Urban geography I: Conceptualisation with and beyond the global-local dialectic

Abstract

In this report, I explore recent work in urban geography to reflect on one of its pervasive ‘thought-infrastructures’: the global-local dialectic. This dialectic features as a prominent force in how we debate, write, and teach cities and urbanization. My aim is not to argue against this dialectic, but to call for greater reflection on it and the work it does, including its increasing resolution through forms of ‘missing middle’. I suggest that de-centring the global-local dialectic may allow greater space for other ways of thinking, writing, and teaching urban geographies – other thought-infrastructures – and set out examples.

I Introduction

A central challenge for urban geography lies with how to draw out generalisations from an urban world of radical difference. As soon as we speak of ‘the city’, ‘the urban’, or ‘urbanization’, we are inescapably caught in the relation between the particular and the global. While this relation for urban geographers is primarily a spatial problematic, it is also one of content and of forms of explanation and description. All of these – space, content, and form – might be ‘assigned’ to either the global or the local in how we analyse, write, and teach. As Parnell and Robinson (2017: 13) have written, any attempt to ‘understand the significance of an urban world bring to the fore the tension between specificity (or difference) and universality in conceptualising the urban’.

In recent decades, debates around how urban geographers might make claims about cities and urbanization have intensified. We see this in debates on the challenge of postcolonial or poststructural thought, or around planetary urbanization, or the potential of comparative methods or conjunctural analysis, or the place of social difference in theory and concept-making (e.g. Addie, 2020; Brenner, 2018; Brenner and Schmid, 2015; Davidson and Ward, 2024; Oswin, 2020; Robinson, 2022; Roy, 2016). These debates have been multi-faceted, searching and generative. A central theme has been the relation between global abstraction and local particularism.

For Martin Murray (2022a: 3, 4), ‘theorising about cities and global urbanism have tended to gravitate toward two opposing poles’: one ‘the “need for an all-encompassing theory” camp’ searching for ‘a single, one-size-fits all theoretical approach to account for urbanisation everywhere’, and another that has chosen to ‘dispense with the committed search for general patterns and commonalities across cases, preferring instead to focus on the particularities of individual urban experiences’. Mark Davidson, 2024 has pointed to a distinction between those advocating global frameworks of the political economies of capitalist urbanization and those opting for a more modest focus on urban particularities. Kate Derickson (2015: 648, 650, 651) has described a divide between ‘Urbanization 1’ – the urban viewed from ‘a more than global scale’ of ‘grand narratives and universal claims’ – and ‘Urbanization 2’, which seeks to ‘locate political possibilities in emergent subjectivities and livelihood strategies’, with a ‘messier and less cohesive story to tell’.

There is a dialectic of global abstraction and local particularism that has come to act as a kind of thought-infrastructure in urban geography. This is a dialectic both of space and explanation (e.g. general patterns, key causes, local description, and unique cases). If dialectics often demand some form of resolution, one way in which this dialectic is increasingly resolved is through a growing focus on what we might call the ‘missing middle’. This is a compromise ground that appears intuitively like the ‘right’ place for urban geographers to be. Two opposing poles are presented and a ‘third way’ must then be the next step: the mid-level, the meso, the mid-range. This mid-level position is often purported to potentially resolve not just a spatial tension between global and local, but the problem of explanation too.

This dialectic of global abstraction and local particularism, and perhaps also its mid-level resolution, is unavoidable. It is also a generative and often relevant part of the research, debate, writing, and teaching in urban geography. It reflects too, at the risk of stating the obvious, actually existing geographical conditions. A great deal of scholarship has carefully demonstrated recurring and generalised global processes of urbanization (e.g. neoliberalisation and its attendant logics and manifestations), just as there is work that illustrates local cases of difference that do not conform with global accounts, whether in relation to the operation of urban economies, cultural conditions, political possibilities, or environmental processes (Jazeel, 2018; Parnell and Pieterse, 2016; Peake, 2016).

I am not going to argue, then, that this dialectic is ‘wrong’ or an unproductive route for making sense of the geographies of cities and urbanization. Nor am I going to argue in favour of abandoning the scalar analytic of global/local/middle as a way of understanding and analysing spatial processes. While I share some of the concerns put forward two decades ago by Sallie Marston et al. (2005) in their provocative ‘Human Geography without Scale’ piece in Transactions of the Institute of British Geographers – particularly the tendency to conflate ‘global’ with causality, structure and the abstract, and local with experiential, difference, and agency – scalar thinking is often a helpful frame for describing how processes work and places change.

My concern instead is to question the prevalence of the dialectic as a framing device and object of thought and debate. It is to ask whether this dialectic might be playing an outsized role in how we conceptualise, describe, explain, write about and teach cities and urbanization, and to argue for a decentring of the preoccupation of ‘space-as-scale’ in urban conceptual work. I do so by posing two inter-connected provocations, both of which hinge on the centrality that the global abstraction-local particularism (GA/LP) dialectic is given. First, does the GA/LP dialectic help or distort how we see cities and urbanization processes? Second, where does the recurring centrality of the GA/LP dialectic leave other ways of thinking about cities and urbanisation? As Austin Zeiderman (2018: 1115, 1123) has suggested, might we worry less about the relationship between the ‘local’ and the ‘global’, and instead ‘loosen this deadlock’ by examining the ‘social lives of our key concepts’, including by focussing more on ‘the boundaries we construct between ourselves and the worlds we study’?

II Sorting the urban from top to bottom

The GA/LP dialectic serves as a go-to coordinating device for thought and analysis: between the generic pattern and the specific case, the recurring and the unique, the explanatory and the descriptive, or the cause and the outcome. It operates to identify approaches to the city and urbanization that are, on the one hand, described – depending on your position – as universalising, overly abstract, global, decontextualised, explanatory, or sweeping, and those that are, on the other hand, portrayed as particularist, case-based, local, descriptive, singular, and sometimes oriented to multiple epistemologies and emphasising difference. Two dialectical relations are, often, collapsed into one: scale (global vs local) and claim (explanation vs description). There have been accusations of ‘global’ universalisms and over-generalisations that are neglectful of history and difference, (e.g. Brenner, 2018; Oswin, 2018; Roy, 2016), and accounts of the other side of the dialectic concerned about what Murray (2022: 4) calls a seeming ‘“anything-goes” celebration of the radical uniqueness of cities’.

We might reflect on what this dialectic opens up and closes off. A degree of simplification is a necessary part of conceptual understanding, but there is also the danger of artifice and caricature. For instance, does the dialectic stretch and exaggerate the differences between so-called GA and LP accounts, rather than identifying commonalities? Might it undermine, for example, shared conceptual commitments to understandings of urban space (e.g. as relational processes), or common efforts to politicise the state of urban conditions and develop possibilities for alternatives (Derickson, 2015)?

Thought-infrastructures operate as a logic of simplification. Without much difficulty, nuanced arguments and claims can be pulled into this or that end of the dialectic. The specificity of different approaches and arguments can be reduced as they are grouped together. This risk is, perhaps, inevitable if we continue to link ‘global’ with terms like ‘abstract’ or ‘grand’, and ‘local’ with terms like ‘grounded’ and ‘particular’. These terms, after all, do not necessarily belong together. Their co-location is a product of the dialectic in action. And yet, as we know, it is perfectly possible for theory or conceptualisation to be ‘global’ and ‘particular’. As feminist and postcolonial scholars have taught us, claims to global comprehension can be profoundly parochial, reflecting narrow intellectual, empirical and theoretical hinterlands (Roy, 2016). Similarly, theory can be ‘particular’ and ‘abstract’, and it can be ‘local’ and ‘grand’ (e.g. arguments that make generalised or universal claims based on one case).

It is perhaps inevitable that the response to the GA/LP dialectic is to seek out different positions in the ‘middle’ that might resolve its contradictions. The question that seems to be continually posed is: are you ‘zooming in’ or ‘zooming out’, to use the terminology Ash Amin (2013) deployed in his discussion of ‘telescopic urbanism’, or locating your work somewhere in-between?

1 The missing middle

A focus on mid-level conceptualisation is not, of course, new. There is, first, geographical mid-level categories that tack this course. The long history of regionally-oriented urban geography is an example. Some of this work has sought to identify political economic processes, knowledge politics and ways of seeing cities and urbanization from different world areas. We see versions of this lineage in research that seeks to depart from the global North/South binary, such as work on the ‘global southeast’, Eastern Europe, or Southern or Eastern Africa (Chelcea et al., 2021; Müller, 2021; Parnell and Oldfield, 2014; Roy and Ong, 2011; Shin, 2021; Yiftachel, 2020). There is also work on urban ‘types’ that sit between global abstraction and local particularism. Murray (2022b), for instance, has focussed on what he sees as sui generis forms of global urbanism that are neither global or local, including: prototypical globalising cities, post-industrial shrinking cities, sprawling megacities, and master planned ‘instant’ cities.

Or, second, we might consider how a series of epistemic concepts and domains operate as ‘bridging terms’ navigating a mid-level position between GA and LP. Mark Davidson, 2024 for example draws attention to terms like ‘world-making’ (Amin and Thrift, 2017; Roy and Ong, 2011) and the ‘repeated instance’ (Jacobs, 2012). There are research problematics that function as a kind of connective tissue articulating the global with the local in cities, including infrastructure, public space, policy mobility, community, and so on (e.g. McCann and Ward, 2011). Some scholars have turned to metaphor to help with this bridging work. Sharon Meagher (2015: 806), for example, uses the metaphor of ‘the weed’ to signal the always-entangled nature of urban, rural, epistemology, and politics ‘between the “universal” and the particular’. The list could go on.

There are, third, broadly methodological routes to the missing middle. Arguments for situational and comparative approaches, for instance, sometimes identify a mid-level position. Using Karl Popper’s commitment to falsifying theory, Davidson (2023: 118) has argued for ‘situational analysis’ to ‘develop middle-range theoretical approaches that produce justified generalising claims about contemporary urbanisation’. From this position, a claim like ‘urban processes are ultimately defined by capitalist accumulation…is highly likely to be falsifiable hypothesis’, as is ‘claiming that all urbanization is singular’ (p. 126). In relation to strategies of comparative urbanism, Jennifer Robinson (2016: 194) has described how comparative urbanism might work with ‘located insights’ that build ‘resonating’ understanding across diverse places, by developing ‘new approaches to understanding an expanding and diverse urban world, building theory from many different starting points, perhaps resonating with a range of different urban outcomes’ (and see Robinson, 2022).

Especially emblematic of efforts to refute the apparent choice between GA and LP accounts is the recent upsurge of interest in conjunctural analysis. As derived from thinkers ranging from Antonio Gramsci, Louis Althusser, Stuart Hall, and Doreen Massey, conjunctural analysis takes different forms, but typically operates as broad methodological and/or analytical guide aimed at identifying key ‘moments’ in which different forces and contradictions combine to create a distinctive impact (Davidson and Ward, 2024; Hart, 2020; Peck, 2023; Woolston and Mitchell, 2024). Jamie Peck (2023: 2, 3, 4) describes a form of conjunctural analysis which ‘digs deeply into particular situations but which tend to be characteristically dissatisfied with immediate, proximate, or otherwise pre-emptive accounts of causality’.

An approach that will ‘reach backward and spiral outward’ beyond the study site, conjunctural analysis finds, he continues (2023: 3, 4), a ‘meso-level’ between ‘overly abstract speculation’ and ‘excessive particularism’, moving ‘between the macro and the micro, the epochal and the everyday, the structural and the contingent, the historical and the quotidian, while disengaging from none of these’ (and see Yeung, 2023 provocative argument for mid-level causality-focussed theoretical explanation). Leitner and Sheppard (2020: 492) have argued that there is a prioritisation of ‘the dialectical relationship between the general and particular’ in conjunctural approaches, while Davidson and Ward (2024: 16) suggest that ‘conjuncturalism represents one possible, and we argue, productive way, to negotiate these differences [between universalising and particularist approaches] by providing geography with a framework capable of developing knowledge attuned to both particularity/regularity and instability/fixity’.

In these different spatial, epistemic, methodological and analytical forays into a middle ground lie generative agendas that promise new approaches and conceptual insights, while reminding us of the different ways in which scalar strategies are politically deployed by states, activists, and others to further or contest hegemonies. The GA-LP dialectic functions to different extents across these, sometimes pushed to the background, at others serving as the central concern or point of departure. My point here is not to focus on one or other of these particular approaches but instead to point out the larger preoccupation with the GA/LP dialectic. One consequence of this preoccupation is that there can be a certain distancing effect between theory/concepts/claims on the one hand, and ‘the field’ on the other. It is not that ‘the field’ ceases to be taken seriously in developing conceptualisations of cities or urbanization – far from it – but that the dialectic can serve to curiously draw our focus both from it and from other ways of thinking, writing, and teaching cities and urbanization that are outside of, or that at least de-centre, the GA/LP dialectic.

III Other thought-infrastructures

How might we ‘see’ cities and urbanization if we shift from the GA/LP dialectic to other ways of thinking, writing, and teaching their geographies? There are, of course, an incredible range of theoretical and political resources and concerns that might be identified here, from other spatial metaphors – networks, extensions, distributions, resonances, and so on – to conceptualisations emerging from sources ranging from political ecology or postcolonialism, to feminist theory or actor-network theory (though in many cases these become repositioned in the GA/LP dialectic). Here, and simply by way of example, I focus on just three alternative thought-infrastructures: the urban itself, circulation, and the spatial lives of concepts.

1 The urban itself

First, the urban itself. Decentring the sifting and sorting of content and explanation into global and local in the GA/LP dialectic could enable greater attention to how concepts emerge from cities and urbanization. Here, there is a disparate body of work that points to an alternative urban thought-infrastructure. It is not that the GA/LP scalar dialectic becomes irrelevant, but that it ceases to be a key force in writing urban geographies. Consider for example work that we might be tempted to group under the ‘local particularism’ banner.

Here, we might consider forms of urban inhabitation that are so often disavowed space and meaning in the city, as Michele Lancione (2023) has shown in his work on home and homelessness. Or, we might follow Tatania Thieme’s (2018) approach as she conceptualises the ‘hustle economy’ in the work of waste labourers in Nairobi, and relates it to youth precarities and socialities. We might, with AbdouMaliq Simone (2022), attend to the ‘urban surrounds’ that surface and recede from social and economic activities, from overhead comments amongst neighbours to the ways in which mobile precariously employed men variously use, plug into, navigate or remove themselves from different kinds of labour. We might learn from Romit Chowdhury’s (2023) linking of anecdotes and ‘character types’ to ideologies of masculinity in urban space in Kolkata. Or, we might pay attention to what Gupte and Shetty (2022: 547) call ‘small forces’: lived experiences, storytelling, and blurred lines between urban form, functions, uses, practices, and imaginaries, including ‘obsessions, rumours, gossip, myths, rituals, aspirations, dreams, desires, idiosyncrasies, solidarities, claims, atmospheres, intimacies, and immensities’.

Given that these are all detailed case study works, it would not be difficult to locate these examples on the GA/MM/LP scalar frame as instances of ‘LP’. Yet, all of these accounts speak out from their cases to and with larger debates in ways that not only sit uncomfortably with a scalar analytic of local/middle/global, but which decentre it in how they write the urban. They each generate conceptualisation from the urban and work with different kinds of claims – explanatory, descriptive, causal, speculative, hesitant, decided, normative, and so on – all of which relate to different spatialities. Here, conceptualisation may even suggest a ‘post-scalar’ analytic, not detached from it, but not defined by it.

In their different ways, they allow the urban itself to shape conceptualisation. Philosopher David Kishik (2015: 95) points to this different thought-infrastructure when he suggests, provocatively, that we consider shifting the balance from asking how our theories and concepts might be applied to the urban, towards a greater attentiveness to how conceptual work might emerge from the urban: ‘It is about time that we let the city change the way we think’. This intriguing vexation is not quite, to use a phrase that has been in circulation in recent years, equivalent to ‘seeing like a city’ (Amin and Thrift, 2017), but is instead a call – and an ultimately impossible and outlandish one – to ‘be like a city’: ‘The ultimate goal is therefore not to be in a city but, as strange as it may sound, to be a city, to let it affect us – with no fear and no remorse – to such an extent that we become it rather than expect it to become more like us’ (Kishik, 2015: 216).

We might short-hand this thought-infrastructure as ‘the urban itself’. Of course, as Kishik intimates the urban does not come to us unmediated. Indeed, this rendering of urban thought-infrastructure chimes with the larger effort to foreground the researcher and greater self-reflexivity in the socio-spatial positionality of concepts (Leitner and Sheppard, 2020), including who is conceptualising them, from where, and with what consequences for understanding (this extends too to the language we use in conceptualisation, including how the specific terms we use and do not use capture different ‘versions’ of the urban [Bodden, 2023; Zhao, 2020]).

Kishik’s point of reference in making this argument is the work of Walter Benjamin. Across significant areas of Benjamin’s work, the urban entered into thought and expression. This includes most notably the montage experimentation in The Arcades Project, a sprawling effort to foreground the incessantly combinatorial juxtapositions and distanciated relations of cities and urbanization (Benjamin, 2003; see Cresswell, 2019). The Arcades is an experiment in ‘being like a city’ through building consistencies and resonances across different sources, staging fleeting encounters, questioning established histories from different angles of vision, finding disruptions, speculating on possibilities, and more.

We could position the Arcades in the GA-LP dialectic, but doing so would miss the different spatial sources and connections the text makes as it entangles, for instance, encounters on a Parisian bus or café, histories of iron construction and street lighting, state redevelopment programmes, changing cultures of seeing urban space, poetry about urban destitution, Marxist analysis of insurrection, and so on. The text is urban in content and form: sprawling, fragmented, disharmonious, intensive, and creating surprising connections and juxtapositions. Of course, the Arcades is a highly distinct, singular piece of work, but the point here is to suggest that decentring scale might allow us to foreground other geographies and concepts as the emerge from and urbanization.

2 Circulation

Second, and following on from this, decentring the GA/LP dialectic might entail turning to thought-infrastructures such as circulation. Here, we might look to work that both uses but at the same time decentres the role of scale in how it analyses and writes space and develops explanation. Ananya Roy (2012), for example, has examined the emergence of forms of market rationality as a Foucauldian apparatus that connects disparate discourses, institutions, knowledges and practices and which collectively act to put ideas and forms of subjectification into circulation across space and time (and see Tsing, 2005; Li, 2007). Roy’s focus is on the circulation of ideas, discourses, policies, ways of thinking and doing, models of urban planning, techniques of financialisation, and so on.

We could position this as an instance of the missing middle. Indeed, Roy utilises a scalar frame as her methodological entry point – the ‘mid level’ technocrat or bureaucrat. At the same time, though, she calls for ‘methodologies of composition’ (p. 36) that can identify the ‘circulatory capacity’ of institutions in the ideas and discourses they peddle, or the ‘circulating scripts’ of how to perform particular techniques of governance or development, or the ‘framings’ of ideas like ‘gender empowerment’ that are shaped and contested by different actors, centres, and trajectories.

In this emphasis on tracking routes, webs, networks, spatial extents, and resonances, we might start in any number of places: the community-based organisation, the statistical model, the disparate epistemic community, the institutional policy frameworks, and so on. These places may well be local, global, or in the middle, but if were to de-centre scale as a thought-infrastructure, we might see in its place other circulatory geographies as we write spatiality and explanation. A very different set of examples might be some strands of literature on urban metabolisms or planetary urbanisation, which write urban geographies by gathering multiple spatial registers, from hinterlands and agricultural zones to bodies, in the unequal circulation of capital, resource, and power in ways that both use but which can decentre scale – including the scale of the city itself (Connolly, 2019; Lesutis and Kaika, 2024).

There are of course spatial concepts beyond circulation (in my own work, for instance, I have attempted to explore ‘resonances’ and echoes in urban form and experience across urban spaces – McFarlane, 2021). Decentring (not disposing) of the scalar GA/LP dialectic can support the kind of attentiveness to different kinds of urban geographies in our thinking, writing and teaching that Geographers have often long argued for (e.g. Massey, 2005).

3 The spatial lives of concepts

Finally, third, we might consider the ‘spatial lives of concepts’ as another alternative thought-infrastructure. Austin Zeiderman (2018: 1122) has argued persuasively for greater attention to what he calls the ‘social lives’ of concepts about the urban (and see Robinson and Roy, 2016). Building on Zeiderman, albeit in a different direction, my suggestion here is that we might attend more directly to the spatial lives of concepts (I use ‘spatial lives’ instead of ‘social lives’ to signal that my argument here is for a decentring of the particular preoccupation of space-as-scale in urban debates in Geography).

As Zeiderman (2018: 1123) reminds us, not only is it the case that our concepts ‘do not belong exclusively to us’, the stakes of concepts of the urban or city or urbanization, and their attendant concerns, are typically higher beyond the pages of our journals and books. Relevant here is how legal and administrative definitions of what/who is and is not ‘urban’ can become the subject of critical interrogation, including by the state and different publics, with implications for often marginalised epistemologies and identities. Describing struggles for Afro-Colombian rights in Buenaventura, Zeiderman (2018: 1122) writes that ‘activists are well aware that these classifications fundamentally shape the rights they have, the protections they are entitled to and the demands they can make on the state’ (and see Davidson and Iveson, 2015, on ‘the city’ as a political category). The imperative here is to attend not just to how 'our' concepts of the urban shed light on cities and urbanization by parsing out their local, mid-level and global dimensions, but to how concepts travel beyond the academy to make their mark on all kinds of spaces and concerns.

If a focus on the GA/LP dialectic can draw our attention to whether or not our conceptualisations are adequately about the urban, a focus on the spatial lives of concepts as they move in and out of cities calls on us to ask how conceptualisations act in the urban, including where they come from, where they move through, and who/what they include and exclude. As a different example, we might take the highly popularised Lefebvrian idea of ‘right to the city’. We might say that this idea is a good example of a ‘mid-level’ position that could serve to resolve the GA/LP dialectic, informed by everyday places but shaped by and oriented to global processes. Again, my position here is not to argue against this but instead to suggest that what makes the right to the city idea compelling is that it is a political claim that does work ‘in’ different contexts, work that decentering GA/MM/LP might help us to better see and explore (of course, this work might itself be scalar, as activists seek to move campaigns from the local to the city, national, or global, but it may also be about other spatial imaginaries and forms).

Hillary Angelo (2017) has argued that we have inherited a 19th century lens – the ‘city lens’ – for thinking about the urban and urbanization, one rooted in particular in the experiences and transformations of cities like urban industrial Berlin, London, Paris, New York, and Chicago – which has consequences for how the idea of the ‘city’ circulates beyond the academy (and see Angelo and Wachsmuth, 2015). Today, the city lens no longer holds quite the same grip in urban theory, given the clamour of research around peripheralisation, overlooked small and medium-sized cities, urbanised hinterlands, the changing relations between the ambiguous categories of the urban and the rural, and planetary urbanisation (e.g. Merrifield, 2013; Ruszczyk et al., 2021; Schmid and Streule, 2023). As their meaning and significance change, so too might their ‘spatial lives’ beyond the academy. How, then, do concepts of cities and urbanization travel, and with what consequences, across places, activities and actors? And how might they change when placed in dialogue with how constituencies beyond the academy understand and use them?

IV Conclusion

I am not, again, arguing for the dispensing with the GA/LP dialectic or the various ways in which it is being recast through mid-level conceptual and methodological approaches, a move which would be neither possible nor desirable. In the three thought-infrastructures I have pointed to in the second part of this piece, the scalar dialectic has been and will continue to be helpful.

Instead, my hope is that the discussion in this piece encourages a ‘pause’ and critical reflection on the performative effects of this dialectic – to what it opens up and closes down, to the space it takes up in urban geographical debate and thinking, and to the kinds of relations between the urban, the researcher, and conceptualisation that might be receiving less attention as a consequence. A call to experiment and expand with ways of thinking, writing and teaching in urban geography about generic patterns, resonances, circulations, networks, recurring instances, the unique, the site, the explanatory, the descriptive, the causal, and so on – experiments that are already ‘out there’ in urban archives within and beyond academia, but which are very often drawn into the gravitation pull of the dialectic.