-demand digital platforms are shaping processes of urbanisation by transforming governance processes, worker subjectivities and consumption practices. However, claims about such transformations risk ignoring the diverse and often underspecified ways that evaluations about platform urbanism are being made. This paper grapples with our incapacities to know platform urbanism, not as pragmatic barriers that can be overcome, but as limits to be reckoned with. Reflecting on fieldwork encounters with people speaking about on-demand platforms from diverse governance, production and consumption perspectives, the paper foregrounds experiences of unknowability, illegibility and ambivalence in platform urbanism. These concepts invite a rethink of the subjectivities involved in evaluating platform urbanism and they provoke questions about the operation of power. The paper argues that attending to these ‘negatives’ provides an alternative counter-political perspective that apprehends both the instability of politics and our practices of judgement. Ultimately, admitting a more aporetic understanding of platform urbanism is not about hobbling our capacities to intervene as urban theorists, but about questioning what intervention might look like and what might be possible.

Keywords:

• digital platforms
• platform urbanism
• embodiment
• gig economy
• incapacity
• infrastructure

Introduction

The intensification of on-demand digital platforms has become a defining concern for urban research over the past decade. The rapid rise of ‘lean’ capitalist platforms for the on-demand mobility of people, goods and services—such as Uber, Deliveroo, and Airtasker—has sparked hot debate about their impact on cities from a range of critical perspectives. ‘Lean’ platforms are intermediaries that bring together consumers and workers, but where risk and labour costs are disproportionately displaced to workers themselves, rather than being the responsibility of a traditional employer (Scholz 2017). Considerable attention has rightly been devoted to critiquing how platforms are exploiting workers and intensifying the precarity that many workers already experience (Richardson 2020; Van Doorn 2017). These critiques have provoked wider questions about how on-demand platforms are transforming processes of urbanisation more broadly, by changing governance processes, worker subjectivities, and consumption practices.

Urban researchers have tackled key ontological questions about what, precisely, platforms are and how they operate; through what digital infrastructures and practices they are mobilised; and they have highlighted the kinds of control they are giving rise to (Richardson 2020; Barns 2019; Lynch 2020; Wiig and Masucci 2020; Graham 2020). However, epistemological questions about how we come to know these transformations have arguably received less attention. Since political claims about platform urbanism are a product of epistemological commitments (Kinkaid 2022), my argument is that the detailed ontological attention to ‘what platforms are’ requires a similarly sensitive epistemological attention to the processes of how we come to know the transformations that are taking place. Responding to Macrorie, Marvin, and While’s (2021) invitation to enhance our theoretical appreciation of platform urbanism, and in the spirit of pursuing the ‘uncharted futures of platform urbanism’ (Fields, Bissell, and Macrorie 2020, 465), this paper develops feminist digital geographical work (Elwood and Leszczynski 2018; Richardson 2018) to advocate for greater epistemological attention to what can be known in platform urbanism and how this knowledge is generated.

The main argument is that assured claims about what is happening in platform urbanism need to be tempered by an acknowledgement that some dimensions of platforms are not so straightforwardly comprehendible. Reflecting on fieldwork encounters with people speaking about on-demand platforms from diverse governance, production and consumption perspectives, the paper foregrounds experiences of unknowability, illegibility and ambivalence in platform urbanism. Contrasting with theories that emphasise entanglement and connection, the paper draws inspiration from cultural geographical theories of negativity to argue that separation and disconnection are an overlooked dimension of platform urbanism. This argument does not negate responsibility to call out situations of suffering and exploitation that are intensified by platforms. Rather, it draws attention to the instability of practices of judgement that are taking place, thereby contributing to our apprehension of the ‘digital in its ambivalences and complexities’ (Fazi 2019, 11). Though the substantive focus of this paper is on-demand platform urbanism, its central arguments speak to wider debates in urban research on how claims about urban transformation are made.

(Un)knowing the actually existing platform city

In popular accounts of the rise of ‘lean’ on-demand platforms there is a tendency to make assured and generalised claims about what, exactly, is currently taking place in cities. For instance, Rosenblat (2018, 5) unequivocally states that companies like Uber have ‘upset the status quo across society’ and ‘ushered in a wave of changes touching most aspects of society, be it family life or childcare arrangements, worker conditions or management practices, commuting patterns or urban planning, or racial equality campaigns and labor rights initiatives.’ Others diagnose an even broader trend of ‘Uberficiation’ happening across urban infrastructures and services—from education (Hall 2016) and finance (Baldwin 2018) to domestic work (Dattani 2021). Elliott argues that such changes are a part of a ‘technological tsunami’ (2019, 25) that characterises our current era of ‘digital revolution’ (2019, 24). Generalised descriptions of new technologies have a long history, from the accelerationism of ‘high speed society’ (Rosa 2013), to older declarations about how disruptive technologies have fundamentally altered perception through speeded up forms of transmission that have reorganised social space (Virilio 1986). Qualities of assuredness and generalisation are also present in foundational writings on the relationship between technology and urban life, from Benjamin’s (1974, 216) diagnosis of ‘a sense perception that has been changed by technology,’ to Simmel’s ([1903] 2012) infamous diagnoses of the overstimulated body, dizzied by the quickening pace of urban life. What unites these accounts are the definitive, even heroic, nature of the claims made about the changing relationship between technology and urban life.

As platforms such as Uber, Deliveroo and Airtasker have proliferated, there are undoubtably urban transformations taking place in many of the domains that Rosenblat (2018) points to. Both politically and pragmatically it is vital that we know how on-demand rideshare platforms are changing the use of public transport; how food delivery platforms are changing the viability of restaurants; as well as the knock-on impacts on the planning and design of cities. These are transformations that we have a responsibility as academics to better understand so that we can respond to the new forms of inequality that are emerging. Assured and generalised claims about platform urbanisms such as those articulated above are certainly one way of ‘untangling the mess’ (Hou and Chalana 2016) of platform urbanism, to reveal what has (thus far) been concealed from us (Felski 2015). Yet such definitive claim-making risks ignoring the diverse and often underspecified ways that evaluations about platform urbanism are being made.

Urban geographers have punctured generalised diagnoses of platform urbanism by insisting that on-demand platforms manifest and evolve in place-specific ways. Contrasting with the universalised ‘placeless’ discourses of some theorists of digital technologies, as well as the often-boosterist claims espoused by platform companies themselves, the specific political contexts of cities shape how platforms manifest locally (Sadowski and Maalsen 2020). Accordingly, like earlier political economy work on cities (Brenner and Theodore 2002), recent work has championed a turn to ‘actually existing’ platform urbanism which zooms in to evaluate the place-specific manifestations of urban platforms (Shelton and Lodato 2019). Connecting with a longer line of work on smart cities, Söderström and Mermet (2020, 6) argue that ‘actually existing smart urbanism is probably to be found first and foremost in the way platforms such as Airbnb, Uber, Deliveroo and the like are reshaping contemporary cities.’ Their contention is that generalised claims fail to account for what is actually happening ‘on the ground’ in different places. Likewise, instead of diagnosing the power of platforms in totalising ways, Van Doorn, Mos, and Bosma (2021) draw attention to processes of ‘actually existing platformization’ that platform companies engage in which they describe as a path-dependent and locally situated process in which platform companies engage in various forms of ‘boundary work’ with other actors seeking to retain and/or gain power. Rather than starting with the technologies themselves, such geographically informed work is grounded in places, ‘their specific populations, resources and problems’ (McFarlane and Söderström 2017, 313). Accompanying this focus on political context, work on ‘actually existing’ platform urbanism is concerned with how technologies are implicated in the contingency of everyday lives, experiences and bodies in situ (Rose et al. 2021; Straughan and Bissell 2021), shedding light on the diversity of ‘actually existing subjectivities’ that exist in the smart, platform city (Rose 2020).

A turn to ‘actually existing’ platforms corresponds with a shift in methods through which claims about platforms are made. This can be seen through calls for detailed ethnographic fieldwork that traces the ‘observable impact of digital platforms’ (Söderström and Mermet 2020, 3). Rather than striving to merely critique discourses spun through platform companies’ promotional materials, Cugurullo (2020, 5) explains how it is through ‘empirical research which can determine, on a case-by-case basis, what novel manifestations […] are permeating the built environment and how cities are responding to them.’ Emphasising the significance of geographical context, Pink and Sumartojo (2018, 850) call on researchers to ‘attend to the actual settings that new automated technologies will become part of, through close ethnographic analysis.’ Highlighting the need to trace specific connections and flows, Graham, Hjorth, and Lehdonvirta call for

more detailed empirical inquiry into flows of value in the sector, and further research about who creates it, who captures it, how flows are being reconfigured and who benefits from these reconfigurations, and about whether we see sustainable or dependent local linkages, knowledge spillovers, and impacts on local economies and communities. (2017, 152)

In short, urban researchers have turned to precision and specificity, not to scale up to make grand claims, but to provoke thinking about what urban platforms are and how they give rise to unique place-based constellations of power.

Yet tracing the precise manifestations of how urban platforms ‘actually exist’ in different settings presumes the possibility of knowing what ‘actually exists.’ There are three challenges to this possibility. First, and pragmatically, there are methodological difficulties of opening the ‘black box’ of platforms owing to their proprietary nature (Fields, Bissell, and Macrorie 2020; Bucher 2016). Second, and ontologically, poststructuralist urban theorists have cautioned that cities are full of unpredictable happenings owing to their overwhelming complexity. Thrift (2014, 15) signals this challenge when he acknowledges that ‘things do not always join up. If there is a vocabulary we really find hard to articulate, then it is surely this one.’ Rather than understanding cities in terms of neatly interlocking systems, such theories have encouraged a turn to the fragmentary and happenchance nature of the urban which makes it difficult to make assured, long-duration declarations about what ‘actually exists,’ given that what actually exists is in flux and thus ‘incomplete’ (Guma 2020). Third, poststructuralist theories of immanence insist that there are many entities that ‘actually exist’ but are often unaccounted for in current urban theory. These theories have expanded our objects of analysis to include ‘virtual’ immaterial entities that have a reality and exert a force but cannot be seen, such as memories, tendencies and desires (Pedwell 2017). Such theories problematise our capacity to trace the ‘observable impact of digital platforms’ (Söderström and Mermet 2020, 3), since what exists in the present has a temporal complexity. As Anderson (2021, 208) points out, ‘the “present” is never fully present; it is not a punctual, separate “now”. It is full of tendencies and latencies; traces of past and present futures which exert some kind of presence.’

These challenges have prompted urban theorists to develop new ways of accounting for what ‘actually exists.’ Rather than tracing the in-situ specificity of materials, practices and events, some have emphasised the significance of ex-situ objects that take looser forms. For instance, through ethnographic work in Jakarta, Simone (2019) recounts how participants drew his attention to how they are attuned to a ‘background’ through a vague sense of something happening ‘out there.’ In contrast to the priority often accorded to the specificity of what ‘actually exists,’ for Simone,

the meanings of details seem to fall away. The events, conditions, and things of their surrounds converge and blend into an always tentative form, a series of strange composites that then suggest different ways of thinking and feeling about the circumstances they face. (2019, 996)

This is a city that pulses to the uneasy beat of vague lures that cannot be easily pinned down, highlighting the significance of indistinct phenomena—such as hunches, rumours and hearsay—that have a reality but are potentially much harder to trace. Contrasting with assured, generalised claims about platform urbanism, this insight suggests that what ‘actually exists’ in platform urbanism is potentially more challenging to articulate than is currently acknowledged.

To recap, generalised claims made about platform urbanism have been extensively critiqued by urban geographers who insist that we need to explore their precise manifestations in different places. This turn to investigate the place-based specificity of urban platforms has encouraged methodological strategies that are designed to capture place-specific detail. Conceptual advances in urban theory have expanded the remit of what ‘actually exists’ in platform urbanism to include immaterial but real phenomenon, as well as the ex-situ forces that take vaguer forms. What unites this work that has expanded the range of objects of analysis for platform urbanism is that they are all additive. Each of these theories seek to extend our inventory of urban phenomena by multiplying the things that exist. What’s more, such an extended inventory relies on a parallel augmentation of our capacities to attune to these multiple phenomena. In response, this paper contributes a different perspective to this set of debates on what ‘actually exists’ in platform urbanism. Rather than focusing on the ‘positive’ expressions of materialities and forces in platform urbanism—positive in the sense that they that have an existence, however vague—the remainder of this paper explores an overlooked ‘negative’ in urban theory—negative in the sense of non-existence, absences and gaps—to develop understandings of platform urbanism in a different direction.

This paper grapples with our incapacities to know platform urbanism, not as pragmatic barriers that can be overcome, but as limits to be reckoned with.

Knowledge incapacities have been highlighted by digital geographers before. These include the opacity of proprietary algorithms in on-demand platforms that pose a barrier to ‘transparency’ (Barns 2019), and the prevalence of machine-to-machine communication that bypasses human consciousness altogether (Ash 2013). However, my concern here relates to our embodied capacities to make sense of platform urbanism through ethnographic fieldwork encounters. Specifically, it extends Bucher’s (2016, 86) acknowledgement that, rather than just a political or technical challenge, ‘opacity, secrecy, and invisibility’ are ‘a basic condition of human life.’ Contrasting with the ‘relational’ turn in urban theory that prioritises the ‘positive’ dimensions of connections, extensions and entanglements in the city (Robinson and Roy 2016), my argument draws succour from a minor strand of cultural geographic theory that emphasises ‘negative’ aspects of situations that are not about connection or communication but rather about irresolvable limits that cannot be transcended (Rose, Bissell, and Harrison 2021; Dekeyser and Jellis 2021). From the utter inscrutability of the future to the unreadability of other people (Rose 2018), this body of work insists that gaps, aporias and absences are an inherent dimension of worldly experience, and yet are too often papered over by our impulse to emphasise connection, encounter and relation in our accounts. These are not blind spots waiting to be revealed by a better method, but ‘ineliminable hollows that shadow all relational activity’ (Rose, Bissell, and Harrison 2021, 17). Crucially, by disrupting ‘the everyday world of connectedness’ (Colebrook 2019, 185), such dimensions enact a ‘cut in relationality, rather than just a mutation of relations’ (Colebrook 2019, 186). This argument invites a more vulnerable understanding of the bodies involved in platform urbanism. As Harrison (2008, 432) summarises, ‘corporeality must also be approached outside its capacities and powers and through its impossibilities and its “not-being-able-to”.’ This is about making space for ‘types of being do not know, that hesitate’ (Harrison 2011, 159). The upshot of these ‘negative geographies’ is an invitation to temper our heroic impulse to be able to tell what is going on, to be an even more ‘modest witness’ (Haraway 1997).

Admitting the non-relational—rather than just focusing on relations—invites a rethink of the power and politics of platform urbanism. On one hand, it repositions our sense of what power is responding to. In this regard, the machinations of power from this perspective are a response to a more primary and existential sense of vulnerability (Rose 2014). On the other hand, it rethinks the efficacy of power. A key insight of this perspective is that things often do not work as well as we claim they do, such that power is not as totalising as we might presume (Rose 2014). From the point of view of re-evaluating power, this perspective shares similarities with Guma’s (2022, 60) insistence that we imagine urban infrastructures as processes that are spatially transient; that progress in nonlinear ways; and that are contingent on their situation, and therefore are ‘not solely [driven] by neoliberal interventions.’ Within work on platforms, this insight has been developed through work on ‘minor’ platform urbanism that has punctured totalising techno-masculinist metanarratives of platform power. Leszczynski’s work in particular attends to ‘everyday instances where platform urbanism is neither frictionless nor inevitably successful’ (2020, 195). Through a focus on the ‘everyday contingencies of the comings-together of platforms, cities, and urban residents’ (2020, 202), Leszczynski traces diverse ‘glitches’ where things don’t go according to plan, such that the glitch ‘materializes and (re)makes platform urbanism in different ways, coming in the form of surprise, casualty, and absence’ (2020, 197). Leszczynski’s work contributes to a broader body of work that acknowledges how digital infrastructures are not used in the ways that power intends (Mrázek, 2002), as well as work on cities in the Global South that ‘destigmatizes infrastructural ambivalences and ambiguities’ (Guma 2022, 62).

Though my argument is aligned with such claims about the incapacities of power, it pushes in a different direction. Where Leszczynski focuses on how politics is made more complex by the ‘comings-together’ (2020, 197) of different agencies and materialities in specific settings, my concern is the politics that results from separations and distances. In the remainder of the paper, I explore how three different forms of the negative—unknowing, illegibility and ambivalence—invite a rethink of the subjectivities involved in evaluating platform urbanism to provoke questions about the operation of power. Admitting a more aporetic sense of relationality is not about hobbling our capacities to intervene as social scientists, but about questioning what intervention might look like and what might be possible.

Methods

The project that this paper draws from set out to explore on-demand platform urbanism in Melbourne. The project was framed through some of the key theories articulated in the previous section that underpin work on platform urbanism. First, the case study approach was motivated by calls to study ‘actually existing’ platforms in specific settings, through attending to place-based practices, events and materialities. Second, through fieldwork with different people across the domains of consumption, work and governance, the project responded to calls to trace the connections between different urban domains. Practically, to recruit consumers and workers to speak with, I undertook a social media survey which garnered 744 and 182 responses, respectively. From these responses, I invited 30 consumers and 30 workers to interview, based on maximising diversity of a set of demographic and geographical parameters as well as respondents’ qualitative answers. To date, I have interviewed 20 people involved in the governance of on-demand, with participants selected based on their job. Interviews were mainly conducted in-person and, though framed through a set of topics, were relatively unstructured and followed the thread of what was most relevant for each person.

A conventional way of analysing interviews is to identify significant themes through the systematic coding of occurrences which can then be cross-referenced with other occurrences to construct an idea about how a person’s experiences relate to a broader place or event (Crang and Cook 2007). Yet recent conceptual moves have questioned the status of knowledge produced through these analytical procedures (Williams et al. 2019). Feminist thinkers have, for some time, argued that interviews are not a transparent window onto a participant’s lifeworld but rather a partial, situated reflection. However, and significantly, recent work that draws attention to both the plasticity of bodies in terms of their ongoing changeability as well as well as their incoherence (Ruez and Cockayne 2021) invites renewed consideration about what, exactly, is produced in an interview and what we might draw from it. Though the people I interviewed certainly talked about their practices, during many of the interviews—and through subsequent reflecting and writing—what became apparent were the challenges of being able to articulate a strong narrative about their imbrication in on-demand digital platforms (Bissell, 2022a). What I sensed to be ‘actually existing’ both in the narratives themselves, and in my comprehension of them, were (in part) gaps and blind spots. The following three sections each explore a different dimension of this negative platform urbanism. The first reflects on interviews with people involved in governance who expressed uncertainties about knowing what was going on. The second reflects on interviews with workers where my capacities to make sense of their situation was challenged. The third reflects on interviews with consumers whose self-knowledge of their practices seemed difficult and unassured. Each section begins with a short reflection of my struggles to make sense of what was going on in interviews. Some use quote snippets, others deploy recollections.

Unknowing platform governance

Policy interviews were mostly straight-laced demonstrations of competence. A domain of interlocking portfolios, customer and market analysis; modelling and prediction phrased through risk, opportunity and compliance. Yet amid the ostensible clarity of procedure and protocol were hollows—some easy to miss, but definitely there. Knowledge was narrated, but there were shadows of perplexity. ‘I get confused,’ said one woman responsible for on-demand futures, after admitting flatly that their department doesn’t really know whether on-demand rideshare is cannibalising public transport. ‘I guess consumers and how they’re using it we’re still quite puzzled about,’ she said. For others, the solid conviction that rideshare is ‘what consumers want’ risked masking acknowledgments that such desires are capricious. ‘I think you can go down a rabbit hole of understanding what drives behaviour change but it’s just like, five minutes later, it’s a different answer,’ a taxi reforms director admits at the end of an interview. Even a rideshare company manager said that ‘sometimes customers don’t know what they want until they get it.’ For platform companies that laud equity, talk of their business models can hit impasses. ‘It’s tricky for us because we want to give everyone access to these earning opportunities,’ says one policy manager, ‘but at the same time, if there’s not the demand on a rider side or an eater side, then what does that mean?’ she adds, shrugging her shoulders. Government policy levers are pulled, but what transpires can be hard to figure out. ‘We’re waiting for the industry to settle down,’ a regulator says reflecting on the rideshare policy changes that have happened over the last five years, ‘before we can have a handle on things.’ Presentations in policy workshops are slick but slips happen in less guarded chats afterwards. ‘They don't really know anything, they’re panicking a bit and reverting to what they know,’ says one woman of her department that is responsible for platform policy, attempting to get a handle on the rapid increase in rideshare vehicles. And all of this during a time that Coronavirus had just descended. ‘God only knows what’s going to happen,’ says a woman in future mobility provision after sharing some figures on rideshare patronage. ‘It just feels like we’re in a space where we don’t really know what’s happening anymore,’ another admits. ‘It feels like it's a bit up in the air.’

Where the spoken expression of on-demand policymakers in interviews tended to mirror the straight lines of procedure and strategy, these moments above eschewed such performative norms. Of course, it could be the case that I was simply talking to the wrong people at the periphery of power. Admittedly, a bird’s eye panoptic gaze is never possible in complex technological systems, and the resulting ‘oligoptic’ partiality of perspective inevitably leads to blind spots (Gregson, Crang, and Antonopoulos 2017). Furthermore, platform companies leverage power from information asymmetries whereby their proprietary data is purposefully withheld from public authorities in many cities, thereby impeding urban governance opportunities (Van Doorn and Badger 2020). Yet each of these expressions above seem to point to more than just knowledge insufficiencies caused by a partiality of perspective. Unknowability took many different forms. For some in government, unknowability manifested in terms of uncertainty over how consumers are using on-demand services and for what precise purpose. Though these expressions of unknowability arguably could be remedied to an extent with enhanced data (though see Pajević and Shearmur 2017), others hint at a more unreconcilable unknowability—such as the capricious nature of what consumers actually want and how such desires change unpredictably. For platform companies, rather than information asymmetries, this more radical sense of unknowability manifested in terms of how to reconcile tensions that inhere in their business models. The final remarks above from people who work in government policy signal an opaque situation, exacerbated by the uncertainties of the coronavirus pandemic.

These expressions are significant because they counter assumptions that are made in platform urbanism literature about the operation of power. It can be tempting to imagine that it is in these capacious spaces of government and platform company policy where power is ultimately negotiated and meted out. This is, after all, where many decisions about on-demand infrastructures happen. And yet these remarks, these seemingly superficial afterthoughts, are a reminder that such activity is always undertaken in the face of ‘holes, black-spots and blindnesses that cover the landscape’ (Rose 2007, 473). Nuancing the politics of platform urbanism from this observation is a reminder that the work of platform governance is often undertaken from a position of not knowing. What this means is that though it is important to identify the presence of power and power relations that are borne from knowledge, this is only ever part of the story. Rather than seeing power just in terms of its successes, such remarks indicate how power must also be conceptualised in terms of its incapacities (Rose 2014). As Rose argues, politics here is the ‘art of living with always incomplete information and with the anxiety of not knowing or seeing what is in front of us, what complexities might arise, what possibilities could be coming’ (2007, 472).

Much has been said of the intrinsic relationality of digital platform infrastructures. The contention is that to properly understand the impact of these new technologies in our cities, we must identify how infrastructures such as on-demand platforms intersect with multiple domains of urban life. Some have called for a ‘whole city’ approach to understanding the how technological infrastructures connect across urban domains, rather than a focus on discrete applications (Macrorie, Marvin, and While 2021). Others insist that new technologies are ‘contextual’ in the sense that they are ‘always part of something else’ (Pink and Sumartojo 2018, 839) that is ‘non-digital centric’ (Pink 2016, 162). Of course, tracing such relations is important for understanding processes of amalgamation and convergence. Yet these accounts of not knowing presented here insist on a parallel attention to relations of non-relation: of separations and limits. Though there is growing appreciation that ‘contemporary cities are full of loose ends and failures of control’ (Thrift 2014, 15), there is also a tendency to suspect that such gaps and failures inevitably become conscripted into opportunities for corporate and state governance (Van Doorn, Mos, and Bosma 2021). Though this may be the case, it also too quickly negates the possibility for politics in platform urbanism to be understood as a faltering responses to situations of radical uncertainty, rather than a brilliant light shining in the darkness.

Illegible platform labour

Interviews with platform workers were hard. It was difficult to witness expressions of suffering. I listened to stories of debt accumulation, house repossession, sliding mental health, people not holding things together, lives that seemed to be spinning out of control. Though at times I felt that there was something gratuitous about provoking expressions of bodies being pushed to their limits, in some respects I was prepared for, even expecting, such accounts; my listening primed through engagement with literatures that prepared me for such sadness. And yet, more challenging, somehow, was hearing stories that didn't seem to make sense to me. What I was less prepared for was witnessing accounts that seemed occluded. One woman told me about working through Airtasker; cleaning jobs here and there across the city that she had to get to on public transport. I began to notice a few refrains for orientation—she said she distrusts men, staring at me cold as she did. She implied that something bad happened to her, and I tried to imagine what this might be, but I couldn’t push further. I asked other questions about her living situation, about likes and dislikes, about her plans for the future, but her answers seemed oblique, as if responding to an entirely different prompt. It felt like a hypnotic dream, with few footholds, and I started to doubt whether I was really there, somehow. If there were threads, I kept losing them. Another woman told me about working for a woman-only rideshare platform. I barely had to say anything during the interview, it was long and freewheeling with no prompts. At some point in her story, it became apparent that she was in a very difficult financial situation and she was homeless. Yet different parts of her stories seemed to loop back on each other, refraining again and again, changing slightly each time in tone and expression, such that it was difficult to make out what was good and what was bad for her, as if her account was undoing itself. A man told me about working full time through Uber Eats. Yet his motivations for doing this given his professional background seemed unclear despite questions that attempted to get to the bottom of this. Even his presence in the room we were in felt somehow unassured. He kept gazing out the window to some distant point, speaking with a soft slow precision which made it difficult to read his disposition.

These encounters were difficult to make sense of, scrambling straightforward evaluations of what was good and what was bad for these people. The aftermath of many other interviews with workers left me feeling defeated; a sense that I had not managed to get to the bottom of things adequately. Of course, it might be that these experiences of incomprehension are just evidence of my own interpretive inadequacy—not working hard enough to connect the dots that might well have been there. And yet the more I sat with the interview recordings and the more I reflected on these encounters, the more that these lifeworlds seemed to recede from view. Though successful interviews have often been understood in terms of the capacity to foster proximity between people (Bondi 2014), palpable in these encounters was a sense of growing space between us. In the absence of locating a ‘smoking gun’ or a red thread that would help to connect some of the disparate scenes that made up these people’s stories, these encounters left me adrift. It’s certainly possible that in some instances things were being purposefully concealed from me, given how some people working in the gig economy have few other employment options owing to restrictive visa status (Van Doorn 2017). Yet to suppose such active concealment would be too simple a trick, reinstating the interviewer as all-knowing spymaster of situations that felt anything but simple. The incoherence of these narratives and kinds of hallucinatory perception that they induced seem to resist unlocking by any master key.

These strange encounters are significant because, in resisting disclosure, they challenge assumptions made in the platform literature, and urban theory more broadly, that workers’ subjectivities can be made sense of. In the absence of definitive evaluations of their work conditions and transparency about their life contexts, these workers seem to sidestep the demand to declare a clear position to me. Barthes (2005) called this kind of negation that is a refusal to evaluate ‘the neutral.’ The concept of ‘neutrality’ is often wielded pejoratively in relation to new urban technologies, as a way of puncturing misguided illusions of their political innocence (Maalsen and Dowling 2020). But here, neutrality understood as ‘an affect of detachment from affection itself’ (Filippello 2019, 340) could be interpreted a kind of defence. For Berlant (2015), this sense of neutral unreadability might offer the weary subject some protection from the depletion that could come from being more invested in a situation (see also Bissell 2022b). Performed consciously or unconsciously, affective withdrawal in this sense offers a kind of minimal tactic that remains available in situations of real constraint or trauma. In this regard, our job as researchers in this situation is to acknowledge how we might be ‘confused but attuned’ (Stewart 2007, 10), where we need to admit more readily that we don’t quite know what’s going on.

Much has been said about the emotional demands of digital platforms for workers and the various tolls that they put on the people involved (Gregory 2021). This reflects a broader turn by urban researchers to better understand the experiential complexities of urban life. Yet where researchers have increasingly looked at bodily practices and their modulating capacities for acting and feeling to better understand contemporary urban processes (Amin and Richaud 2020), this assumes that people and their lifeworlds are readable in some way. Though more generalised declarations have been made about how digital platform work can create specific kinds of depleted subjectivity (Gregory and Sadowski 2021), there is often an assumption that we can indeed make sense of these subjectivities. Work subjectivities are complex, and much has been said of the tensions and contradictions that inhere in subjectivities where people seemingly remain attached to lives that wear them out (Berlant 2011). Yet reflecting on the encounters with workers above indicates that our confidence in diagnosing such attachments might at times be misplaced (cf. Anderson 2022); such affectivities might not be so easy to diagnose.

Ambivalent platform consumption

Interviews with consumers felt easier, less emotionally intense. Some things were said with a definiteness of expression, easy to comprehend. People reflected on food ordering habits and preferences, bad experiences of having to wait for rideshares, the practical or aesthetic differences between platforms in terms of their cost or useability, noticing how different things were available in different parts of the city. Indeed, such declarations on the face of it seemed to reveal ‘what people want.’ And yet even these interviews were replete with strange gaps that seemed to undo my confidence in making definitive claims. One man talks about his evening walk home from work through the CBD. He says that he orders food once or twice a week through Uber or Deliveroo. And yet whether he does this seems utterly contingent, in the moment. Sometimes he’s tired, sometimes not. Never a calculated decision made in advance, ordering seems a capricious affair, motivated by vague lures and barely conscious perceptions, and sometimes cancelled when he arrives home if ingredients in his pantry catch his eye. Another man says he orders food perhaps once a week. He says that he can spend a few hours flicking through options on his phone, sometimes while at work in his café job just to pass the time, and then sometimes half-heartedly in the evening while doing something else. But he says he gets tangled up in knots of indecision, unsure about what to have, or even whether he wants it, often not ordering anything at all. A woman tells me she’s tired, exhausted by her job, her commute, other people’s perceptions of her, being on demand for others. Sometimes she orders Uber Eats, especially when she’s feeling depressed and can’t be bothered to cook. Sometimes it’s ok, but she says she doesn’t really like the food. It’s expensive, unhealthy, over-packaged. Even the brief encounter with the delivery rider seems to provoke intense shame in her. She says she’s not even hungry, it’s not even a guilty pleasure.

These encounters draw attention to the ambivalence of people’s attachments to on-demand services. Rather than assuming that people simply like or dislike using such services, or get ‘recruited’ into practices involving on-demand that then remain relatively obdurate (Shove and Pantzar 2007), what these glimpses reveal is a much more complex sense of the consumer subjectivities imbricated in on-demand platforms. Through these interviews I got a sense of people actively wrestling with their use of these services, and always in the context of a wider force field of relations involving other dimensions of their lives. Methodologically, this observation reveals the value of qualitative interview techniques for drawing out this richness and complexity of how consumers relate to using on-demand platforms, providing a space in which to explore how engagement with on-demand platforms is neither straightforward nor static. In this regard, interviews also drew out the temporal dimensions of people’s relationships with on-demand platforms, both in terms of how decision making is a fickle process that takes place in unpredictable ways during the day rather than a discrete event (Zhang 2018), but also that orientations to on-demand platforms evolve incrementally over time and can reach tipping points (Bissell 2020).

These ambivalences are significant because they counter assumptions that consumers’ engagement with on-demand platforms is simply prescribed by the powerful strategies of platform companies. Digital platform infrastructures clearly do not operate on consumers in a deterministic—or even necessarily probabilistic—manner (cf. Adey 2008). But from a performative perspective, neither is there clear counter-agency ‘from below’ that would require the existence of strongly sovereign selves. Rather, the modes of evaluation that take place for people are complex because, as work on performativity has emphasised, people are affectively and emotionally incoherent (Najafi, Serlin, and Berlant 2008), meaning that engagement with these platforms will necessarily be incoherent too. Since the affective turn invites us to appreciate the ‘digital in its ambivalences and complexities’ (Fazi 2019, 11), rather than producing simply joy or sadness, these interview reflections point to a much more ambiguous oscillation of feeling for those involved (Gammerl, Hutta, and Scheer 2017). As these interview reflections indicate, this involves taking seriously the various impasses and undecidabilities that people sit with and navigate, where even our ability to be able to evaluate what enlivens or depletes us is not assured. This understanding of limits is the ‘muddled middle ground between affirmation and negation’ (Ruez and Cockayne 2021, 95) that other research techniques often seek to short-circuit. It is based on an understanding of subjectivity where subjects are not transparent to themselves, such that knowledge is always contingent and incomplete.

Much has been said about what on-demand digital platforms are supposedly doing to consumers. From the creation of dependencies to anxieties about deskilling (cf. Stiegler 2010), such diagnoses have inevitably generated anxiety about the broader socio-spatial impacts that on-demand platforms are having on cities. Yet admitting ambivalence ruptures these clearly formulated diagnoses. Though it can be tempting to focus on statistical trends of service use or even the regularities that people invoke in and across interviews, to generate a picture of consistency betrays the much more changeable situation that exists at the level of people’s indeterminate capacities to affect and be affected. This is not just about appreciating the diversity of ways that people incorporate on-demand platforms into their everyday lives, but it is rather about admitting that the way that people use these services is not straightforwardly knowable. In contrast to more generalised declarations that new technologies are creating new subjectivities (Thrift 2014), this acknowledgment compels us to question simple appeals to affects of convenience, ease and immediacy that are presumed to drive consumer practices in this space, as well as claims by digital platform companies and urban governance that they know ‘what people want.’ If we accept that ‘much of the world is vague, diffuse or unspecific, slippery, emotional, ephemeral, elusive or indistinct, changes like a kaleidoscope, or doesn’t really have much of a pattern at all’ (Law 2004, 2), then our job is not to create clarity, but rather to acknowledge the ‘vagaries of affective life’ (Anderson 2014, 7) and the ‘ambiguous existence of collective affects’ (Anderson 2014, 19) relating to platform use.

Discussion

The rise of ‘lean’ capitalist digital on-demand platforms over the past decade is altering consumption, production, and governance in many cities. There are undoubtably transformations taking place to consumer habits, work opportunities and conditions, and policy frameworks. Furthermore, such transformations are likely inducing wider flow-on effects in cities, especially in the domains of urban transportation as well as food consumption and production. From a political economy perspective much has been said about how the emergence of lean platforms has altered the contours of power in cities. There are compelling arguments about how platforms wield power over both workers themselves (Richardson 2020) as well as local and state governments (Gorwa 2019). Furthermore, the provision of on-demand services can induce further inequalities. Food delivery platforms can threaten the viability of restaurants (Nhamo, Dube, and Chikodzi 2020). Rideshare platforms can threaten the viability of public transport (Mulley and Kronsell 2018). These rapidly changing urban transformations require our attention as there are important social justice consequences. Yet what these urban transformations actually are and how we diagnose them remains an open question. This paper cautions against impulses to make definitive claims about how cities are changing.

The paper has built on work in minor urbanism and beyond that highlights the challenges of pinning down cities, even through a sensitivity to what ‘actually exists.’ Holding onto rapidly changing objects of analysis is an endemic challenge since there are always multiple forces at play in the urban kaleidoscope (Amin and Thrift 2017; Guma 2020). This means that ‘objects are always looser than they appear. Objectness is only a semblance, a seeming, a projection effect of interest in a thing we are trying to stabilize’ (Berlant 2016, 394). Our embodied evaluations of these objects introduce further complexity, both in terms of multipliying our objects of analysis, and complicating our sense of their spatialities (Simone 2019). All of which demands more fragmentary accounts of urban life that seek to undo confident declarations in heroic accounts of the city—though of course, ironically, even a fragmentary account can risk being heroically presented (Linz and Secor 2021). This perspective has been developed in research on different dimensions of platforms, both through interrogations of ‘actually existing urbanism’ (Söderström and Mermet 2020), as well as work on glitchy failures (Leszczynski 2020). Such work emerges from the perspective of a material world of relations, abundance and connection, emphasising the failures, inoperability, and doubt that can emerge in complex systems. Expanding our understanding of ‘how digitality (re)produces power’ (Elwood and Leszczynski 2018, 630), this work expands our understanding of the agents of change in on-demand platforms through a focus on interruptions, diversions, and transformations. Charting what happens when the smooth logistics systems imagined by engineers and designers rub up against the messy materialities of the city, such work emphasises how the dominant drivers of change are not just the technology companies behind on-demand platforms.

Where this paper’s argument differs is that much of this work prioritises relations and connections, rather than disconnections and limits. In response to arguments about the need to open the ‘black box’ of digital platforms with the aim of revealing what, exactly, is going on, this paper has underscored the epistemological challenges associated with making sense of platform urbanism (Bucher 2016), not just the platforms themselves in terms of their algorithms but also their effects. The epistemological problem that links these discussions concerns the limits of knowability of on-demand platforms—not in the practical or institutional sense of knowledge being purposively withdrawn, but rather in terms of a much more ontological understanding of withdrawn entities that cannot be known. Discussion of fieldwork encounters in each of the three domains has shown different configurations of this problem. Interviews with people involved in platform governance drew attention to the admittance of not knowing. Interviews with platform workers drew attention to the unknowability of others. Interviews with consumers drew attention to how people navigate incoherences of knowing. Though each domain reflected one of these negative configurations most acutely, they are not mutually exclusive. Each of these configurations of unknowing might be a characteristic of any encounter. These epistemological challenges are not methodological shortcomings that could be remedied through more extensive fieldwork, but rather an existential problem to be reckoned with. They each puncture our capacity to claim with confidence that we know what platforms are doing.

In articulating these challenges, this paper insists on the value of work in cultural geography on the negative for urban theory. Instead of foregrounding the complicating abundance of relations through things that ‘actually exist,’ the empirical moments in this paper foreground impasses, gaps and distances. These negatives are an ‘alterity that resists being drawn into a relation and refuses attempts to bring it into knowledge’ (Shubin 2021, 66). The empirical moments presented in this paper insist on grappling with the irreconcilable non-relations that puncture cities (Harrison 2008). These theories question the remit and possibility of knowing ‘actually existing’ urbanism, including alternatives, counters, and resistance to grand metanarratives, as well as the challenges of making claims about what ‘actually exists’ and the transformations that might be taking place. As such, this negative perspective invites consideration of new ways of writing about platform urbanism. Where others have developed vocabularies and grammars that pay attention to the background, the elsewhere, the ‘off-scene’ (Simone 2019), the perspective developed in this paper invites different ways of writing the city that are potentially more aporetic—where everyday life is not a site of knowability, transparency and revelation; but of uncertainty, illegibility and ambivalence.

To be clear, my argument is not that we can’t know anything about platform urbanism. Rather it seeks to temper an impulse evident in research on platform urbanism to make overly conclusive claims about the nature of transformations. There are two implications that this argument has for future research on platform urbanism. First, admitting the negativities of unknowability, illegibility and ambivalence invites us to be more hesitant about the claims that we can make about platform urbanism. Such an invitation aligns with thinkers who have advocated for more speculative modes of enquiry. Anderson (2021, 208) invites us to make claims that are more conditional, such that ‘that this could have happened, or perhaps this might be happening,’ a practice that respects the spacing between us and our object of analysis. Knox (2017, 369) similarly calls for a ‘material diagnostics’ which is ‘a kind of questioning, interrogating, tracing, supposing.’ Such speculative suggestions are highly compatible with the perspective developed in this paper, since speculation might involve reflecting on our own analytical habits, both in terms of our faith in the claims of others or in our own convictions. As Ahmed (2017, 7) underscores, it is important to ‘temper the strength of our tendencies with doubt; to waver when we are not sure, or even because we are sure.’ In this regard, where Guma (2020) advances the concept of incompleteness as an ontological frame for thinking about the nature of urban processes, I suggest that incompleteness is also a very useful epistemological frame for evaluating the status of knowledge claims about cities. Yet how to balance this call for unclarity with our responsibility to call out situations of exploitation and systematic injustices with conviction? Critics might argue that such a disposition is futile, irresponsible even, as it lets platform companies off the hook. Such critique might be particularly warranted when platforms themselves are using indeterminacy as a resource (Van Doorn, Mos, and Bosma 2021). Surely there is a need to not be so ambivalent in this respect about what is going on (Linz and Secor 2021)?

In response—and the second implication of this argument for future research on platform urbanism—is that admitting the negative invites reconsideration of the sites of power in platform urbanism, and what power is responding to. As Fields, Bissell, and Macrorie (2020, 462) suggest, ‘while taking seriously the need to apprehend the geographical political economy and the distributional consequences of urban-digital entanglements, geographers also seek to theorize a counter-politics that is not rooted exclusively in resistance or antagonism.’ In this regard, the negative provides an alternative counter-political perspective that apprehends both the instability of politics and our practices of judgement. This is especially pressing given that ‘there is no guarantee that in struggling for justice we will ourselves be just’ (Ahmed 2017, 6). Attention to these ‘embodied, affective engagements with infrastructure allows us to unpack something of the unstable or uneven quality of politics’ (Knox 2017, 368) in the city and, I would add, our difficulties in arriving at such judgements. Where the question of politics in platform urbanism has typically remained focused in on sites of governance—such as the regulatory negotiations that happen between platform companies and city governments, for example—this paper retreats from privileging this site of politics by arguing that what is also being negotiated within platform urbanism are a series of negatives. Fieldwork encounters with people in platform governance, workers, and consumers each gave a sense of people and organisations grappling with blind spots and impasses. Furthermore, fieldwork itself was a process replete with blind spots and impasses. In short, politics in platform urbanism therefore needs to be appreciated more widely as the myriad processes of responding to and grappling in an ongoing and incomplete way with these primary negatives—uncertainties, illegibilities and ambivalences. Rather than claiming to know politics in advance where the contours of power have already been decided on, these blind spots and impasses indicate that things might not be as they seem—and situations might be otherwise, already. And yet far from noise to be removed, from a research perspective, these negatives are the voids that compel us to attempt to make sense in the first place, to reach out, to enquire, in the knowledge that they will always fall short.

Acknowledgements

Thank you to the two anonymous reviewers and the City editorial team for their helpful suggestions. Thank you also to audiences at the University of Wollongong and Guangzhou University whose comments and questions helped to develop aspects of the argument.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by the Australian Research Council [grant number FT170100059].

Notes on contributors

David Bissell

David Bissell is Professor of Human Geography at the University of Melbourne.