Environment and Planning D: Society and Space 2011, volume 29, pages 649 ^ 671

doi:10.1068/d4710

The city as assemblage: dwelling and urban space

Colin McFarlane
Department of Geography, Durham University, Science Site, Durham DH1 3LE, England;
e-mail: colin.mcfarlane@durham.ac.uk
Received 12 March 2010; in revised form 20 July 2010; published online 3 June 2011

Abstract. In this paper I consider what `assemblage' might offer a conception of the city. Although
assemblage is gaining currency in geography and beyond, there has been little effort to consider how it
might be conceptualised and what its specificity might be. In offering a conceptualisation of assem-
blage, I bring assemblage into conversation with particular debates around dwelling and argue, first,
that assemblage provides a useful basis for thinking of the city as a dwelling process and, second, that
it is particularly useful for conceiving the spatiality of the city as processual, relational, mobile, and
unequal. Despite their distinct intellectual histories, I suggest there is a productive debate to be had by
bringing assemblage and dwelling into dialogue. I examine some of the ways in which assemblage and
dwelling might interact and reflect on particular moments of fieldwork conducted in Sa¬o Paulo
and Mumbai and on diverse examples ranging from `slum' housing to urban policy and mobility.

Introduction
In an afternoon in August 2008 I accompanied Gabriella, a social worker, as she
visited projects sites in Sa¬o Paulo's large favela, Paraisöpolis. Paraisöpolis is a settle-
ment of some 70 000 people located on the steep hills of southern Sa¬o Paulo. Adjacent
to the settlement is Morumbi, an elite gated suburb that towers above the red brick
neighbourhood (Caldeira, 2000). Late in the afternoon Gabriella took a detour to
show me what she described as an entirely unique house. She was not exaggerating.
The man who came to the spectacular arched doorway knew Gabriella and narrated
the story of how he built the house through amalgamating coloured stone and just
about any object that came to hand, constructing the house from the sea of everyday
life. The roof was made up, amongst other things, of discarded pieces of plastic, old
shoes, an iron, children's toys, mugs, kitchen utensils, and all manner of other materi-
als, cemented into an intricate arrangement of decorative, coloured stones. He patted
the walls and ceiling proudly and gestured over his shoulder, revealing an interior that
was even more impressive than the exterioröa rabbit warren of smoothed and angular
stones and archways assembled through a wide range of materials, generating a prism
of light and dark as the sun shone here and there. On the roof is a garden used by his
children that continues the theme. The house was in constant construction, he said,
and the project had taken years so far. Figures 1 and 2 show some of this complex
assortment.
I should say at this point that I do not wish to romanticise this house. The house,
creative as it is, is nonetheless situated within a poor, precarious neighbourhood
where people struggle to make a safe and sustainable livelihood and are sometimes
victims of appalling abuse and violence. I want to use this example as a heuristic that
opens a wider ontological argument about the city: the city as assemblage. If this
remarkable house offers a vivid and particular example of sociomaterial assembly, it
nonetheless prompts the question of what a more general conception of the city as
assemblage might afford, of what these practices of gathering, composition, alignment,
and reuse might bring to how urbanism is conceived. What might such a conceptualisation
650 C McFarlane

Figure 1. [In colour online.] House of Paraisöpolis. (source: author's collection).

Figure 2. [In colour online.] Ceiling of house (source: author's collection).

of the city entail? In responding to these questions, the argument is presented in two
stages. First, I remain with the domain of low-income housing to make the argument
that if housing is a doingöif it is dwelt or inhabited as much as it is builtöthis
dwelling is a form of urban assembly. This characterisation of housing provides a basis,
I suggest, for thinking of the city itself as a gathering process. Second, and building on
this, I argue that the concept of assemblage is particularly useful for grasping the
spatially processual, relational and generative nature of the city, where `generative'
refers both to the momentum of historical processes and political economies and to
Dwelling and urban space 651

the eventful, disruptive, atmospheric, and random juxtapositions that characterise

urban space.
In making this argument, I bring a theory of assemblage into dialogue with a
conception of dwelling. If dwelling has experienced some theoretical resurgence in
geography (eg Elden, 2001; Harrison, 2007; Jacobs and Smith, 2008; Kraftl and
Adey, 2008; Obrador-Pons, 2003), assemblageöwhile in frequent useöremains under-
theorised. And yet there are a remarkable set of similarities between the two notions,
notwithstanding their quite distinct theoretical ancestry. For Obrador-Pons (2003),
writing in the context of Mediterranean tourism, dwelling functions as a metaphor that
draws attention to the contingent, practical, and sociomaterial nature of tourist activity,
thereby revealing the multiple constitutive spatialities that produce tourist practice.
Indeed, he argues (2003, page 51), dwelling exists through the constant and practical
weaving of human and nonhuman assemblages: ``Dwelling does not take place in
confrontation with the spatial and the material, rather it is always spatially situated
and in constant and reciprocal relationship with human and non-human networks thus
creating different types of assemblages'' [and see Jacobs and Smith (2008) on housing].
But despite the ostensible similarities, the relationship between dwelling and assem-
blage has not been examined. This is not in and of itself surprising given, first, their
different theoretical histories and, second, the relatively recent arrival of assemblageö
as what? a concept, a sensibility, an orientation?öto geographical debate. Nonetheless,
the radical relationality often found in more recent concepts of dwellingöthe shift that
Paul Harrison (2007, page 626) highlights from humanistic accounts of dwelling as
belonging and wholeness to conceptions of dwelling as posthuman or transhumanismö
holds a mirror to some recent uses of assemblage. As Harrison makes clear in his
account of dwelling in the thought of Heidegger and Levinas, dwelling often functions
as a `middle term' held between subject and world, inside and outside, private and
public, and thereby ``names the binding and the manner of this binding'' (2007,
page 628). But if dwelling is a way of naming relations, those relations always involve
a necessary and constitutive spacing; as Harrison (2007) describes it, any discourse on
dwelling is bound up with reckoning the relation to the other. Whatever differences and
uncertainties hover around the notion of assemblage, what is clear is that it similarly
operates both as a middle term and as a way of approaching, understanding, or
representing relations between multiple actors that are variously present or absent,
near or far, interior or exterior, human and nonhuman (eg Dovey, 2010).
In examining what assemblage might offer a conception of urbanism, then, I explore
an encounter between assemblage and dwelling. I contend that this is a fruitful
dialogue. For example, while there is in the concept of dwelling, even despite its
transhuman framing, a tendency to return to the centrality of the individual, assem-
blage more closely orientates attention towards agentic distributions. It is not that
assemblage necessarily decentres the human, but that it necessarily draws attention
to the constitutive human ^ nonhuman multiplicity of relations. At the same time,
however, dwelling brings to assemblage a means for thinking through how assemblage
actually takes place: ie, the processual everyday practicalities of dwelling highlight the
very acts of assembling themselves. In other words, dwelling offers assemblage a focus
on its verböie, on assembling örather than the noun, the assemblage. At stake in this
conjuncture is a particular conception of how urbanism is, or, more specifically, what
it means to consider the city as a place that is not just inhabited but which is produced
through that inhabiting. Dwelling names the constitutive processes of assemblage,
while assemblage is the spatiality of dwelling. In this respect, one important corrective
that assemblage brings to dwelling is the incessant mobility and incipient trans-
locality that defines assemblage and which offers a contrast to the often bounded,
652 C McFarlane

`earthy' baggage of connotations that dwelling carries (Hinchliffe, 2003). I present this
argument less as an attempt to argue against other conceptualisations of the city but
more as a problem-space, critically reflecting on what an ontology of city-as-assemblage
might afford. The remit of the paper, then, is by necessity wide-ranging: in attempting
to consider some of the questions and contributions assemblage might bring to the city,
I will encounter a diversity of urbanisms, from housing and policy to everyday experi-
ence. Before continuing, it is worth making a brief general statement on how assemblage
intersects with the city.
Cities are produced through processes of uneven development based on rounds of
accumulation, commoditisation, and particular geographies of biased investment and
preference that produce unequal processes of urbanisation. This historical process of
accumulation and dispossession has to be actively produced öurbanism is an unequal
achievement, and in that achievement, the past, present, and future of the city are
constantly being brought into being, contested, and rethought. Assemblage intersects
with the city precisely on this theme of making urbanism. Urbanism is produced
through relations of history and potential, ie of the multiple and interrelated tempora-
lisationsöof capital, social relations, cultures, materials, and ecologiesöthat produce
the city, but that have been and continue to be resisted and subject to alternative
possibilities. Here, assemblage does not separate out the cultural, material, political,
economic, and ecological, but seeks to attend to why and how multiple bits-and-pieces
accrete and align over time to enable particular forms of urbanism over others in ways
that cut across these domains, and which can be subject to disassembly and reassembly
through unequal relations of power and resource. For example, urban policy is
assembled not just through structures of political economy, but through particular
atmospheres of reception in the boardroom or coffee room, the materiality of policy
documents themselves (eg the agentic force of the texts and their visuals and modes of
presentation), serendipitous moments and juxtapositions, and forms of friendship or
conflict, all of which operate with different and contingent forms of power and impact.
If urbanism is an achievement, it is also a crucial domain of the possible öie of
imagining different forms of urbanismöwhether progressive, radical, conservative, or
exclusionary. Assemblage orientates the researcher to the multiple practices through
which urbanism is achieved as a play of the actual and the possible, and as such
resonates with the broader histories of critical theory and critical urbanism [there
is not space to elaborate these connections here, but for a discussion of this, see
McFarlane (forthcoming-a)]. Its object is therefore the emergent, the processual, and
the material, the city that is being made through the active and disparate labour
and resource that aligns it particular ways and that is constantly subject to being
imagined and lived differently.

Conceptualising assemblage
We can think about assemblage in a variety of ways. There is a general conception of
assemblage as a descriptive emphasis of how different elements come together. This
broadly nonconceptual sense of assemblage contrasts with a more explicit rendering
of assemblage as idea ö a name for relations between objects that make up the world,
an ontology of assemblage ö which then requires content specification. And there is
a notion of assemblage as an approach, an orientation to an object that operates as a
way of thinking the social, political, economic, or cultural as a relational processuality
of composition and as a methodology attuned to practice, materiality, and emergence.
These are not necessarily mutually exclusive positions; we can think of assemblage as
both orientation to the world (eg a form of thinking about urban policy production)
and as an object in the world (eg an urban policy, house, or infrastructure). In offering
Dwelling and urban space 653

a conception of the city as assemblage, I am working with both these senses of

assemblage as orientation and assemblage as object. But, as part of both these
positions, I am also thinking of assemblage as broadly political (ie as a way of
thinking about not just how cities are) but how cities might be otherwise (ie assem-
blage as a means of continually thinking the play between the actual and the possible)
(McFarlane, forthcoming-a).
As a general currency, assemblage is increasingly used to connote, expansively,
indeterminacy, emergence, becoming, processuality, turbulence, and the sociomater-
iality of phenomena. It is, then, part of a more general reconstitution of the social
field as materially heterogeneous and practice based (De Landa, 2006). As a descrip-
tive term for transgressing modernist dualisms like nature ^ culture, body ^ technology,
or physicalöpolitical, it often functions not just as the name for social processuality
and relationality, but as a style of knowledge production alert to compositional align-
ment and realignment (Phillips, 2006). In urban geography, assemblage has been
deployed in various ways: as a descriptor of sociomaterial transformation in accounts
of urban socionatures, cyborg urbanisms, or urban metabolisms (eg Gandy, 2005;
Swyngedouw, 2006); as a means for thinking through the contribution of actor-
network theory for rethinking the city [eg Far|¨ as's (2009) usage of assemblage as a
basis for decentering the city and rendering urbanism as a multiplicity of processes
of becoming, sociotechnical networking, and heterogeneous collectivity, building on
Deleuze and Guattari's (1981) notion of agencementöthe alignment of different elements];
and in relation to urban policy mobilities, including the relations between travelling
policies and their localised substantiations (eg Allen and Cochrane, 2007; McCann and
Ward, forthcoming; McGuirk and Dowling, 2009; Ong, 2007; Sassen, 2007).
In these very different usages, assemblage is generally deployed as a descriptive
term that signals a relational processuality of composition, whether social ^ material,
natural ^ cultural, or near ^ far. While I would wish to resist the temptation to view the
disparate usages of assemblage as an historical common field, there are, as Venn
(2006a, page 107) argues, a set of emphases that many uses of the concept share,
including ``adaptivity rather than fixity or essence ... co-articulation and compossibility
rather than linear and discrete determination ... and the temporality of processes''.
Assemblage, then, has become a vocabulary for describing the productivist alignment
of different sources, but is rarely itself an object of conceptual elaboration. As a
general definition I use the Deleuzian conception of assemblage as ``a multiplicity
constituted by heterogeneous terms and which establishes liaisons, relations between
them'' (Deleuze, 2007, page 52). For Deleuze, the only unity of assemblage is that of
``co-functioning: it is a symbiosis, a `sympathy'. It is never filiations which are impor-
tant but alliances, alloys; these are not successions, lines of descent, but contagions,
epidemics, the wind'' (page 52). This means that urban actors, forms, or processes are
defined less by a pregiven definition and more by the assemblages they enter and
reconstitute. The individual elements define the assemblage by their cofunctioning,
and they can be stabilised (territorialised or reterritorialised) or destabilised (deterri-
torialised) through this mutual constitution. But this is not to say that an assemblage
is the result of the properties of its component parts. It is the interactions between
components that form the assemblage, and these interactions cannot be reduced to
individual properties alone. As a form of spatial relationality, assemblage thinking is
attentive to both the individual elements and the agency of the interactive whole, where
the agency of both can change over time and through interactions. The changing
nature of assemblages through interactions is one of the ways in which, as Manual
De Landa (2006, page 10 ^ 11) has argued, assemblages operate as wholes characterised
by ``relations of exteriority''. The other sense in which assemblages are characterised by
654 C McFarlane

relations of exteriority is that component parts may be detached and plugged into
a different assemblage in which its interactions are different. My reading of urban
assemblage departs from three broad starting points.
First, as an orientation, assemblage is an attempt to emphasise that urban pro-
cesses are more than just the relations between sites or actors. Assemblage thinking
emphasises the depth and potentiality of sites and actors in terms of their histories, the
labour required to produce them, and their inevitable capacity to exceed the sum of
their connections (McFarlane, 2009; forthcoming-a). By `depth' I am referring both to
the crucial role of urban histories in shaping trajectories of urban policy and economy,
habits of practice, and ways of going on and to the intensity and excessiveness of the
moment, the capacity of events to disrupt patterns, generate new encounters with
people and objects, and invent new connections and ways of inhabiting everyday urban
life. `Potentially' connects to both these meanings of depth öfor instance, in the
potential of urban histories and everyday life to be imagined and put to work differ-
ently, whether in the form of blueprints, models, or hope for a better city or in the
capacity of random connections to generate the possibility of new encounters, spaces,
and collectives. Potential signals the relation between the actual and the virtual cityö
between what ostensibly is and what might be or could have beenöand thereby speaks
both to the urban imagination, the sense of possibility that the city can generate under
varying conditions of restraint and inequality, and the relations between past, present,
and future. But as a means for thinking urban relations, assemblage draws attention
less to an ecology of relations and more to the particular urban alignments formed
through processes of gathering, dispersion, and change. At particular moments and for
certain durations, different exteriorities can enter into the constitution of assemblages,
only to change or disperse at a different time or from a particular `angle of vision'
(Li, 2007; De Landa, 2006). If elements of this reading of assemblage connect with
certain conceptions of network, it is worth highlighting the broad contrasts between
assemblage and the notion of network.
In relation to actor-network theory (ANT), for example, assemblages are relations
not just of stability and rigidity, but of excess, flux, and transformation. This is not to
say, of course, that ANT studies only emphasise rigidity and stability, but that the
emphasis is often on these forms. As Ong (2007, page 5) argues in relation to assem-
blage and neoliberalism: ``Although assemblage invokes nexus, it is radically different
from concepts such as `network society' or `actor network theory' that seek to describe
a fully fledged system geared toward a single goal of maximization ... . The space of
assemblage is the space of neoliberal intervention as well as its resolution of prob-
lems of governing and living.'' Assemblage connotes transformation, or the work of
reassembling, thereby focussing attention on the possibility of invention. Invention
here operates not necessarily as something new but, as Barry (2001, pages 211 ^ 212)
has argued, as arrangements in which objects or devices become ``aligned with
inventive ways of thinking and doing and configuring and reconfiguring relations
with other actors ... . What is inventive is not the novelty of artefacts and devices in
themselves, but the novelty of the arrangements with other objects and activities within
which artefacts and instruments are situated, and might be situated in the future.''
Assemblage implies a greater conceptual openness to the unexpected outcomes of
disparate intentions and activities.
However, the conception of assemblage offered here is not one that is intended to
function as an alternative to the notion of network in ANT. Indeed, if ANT has often
been preoccupied with the stabilisation of networks, the focus has been increasingly
on an ``ethos stressing fluidity, transformation and ambivalence'' (Van Loon, 2006,
page 310). But as Legg (2009) points out, even ANT's main protagonist, Bruno Latour,
Dwelling and urban space 655

has distanced himself from the persistent tendency to connote network with rigidity
and to undermine the complexity of relations such as structure/complexity or human/
nonhuman. If assemblage differs from some ANT readings of network in that it
attends both to change and rigidity, it nonetheless exists in similar conceptual terrain
attempting to confront the complexity of sociomaterial relationality.
Second, as a conception of urbanism, urban assemblages are not simply a spatial
category, output, or resultant formation, but signify doing, performance, and events.
There is no necessary spatial template for assemblage; the spatiality of assemblage is
that of sociomaterial alignment, which brings into view a range of spatial forms, from
those generated by historical processes of capital accumulation and social polarisation
to random juxtapositions and disruptive events and predictable daily and nightly
rhythms of activity, atmosphere, and sociability. At different moments in time, relations
within and between sites and actors may require different kinds of labour and are more
or less vulnerable to collapse, or to reassembling in different forms. As Bennett (2005,
page 461) points out, drawing on Deleuze and Guattari (1986), this underlines the
agency not just of each member of the assemblage but of the groupings themselves:
the milieu, or specific arrangement of things, through which forces and trajectories
inhere and transform. In this reading the different `parts' of assemblage do not interact
atomistically but as coconstituting relations that define one another. For all that this
underlines the temporality of assemblage, it is worth noting here that for some
Foucauldian scholars the temporality of assemblage is that of the ephemeral rather
than the longue durëe (eg Legg, 2009; Li, 2007; Rabinow, 2003). For Rabinow (2003,
page 56) assemblage is an ``experimental matrix of heterogenous elements, techniques
and concepts'' that disappears in years or decades rather than centuries. Longer lasting
`problematisations'öwhat he calls `grander problematisations' öconnote less a sense
of emergence and more a sense of resultant formation, and their form is that of
the apparatus or dispositif (see Legg; 2009; Li, 2007). In contrast, while assemblage
does emphasise openness and emergence, I do not assign any particular temporality
to assemblage nor any necessary level of stability. Rather than opposing assemblage to
apparatus, I prefer to think of how forms of power, rationality, and intelligibility
structure and enclose urban assemblages, or öto use a Deleuzian language of assem-
blageöterritorialise, deterritorialise, or reterritorialise (De Landa, 2006; Dovey, 2010).
Assemblages can be made singular through the action of particularly powerful agendas
or groups, even if there is always, to varying extents, the potential to be otherwise.
Elements are drawn together at particular conjunctures only to disperse or realign.
What this begins to outline is a conception of the city as multiple assemblages of
actual and virtual urbanisms located in emergent material practice, and which are not
characterised by any necessary pregiven spatial or temporal templates.
Third, and finally, as a concept the urban assemblage is structured, hierarchised, and
narrativised through profoundly unequal relations of power, resource, and knowledge.
Rather than a kind of crude opposition to structural hierarchy, the spatialities and
temporalities of urban assemblages öfor instance, in relation to policy or development
formationsöcan be captured, structured, and storied more effectively and with greater
influence by particular actors or processes than by others. As the examples cited
above of Gandy (2005) and Swyngedouw (2006) have in their different ways vividly
demonstrated, power, political economy, and sociocultural exclusion are central to
how sociomaterial assemblages are produced. For example, Gandy (2005) shows how
the cyborg figure allows a critical purchase on connections between body, technology,
exclusion, and violence. Whether in the functioningöor nonfunctioningöof infrastruc-
tures as life-support systems; or in the sociomaterial militarisation of society through
the technologically enhanced urban soldier, the destruction of civic infrastructure, or the
656 C McFarlane

radical extension of surveillance technologies through cities (Graham, 2010); or in the

``decyborginization'' to bare life of the marginalised through violence, impoverishment,
and disease, Gandy (2005, page 32) shows how certain forms of urbanism have the
power to destroy, reduce, and enable particular forms of urban life. Swyngedouw (2006),
in his critical elucidation of urbanisation as the deterritorialisation and reterritorial-
ization of metabolic flows, argues that unequal relations of power allow particular
actors to defend and create their own urban environments along lines of class, ethnic-
ity, race, and gender. As he writes: ``Under capitalism, the commodity relation and the
flow of money veils and hides the multiple socioecological processes of domination/
subordination and exploitation/repression that feed the urbanization process and
turn the city into a metabolic socio-environmental process that stretches from the
immediate environment to the remotest corners of the globe'' (2006, page 106).
Having outlined a conception of assemblage ö as productive and unequal relations
of depth and potentiality ö in the next section I consider dwelling and its relation to
assemblage thinking.

City dweller
The house in Paraisöpolis briefly described earlier focuses our attention on urban
materialism. In relation to the house, this is a materialism of the things themselves:
what materials might be used for, how they might function out of place, whether
they are `useful' in this context, and the sheer improbability of their juxtaposition
(see Bennett, 2010). Our attention focused, these materials are nonetheless made
strange. In this urban uncanny, the materials relate only because of their alignment
as part of the roof or walls of the house. In seeing these materials used unexpectedly,
we are both reminded of their more typical functions (eg that of an egg-box or an iron)
while clearly seeing that they are defined by their new context as interrelations that
constitute part of the roof. In short, the individual artefacts are defined by a potential-
ity that exceeds their more familiar and usual functions, while also constituting a new
whole. This is central to the production of urban assemblages: that the parts constitute
a whole but nonetheless exceed that whole.
But the materiality of the house also casts up another immediate sense: that of
the builder and his or her own motivations and skills. The house was built by an
artist who has designed other areas of Paraisöpolis, including a public stairwell
(figure 3) for the settlement in a style reminiscent of Gaudi's colourful nonlinear
designs in Barcelona. We are presented, then, with a sense of the labour and craft
that has been put into this house. The seemingly random assortment of materials
suggests that the construction was built through embodied immersion rather than
abstract contemplation. In other words, the very materiality at work in this house entails
a way of seeing that is a haptic rather than a purely optical process. As Marks (1988,
page 332, cited in Gregory, 2000, page 303) has written, ``Because haptic visuality draws
upon other senses, the viewers body is more obviously involved in the process of seeing
than is the case with optical visuality.'' What this opens up, I suggest, is the question
of how human ^ material relations function in the urban environment, and what these
relations reveal is the urban as a process of dwelling-through-construction öof the city
as multiple, unfolding habitation.
Housing is often an incremental process of gradual manipulations within the urban
environmentö``tinkering'', to borrow from Annemarie Mol (2008), housing-as-a-verb,
in John Turner's (1972) memorable description. As literature on the material culture of
the home has demonstrated, home-making in its most general sense is a cumulative
process of assembly (eg Jacobs and Smith, 2008; Noble, 2004). As Jacobs and Smith
(2008, page 517) argue, the house is a set of contingent sociomaterial orderings, the
Dwelling and urban space 657

Figure 3. [In colour online.] Public stairwell, Paraisöpolis (source: author's collection).

constitutive geographies of which extend far beyond the territory of the house: ``The acts
of `housing' and `dwelling' are a coproduction between those who are housed and the
variant technologies that do the work of housing: ornaments and decorations, yes,
architecture and bricks and mortar, sanitation and communication technologies, too,
but also housing policies and practices, mortgage lending and insurance, credit scores,
and all the other lively `things' of finance.'' For Jacobs and Smith (2008, page 518), part
of the challenge here is to do away with the home ^ housing binary in order to focus
more upon the dispersed logics, practices, meanings, and experiences that perform
`home' as an ``assemblage of dwelling''. In one of Heidegger's (1971) descriptions, dwell-
ing is not just about engineering, architecture, or techne (in the Greek conception of
`letting appear'), but the raising of locations and joining of their spaces through
gathering, or assembly. What matters most about dwelling, as Heidegger suggests
in this instance in relation to housing, is that people must learn to dwell. But as the
example from Paraisöpolis suggests, the learning of dwelling is structured by stark
geographies of inequality.
For example, one nongovernmental agency working on low-income housing based
in Mumbai, where I do most of my research, describes a process of incremental
housing that takes place in the context of often severe urban poverty (Alliance, 2008,
page 6). In this passage, they describe a family of migrants who have just arrived in the
city and who have found a location for housing:
``They start out by putting up plastic sheets on poles, under which they sleep at night
and pack up during the day. Over time, corrugated metal sheets replace the plastic,
which become the walls and ceiling of the shack, to be later replaced by bricks and
mortar. Families sleep inside, outside and on the roofs of the shack. Gradually, the
roof becomes the first floor, as metal sheets are put up as walls. These eventually
658 C McFarlane

become concrete as well, with ladders or narrow, steep staircases leading up from
the outside. Further investments are put into obtaining amenities like water, elec-
tricity and drainage. The additional floors are often rented out to other migrants,
thus increasing the income of the original family. Depending on the need, conges-
tion, and rate of growth of the slum and the families, slum dwellers continue
building these incremental levels. However, they tend to be dangerous because of
the lack of proper construction materials and techniques. A lot of investment goes
into building these houses and upgrading them over time öthe cost of the material,
construction, maintenance and repair are considerable for the limited means of the
slum dwellers.''
This trajectory of incremental housing does not, of course, apply to all people
living in poor neighbourhoods in Mumbai, let alone anywhere else, and the trajectory
itself is mediated by class, caste, religion, ethnicity, gender, and other factors. The
narrative also hides the various actors that the poor must negotiate in order to get
access to different infrastructures and services, not to mention the labour and costs
involved. It masks the multiple processes through which, as AbdouMaliq Simone
(2008a, page 13) has written of African cities, ``actors come to make their mark on
collective transactions and the way in which idiosyncratic constellations of such actors
provide a workable balance between the provisional and incessantly mutating practices
required to viably `make do' in most African cities and a sense of order, if only
temporary.'' In other words, this narrative underplays the ways in which people inhabit
the assemblageöie how they live through the varying forms of porosity and closure of
the assemblage, including the possibilities that assemblage opens but which are not
part of its current alignment. (1) This notwithstanding, the central point that I want to
take from this gradual process of urban dwelling is that of urbanism as a process of
ongoing dwelling as assembly.
This process of sociomaterial engineering involves the translation of various mate-
rials into new uses over time, including roofs that become, first, floors for sleeping
then, later, spaces for renting out and ladders that shift from being access points to the
roof to stairs for a new family living in a newly built shack on the roof. In each step in
this ``chain of translation'' (Latour, 1999), materialsösuch as the ladder or the corru-
gated metal sheetsöoperate as functional systems that coordinate different domains,
from the spaces of the shack to the aspirations and desires of the inhabitants and the
availability of money and materials.
The perception of objects pertains to what objects might offer in the context of a
precarious and impoverished everyday lifeöas Cullen and Knox (1982, page 285)
argued in a searching paper on the self and the city, ``we do not discover the `usability'
of things by observing them, or by observing them and establishing their properties ö
but by the `circumspection of the dealings in which we use them' (Heidegger, 1962,
page 102)''. These `learnt objects' are not just practically useful because of what they
might afford for new relations, they also play an active part in conditioning the
possibilities of urban life. Dovey (2010) describes this processual multiplicity through
a distinction between extensive and intensive multiplicity, drawing on Deleuze and
Guattari (1987). An `extensive multiplicity' is where the constituent parts are defined
by their spatial extension and are unaffected by new additions, whereas an `intensive
multiplicity' is a multiplicity that can, to different extents, be changed by new additions:
``A house, neighbourhood or city is an intensive multiplicity. When different people
move in, new buildings or rooms are added, the sense of the larger place changes''
(Dovey, 2010, page 27).

(1) I am grateful to one of the anonymous reviewers for this point.

Dwelling and urban space 659

While the shack is built, then, it is a practice more accurately described as dwelt.
As habitat, the incremental shack is made through the current of precarious life
activities. Notwithstanding its pejorative connotations, it is in this sense that the
ubiquitous term `slum dweller' resonates. But this dwelling also offers a window
into the city more generally: the city as an ongoing chain of sociomaterial trans-
lations that affords different possibilities at different times under deeply unequal
conditions. The everyday spatialities of the people ^ object ^ environment relations
in incremental shack construction are a form of dwelling produced through the
labour of assembly. Incremental housing demonstrates that, contrary to how so-called
`informal settlements' are often perceived, improvisation is not simply ad hoc but the
product of tinkering and tweaking through urban assemblages over time [see Dovey
(2010, chapter 6) for an example]. This is the case for cities more generally, as Graham
and Thrift (2007) have shown in relation to the forms of hidden learning and adapta-
tion that go on daily in the maintenance and repair work of urban infrastructures,
from roads and electricity to water and sanitation. As they argue, it is through these
often labour-intensive sociotechnical processes of maintenance and repair that the
urban world often appears to us `ready-to-hand'.
Indeed, incrementalism, as laborious and historical accretion, is an important form
of dwelling as assembly and is common to a whole range of urban processes and
forms, from housing and policy to infrastructure and culture. As Simone (2008a,
page 28) has argued, cities:
``no matter how depleted and fragmented, still constitute platforms for trajectories
of incrementalism. Houses and limited infrastructure are added onto bit by bit; the
mobilization of family labor buys time for a small business to grow; migration is
used as an instrument to pool together savings in order to start a new economic
activity; mobile work crews are formed to dig wells, help with construction, or
deliver goods until they make enough contacts to specialize on one particular
activity.''
These forms of urban dwelling are stratified, unequal, and controlled. For instance,
some groups, Simone (2008a, page 28) continues, ``are able to organize labor, money,
and contacts to finish roads, complete water reticulation projects, or electrify their
compounds and neighborhoods, while others languish''. How, then, might this process
of dwelling as assembly be conceived? To begin to respond, we must consider the
specificity of dwelling in more detail.
The education of attention
In The Perception of Environment Tim Ingold (2000) links debates in ecological psychol-
ogy to those in phenomenologyöparticularly in the work of Heidegger (1961) and
Merleau-Ponty (2002)öin the premise that people are before all else beings-in-the-
world, although here the emphasis is not just on the person in the world, but the person
as part of the process of coming-into-being of the world. Here, the perceiving agentö
such as the Paraisöpolis sculptor with which the paper beganöis an embodied presence,
where bodily action is not merely symbolic of cultural meaning, but generates its own
set of meanings through embeddedness in sociomaterial contexts. If this is one of
Merleau-Ponty's preoccupations in [2002 (1945)] Phenomenology of Perception, Ingold's
argument shows how learning öas a changing form of perceptionöis centrally imma-
nent in the perceptual engagement of person and environment. From this perspective,
a process like house construction occurs through different and shifting ways of attuning
perception to objects and events in a process of ongoing immersion.
This immersion, which Ingold (2000, page 153) calls a ``dwelling perspective''
inspired by Heidegger and phenomenology, is ``an inescapable condition of existence''.
660 C McFarlane

In contrast to the view that people explicitly and implicitly construct the world in
consciousness before they act in it, organising sensory data through culturally specific
lensesöthe `building perspective'öIngold proceeds from practice theory [a disparate
body of work that, notwithstanding the differences within it, might include thinkers
as diverse as Heidegger, Wittgenstein, Bourdieu, and Foucault, to Butler, Taylor, and
Schatzki; see Harrison (2009), Schatzki (2001), Shove et al (2007)]. If the building
perspective states that worlds are made prior to being lived in, the dwelling perspec-
tive asserts that worlds are made, whether in imagination or `on the ground', ``within
the current of their life activities'' (Ingold, 2000, page 154). Ingold takes this perspec-
tive to architecture and buildings themselves and argues that buildings are made
through human involvement and are thereby continually under construction [and
see Kraftl and Adey (2008)]. One implication is that meaning, for instance, in
relation to a house, is ``immanent in the context of people's pragmatic engagements''
with the building; meaning is located in the relational contexts of people's ``practical
engagement with their lived-in environments'' (Ingold, 2000, pages 154, 168). As
Obrador-Pons (2003, page 49) writes of the Heideggerian dasein [being-there], dasein
``is always already amidst-the-world. Our involvement, that is, our way of dwelling in
the world is mainly practical not cognitive. Being-in-the-world is an everyday skilful,
embodied coping or engagement with the environment.''
Perception, then, creates knowledge that is practical because it is based on whatever
activity the person is currently engaged in: ``to perceive an object or event is to perceive
what it affords'' (Ingold, 2000, page 166; emphasis in original). As Lingis (1996,
page 14, cited in Harrison, 2007, page 631) argues through Heidegger: ``To see some-
thing is to see what it is for; we see not shapes but possibilities.'' As central to dwelling,
learning, then, involves not just technical competence, but developing forms of related-
ness to objects. The ways in which we know, learn, and build the city is not a trans-
mission of information, but what Ingold calls, after psychologist James Gibson, an
education of attention (Gibson, 1979, page 254; Ingold, 2000, pages 166 ^ 167). The
world, argues Ingold (2008, page 1797) in a later essay, is not just occupied by already
existing things, but inhabited, ie ``woven from the strands of their continual coming-
into-being''. We might consider, for instance, how urban infrastructure comes to matter
through particular uses and practices, as Susan Leigh Star (1999, page 380) argues in
relation to water infrastructures: ``The cook considers the water system as working
infrastructure integral to making dinner. For the city planner or the plumber, it is a
variable in a complex planning process or a target for repair.'' We are not particularly
far away here from Bourdieu's (1977) work on habitus: the ways in which people acquire
specific dispositions and sensibilities over time that are particular relations to their
environment, thereby enabling and inhibiting different kinds of learning and action.
Ingold (2000) traces this education in the production of skill as a practice-based form
of fine-tuning. If this renders dwelling in relation to the sort of craft ability that the
Paraisöpolis sculptor portrays in the built form, the implication is not that dwelling
requires skilled craft but that dwelling is the form of inhabitation.
There are two points I want to raise here about the relations between assemblage
and dwelling. First, for all that this account of dwelling emphasises the distributive
nature of agency, the tendency is to return to the centrality of the humanöof human
perception and action, albeit situated in a broader ecology. What assemblage offers
here is a stronger focus on the materialities and histories of human ^ nonhuman align-
ments that produce cities. Second, placing dwelling in the context of a theory of
assemblage provides a basis not just for describing the distributive nature of agency in
Ingold's account of dwelling, but it also disrupts any idea of dwelling as a kind of holism.
Precisely because of the sense of openness and disruption that assemblage brings, the
Dwelling and urban space 661

idea of dwelling as an ecological whole falls away against a conception of assemblage

as gathering and dispersion, actual and virtual, stable and disruptive. As a composi-
tional unityönot necessarily of internal coherence but of elements aligned together
(Li, 2007; McGuirk and Dowling, 2009)öassemblage connotes a state of becoming
and thereby of depth and potential. If the specificity of dwelling is the ongoing
education of attention to the urban environmentöa means of learning the city through
everyday engagement (McFarlane, forthcoming-b)öassemblage emerges not just as the
spatiality of this education of attention but as a means for illuminating the constitutive
role of nonhuman actors in urbanism.
Indeed, and with a view to the next section, dwelling itself can be translated from
localised lifeworld into a mobile representation. For example, a `slum' neighbour-
hood like Paraisöpolis is constantly being translated öterritorialised, deterritorialised,
and reterritorialisedöthrough different circuits within the Sa¬o Paulo municipality
or through international aid agencies. We might consider, for instance, how particular
translations produce material representations as an `inscription', a process of translation
that Latour (1999, page 306) refers to as the ``types of transformations through which
an entity becomes materialized into a sign, an archive, a document, a piece of paper,
a trace''. These inscriptions ösuch as documentary data about Paraisöpolisöoften take
the form of quantitative, empirical knowledges that by virtue of their codified form can
circulate different urban assemblages. Figure 4, for example, reveals one inscription
by the organisation Gabriela works for. Here, a photo montage provides an illustration
of an upgrade project in Paraisöpolis that is used to market the organisation's work
amongst the municipality or international donors. These kinds of inscription have
the potential both to draw different actors into an urban assemblage öpotentially
reforming that assemblage in the process öand to influence the terms of debate within

Figure 4. [In colour online.] Development project Paraisöpolis (source: author's collection).
662 C McFarlane

that assemblage, for instance, around the sorts of urban development projects that
might take place. They take a form of urban dwelling and translate it into a form
that allows it to travel and potentially act at a distance öin other words, assemblage
can mobilise the recalibration of dwelling as a mobile representation. This mobility
and translocality of assemblage brings a wider spatial matrix to how dwelling might
be understood, important considering the bounded, leaden-footed connotations that
dwelling can sometimes bring.
This inscription of Paraisöpolis represents one attempt to educate attention through a
particular kind of dwelling in Paraisöpolis: a form of urban becoming through incre-
mental, step-by-step favela improvement that is visually represented. It is one means of
communicating to and enrolling support from an `outside' world about this organisa-
tion's capacity to assemble the urban, and as such points to one of the ways in which
urban assemblages can disrupt dualisms of `interior' and `exterior' through translation.
If assemblage is the spatiality of urban dwelling, its unfolding and translations raise the
question of how this relational, coconstituted spatiality might be conceived. The next
section addresses the spatiality of urban assemblages more specifically by moving
beyond the specificity of housing to consider the spatial composition of assemblage
through different forms of inhabiting, including mobile forms.

Spatialities of urban assemblage

Where does assemblage leave a conception of urban space? In particular, what does the
conversation between dwelling and assemblage bring to conceptions of space? What is
immediately clear from the preceding discussion is that urban space must be first and
foremost understood as a relational constitution, ie urban space is produced by
assemblages ösociomaterial alignments, sometimes stable, sometimes precarious ö
that make up the continuities and discontinuities of urban life. These assemblages
take various spatial forms and functions, of which three in particular stand out. First,
assemblages can structure the possibilities of dwelling. They are not simply produced
by dwelling practices but curtail the possibilities of dwelling, meaning they can be
enabling or disruptive. Housing or infrastructure materials within informal settlements,
for instance, are central to how many people experience and cope with urban life and
poverty. Second, assemblages can, nonetheless, exceed these structural confines. This
excess takes at least two forms: as surplus both to given fields of possibility and to that
which is immediately present. They can generate random juxtapositions, they can be
virtual, imagined but not realised, and they can be seen, felt, heard, and noncognitive.
Third, assemblages have no necessary temporal or spatial template: they can be fleet-
ing, lasting, instantaneous, or laboured, and they may be near or distanciated, present
or absent.
There is, of course, a long history of urban thinking that seeks to capture the
liveliness of space that these three issues point to. From the dizzying, jarring, and
noisy scenes of movement that fill Walter Benjamin's vivid descriptions of cities like
Berlin, Naples, or Paris (eg Bullock and Jennings, 2004), or Henri Lefebvre's account
of the generative nature of space created through capitalist production or the dwell-
ing, passion, and action of everyday life, including through his (2004) project of
rhythmanalysis, to the politics of Guy Debord's antispectacle everyday urbanism
(eg McDonough, 2009), and Manuel Castells's (1983) early interventions on the possi-
bilities of grassroots movements to reshape urban spaces. Outside of the `Western
urban canon' there is an equally rich history of accounts of everyday urban spatialities.
Edgar Pieterse (2008, page 113), for instance, examines work that traces the ``below-
the-radar sets of small actions to gain a little piece of pavement, or a few square foot
of floor space, or illegally tapped kilowatts from a government-owned power line, or
Dwelling and urban space 663

enough invisibility to duplicate copyrighted goods for sale at informal markets'', while
Simone (2008b, page 200) plots how a range of `everyday transactions' ``facilitate, even
at difficult and uncertain costs, the capacities of diverse urban residents to continu-
ously make and adapt to conditions that keep the vast heterogeneities of urban lifeö
its things, resources, spaces, infrastructures and peoplesö in multiple intersections
with each other'' [and see Bayat (1997)]. In these different accounts, urbanismöto para-
phrase Ingold (2008, page 1808)ödoes not so much exist as occurs; it is a sociomaterial
achievement continuously remade through different encounters, labours, and mobilisa-
tions. These accounts focus on the liveliness of urban dwelling as a process of forced
improvisation in contexts of often profound urban inequality.
In my own research in Mumbai I am often struck by the difficulty of describing
the spatial discontinuities of the city's eventful and disruptive everyday encounters, the
changing daily and nightly rhythms of recurring practices and random thrown together-
ness, the juxtaposition of cacophonies of noise, exhaustion, and peace that characterise
so many of its neighbourhoods, and the atmosphere of precarious sociability that is
generated in the city's streets in often sedate evenings ö in short, some of the ``key
qualities of a city-type ambience, of `citiness' '' (Healey, 2002, page 1780). Indeed, if
dwelling is an education of attentionöan atunement of perception through the urban
environmentöit is through combinations of tactile, sensual, and explicit knowledges
that we come to learn and relearn the city through dwelling, what Ingold and Kurtilla
(2000, page 189) refer to as a ``multisensory awareness of the environment'' central to
``spatial orientation and the coordination of activity'' (see Crang, 2001; Middleton,
2009). Assemblage thinking is alert, then, to a temporal dynamism, where the urban
street, for instance, is a place of changing density and quietude, altering functions
(shopping, protest, sociality, passage, etc), growth and death, changing aesthetics, and
the ``flows of life, traffic, goods and money that give the street its intensity and sense
of place'' (Dovey, 2010, page 16).
The challenge here is to come to terms with the spatiality of urban assemblage as
a constant relational coproduction in which the possibilities for dwelling the city are
altered, generated, and delimited across multiple temporalities. Importantly, rather
than figuring this spatiality as localised and bounded, assemblage brings to a con-
ception of dwelling a processual and mobile conception of urban spatiality. This
mobility and translocality of dwelling is crucial, and it unsettles the heavy connota-
tions of localism and rootedness that dwelling carries. Indeed, Ingold (2008) has
come to regret his emphasis on dwelling, preferring now to speak of inhabiting. Citing
Steve Hinchliffe's (2003) concerns with the notion of dwelling, Ingold acknowledges
that despite his effort to transcend a notion of dwelling as a kind of territorial and
earthly romanticism, ``it is, nevertheless, true that the concept of dwelling carries a
heavy connotation of snug, well-wrapped localism'' (2008, page 1808). But rather than
dismiss the notion of dwelling, another route is to consider how we might rethink
dwelling through mobility. Again, the conceptual openness of assemblage to multiple
spatialities and to no one particular spatial template offers an important corrective
here.
Mobile dwelling
The city is increasingly conceived of as relational and coconstituted (eg Amin and
Thrift, 2002; Massey, 2005; 2007; 2011). The starting point of this relational perspective
is that it is impossible to understand cities as territories prior to their engagements with
other places (Massey, 2011). As Ash Amin (2002, page 397) has put it, the city ``is a place
of engagement in plural politics and multiple spatialities of involvement''. This has been
demonstrated, for example, through literature on urban policy travel, for instance,
664 C McFarlane

in studies of the mobility of particular policies, ways of thinking the city, imaginaries
of urban futures, urban plans, or military urbanisms [see, for contrasting examples,
Graham (2010), King (2004), McCann and Ward (2011), Nasr and Volait (2003)].
Emerging work on what Eugene McCann calls `urban policy mobilities' (eg McCann,
2007; McCann and Ward, 2009) is one important example here. This disparate work
has considered, for instance, how certain cities draw on particular policy discourses
of urban redevelopment in their plans for the city, whether in the circulation of
urban policy knowledges, discourses of `knowledge cities', or neoliberal, revanchist
and punitive ideologies (eg see McCann and Ward, 2009; MacLeod, 2002; Peck,
2005; 2006; Peck and Theodore, 2010; Smith, 1996; Swanson, 2007; Wacquant, 2008;
Ward, 2006).
If travelling urbanism is a far from new phenomenon, as McCann and Ward (2011)
argue, urbanism is nonetheless increasingly assembled through a variety of sites,
people, objects, and processes. For McCann and Ward, the relational mobilities of
urbanism and the territorial histories of specific places coconstitute, and they are
best understood as urban assemblages. One challenge here, then, is unpacking the
material geographies through which cities are actively reassembledöa process that
occurs most profoundly in planning and policy contexts, and one which is increasingly
a process of translocal encounters. As McCann and Ward (2011) argue, contemporary
policy making increasingly involves a scanning of the policy landscape and churning of
ideas from elsewhere, including through publications and reports, websites, blogs, the
media, contacts, and word-of-mouth as `best practices' and `off-the-shelf ' policies travel
translocally. These policy mobilities interact with and are translated through local
histories and policy contexts in complex ways, from the circulation of revanchist
policies through cities like New York to Sa¬o Paulo or the travelling free-market ideo-
logies propagated by think-tanks like the Manhattan Institute or Heritage (Peck, 2006)
to variants of urban entrepreneurialism drawn from seductive `success' stories like
Shanghai and Singapore imported to Mumbai or Delhi, or the widespread ``sale of
community and boutique lifestyles'' that accompanies the `new urbanism' movement
for city centre remodelling (Harvey, 2008, page 32).
One challenge here lies in conceptualising these mobile formations in ways
that move beyond spatial distinctions of integrating parts such as local ^ global or
territorial ^ relational, to a conception of the city as a dwelling process that aligns
multiple space-times of knowledge, ideas, materials, resources, and people. As Urry
argues, the increasingly mobile nature of the worldöand of cities in particularö
requires a recontextualisation of dwelling from the sorts of hunter ^ gatherer contexts
that Ingold has often described: ``The emergence of new, often more or less instanta-
neous, mobilities mean that the patterns of dwelling described by Ingold require
extensive reconceptualisation'' (2000, page 136). Urbanism is increasingly dwelt in
and through differential mobilities. Assemblage thinking is one way of attempting to
overturn the spatially rooted sense of dwelling and the ``false naturalism'' that equates
locality with a stable sense of being (Rajchmann, 1998, page 86; Dovey, 2010, page 24).
A key concern here lies in elucidating the forms of power at work across distances
as urban policy travels. This attends to the third dimension of assemblage I outlined
early in the paper, ie the crucial processes through which assemblages are structured,
hierarchised, and narrativised through profoundly unequal relations of power,
resource, and knowledge. John Allen's (2004) work on power is particularly instructive
here. Allen's work reminds us of the need not to confuse the capacity to do something
with the actual exercise of power. As Allen (2004) has argued, power's operation
across space is provisional rather than determined in advance because what happens
between `here and there' makes a difference to the workings of power. What this
Dwelling and urban space 665

underlines is the importance of attending to the geographies of sociomaterial agency,

of how different forms of power function in the constitution of urban assemblages.
Allen (2004) points to a range of powers here, including domination, authority,
manipulation, inducement, coercion, seduction, and instrumental and associational
powers, all of which are different in their character and reach. For example, a Power-
Point presentation by an urban policy consultant might have an immediate effect
based on seduction ö where the possibility of rejection is central to its efficacy ö and
inducement ö the use of incentives, such as predictions of urban growth ö while a
World Bank conditionality loan to an urban municipality might entail coercion
through monitoring and target-setting. Each of these forms of power can work
simultaneously and can have different geographies ö for example, authority may
work particularly well at close contact because of a need for recognition, while
seduction might work well across large distances, such as in the attractiveness of urban
design or visuals of urban futures. Here, the challenge is to attend to the material
contexts of particular places and to the range of spatialities that connect `here' and
`there', because they are of substantive importance to how power takes shape. One
task, then, is to expose the different ways in which power functions through urban
assemblages across space and delineates particular forms of dwelling through an
ongoing range of inclusions, exclusions, and contestations.
As important as the critical policy mobility literature is, both the policy mobility
literature and indeed the mobilities literature more generally have tended to focus on
the mobilities of elite groups. To be sure, the mobile dwelling through assembly that
increasingly characterises the work of urban policy-makers can have a particularly
powerful role in shaping the nature of urban life and justice, but there is a risk in these
debates of implicitly producing a binary of globally mobile capital and policy set
against rooted, bounded poor urbanites. In other words, the mobility of urban assem-
blage is assigned to the globe-scanning translocal policy elite, while the city's neglected
poor are cast the territorially restricted role of dwelling. In contrast, we might think
the urban mobility of the city through, for instance, the impoverished migrants who
construct the low-income housing described earlier in the paper, and who themselves
must assemble particular urbanisms in order to get by. Here, the politics of mobility
is thrown into sharp relief, for instance, in the ways in which `pavement dwellers'
are singled out in Mumbai through an elite and legalistic discourse of mobility
and transiency, where mobility is enrolled in a deeply immoral justification for the
demolition of `slum' settlements (McFarlane, 2004).
From the perspective of the policy elite of Mumbai who are desperately trying
to transform Mumbai into a so-called `world-class city' that explicitly apes the
politicocorporate agenda of cities like Shanghai or Singapore, pavement or slum
dwellers exhibit the wrong sorts of mobilities. But if the spatialities that the urban
poor assemble through dwelling are often at odds with politicocorporate attempts to
assemble the elite city, we need also to be careful of avoiding the conceptual trap of
imagining the spatialities of policy makers as `global' and those of the poor as `local'
(eg Featherstone, 2008). Movements like Slum/Shack Dwellers International (SDI), for
instance, which began in part through the activism of Muslim women living on the
pavements of central Mumbai and which now stretches over twenty countries, reveals
some of the ways in which `slum dwellers' form what I have elsewhere called `translocal
assemblages' (McFarlane, 2009). These translocal assemblages are constituted by
mobilities that connect particular slum activist ideas and strategies and translate
them through local sites and histories. Through what SDI leaders call `horizontal
exchanges', groups of the urban poor travel from one informal settlement to another
in what amounts to an informal learning process around low-income housing
666 C McFarlane

construction that moves through cities as different as Cape Town, Phnom Penh, and
Mumbai. SDI's work constitutes relational assemblages that combine the codified
and tacit, the social and the material, and the `here' and `there', the object of which is
to discuss, formulate, and practice better forms of urban dwelling in the domains of
housing and infrastructure. SDI's work involves the education of attention through
learning in groups around housing construction; it is a form of urban dwelling con-
stituted through mobility and territorialisation. Like the inscription of Paraisöpolis
described earlier, SDI creates representations of urban dwelling and makes them mobile
in order to act at a distance through translocal assemblages.
To echo Doreen Massey (2005, pages 100, 189), the conception of space at work in
these translocal urban assemblages is less space as resultant formation and more space
as a ``multiplicity of stories-so-far'', the ``openended interweaving of a multiplicity of
trajectories (themselves thereby in transformation), the concomitant fractures, ruptures
and structural divides''. If pavement and slum dwellers on the streets of Mumbai
are singled out for harassment and demolition partly because the state and media
constructs them as mobile and transient, SDI responds by mimicking övia different
speeds and routesöthe translocal spatialities of globe-trotting urban policy consultants
as a basis for a more progressive, socially just urbanism. But rather than the slick
PowerPoint presentations of Richard Florida's (2005) elitist `creative city' thesisö
which has, notes Peck (2005, page 740), taken on a ``seductive'' appeal for civic leaders
``from Singapore to London, Dublin to Auckland, Memphis to Amsterdam''öSDI
seeks to develop (not, to be sure, without its own inequalities in power relations or
resource and information control) a more reflective, dialogic form of the education
of urban attention rooted in solidarity and exchange, even if that form of solidarity
is to a particular and sometimes conservative conception of the poor as `skilled' and
`entrepreneurial' (McFarlane, 2004; forthcoming-b).
But this argument about mobile dwelling through assemblage can be pushed
further. Behind this discussion of mobility, space, and the city is the broader reality
of displacement as constitutive of the urban experience. However we might conceive
contemporary globalisation, one of its defining features is the intensification of dis-
placement of people through migration between cities. Cities have long been transit
points of cultural exchange, but in many cities this very mobility has come to be viewed
as a `problem' that needs to be addressed (Venn, 2006b). Assemblage offers one spatial
image of cities as incubators of mobile socialities, where the reterritorialisation of the
city is predicated on multiple durations of spatiotemporal displacement. The city
emerges here as a historically relational composite of cohabiting spatiotemporalities
where the condition of possibility of dwelling rests in part upon this very discrepant
mobility. As Venn (2006b, page 47) writes of the material geographies of diasporic
groups within cities, the unfolding of urban lifeworlds emerges from long histories of
pluralised mobility, ``evidenced in the everyday in the decor and style of houses and
buildings, or indeed in graffiti, and, generally, in the translation of place through their
reiteration (their repetition with-and-in difference) in a new spatialityösay, China
Town in Los Angeles, or Little India in Singapore, New England in New Yorköis a
form of memorialisation and socialisation of living space''. Far from conceiving this
constitutive mobility of dwelling as a problem, Venn's account stresses the advantage of
the assemblage concept for envisaging disequilibrium as normal and of dwelling in the
city as an ongoing spatiality of translation and mutation.
Dwelling and urban space 667

Conclusion
If assemblage is increasingly used as a descriptor of urban production and change,
there has been little attempt to consider what assemblage might specifically offer a
conception of the city. In this paper I have offered three broad orientations and
conceptualisations that assemblage emphasises. First, assemblage is an attempt to
emphasise that urban processes are more than just the relations between sites or actors.
In contrast to accounts of urban `networks', assemblage places greater emphasis on the
depth and potentiality of sites and actors in terms of their histories, the labour required
to produce them, and their inevitable capacity to exceed the sum of their connections.
This is assemblage as an orientation. Second, urban assemblages are not simply a
spatial category, output, or resultant formation, but signify doing, performance, and
events. Again, there is no necessary spatial template for assemblage; assemblage calls
into view a range of spatial forms, from those generated by historical processes of
accumulation and sociospatial polarisation or the distanciated mobilities of travelling
policies to random juxtapositions and disruptive events and predictable daily and
nightly rhythms of activity, atmosphere, and sociability. This is assemblage as a
conceptualisation of urban objects, such as policy, social movements, or everyday
urban culture. Third, and finally, urban assemblages are structured, hierarchised, and
narrativised through profoundly unequal relations of power, resource, and knowledge.
Again, this is assemblage as a conceptualisation of urban objects. In short, then, I have
offered a reading of assemblage as productive and unequal relations of depth and
potentiality and considered it through a range of urban examples, from housing
and the everyday to policy and displacement. The city emerges here as a series of
more or less open assemblages, structured by a range of forms of power, capital,
discourse, and groups but always exceeding those structures and always with differ-
ential capacities to become otherwise. In doing so, I have argued that debates on
dwelling are particularly instructive.
Despite their distinct intellectual histories, assemblage and dwelling have much in
common. They share, in their more recent formulations, an emphasis on the processual
composition of the world through sociomaterial practices. The interpretation of dwell-
ing offered in this paper, which departs from a reading of Ingold (2000) in particular,
draws specific attention to urban living as an education of attention, an atunement of
perception to what urbanismöin conditions of often extreme inequality of different
sortsömight enable and delimit and to how people might negotiate it. If dwelling in
this conception recentres the human, assemblage insists upon the sociomaterial dis-
tributions through which the human enters into the constitution of urbanism. While
dwelling draws attention to how assemblages are constituted öie as a particular mode
of assembling öthe role of assemblage here is to shift attention away from the central-
ity of the human to the agentic role of sociomaterial allocation that make cities. But,
importantly, this does not leave dwelling as a kind of holism. Instead, assemblage
provides a basis for not just describing the distributive nature of agency in Ingold's
account of dwelling, but for infusing it with a sense of openness and disruption, actual
and virtual, stability and transformation. In addition, if dwelling connotes a bounded
localism, assemblage offers a spatial corrective through an openness to how relations
are gathered together, aligned, transformed, or dispersed. There are three summary
points, then, to make in relation to this encounter between assemblage and dwelling.
First, dwelling names the everyday practicalities that constitute assemblages, and
the perceptual fields that are both shaped by and in turn alter those practical enactments.
Second, assemblage positions the humanöa category that dwelling, if not in terms of its
content but as a mode of analysis, continually recentresöwithin broader sociomaterial
distributions that produce urbanism. And third, as the spatiality of dwelling, the focus
668 C McFarlane

on mobility and translocality that assemblage brings disrupts the bounded, grounded
baggage that dwelling connotes, opening a wider imaginary of urban spatial topology.
What, then, might be the implications of this conception of the city as dwelling
through assemblage? In closing, I wish to highlight three implications in particular.
First, an ontology of the city as assemblage insists upon space as a key constitutive
of urbanism. The focus on making space through sociomaterial dwelling, including
transforming spatialities of inhabitation that function through an education of atten-
tion, or mobile forms of dwelling represented through policies, blueprints, models, and
migration, positions space as far more than a simple variant of or supplementary to
urbanism, but as central to how it is produced, lived, structured, and remade. Second,
and following this, urbanism exists only through the process of inhabiting the city,
where inhabiting refers both to everyday forms of education of attention and to the
mobile constitution of urbanism, which are the product of historical accretion and
alignment. The city, then, is inhabited, but is also more importantly a multifarious
set of processes of inhabiting, of making and remaking urbanism through sociomateri-
alities of near and far, actual and virtual, the everyday and the long durëe. The city's
continual coming-into-being signals urbanism as a kind of achievement, but one that is
deeply unequal and dependent on a whole range of exclusions, resources, histories, and
forms of power. For urban geographers, one central question here is how the changing
geographies of urban inhabitation are produced, lived, contested, and transformed.
Third and finally, a conception of the city as assemblage prompts the question of
who and what has the capacity to assemble the city. Urban politics is framed in large
part through different ideologies of urban dwelling, which have consequences for
both the range of possibilities for dwelling and the sorts of dwelling that come to
be valued. For instance, current debates about the `creative' or `smart' city value
particular exclusive groups, spaces, and forms of urban development, particularly
around well-educated elites living in premium residential spaces and working in
high-end service economies, including in particular science, technology, research,
media, and finances. For urban geographers, a conception of assemblage can serve
to expose which groups and ideologies have the greater capacity to render urbanism in
particular ways over others, and therefore offers a ground for thinking how the city
might be assembled differently. At stake here is the critical relationship between the
actual and the virtual city, between the city that is and the city that might have been
or that might otherwise arise.
References
Allen J, 2004, ``The whereabouts of power: politics, government and space'' Geografiska Annaler,
Series B 86(1) 19 ^ 32
Allen J, Cochrane A, 2007, ``Beyond the territorial fix: regional assemblages, politics and power''
Regional Studies 41 1161 ^ 1175
Alliance 2008 Citywatch: A SPARC ^ NSDF ^ Mahila Milan Publication (SPARC, Mumbai)
Amin A, 2002, ``Spatialities of globalisation'' Environment and Planning A 34 385 ^ 399
Amin A, Thrift N, 2002 Cities: Reimagining the Urban (Polity Press, Oxford)
Barry A, 2001 Political Machines: Governing a Technological Society (Athlone, London)
Bayat A, 1997 Street Politics: Poor People's Movements in Iran (Columbia University Press,
New York)
Bennett J, 2005, ``The agency of assemblages and the North American blackout'' Public Culture
17 445 ^ 465
Bennett J, 2010 Vibrant Matter: A Political Ecology of Things (Duke University Press, London)
Bourdieu P, 1977 Outline of a Theory of Practice (Cambridge University Press, Cambridge)
Bullock M, Jennings M W, 2004 Walter Benjamin: Selected Writings, Volume 1, 1913 ^ 1926
(Harvard University Press, Cambridge, MA)
Caldeira T, 2000 City of Walls: Crime, Segregation, and Citizenship in Sa¬o Paulo (University of
California Press, Berkeley, CA)
Dwelling and urban space 669

Castells M, 1983 The City and the Grassroots: A Cross-cultural Theory of Urban Social Movements
(University of California Press, Berkeley, CA)
Crang M, 2001,``Rhythms of the city: temporalised space and motion'', in Timespace: Geographies
of Temporality Eds J May, N Thrift (Routledge, London) pp 187 ^ 207
Cullen J, Knox P, 1982, ``The city, the self and urban society'' Transactions of the Institute of
British Geographers, New Series 7 276 ^ 291
De Landa M, 2006 A New Philosophy of Society: Assemblage Theory and Social Complexity
(Continuum, London)
Deleuze G, Guattari F, 1981, ``Rhizome'' Ideology and Consciousness 8 49 ^ 71
Deleuze G, Guattari F, 1987 A Thousand Plateaus (University of Minnesota Press, Minneapolis,
MN)
Deleuze G, Parnet C, 2007 [1977] Dialogues II (Columbia University Press, New York)
Dovey, K, 2010 Becoming Places: Urbanism/Architecture/Identity/Power (Routledge, New York)
Elden S, 2001 Mapping the Present: Heidegger, Foucault and the Project of a Spatial History
(Continuum, London)
Far|¨ as I, 2009, ``Introduction: decentring the object of urban studies'', in Urban Assemblages:
How Actor-network Theory Changes Urban Studies (Routledge, London) pp 1 ^ 24
Featherstone D, 2008 Resistance, Space and Political Identities: The Making of Counter-global
Networks (Wiley-Blackwell, Oxford)
Florida R, 2005 Cities and the Creative Class (Routledge, New York)
Gandy M, 2005, ``Cyborg urbanization: complexity and monstrosity in the contemporary city''
International Journal of Urban and Regional Research 29 26 ^ 49
Gibson J, 1979 The Ecological Approach to Visual Perception (Houghton Mifflin, Boston, MA)
Graham, S, 2010 The City Under Siege: The New Military Urbanism (Verso, London)
Graham S, Thrift N, 2007, ``Out of order: understanding repair and maintenance'' Theory, Culture
and Society 24 1 ^ 25
Gregory, D, 2000, ``Cultures of travel and spatial formations of knowledge'' Erdkunde 54 297 ^ 309
Harrison P, 2007,``The space between us: opening remarks on the concept of dwelling'' Environment
and Planning D: Society and Space 25 625 ^ 647
Harrison P, 2009, ``In the absence of practice'' Environment and Planning D: Society and Space
27 987 ^ 1009
Harvey D, 2008, ``The right to the city'' New Left Review number 53, 23 ^ 40
Healey P, 2002, ``On creating the `city' as a collective resource'' Urban Studies 39 1777 ^ 1792
Heidegger M, 1962 Being and Time (Harper and Row, New York)
Heidegger M, 1971, ``Building dwelling thinking'', in Poetry, Language, Thought Ed. A Hofstadter
(Harper and Row, New York) pp 143 ^ 162
Hinchliffe S, 2003, `` `Inhabiting': landscapes and natures'', in Handbook of Cultural Geography
Eds M Domosh, S Pile (Sage, London) pp 207 ^ 225
Ingold T, 2000 The Perception of the Environment: Essays in Livelihood, Dwelling and Skill
(Routledge, London)
Ingold T, 2008, ``Bindings against boundaries: entanglements of life in an open world'' Environment
and Planning A 40 1796 ^ 1810
Ingold T, Kurtilla T, 2000, ``Perceiving the environment in Finnish Lapland'' Body and Society 6
183 ^ 196
Jacobs J, Smith S, 2008, ``Living room: rematerialising home'' Environment and Planning A 40
515 ^ 519
King A D, 2004 Spaces of Global Cultures: Architecture, Urbanism, Identity (Routledge, London)
Kraftl P, Adey P, 2008, ``Architecture/affect/inhabitation: geographies of being-in buildings''
Annals of the Association of American Geographers 98 213 ^ 231
Latour B, 1999 Pandora's Hope: Essays on the Reality of Science Studies (Harvard University
Press, Cambridge, MA)
Lefebvre H, 2004 Rhythmanalysis: Space, Time and Everyday Life translated by S Elden, G Moore
(Continuum, London)
Legg S, 2009,``Of scales, networks and assemblages: the League of Nations apparatus and the scalar
sovereignty of the Government of India'' Transactions of the Institute of British Geographers,
New Series 34 234 ^ 253
Li T M, 2007, ``Practices of assemblage and community forest management'' Economy and Society
36 263 ^ 293
Lingis A, 1996 Sensation: Intelligibility in Sensibility (Humanities Press, Atlantic Highlands, NJ)
670 C McFarlane

McCann E J, 2007, ``Expertise, truth, and urban policy mobilities: global circuits of knowledge
in the development of Vancouver, Canada's `four pillar' drug strategy'' Environment and
Planning A 40 805 ^ 904
McCann E, Ward K, 2009, ``Relationality/territoriality: toward a conceptualization of cities in
the world'' Geoforum 41 175 ^ 184
McCann E, Ward K (Eds), 2011 Mobile Urbanism: City Policymaking in the Global Age (University
of Minnesota Press, Minneapolis, MN)
McDonough T (Ed.), 2009 The Situationists and the City (Verso, London)
McFarlane C, 2004, ``Geographical imaginations and spaces of political engagement: examples
from the Indian Alliance'' Antipode 36 890 ^ 916
McFarlane C, 2009, ``Translocal assemblages: space, power and social movements'' Geoforum 40
561 ^ 567
McFarlane C, forthcoming-a, ``Assemblage and critical urbanism'' City
McFarlane C, forthcoming-b Learning the City: Knowledge and Translocal Assemblage (RGS-IBG
Blackwell-Wiley, Oxford)
McGuirk P, Dowling R, 2009, ``Neoliberal privatisation? Remapping the public and the private
in Sydney's masterplanned residential estates'' Political Geography 28 174 ^ 185
MacLeod G, 2002, ``From urban entrepreneurialism to a `revanchist city'? On the spatial injustices
of Glasgow's renaissance'' Antipode 34 602 ^ 624
Marks L, 1988, ``Video haptics and erotics'' Screen 39 331 ^ 348
Massey D, 2005 For Space (Sage, London)
Massey D, 2007 World City (Polity Press, Cambridge)
Massey D, 2011, ``A counterhegemonic relationality of place'', in Mobile Urbanism: City
Policymaking in the Global Age Eds E McCann, K Ward (University of Minnesota Press,
Minneapolis, MN)
Merleau-Ponty M, 2002 [1945] Phenomenology of Perception translated by Colin Smith (Routledge,
London)
Middleton J, 2009, `` `Stepping in time': walking, time, and space in the city'' Environment and
Planning A 41 1943 ^ 1961
Mol A, 2008 The Logic of Care: Health and the Problem of Patient Choice (Routledge, London)
Nasr J, Volait M (Eds), 2003 Urbanism: Imported or Exported? (Wiley Blackwell, Oxford)
Noble G, 2004, ``Accumulating being'' International Journal of Cultural Studies 7 233 ^ 256
Obrador-Pons P, 2003, ``Being-on-holiday: tourist dwelling, bodies and place'' Tourist Studies 3(1)
47 ^ 66
Ong A, 2007, ``Neoliberalism as a mobile technology'' Transactions of the Institute of British
Geographers, New Series 32 3 ^ 8
Peck J, 2005, ``Struggling with the creative class'' International Journal of Urban and Regional
Research 29 740 ^ 770
Peck J, 2006, ``Liberating the city: between New York and New Orleans'' Urban Geography 27
681 ^ 713
Peck J, Theodore N, 2010, ``Mobilizing policy: models, methods and mutations'' Geoforum 41
169 ^ 174
Phillips J, 2006, ``Agencement/assemblage'' Theory, Culture and Society 23 108 ^ 109
Pieterse E, 2008 City Futures: Confronting the Crisis of Urban Development (Zed Books, London)
Rabinow P, 2003 Anthropos Today (Princeton University Press, Princeton, NJ)
Rajchmann, J, 1998 Constructions (MIT Press, Cambridge, MA)
Sassen S, 2007 Territory, Authority, Rights: From Medieval to Global Assemblages (Princeton
University Press, Princeton, NJ)
Schatzki T, 2001 The Practice Turn in Contemporary Theory (Routledge, London)
Sheller M, Urry J (Eds), 2004 Tourism Mobilities: Places to Play, Places in Play (Routledge, London)
Shove E, Watson M, Hand M, Ingram J, 2007 The Design of Everyday Life (Berg, Oxford)
Simone A, 2008a, ``Emergency democracy and the `governing composite' '' Social Text 26(2) 13 ^ 33
Simone A, 2008b,``The politics of the possible: making urban life in Phnom Penh'' Singapore Journal
of Tropical Geography 29 186 ^ 204
Smith N, 1996 The New Urban Frontier: Gentrification and the Revanchist City (Routledge, London)
Star S L, 1999, ``The ethnography of infrastructure'' American Behavioural Scientist 43 377 ^ 391
Swanson K, 2007, ``Revanchist urbanism heads south: the regulation of indigenous beggars and
street vendors in Ecuador'' Antipode 39 708 ^ 728
Swyngedouw E, 2006, ``Circulations and metabolisms: (hybrid) natures and (cyborg) cities''
Science as Culture 15 105 ^ 121
Dwelling and urban space 671

Turner J F C, 1972,``Housing as a verb'', in Freedom to Build Eds J F C Turner, R Fichter (Macmillan,

New York) pp 148 ^ 175
Urry J, 2000 Sociology Beyond Societies: Mobilities for the Twenty-first Century (Routledge,
London)
Van Loon J, 2006, ``Network'' Theory, Culture and Society 23 307 ^ 314
Venn C, 2006a, ``A note on assemblage'' Theory, Culture and Society 23 107 ^ 108
Venn C, 2006b, ``The city as assemblage: diasporic cultures, postmodern spaces, and biopolitics'',
in Negotiating Urban Conflicts: Interaction, Space and Control Eds H Berking, S Frank, L Frers,
L M Low, S Steets, S Stoetzer (Transaction, London) pp 41 ^ 52
Wacquant L, 2008,``The militarization of urban marginality: lessons from the Brazilian metropolis''
International Political Sociology 2 56 ^ 74
Ward K, 2006, `` `Policies in motion', urban management and state restructuring: the trans-local
expansion of Business Improvement Districts'' International Journal of Urban and Regional
Research 30 54 ^ 70

ß 2011 Pion Ltd and its Licensors

Conditions of use. This article may be downloaded from the E&P website for personal research
by members of subscribing organisations. This PDF may not be placed on any website (or other
online distribution system) without permission of the publisher.