International Journal of Drug Policy 33 (2016) 15–20

Contents lists available at ScienceDirect

International Journal of Drug Policy

journal homepage: www.elsevier.com/locate/drugpo

Commentary

Assemblages, territories, contexts

Cameron Duff *
School of Management, RMIT University, Australia

A R T I C L E I N F O A B S T R A C T

Article history: Human geographers have been at the forefront of efforts across the social sciences to develop ‘‘assemblage
Received 2 June 2015 thinking’’, applying and extending this model in a series of highly original empirical studies. This
Received in revised form 8 September 2015 commentary assesses some of the conceptual, methodological and procedural implications of this research
Accepted 14 October 2015
for contemporary drug studies. I will argue that the most useful way of approaching assemblage thinking in
the analysis of drug problems is to focus on the ways assemblages draw together social, affective and
Keywords: material forces and entities. I will briefly review these three nodes before indicating how their analysis may
Assemblage
inspire novel empirical assessments of drug assemblages. I will conclude by exploring how the assemblage
Deleuze
may replace the ‘subject’ and ‘social context’ as a discrete unit of analysis in drug studies.
Affective geographies
Context ß 2015 Elsevier B.V. All rights reserved.
Drugs

As an object of social science inquiry, the consumption of alcohol
and other drugs (AOD) is almost always situated as a problem with
specific personal, social, economic and political consequences. This
is as true of epidemiological research that seeks to clarify the
incidence and prevalence of AOD use and its sequela in a given
population, as it is of social research that seeks to understand this
consumption by way of its cultural and political aspects (Fraser,
Moore, & Keane, 2014). What these approaches share, beyond the
articulation of particular kinds of health and social problems, is an
epistemological commitment to the ontological separation of
individuals from the social contexts, and the differentiation of drug
objects from cultural practices of consumption. Each approach
acknowledges the role of social factors in shaping how alcohol and
other drugs are used, as well as the problems associated with this
consumption, and so each approach is left with the challenge of
explaining how these factors actually mediate consumption in
particular instances (Fitzgerald, 2015). Bruno Latour (2005:219)
calls this the problem of ‘‘action at a distance’’. How, in other words,
do social factors held to be distal or remote from events of AOD use –
examples may include cultural norms that govern consumption
practices, public policy arrangements, legislation and its enforce-
ment, drug market dynamics or economic fluctuations – actually
transform the ways substances are consumed in a given setting? It is
* Correspondence to: School of Management, RMIT University, 445 Swanston
Street, Melbourne, Victoria 3000, Australia. Tel.: +61 3 9925 5920.
E-mail address: cameron.duff@rmit.edu.au
not enough, Latour adds, to identify associations between these
phenomena at a population level. This only yields a probabilistic
logic in which factors are more or less likely to mediate AOD use.
What is needed is a method for tracing how diverse actors, both
distal and proximate, actually intervene in events of AOD use and
somehow make a difference (Duff, 2013).
One way to do this is to dispense with the notion of distal and
proximate actors altogether, to rescind the ontological separation
of behaviours from their social contexts, and to revoke the idea of
discrete actors and forces mediating each other’s behaviour. As
Gomart and Hennion (1999) would have it, the aim is not to look at
‘who acts’ but ‘what occurs’. This paper examines the extent to
which emerging notions of ‘‘assemblage thinking’’ (Marcus & Saka,
2006) may assist with this goal, and the ways this thinking may
then be applied to studies of alcohol and other drugs. To this end, I
will briefly review recent applications of assemblage thinking in
human geography (Anderson & McFarlane, 2011) for insights into
how this approach may inform novel investigations of AOD use.
However, I will start by clarifying what I think the major benefits of
adopting this approach may be for studies of AOD use.
The assemblage as a novel unit of analysis
It should prove useful to introduce assemblage thinking by way
of its contrasts with more conventional methods of social science
inquiry, and their adoption in contemporary drug studies (see Duff,
2014). Consider the following account of a young person’s AOD use,
and its temporal and spatial trajectories:

http://dx.doi.org/10.1016/j.drugpo.2015.10.003
0955-3959/ß 2015 Elsevier B.V. All rights reserved.
16 C. Duff / International Journal of Drug Policy 33 (2016) 15–20
Simon began drinking at 14 following the divorce of his parents.
He goes to live with his Dad who is often absent from home. He
sees his Mother and sister rarely. Most of his friends drink
heavily too, although Simon often drinks alone. Simon begins
having problems at school, turning up less frequently, prefer-
ring to hang out with friends in parks. He starts smoking
cannabis and is soon offered ice [methamphetamine]. At
17 Simon is hospitalised following a violent incident at a party.
He presents for drug treatment.
This account is drawn from an ethnographic study of
methamphetamine use conducted in Melbourne, Australia (see
Duff & Moore, 2015). Ordinarily, this report might be read as a
reasonably coherent statement of ‘Simon’s’ drug problem (or
problems given the appearance of several individual substances in
this account), and some of the factors that might be said to have
mediated this problem, perhaps even caused it. Reflecting the
purview of its method of articulation, several factors are jumbled
together in this account of Simon’s AOD use; the divorce of his
parents, a change in his domestic arrangements and the subse-
quent estrangement of Simon from his mother and sister, possibly
his father too; a change in Simon’s peer networks as he disengages
from school; the prevalence of heavy, episodic alcohol use in his
peer group; the initiation of methamphetamine use and a violent
incident at a party; Simon’s enrolment in drug treatment.
Conventional social science analysis of these data would likely
concentrate on Simon’s consumption of alcohol and his rapid
transition to cannabis and methamphetamine use, such that Simon
becomes both the locus and subject of a discrete drug problem.
Simon has a drug problem and so he presents for drug treatment.
While several factors in Simon’s social context apparently
contribute to this problem – such as the breakdown of his parent’s
relationship, changes in Simon’s domestic arrangements, or the
effect of widespread AOD use in his peer group – the focus must
remain with Simon given that he is the one receiving treatment for
his problems. Always, conventional analysis is drawn back to the
subject, given both its methodological familiarity and its apparent
liability to correction.
So what of the various social factors described in this report of
Simon’s drug problems? These factors are normally granted some
mediating role, with the predilections of theoretical preference
determining which receive the greatest salience. Perhaps the
divorce of Simon’s parents demands the greatest attention in this
regard, or the normalisation of recreational drug use in Australian
youth cultures, or problems with public schooling in Victoria. The
point is that conventional analyses of AOD use make a series of
attributions of agency in problematising particular kinds of
behaviour (Fraser et al., 2014). First, (human) subjects are ascribed
particular kinds of effective agency (capacities for action,
intentionality, purpose, volition, and so on), and then broader
social and/or political factors are accorded their measure of
mediating force. The latter may include social factors or cultural
norms within peer groups, trends in parenting and changing
attitudes towards AOD use, shifts in drug markets with subsequent
changes in the availability of specific substances, or changes in the
ways schools respond to the incidence of drug problems in the
student body. These factors ostensibly mediate the incidence of
problems – they make a difference somehow – and so each may be
said to have some measure of agency. Yet no matter how much
these contextual or structural factors are said to mediate patterns
of AOD use, attention is inevitably drawn back to the subject of this
use as the primary ground of the articulation of drug problems.
Individuals have drug problems after all. This tendency may be
observed in virtually all prevailing analyses of drug use, popular
and more technical, from self-help and 12 step narratives, to
popular discourses of addiction, contemporary neuroscientific
accounts of psycho-pathology and sociological renderings of the
social contexts of consumption (Fitzgerald, 2015; Fraser et al.,
2014; Keane, 2002). All reify the subject of consumption, even as
they endorse the role of select social factors in this use.
For all the effort to highlight the manifold risk environments
(Rhodes, 2002) that subtend drug problems, little progress has
been made in articulating how these environments may be
transformed to act differently, to reduce risk and to reduce the
incidence of drug problems. This lack of progress, incidentally, is
less the fault of scholars interested in developing such lines of
inquiry, and more a reflection of the scale of the challenge. So
engrained is the habit of treating individual human subjects as the
agents of their own biographies, as the authors of their own
choices, it appears that no amount of attending to the dynamics of
power, social structure or context is ever enough to overcome it. As
a result, when it comes time to account for what might be done
about problems like AOD use, it is almost always the individual
agent that receives the greatest attention (Fraser et al., 2014;
Weinberg, 2013). The agency individual’s exhibit is familiar, and
the social sciences have recourse to varied technical apparatuses
for identifying this agency and tracing its effects (Latour, 2005).
The agency of nonhuman, or ‘‘more-than-human’’ forces such as
contexts or power, is much more difficult to articulate and
investigate empirically (DeLanda, 2006). While the social sciences
abound with reports of the force of social factors, agreement about
how these forces act, and how they may be made to act differently
is rarely obtained (Duff, 2014). Social scientists talk about the force
of context, but scarcely know how to change it.
I want to argue is this paper that one of the major reasons for
this difficulty is the intransigence of the ontological and
epistemological foundations on which it rests. For as long as
individuals are abstracted from their practices and relations – for
as long as the individual subject of AOD use is held to be
ontologically separate from and prior to the contexts of this use – it
will always be easier to defer to conventional understandings of
the force of human agency, and to therefore make individuals
mainly responsible for the events that befall them. As a result,
analysis of the social dimensions of phenomena such as AOD use
will always struggle to match the sophistication, popular
awareness and political utility of accounts that privilege the
agency and responsibility of the individual subjects of this
consumption. A quick scan of popular understandings of addiction,
and their foundations in both ‘natural’ and ‘social’ scientific
problematisations of drug use, ought to be enough to carry this
claim (Fraser et al., 2014; Keane, 2002). This is precisely the
ontological, political and empirical challenge that the assemblage
addresses; how to account for all the factors, human and
nonhuman, individual and social, that mediate or transform a
given phenomenon? (DeLanda, 2006) Assemblage thinking starts
by dismissing the ontological differentiation of subjects and
objects, individuals and contexts, and focuses instead on how
action or agency is generated in encounters. From this perspective,
there is simply no such thing as an individual body or subject, and
no such thing as a reified social context, for these phenomena are
always, already a function of many different things acting together
(DeLanda, 2006). It is for this reason that Deleuze and Parnet
(1987:51) conclude that ‘‘the minimum real unit is not the word,
the idea, the concept, or the signifier, but the assemblage’’.
Responding to these provocations is the main objective of all
‘‘embodied and affective geographies’’ (Jayne, Valentine, & Hollo-
way, 2010) of alcohol and other drugs. The goal across these
emerging geographies is to account for what actually happens in a
given event of AOD consumption, who or what acts in and through
these events, and the complex or ‘‘emergent’’ causalities that might
explain the incidence and prevalence of either safer or harmful
events of consumption (see Dilkes-Frayne, 2014; Race, 2014).
 C. Duff / International Journal of Drug Policy 33 (2016) 15–20
What is perhaps most interesting about these geographies
however, is the extent to which they have managed to sustain
the ontological eschewal of the agent/structure, individual/context
dyads that structure so much empirical analysis in the social
sciences. This, in my view, is precisely what is needed in
contemporary studies of AOD use if the ontological privileging
of the subject is to be overcome such that a broader sweep of
human and nonhuman forces may be acknowledged. Only then
may it be possible to imagine responses to drug problems that
don’t inevitably reify the subject of such use, while overwhelming
that subject with responsibility to fix these problems. In what
follows, I will first briefly describe some of the key features of
assemblage thinking before indicating how this thinking has been
adopted in recent studies in human geography. Reflecting on these
approaches, I will argue that the most useful way of approaching
assemblage thinking in the analysis of drug problems is to focus on
the ways assemblages draw together social, affective and material
forces and entities. I will briefly review these three nodes before
indicating how their analysis may inspire novel empirical
assessments of drug assemblages. I will conclude by exploring
how the assemblage may replace the ‘subject’ and ‘social context’
as a discrete unit of analysis in drug studies.
What is an assemblage?
The ‘‘realist ontology’’ (DeLanda, 2006:3–4) that informs the
analysis of assemblages does not abandon the subject, much less
the realities of social life, yet it does refuse to accept either subjects
or contexts as ontological foundations for empirical inquiry.
Subjects and their social interactions are not ‘‘given’’ in experience
as ontological invariants expressive of a particular set of ‘‘essences’’
or qualities (DeLanda, 2006:1–5). Rather, both subjects and the
social lives they participate in are the product of a more
fundamental set of relations, affects, events and processes. Hence
the interest among philosophers of the assemblage, such as Gilles
Deleuze, Manuel DeLanda and Bruno Latour, in the ontogenesis of
subjects and social organisation; their coming into being. If neither
subjects nor contexts are given in experience, they are nevertheless
assembled, organised or ‘‘bundled together’’ in unique arrange-
ments of relations, forces, matter, affects, signs and spaces (see
Latour, 2005:64–69). Subjects and contexts are made in experience
in and through the emergent coming together of heterogeneous
materials, forces, spaces, signs and bodies. This explains why
thinkers like Deleuze and Latour focus on the ontogenesis of form,
rather than the emergent forms themselves. This also explains why
the assemblage ought to be regarded as a unique unit of analysis
for empirical inquiry, rather than subjects and contexts, insofar as
the analysis of assemblages is intended to explain how particular
contexts and/or subjectivities actually hold together in experience
(Duff, 2014:128–132).
Human geographers have been among the most active of social
scientists in the adoption and development of ‘assemblage
thinking’, applying and extending this approach in a series of
highly original empirical studies (Anderson & McFarlane, 2011).
Reviewing the focus and diversity of this research, Anderson and
colleagues (Anderson, Kearnes, McFarlane, & Swanton, 2012:171–
72) note that ‘assemblage’ has been used in recent geographical
research in three ways; as a ‘‘descriptor’’ of particular kinds of
observable social forms; as an ‘‘ethos’’ that provides a particular
orientation to the analysis of social problems; and finally as a
‘‘concept for thinking the relations between stability and
transformation in the production of the social’’. I should like to
focus on this third dimension as I think it provides the most useful
orientation to empirical studies of the assemblage (Dewsbury,
2011), while also helping to indicate how assemblage thinking may
be developed in drug studies. As a concept, assemblage provides a
17
means of analysing the production of drug problems in ways that
refuse to reify the subject of such problems. It eschews the
reduction of drug problems either to deficits or ‘unruly’ passions
evident in certain individuals, or to the ‘maladaptive’, oppressive or
unhealthy effects of particular social contexts. Instead, assemblage
thinking serves to emphasise the real conditions in which drug
problems emerge by way of the entire cast of human and
nonhuman, distal and proximate forces at work in such problems.
The point is that problems such as alcoholism, addiction,
dependency or drug misuse – the distinctions here really do make a
different as we shall see – are personally and socially contingent,
meaning that they ought to be regarded as the outcome of
relations, practices, forces and processes with discrete spatial and
temporal characteristics (see DeLanda, 2006). Accordingly, the
individual incidence of drug problems may be treated as an effect
of heterogeneous entities (bodies, technologies, practices, rela-
tions), operating at varying spatial and temporal scales, that are
each subject to processes of stabilisation (territorialisation in
Deleuze’s terms) and transformation (deterritorialisation). Such is
the realist ontology that underpins assemblage thinking insofar as
this reasoning emphasises the importance of interrogating the real
conditions of emergence by which problems and their subjects and
sequela are produced.
Underscoring this ontology is a ‘‘tetravalent’’ (Deleuze &
Guattari, 1987:88) model of the assemblage and the forces of
territorialisation and deterritorialisation by which forms and
structures emerge and subside. It is worth briefly reviewing this
model to properly introduce the discussion to follow regarding
how assemblage thinking may inform novel empirical investiga-
tions of AOD use. This model comprises two axes (one horizontal
and one vertical) that combine to describe four inter-dependent
processes (or valences). The horizontal axis draws together ‘‘forms
of content’’ including ‘‘bodies, actions and passions, an intermin-
gling of bodies reacting to one another’’, with distinctive ‘‘forms of
expression’’ or ‘‘acts and statements’’ that are attributed to
corresponding forms of content (bodies and passions) in ways
that moderate their ‘‘scope of activity’’ (Deleuze & Guattari,
1987:88). All assemblages combine bodies and statements in the
creation of an intensive functional identity. An example might be
an assemblage of bodies, objects and spaces in an inner-city club,
or assembled in a drug consumption room. Each assemblage may
be characterised along one axis by forms of content (bodies, human
and nonhuman), and along another axis by forms of expression
that are both about, and potentially enunciable by, these bodies.
Each axis generates modes of individuation (or identity) for the
assemblage.
Assemblages are also characterised by a second, vertical axis,
comprising forces of territorialisation (stability) and deteritoria-
lisation (transformation or ‘lines of flight’). An obvious example
concerns the ways assemblages draw together material resources
in the deterritorialisation and reterritorialisation of place
(DeLanda, 2008). All assemblages create a territory in other words.
Yet the material elements that comprise territories cannot be
regarded as fixed in that the material elements available for the
work of territorialisation are always in motion, even if this motion
is often imperceptible. Examples include the geological motion of
tectonic plates, but also the slow transformations that characterise
the built environment of any urban street. Each of these
assemblages combines materials in the territorialisation of place,
just as this assembling is subject to countervailing forces of change
and disruption. In the first instance this involves the selection and
combination of materials out of which discrete territories are
composed, thereby establishing a ‘‘stable functional structure’’
(form of content) for the elements so combined (Deleuze &
Guattari, 1987:41). Yet this process of assembling and selection is
never completed or fixed insofar as it describes a tendency towards
18 C. Duff / International Journal of Drug Policy 33 (2016) 15–20
stabilisation rather than the final achievement of this state. For
DeLanda (2008:164), this means that all material entities, forms,
spaces and territories must be regarded as ‘‘objectively change-
able: they may undergo destabilising processes affecting their
materiality, their expressivity or both’’. This is why Deleuze and
Guattari emphasise processes of deterritorialisation and reterri-
torialisation, in that all material forms, all assemblages, remain
fluid and unstable (objectively changeable) according to the
historical, political, social and economic forces applied to, or
expressed through, them as they change.
Reviewing Deleuze and Guattari’s model of the assemblage, and
its utility for geographical research, Dewsbury (2011:149) high-
lights its capacity to ‘‘fit together all the ways in which the world is
now characterised by flows, connections and becomings, whose
functioning logic is more about folds than structures, more
complex than linear, more emergent than totalizing’’. On the
basis of this assessment, and my own reading of the assemblage
thinking of Deleuze, DeLanda and Latour (Duff, 2014), I would
stress that the most useful way of approaching the tetravalent
properties of the assemblage in novel studies of alcohol and other
drugs is to concentrate on the ways assemblages draw together
social, affective and material dimensions. Assessments of a given
‘‘drug assemblage’’ (Duff, 2014:128–132) ought to focus on how
social, affective and material dimensions are made to hold together
in particular instances. I will briefly review these three dimensions
before indicating how their analysis may inspire novel empirical
investigations of drug assemblages.
Assemblages: social, affective and material
The production of social life provides obvious examples of how
the assemblage may be used as a novel unit of analysis for
contemporary drug studies. Social life is almost always char-
acterised in terms of processes that bring together diverse entities
in some kind of shared or collective experience (DeLanda, 2006:
52–57). This understanding of the ways sociality is comprised by
entities and their collective experiences is not so different from
Deleuze and Guattari’s interest in the ways heterogeneous
elements (bodies, affects, signs, spaces, objects, forces) combine
in assemblages. According to Deleuze and Guattari (1987), the
forces by which sociality is enacted may be said to include the
asubjective desires which conjoin bodies (human and nonhuman)
in social interaction; the affects generated in such interactions,
along with the modulations in the power of acting of the bodies so
assembled; the beliefs that galvanise practical action in social
contexts, such as the beliefs that lead bodies to assemble in pursuit
of political, economic and/or social goals; as well as the power
relations involved in efforts to regulate the conduct of the varied
bodies assembled in a social mass (see also DeLanda, 2006). Each of
these forces combines in the assembling of any social entity,
encounter or social context. They are at work, for example, in the
forces assembled in crowds in the night-time economy, and in the
social settings that comprise this economy and in which alcohol
and other drugs figure as prominent objects of social interaction
and exchange (Jayne et al., 2010).
The relationality that is central to the sociality generated in
assemblages may be understood by way of Deleuze and Guattari’s
(1987) discussion of affect. All drug assemblages should be
regarded as affective entities inasmuch as affective processes
are at least partially responsible for the formation of the
assemblage. Affect is understood here in two distinctive ways.
First, affect describes an array of feeling states such as anger,
shame, fear, sorrow or happiness. Each of these states corresponds
with a specific feeling such that envy, for example, is experienced
as a qualitatively different condition than anger or sorrow.
However, Deleuze (1988:49–50) stresses that affects convey
something more than a simple concatenation of feeling states.
Affects also constitute the body’s ‘‘power of acting’’; its unique
capacity to affect (and be affected by) the world of bodies and
things that it encounters. Deleuze (1988) insists that every
encounter subtly transforms the body’s affective orientations,
either to enhance that body’s power of acting or to diminish it. This
affective modification involves a transfer of power, capacities or
action-potential between bodies (Deleuze, 1988:48–50). The body,
itself a complex assemblage of simple elements both human and
nonhuman, may in this way be characterised by the modifications
in its power of acting that result from the encounters such a body
experiences, or becomes capable of experiencing.
The focus on encounters introduces the need to examine the
material aspects of assemblages, both in terms of the materiality of
bodies, but also the varied material infrastructures that character-
ise ‘‘real experience’’ (Deleuze, 1994). I have already noted the
extent to which assemblages function by way of the generation of a
unique material territory, yet assemblages also combine other
kinds of material forms. Assemblages draw together material
resources in the stabilisation of discrete material forms such as
places, technologies and objects, along with what DeLanda
(1997:27–29) calls the varying material ‘‘exoskeletons’’ that frame
the human body (such as clothing, hardware, buildings, modes of
transport etc.). These material resources are folded into the
assemblage giving it a functional structure that will remain
relatively stable for as long as this folding process is not disrupted
too significantly (by rupture, shock, line of flight etc.). Put another
way, the processes of selection and combination by which material
elements coalesce in assemblages necessarily entail the expression
of a series of explicit functions, capacities and forms. This latter
process establishes (or seeks to determine) the material function,
meaning, purpose or identity of a given assemblage (Deleuze &
Guattari, 1987:88). An interesting example concerns the formation
of crowds and the material and expressive processes involved in
the distinctions drawn between peaceable assemblies, insur-
rectionary mobs, incipient social movements and so on. Yet as I
have noted, material processes immanent to the assemblage are
never stable, as they are affected by varying processes of change or
deterritorialisation. Too much flux and the assemblage and its
relations unfold and are replaced by other social, affective and
material forms.
Drug assemblages
So how might assemblage thinking be put to work in the
analysis of alcohol and other drugs, and their forms, patterns and
consequences? The first point is to emphasise how the assemblage
may serve as a unique unit of analysis for empirical research. Above
all else, assemblage thinking emphasises the significance of
relations, affects and materials in the conditioning of AOD use,
rather than the subjects, agents, structures and forms that populate
more conventional social research (Dewsbury, 2011:148–150).
Deleuze insists that agents and structures only make sense in
terms of their relations, and the affective transmissions (capacities
to affect, transform or act on other entities) in which specific
relations are enacted, and by which specific capacities emerge.
Entities such as subjects, agents, practices, norms, groups or
collectives, structural forms and organisational processes are a
function of particular relations, affects and materials, rather than
their source or cause. This does not mean that structures, subjects
or organisations do not have significant social effects, only that the
relations, affects and events in which such entities emerge ought to
be the primary focus of empirical analysis. Relations obtain
between entities in their encounters, which modify the affects or
capacities these entities may together exercise. Of course, these
entities are always, already a function of earlier encounters and
 C. Duff / International Journal of Drug Policy 33 (2016) 15–20
their relational and affective valances, just as they will continue to
change with future encounters.
Certainly, this language may be confounding at first but it is
consistent with the idea of refusing to accept either subjects or
structures as either ontological conditions or discrete units of
analysis for AOD research. Deleuze argues that empiricism
traditionally relies on the figure of subjectivity (the intentional
agent) to make sense of social life even though the empirical
conditions of the subject’s emergence are not given. By way of this
aporia, subjectivity is presupposed as a transcendental condition of
sensible experience, as the necessary foundation for empirical
inquiry. Deleuze (1994), in contrast, treats subjectivity as an effect
of sensible experience, not its cause, proposing an ontological
model of subjectivation, of the subject’s production in an
assemblage of forces. This is why, in the context of empirical
studies of alcohol and other drugs, assemblage thinking requires
the individual subject to be removed as a unit of analysis and
replaced with a logic of relations, affects and materials, and the
assemblages in which they form and circulate.
One of the most significant implications of this reasoning for
studies of alcohol and other drugs is the challenge it presents to
conventional understandings of choice and personal responsibility. I
mentioned in the introduction to this paper how endemic notions of
individual responsibility are in discussions of drug problems, and
how often these notions overwhelm understandings of the social
contexts of drug problems. The relationality that defines assem-
blages makes it impossible to reliably attribute intentions, desires or
preferences to human agents alone, for intentions and desires are an
emergent effect of encounters between entities and the affects they
generate. As an obvious example of this point, the desire for drugs
cannot exist without an object of this desire, which means from the
perspective of assemblage thinking that the desire for drugs is
partially a function of drugs themselves (Fraser, Valentine, &
Roberts, 2009:124–126). It is not the subject alone that desires
drugs, which he or she then procures and consumes; desire is an
affective function of encounters between drugs and bodies in an
assemblage of forces. This also means that the linear causality with
which consumption behaviours are typically characterised, where-
by a coherent subjective intention is posited as the temporally and
spatially antecedent cause of a subsequent consumption event, must
be abandoned both for failing to grasp the character of subjectivity
itself, and for misunderstanding the array of agentic forces that
participate in any given episode of consumption (Duff, 2014:142–
48). Each episode draws together bodies (human and nonhuman),
spaces and settings, objects, technologies and materialities in an
assemblage that determines the character, nature and effects of this
consumption. Identifying which of these entities may be more or less
active in this drug use, and more or less amenable to manipulation,
cannot be determined in advance of empirical analysis of a given
drug assemblage. This also means that responsibility for drug
problems may not be attributed to individual bodies but must
instead be distributed throughout the emergent relations that
characterise assemblages.
These arguments stand in sharp contrast to the ontological
assumptions that underpin much of contemporary drug policy,
particularly regarding the role of intentional subjects in the
consumption of alcohol and other drugs, and its consequences.
Contemporary drug policies typically rely on a static account of
subjectivity and a linear model of the causal relations that may be
said to mediate drug use. Yet as a number of scholars have argued,
this approach struggles to accommodate the array of human and
nonhuman entities that mediate each episode of use (see Moreno &
Wilton, 2014). It struggles in particular with the idea of dynamic,
nonlinear and emergent modes of causality in which entities
establish causal relations with other entities only as a result of
particular spatial and temporal configurations, and not as a result
19
of their essential, agentic properties. Outside of these configura-
tions, individual entities may exhibit different causal properties,
perhaps even none at all. Kane Race (2014:304–306) has recently
applied the notion of emergent causality to explore the role of
‘‘sniffer dogs’’ in police operations in Sydney. Despite claims that
the use of sniffer dogs has led to reductions in drug related crime in
Sydney, Race examines spatial and temporal configurations
(assemblages) in which the presence of dogs has had no impact
on drug use, and others in which the impact was contrary to the
intended outcome, further complicating causal explanations of the
role of sniffer dogs in local policing.
This suggests once again that causality is not a necessary
outcome of a given entity’s agentic properties, rather it is established
in encounters between entities whereby the direction of causality
may differ from one assemblage to another, just as it may differ from
one encounter to another. Race’s (2014:320–322) analysis indicates
how the presence of sniffer dogs in a given assemblage instantiates
one set of causal relations transforming drug use practices, while in
another assemblage (in another space, in another time) it exhibits no
significant causal effects at all. The point is that casualty cannot be
determined on the basis of a priori assumptions about the behaviour
of a given set of entities but must be confirmed in real experience as
casual relations emerge in encounters between bodies (human and
nonhuman), objects and practices. Of course, this is simply one more
way of saying that assemblage thinking furnishes a novel unit of
empirical analysis for contemporary drug studies. It may also open
up new ways of interrogating the social aspects of drug use, and their
manifestations in particular places (or contexts). The goal, in each
respect, is to uncover the bodies, objects and spaces that participate
in drug use events so that each entity may be given its due in
assessments of how drug use may be made safer, less harmful. This is
the assemblage thinking that may yield a novel harm reduction
praxis. Yet, it may also offer a way out of interminable debates
regarding the relative onus of agents and structures, individuals and
contexts, in the production of drug problems. From the perspective
of the assemblage it simply makes no sense to speak of an
individual’s drug problem given how many other forces will be
active in the articulation of this problem, including peers, family
members, outreach workers, drug objects and paraphernalia,
money, gifts and so on. However, it is equally nonsensical to assert
that power, context or structure are the ‘real causes’ of drug
problems, given the capacities that assemblages avail to individual
bodies. Hence, it is not a question of imagining some ‘meso’ level of
social interaction in which agents and structures interact in the
travails of practice (see DeLanda, 2006). As Deleuze (1994) would
have it, the only way between agents and structures is by way of the
assemblage; and the only way between debates about individuals
and power in the production of drug problems is by way of the
assemblage too. Assemblages experience drug problems not (just)
individuals or social contexts. A harm reduction praxis of and for the
assemblage awaits its proper articulation in policy as in practice.
Acknowledgements
I thank Stewart Williams for thoughtful and productive advice
on an earlier version of this commentary. This research was
partially funded with the award of a Vice-Chancellor’s Senior
Research Fellowship at RMIT University.
Conflict of interest statement: The author declares that there are no
conflicts of interest.
References
Anderson, B., Kearnes, M., McFarlane, C., & Swanton, D. (2012). On assemblages and
geography. Dialogues in Human Geography, 2(2), 171–189.
Anderson, B., & McFarlane, C. (2011). Assemblage and geography. Area, 43(2), 124–127.
20 C. Duff / International Journal of Drug Policy 33 (2016) 15–20
DeLanda, M. (1997). A thousand years of nonlinear history. New York: Zone Books.
DeLanda, M. (2006). A new philosophy of society: Assemblage theory and social complexi-
ty. London: Continuum.
DeLanda, M. (2008). Deleuze, materialism and politics. In I. Buchanan & N. Thoburn
(Eds.), Deleuze and politics (pp. 160–177). Edinburgh: Edinburgh University Press.
Deleuze, G. (1988). Spinoza: Practical philosophy. San Francisco: City Lights.
Deleuze, G. (1994). Difference and repetition. London: Athlone Press.
Deleuze, G., & Guattari, F. (1987). A thousand plateaus: Capitalism and schizophrenia.
Minnesota: University of Minnesota Press.
Deleuze, G., & Parnet, C. (1987). Dialogues. London: Athlone Press.
Dewsbury, J. (2011). The Deleuze–Guattarian assemblage: Plastic habits. Area, 43(2),
148–153.
Dilkes-Frayne, E. (2014). Tracing the ‘‘event’’ of drug use: ‘‘Context’’ and the co-
production of a night out on MDMA. Contemporary Drug Problems, 41(3), 445–479.
Duff, C. (2013). The social life of drugs. International Journal of Drug Policy, 24(3),
167–172.
Duff, C. (2014). Assemblages of health: Deleuze’s empiricism and the ethology of life.
Dordrecht: Springer International.
Duff, C., & Moore, D. (2015). Embracing and evading normality: Estrangement and
ambivalence in the accounts of methamphetamine consumers. Critical Public
Health, 25(4), 488–503.
Fitzgerald, J. (2015). Framing drug use: Bodies, space, economy and crime. London:
Palgrave Macmillan.
Fraser, S., Valentine, K., & Roberts, C. (2009). Living drugs. Science as Culture, 18(2),
123–131.
Fraser, S., Moore, D., & Keane, H. (2014). Habits: Remaking addiction. London: Palgrave
Macmillan.
Gomart, E., & Hennion, A. (1999). A sociology of attachment: Music amateurs, drug
users. The Sociological Review, 47(S1), 220–247.
Jayne, M., Valentine, G., & Holloway, S. L. (2010). Emotional, embodied and affective
geographies of alcohol, drinking and drunkenness. Transactions of the Institute of
British Geographers, 35(4), 540–554.
Keane, H. (2002). What’s wrong with addiction? Melbourne: Melbourne University
Press.
Latour, B. (2005). Reassembling the social: An introduction to actor-network theory.
Oxford: Oxford University Press.
Marcus, G., & Saka, E. (2006). Assemblage. Theory, Culture & Society, 23(2–3), 101–106.
Moreno, C., & Wilton, R. (Eds.). (2014). Using space: Critical geographies of drugs and
alcohol. New York: Routledge.
Race, K. (2014). Complex events: Drug effects and emergent causality. Contemporary
Drug Problems, 41(3), 301–334.
Rhodes, T. (2002). The ‘risk environment’: A framework for understanding and reduc-
ing drug-related harm. International Journal of Drug Policy, 13(2), 85–94.
Weinberg, D. (2013). Post-humanism, addiction and the loss of self-control: Reflections
on the missing core in addiction science. International Journal of Drug Policy, 24(3),
173–181.