In this chapter we try to explore how the invention of complexity as an event in the academic understanding of dynamics, flows, emergence, and adaptivity may impact on the field of urban governance. We will introduce DeLanda’s assemblage theory of social complexity to be followed by our arguments how planners could use this conception for a better understanding of their ‘reality’.
This is a draft version. For citation please refer to
Joris Van Wezemael (2010): Urban Governance and Social Complexity, in: Gert de Roo & Elisabeta da Silva (Eds): A Planner’s encounter with complexity. Ashgate, Farnham, pp. 283-308
We will therefore discuss various fields such as social movements, hybrid government/non-government assemblages, political feed-forward loops, socio-technical collectives, firms and networks, spatial individuals and assemblages of location-networks. We will develop the argument that there is no bird’s eye perspective and no master process in urban governance; the only position to view, affect, or even think an assemblage is another assemblage. At the end of the chapter we introduce concepts of conversion and temporality, which yet have to be explored empirically.
19.1 Introducing complexity and the aims of urban planning
We see the invention of complexity with Isabelle Stengers as an event. Through a transformation in scientific practices in several dimensions this event puts into variation what appears as ‘given’, and puts at risk what counts as the object of knowledge (Stengers, 2000). This chapter asks how complexity may impact on the field of urban governance. Today, complexity is used in a metaphorical (Thrift 1999) as well as in a conceptual sense throughout many disciplines; in planning and in urban governance the metaphorical use is predominant. Metaphorical use, however, does not take complexity as an event and therefore does not push the limits of how we think our topics and problems. However, in recent years, authors such as Brian Massumi (Massumi, 2002), Manuel DeLanda (DeLanda, 1997, 2002, 2006), or Adrian Mackenzie (Mackenzie, 2002) introduce non-metaphorical accounts into social sciences.
A number of authors, which engage with complexity and/in social sciences, draw on Gilles Deleuze’s work on the virtual. It provides a non-metaphorical understanding of creation on the basis of productive difference. The use of Deleuze’s oeuvre as a shared background eases the combination of the work of different authors since it provides a largely shared ontological (or maybe more appropriate: ontogenetic) basis. Manuel DeLanda is working towards a philosophy of generation and elaborates on the key concept of assemblage. In his latest book DeLanda further develops his philosophy of generation and explicates a theory of social complexity on the basis of Deleuzes’ (partly with Guattari) fragments of an assemblage theory (DeLanda 2006a/b).
We suggest that assemblage theory can be very helpful since it establishes a great number of connections between more traditional approaches in social sciences (e.g. the work of Giddens, Bourdieu, Tilly, Weber) on the one hand and intensive sciences and complexity theories on the other. The starting point for the assemblage approach is DeLanda’s critique that many social phenomena or entities beyond the micro or the macro such as interpersonal networks, organisations, cities and regions, etc. lack a proper ontological basis and are wrongfully either reduced to a micro-, macro-, or a meso-position. In most cases their ontogenetic status is simply ignored, which means that an implicit ontology is uncritically accepted as in Network Governance (Sørensen & Torfing, 2004). In his attempt to solve this shortcoming DeLanda gets rid of totalities and essences and also of some unnecessary mystifications in social sciences.
As a brief outline of the assemblage approach we would like to highlight three aspects: (1) Assemblage theory criticises oversimplifications that attribute causes to posited systems, (2) it calls for an adequate ontological conceptualisation of entities which draws on the processes that produce them, (3) and it highlights those theories, which begin with ‘actual’, physically perceived systems, do not adequately explain the origins of those systems.
As a first aspect DeLanda’s major concern refers to oversimplifications that attribute causes to posited systems such as ‘modernist planning’ or ‘urban governance’ without describing the causal interaction of their parts, which would change in different contexts. DeLanda’s theory owes a lot to population thinking rather than to the worldview of typologists. Population thinking argues that variation is the fuel of any kind of evolution (see also Bertolini in this book). For population thinkers heterogeneity is the state we would expect to exist spontaneously under most circumstances, while homogeneity is a highly unlikely state which may be brought about only under very specific selection pressures, abnormally in space and time. Moreover, while the typologist thinks of the genesis of form in terms of the expressions of single types (the European City, the American Suburb, the Garden City, the Medieval City, etc), for the populationalist the forms always evolve within collectives (DeLanda, 2002)48) and they display variation. Therefore the populationalist stresses the uniqueness of everything: whereas for the typologist, the type is real and variation is an illusion, for the populationist the type is an abstraction and only the variation is real (Mayr, 1978). Population thinkers use probabilistic terms (rather than linear causality) in order to overcome homogenising typology. We suggest assemblage theory as a way of theorising populations (of human subjects, social groups, organisations, cities, etc) by means of conceptualising the generative processes from which they emerge.
The focus on generative processes and emergence introduces an important second aspect of assemblage theory. It challenges the ontological solution of hitherto social theory, which did not properly conceptualise the ontological status of ‘intermediate’ levels such as social movements, institutional organisations, social networks or city regions, and instead reduces them to a micro, macro, or lately also to a meso-level:
(1) One family of solutions to the linkage of the micro and the macro has been a strategy to reduce one to the other.
(2) Classical sociology in a Durkheimian or Parsonian tradition as well as many forms of Marxist social theory do not pretend that individuals would not exist, but they assert that society makes the individual. On the basis of socialisation and thus the internalisation of the societal rules – traditional regulations and cultural values – individual persons are an epiphenomenon to social structures (DeLanda 2006a). This macro-sociological reductionism was heavily criticised since the 1960’s by what we can call (2) a micro-sociological reductionism, which views society as a mere aggregate of social action. Similar arguments apply to neoclassical economics. Society as a resulting macro entity thus does not have emergent properties of its own; it is an epiphenomenon from the point of view of phenomenological experiences or choices of (bounded) rationality, respectively.
(3) A third reductionist position has gained popularity more recently in governance related topics such as planning (Healey, 2006b; Healey, 2004). Authors such as Giddens (1984) or Bourdieu (1979) point out that both agency and structure are mutually constructed by practice – they are both sides of a single coin. The constitution of society thus can be reduced to social practice as ultimate reality, or to the habitus as a master process.
In either of the reductionist strategies the generative processes that govern systems and also the complex interactions among their parts (which give rise to the system – ‘emergence’) remain hidden below the shadow of final products.
Assemblage theory is based on the understanding that theories, which begin with ‘actual’, physically perceived systems, do not adequately explain the origins of those systems. This refers to the third key aspect of assemblage theory: the subordination of the term ‘real’ to the terms ‘virtual-actual’. What we usually understand as the ‘real world’ in an assemblage perspective is merely the experience of the ‘actual’ state of final products; they are devoid of their virtual becoming. In order to remedy the situation DeLanda (2002) introduces the Deleuzian virtual as a deeply materialist concept. For this he borrows concepts from complexity and chaos theory, as well as from differential geometry and other fields. He points a way out of the dichotomy between either taking the actual for the real (and the virtual for the non-real), or following the linguistic turn and view the world as a mere social construction. Assemblage theory, we could say, takes the virtual as the more-real and thus focuses on the (virtual) generative processes which produce systems and the actual world which we can (more or less) perceive.
Let us be very clear about this: in our perspective the ‘real’ includes, in addition to the physically perceived, sensuous world of the actual, the virtual properties inherent in assemblages and the intensive processes that select and animate them. This composes a sort of passage from the virtual to the actual. The virtue of a Deleuzian ontology of processes is that, in considering ‘the virtual’ alongside ‘the actual,’ it is able to explain generation (morphogenesis as the birth of metric space, but also the production of places), and bridge virtual and actual viewpoints. (Pease, 2005)
Complex systems as well as assemblages display behaviour that results from the interaction between components and not from characteristics that are inherent to components themselves. The according concept of emergence is at centre stage in assemblage theory, but this theory rejects the idea that systems be perceived as a ‘whole’, which then would explain the behaviour of component parts. DeLanda’s (DeLanda, 2002) reference to nonlinear models and their multiple attractors in assemblage theory define a world capable of surprising us through the emergence of unexpected novelty, a world where there will always be something else to explain.
Assemblage theory implies self-organisation as a direct consequence of the lack of a top-level of organisation. Since there is no pre-supposed global level or god’s eye perspective from the ‘top-level’ of a system, the only position to think, view or affect an assemblage is another assemblage. However, its conceptual development stems from differential geometry (manifold), vector field theory, and the theory of groups, rather than from systems theory.
In this chapter we firstly conceptualise the entities that regional sciences deal with (interpersonal networks, organisations, cities and regions etc) by means of assemblage theory. We will illustrate the concepts by referring to examples from the Swiss context of urban governance and planning. Thus the first aim is an inquiry into the (ontological) status of the ‘ingredients’ of urban governance and spatial planning by means of assemblage theory. Then we moot a concept to re-formulate the problem of governance and strategic spatial planning in a non-metaphoric complexity approach in order to re-think attempts of steering from the field of decision-making in planning and urban governance as ontogenetic modulation.
19.2 Urban regions, governance and complexity
Apart from the perspective on complexity as an event there is a growing belief that the world we deal with really has become more complex itself, rendering linear/Eucledian modes of social sciences increasingly inadequate for the task. As Law & Urry rightly state social and physical changes in the world need to be paralleled by changes in the methods of social enquiry (Law & Urry, 2004). On the planning agenda there are e.g. various aspects of increased global-local interplay, which affect the management of metropolitan areas; another example refers to Edge-Cities with their tangential connections, which interfere with the traditionally radial built transportation networks and pose serious challenges for the governance of urban areas. This is especially the case in areas with small political entities as in Switzerland. The intensification of societal complexity in recent decades deepens systemic interdependencies across various social, spatial, and temporal horizons of action (Jessop 1999, 1) and calls for new modes of governance. Governance networks are believed to provide advantages if compared to hierarchical government organisations. Sørensen and Torfing (2004) define them as relatively stable horizontal articulation of interdependent, but operationally autonomous actors, who interact through negotiations, which take place within a regulative, normative, cognitive and imaginary framework, that to a certain extent is self-regulating, and which contributes to the production of public purpose within and across particular policy areas.
The present phase of re-orientation in planning and related disciplines such as economic, regional and urban geography may reflect these attempts as paralleled changes in the ways of doing research. A relational perspective (see Amin, 2002, 2004; Healey, 2006a; Massey, 2005; Thrift, 1996, 2000) views cities and regions as agglomerations of heterogeneity which are “locked into a multitude of relational networks of varying geographical reach” (Amin, 2004). Furthermore, in urban governance literature there is a lot of vocabulary used, which more or less directly points towards complexity (e.g. relational complexity, unpredictability, self-regulation of regions, heterogeneity, creativity, experimental practice etc). However, as mentioned before, the use of ‘complexity-terms’ within planning and the regional sciences is largely metaphorical. Furthermore, many concepts which are used do link back to reductionist positions (DeLanda, 2006a).
19.3 Links in regional sciences
Assemblage theory puts the topics of regional sciences at the very heart of social complexity. There are innumerable links, which connect regional studies, urban governance etc to complexity. We will illustrate this and refer to relevant issues in our outline of assemblage theory below. In several academic discourses there is a growing awareness of dissatisfactory theoretical frameworks. As Sorensen and Torfing (2004) state, in the case of network governance “most scholars in the first generation [of network governance research] tended to borrow concepts and arguments from other scholars in the field, thus producing a somewhat eclectic and confusing theoretical landscape. […] when people refer to concepts and arguments developed in a different and even contradicting theoretical context, it is problematic”. Indeed, the theoretical frameworks juxtapose concepts, which are based on different ontological strategies. The assumed ontological status of ‘intermediate’ levels between a micro and a macro such as social movements, institutional organisations, social networks or cities and regions is a key problem of urban governance research (Van Wezemael, 2006). The linkage of ‘individual and society’, ‘agency and structure’, ‘choice and order’ is one key question for any social ontology and thus for the respective social theory (DeLanda, 2006a, 2006b). However, it is a paramount question for regional sciences because their key concern is the linkage of manifold heterogeneous agents (e.g. social groups, private organisations, firms, public institutions, quangos etc) in spacing and their interplay in a context of various kinds of ‘negotiation’ (see later).
Even regional sciences, which predominantly deal with the hybrid forms of socio-technical entities, modes of regulation and the question of building regions do fall into the trap of uncritically accepting implicit ontologies (this is particularly true for some literature on regional innovation models such as in Cooke (1996)). For an excellent survey of the field of ‘regionalism’ see Moulaert & Mehmood (2007). The tools of analysis often are mixed with normative concepts. Although to a varying degree, various schools conceptualise regional development and the attempts of steering it. For example regulation theory has identified the importance of the hybrid intermediate-scale entities in socio-technical, politico-institutional, and economic respects. Evolutionary economics thinking and more elaborated approaches to path dependency were introduced through the literature on innovation systems (Dosi, 1988, 1994; Lundvall, 1992).
We argue that taking on ontological questions as outlined above is the very basis for an adequate academic engagement with urban governance. The invention of complexity can boost this debate.
19.4 Assemblage theory and the ‘ingredients’ of urban governance
The theory of social assemblages and the processes that create and stabilise their historical identity opens up towards a great number of Deleuzian concepts. Its core ideas, however, are rather straightforward. The predominant conceptualisation of the relations between parts and wholes in social sciences can be addressed as ‘relations of interiority’. Here, the parts are constituted by the very relations they have in the whole. In contrast, assemblages, are made up of parts which are self-subsistent and articulated by relations of exteriority (here DeLanda draws on Deleuze’s (1991) reading of David Hume). In this perspective organised beings are the result of large numbers of relations between parts, which have no significance on their own: “A flash of red, a movement, a gust of wind, these elements must be externally related to each other to create the sensation of a tree in autumn” (NN, 2006).
Every relation thus has a localised motive, not a transcendent one: there is no Aristotelian ideal type or essence of a tree in autumn (or a governance network or the European City) – only an immanent world of relations, from which such entities may emerge by means of exterior relations of their component parts. Exterior relations therefore may change without their terms changing (Deleuze & Parnet, 2002). The relations in assemblages are not logically necessary but contingently obligatory; an entity therefore is never fully defined by its relations. Whereas necessary relations could be investigated by thought alone, contingently obligatory ones involve a consideration of empirical questions (DeLanda, 2006b: 11). Exteriorly joined components remain certain autonomy from the whole they compose, and they are neither mutually constituted nor fused into a seamless whole. As opposed to relations of interiority, the parts which make up an assemblage are self-subsistent. It “is always possible to detach an entity from one particular set of relations, and insert it instead in a different set of relations, with different other entities” (Shaviro, 2007). Assemblages are secondary to the relations of their constituent components. Larger assemblages emerge from the interactions of their component parts.
DeLanda’s ontology really rather is an ontogenesis. Since there is no pre-given identity which could be drawn form an Aristotelian essence a thing is determined by what it can do (rather by what it ‘is’). Assemblage theory is a thoroughly relational, non-metaphorical (meta-) theory of creation. As soon as assemblages emerge on the basis of their component’s connections they start providing resources for their components as well as constraining them. This introduces a top-down aspect into the so far bottom-up approach.
On the basis of the argument so far it becomes clear that in an assemblage approach there are many levels of emergence, which cannot be reduced to naturally ‘given’ entities (such as ‘society’ or the ‘individual’). Assemblage theory moves us below and above the subject. The move below the subject stems directly from Hume, but it also connects to the ‘society of mind’ thesis in cognitive sciences. The subject becomes an intermediate level of organisation. In DeLanda (2006b) an individual person, a population of individual persons, friendship networks, interpersonal networks, organisations, inter-organisational networks (clusters etc), cities, and territorial states are all conceptualised as assemblages on different ‘scales’, which do not differ in ontologenetic status: they are all historically produced, unique individuals.
The world of assemblages is populated by hybrids: human and non-human, technological and social components give rise to entities by means of their connections. The social practices are inseparable from the material roles of their component parts. E.g. a social movement usually consists (besides its counter-movement) of both populations of interpersonal networks and of organisations (both government and non-government organisations). E.g. the social movements in Switzerland, which urge for an alternative traffic policy and which are component parts in governance networks regarding metropolitan transport systems, produced an association in 1979. Today we can find a hybrid of a number of local (more or less organised) movements which are based on interpersonal networks and at the same time an institutionalised lobby organisation on various institutional levels of the federal state in Switzerland. Similarly, government hierarchies at all jurisdictional scales form networks with nongovernmental organisations in order to be able to implement centrally decided politics. E.g. the Swiss Federal Office for Spatial Development (ARE) needs non-governmental or para-governmental association like ‘glow. Das Glattal’ (Oehler, 2005) or ‘Idee Seetal’ in order to develop an intercommunal governance project in the metropolitan area of Zürich and Lucerne, respectively. Further the Federal Office needs them in order to implement and to re-formulate its metropolitan politics (see Agglomerationspolitik des Bundes, ‘Modellvorhaben’).
The ‘dimensions’ of assemblages
An assemblage can be defined along two analytical dimensions: (1) the roles which assemblages play and (2) the processes which stabilise or destabilise their identity.
Dimension 1: The role assemblages may play can be material (infrastructure needed for communication, labour put into the moderation of conflict in an organisation etc; see below) or expressive (DeLanda 2006a) and usually they are a combination of both. Expressive components may
- rely on specialised vehicles for expression (e.g. coded systems such as language, genetic code), or;
- be directly expressive (e.g. architecture of a building, gesture of a subject, skyline of a city; raditionally but wrongly lumped together under the label ‘symbolic’ (DeLanda, 2006a)).
The focus on the roles, which assemblages really play transgresses the distinction of traditional concepts like ‘facts’ versus ‘values’ and points out that both do not exist as such beyond the concrete relations in assemblages.
Along this first dimension are specified the roles which component parts may play (from a purely expressive role to a purely material role, and predominantly mixtures of the two). The material parts of a social movement consist of the energy and labour involved in maintaining its relations, patching together provisional coalitions, negotiating which of the numerous agendas brought forward by the participants will be mounted as collective action, and hiding internal struggles from public view. The communication hardware used in stabilising the relations play a material role, too (the deprivation from computers or telecommunication hard- and software in totalitarian regimes or the censorship of the world wide web e.g. in the People’s Republic of China illustrate the importance of such material roles for politically relevant networks). Furthermore the production of outputs such as brochures etc and running an office makes part of the material components. In order to affect its audience and be a legitimate claim-maker both in the eyes of its rivals and the government a movement has to be respectable, unified, numerous and committed. Although these possessions can be expressed linguistically (e.g. by publishing a webpage with the number of associated individuals and institutions) it will be displayed more convincingly if a large crowd congregates peacefully in a place in town. (DeLanda 2006b, 60).
DeLanda argues that the processes in which specialised expressive media (e.g. language) can be seen as an additional dimension or a differentiation of the expressive roles in social assemblages (3rd dimension, or dimension 2.1). Specialised expressive media consolidate and rigidify the identity of the assemblage (e.g. by shared stories or myths of origin) or allow the assemblage certain latitude for more flexible operation while benefiting from e.g. linguistic resources in processes of coding and decoding. (DeLanda 2006b, 18-19). Thus an expressive role should not be confused with a linguistic one. In an assemblage perspective e.g. populist movements cannot be analysed merely, not even mainly, by means of their political (linguistic) argument. They are too ludicrous to imagine without their colours, the lights, the music etc as elements which form assemblages with sub-personal ‘agents’ (in the sense of the ‘society-of-mind’ thesis, see above) as well as many assemblages ‘above’ the individual level. However, these non-linguistic expressive components must not be lumped together under the label symbolic – the roles they play in order to affect and be affected and to stabilise and maintain the identity of the individual assemblage have to be examined empirically and experimentally.
Dimension 2: Stabilising and destabilising the identity of an assemblage
- Stabilising = (re-)territorialisation (sharpening its borders, homogenising its components, etc).
- Destabilising = deterritorialisation (free up fixed relations).
The second dimension characterizes processes in which these components are involved: processes which stabilise or destabilise the identity of assemblages. Interpersonal networks, social movements or associational organisations can be stabilised by conflict with their opponents. For instance, advocates of an expansion of public transport can be more or can be less radical in their claims, which is likely to generate internal struggles in the respective movement. However, when confronted with an initiative of the road traffic lobby they will unmistakably be ‘one voice’ in favour of their core requests. Stories and categories play important boundary-defining roles, as DeLanda shows with regard to the work of Charles Tilly (1999), but these are real group boundaries not just phenomenologically experienced or linguistically constructed borders. Therefore not the linguistic label of a category deserves our main focus of interest but the outcome of sorting processes of inclusion and exclusion that produce concretely bounded groups. Let us explain this point.
Take ‘foreigners’ which cannot participate in the direct democratic processes in Switzerland (even if they constitute the majority of the population of a borough). Not the category ‘foreigners’ produces this exclusion, but a real process of differential inclusion and exclusion works as a sorting process regarding the access to formal democratic political power. The category ‘foreigner’ merely catalises the process and sharpens the border of the real boundary. The linguistic category territorialises the exclusion of a group of people. Or take the segregation of foreigners into an area (which will be denominated as a ‘migrant area’). However, this refers not to linguistic categories but to the enforcement of real categorical boundaries. Thus e.g. the housing market should be viewed as a real sorting process of people and life-chances. The change in relations of the components of that sorting machine can deterritorialise an existing order and thus modify the outcomes of the sorting process (see Van Wezemael, 2000). Naming or coding of neighbourhoods can sharpen the neighbourhood’s borders and the overcoding of the residents in the neighbourhood with categories such as e.g. foreigners can catalyse the enforcement of uneven life chances. Specialised expressive media thus rigidify the identity of the assemblage, however they do not produce it.
Assemblages avoid reifications
Assemblage theory makes clear that government organisations in terms of internal heterogeneity and complexity do not differ from economic organisations. Joint action by many (heterogeneous) governmental organisations is objectively complex and problematic and not something that can be taken for granted. Concepts such as ‘the state’ not only tend to reify generalities but they veil the relations of exteriority which exist among the heterogeneous organisations that form a government hierarchy (DeLanda 2006b, 85).
Political feed forward loops can be dealt with as ‘autocatalytic dynamics’ (DeLanda, 1997, 49-52), which foster new developments on the basis of their self-stimulating dynamics. The process of formulation-implementation-reformulation is reflected in the Swiss-German term rollende Planung (a basic concept in Swiss spatial planning). However, such feed-forward loops do not occur on an idealised plane of policy formulation and implementation, rather they move in relational settings with non-governmental organisations and they always work on the basis of the various material and expressive roles their components may play. Thus concrete government organisations are precarious entities, which always intertwine with a number of non-governmental entities in order to be effective.
In an assemblage view the difference between ‘government’ and ‘governance’ (Healey, 2006b) clearly is not a fundamental one. However, the component parts of the relational entities may play different roles and thus actualise a set of specific temporary solutions to the ‘problem of governance’. Whereas the hitherto governance debate generally locates governance as an alternative ‘to market anarchy and organisational hierarchy’ (Jessop, 1999), an assemblage approach stresses the necessary intertwining of manifold organisations and networks, which all are socio-technical collectives rather than pure ‘political’ or ‘economic’ entities. Thus they do not merely ‘use’ material resources and technology in order to follow their goals. Furthermore, assemblage theory makes clear that hierarchies and networks display similar relations of exteriority on different levels of their relational constitution.
Another crucial aspect of an assemblage perspective is the distinction of the assemblage(s) that produce the hierarchy of a federal government (as in the Swiss case) on the one hand, and the territorial entity which it controls, on the other hand. The territorial entity includes populations of organisations of various kinds, populations of persons and interpersonal networks, cities, regions, provinces etc. The governmental hierarchies are intertwined with those populations and do not refer to a separated or ‘upper’ level.
19.4 Non-metaphorical creation
DeLanda criticises the linguistic turn in social sciences as “the worst possible turn” (DeLanda, Protevi, & Thanaem, 2006, 8) and he disapproves of the predominant social constructivist perspectives, which use the term ‘construction’ in the sense of how our minds ‘construct’ the world of appearances via linguistic categories. The processes that literally – not phenomenologically – create a region (processes of morphogenesis or individuation), exceed the constructions of linguistic categories. Similarly, politics is not about the construction of categories, as the example of segregation above may illustrate. New technologies ‘space’ areas long before they are coded by means of categories (see below). The linguistic coding of clusters etc may stabilise a setting and increase the homogeneity of a location (‘Motor City’, ‘Silicon Valley’). However it cannot create it as many attempts to ‘create’ economic clusters on the basis of a decision of a city marketing or planning agency shows (for a Swiss example see e.g. ‘winlink’). Social-constructivists falls short because they use a merely metaphorical concept of the term ‘creation’ and ignore the non-linguistic roles that component parts in assemblages do play (DeLanda, Protevi, & Thanem, 2006). They furthermore neglect the respective de/re-territorialising tendencies.
Almost unperceived by many, a sort of linguistic idealism has become a predominant paradigm for most social scientists. The inherent anthropocentrism creates fundamental difficulties when dealing, as regional scientists do, with socio-technical and environmental issues. In their conceptualisations of urban regions and their (political) steering most contemporary authors follow a social-constructivist strategy (although hardly ever in an explicit way). Amin’s (2004) summoning up of regions and Healey’s (2006a, 2006b) relational complexity are only two recent and influential examples. In the sense of Gidden’s (1984) double hermeneutics this also establishes a broadly shared attitude in many practioner’s communities.
As explained above, the roles components can play in an assemblage are combinations of material and expressive ones. Whereas the material ones refer to the whole repertoire of causal interaction (and should be treated as such), the expressive ones typically involve catalysis. The mechanisms to synthesize assemblages therefore include (mostly non-linear) causality as well as, in the case of social assemblages, reasons and motives. For the latter DeLanda draws on Max Weber’s concept of Verstehen in order to stress out that ignoring the hybrid nature of social mechanisms (we mean the combination of causes, reasons, and motives which must not be reduced to only on of them) can be a source of misunderstanding and mystification in social sciences.
What can we learn from this for decision-making? DeLanda explains that social activities in which means are successfully matched to ends are traditionally labelled as ‘rational’. However, this label obscures the fact that the activities involve problem-solving skills of different kinds and not a single mental faculty like ‘rationality’. Thus the explaining of successful solutions of practical problems will involve consideration of relevant causal events in socio-technical collectives, and not just calculations in the actor’s head (DeLanda, 2006a, 24). The capacity of human beings to be affected by linguistic triggers consequently demands (besides a number of physiological preconditions such as e.g. a nervous system) explanations, which include reasons for acting (e.g. referring to traditional values or personal emotions) and motives (choices and goals, matching means to ends). Whereas matching means to ends will be at the forefront in situations when intensity is high (e.g. in a crisis situation when one must solve a new planning problem), traditional routines tend to dominate in situations with a lower intensity (reproduction of the existing social order). (DeLanda, Protevi, & Thanaem, 2006, 8) In such a schema of high-versus-low-intensity-situations we can talk about degrees of freedom of human action. Assemblage theory means a relevant approach to issues of strategic spatial planning and urban governance because it takes ‘creation’ at face value and qualifies the role of language in respect of other expressive as well as to material roles that components play. Let us illustrate our line of argument with a Swiss case as discussed by our colleague Alain Thierstein.
Spatial development ‘in secrecy’: creation beyond categories
As part of the Interreg III B project ‘Polynet’ Thierstein et al. (2006) investigate what they call ‘spatial development in secrecy’ (‘Raumentwicklung im Verborgenen’). In the process of coping with distinct knowledge/skill-resource dependencies enterprises in the so-called knowledge-economy (which mainly includes finance and consulting industries, high-tech branches, life-sciences) tend to display specific material and expressive roles (see above), which their component parts play. Expressivity of an organisation’s intentions can be explicitly phrased and thus linguistically coded. However, usually they will be a manner of assessments of strategic significance (and not: signification). (DeLanda, 2006a, 81).
Requested premium services are a limiting resource to organisations in the knowledge-economy. Since they are scarce and unevenly distributed, specific economies of agglomeration emerge. The production of knowledge-intensive goods (which are ‘assembled’ from a variety of highly specialised material and non-material elements) critically depends on specific material and expressive roles of their component part. This includes personal interaction, face-to-face communication (‘trust’) and interpersonal networks (‘strength of weak ties’, see Granovetter 1973). The specific roles modulate relations between locations and thus relational spaces, which deviate from the outcome of the firm’s hitherto location-strategies. Firms choose (and eventually create) a set of specific locations in order to solve the problem of sets of resource-dependencies in highly specialised domains. Components of the organisation are at the same time part of socio-technical networks in the region (relations of highly specialised staff, laboratories, sets of legal rules, natural resources used, public infrastructure such as fire-optic cables and airports, etc) and of rather dispersed ‘in-house’ networks, which connect the various component parts of the organisation as an assemblage.
The populations of socio-technical collectives (parts of firms, transportation infrastructure, shared resources, government organisations etc) give rise to another scale of assemblages: relational spatial individuals. They provide resources (beneficiary effects) and constraints (agglomeration costs) for their component parts. Since component parts of an assemblage usually are also simultaneously parts of a number of other assemblages and therefore connect (to) them, modulations in one population of assemblages introduces further modulations to sets of relations between elements in other populations of assemblages. This of course launches ongoing ontogenetic transformations throughout populations of assemblages, which each emerge differently from the modulated relations. At this point it becomes obvious that assemblages are only meta-stable.
Figure 1 illustrates the emergence of spatial individuals, which specifically relate to other localised individuals and relationally produce assemblages of the knowledge-economy on various spatial scales. The relations of the firms are mediated across a large variety of distances by using different kinds of transportation and communication technologies. Eventually this produces simultaneous changes to assemblages on various scales. The change in the kind of resource dependencies in the knowledge-economy produces specific relations between metropolitan areas on a European level (pentagon London – Paris – Milano – München – Hamburg), but it also de/re-territorialises the relations of Swiss locations and clearly conflict with key strategic goals of Swiss spatial development.
Figure 19.1 Specialised agglomerated economies in Swiss knowledge industries
Source: © Thierstein (2006)
The modulation of relations (on the basis of inventions and innovations in local and remote socio-technical collectives) re-territorialises Swiss location networks. Value added chains in knowledge economy seriously produce sorting processes, which de/re-territorialise spatial individuals. The analysis of Thierstein (2006, 14-15) detects a ‘gap’ in the perception of the relevant political institutions regarding spatial development in Switzerland. This means that the (invisible) processes which produce places have changed while planners only focus on ‘actual’, physically perceived changes. They thus do not grasp the origins of the dynamics, tend to miss the changes all together and increasingly render themselves unfit for their very task.
Thierstein et al. (2006) believe that the way ahead is to perceive the ‘problem’ of the so-called spatial development in secrecy differently, and to make planners more aware of changing chains of value creation in the knowledge-economy. However, we must keep in mind that we must not reduce the whole set of material and expressive relations to mere ‘perception’. On the basis of our argument so far, we would rather say that the material and expressive roles of the counterparts in governmental/non-governmental hybrids do really have to change. While government organisations need other organisations in order to implement their policies, the same relational settings may prevent alternative relations from emerging. Therefore spatial policies as real relational networks have yet to be modulated.
This can be illustrated with the federal metropolitan politics (‘Agglomerationspolitik’), which puts a lot of energy and labour into the formation of assemblages with networks such as the mentioned ‘glow. Das Glattal’. It awaits its implementation because of feed forward relations between sets of nested assemblages in the relational networks of policy-formation and implementation in Switzerland.
Conclusion of the case
As a conclusion of this digression into knowledge-economy and spacing in Switzerland we can state that (1) new modes of dealing with resource dependencies in the driving knowledge-economy is modulating sets of relations, which again drive spatial development in Switzerland. And that (2) there is a lacking ‘conversion’ between the domain of knowledge-economy (populations of firm networks and their connections in agglomeration economies) and the domain of urban governance (concrete government/non-government assemblages with regard to strategic spatial development).
However, on the basis of the arguments so far the problem cannot be reduced to perception and negotiation. Rather in the outlined perspective ‘governance’ would relate to a plane of potential conversions between the populations of firm networks and their connections in agglomeration economies, and the concrete government/non-government assemblages with regard to strategic spatial development. We could say: ‘governance’ builds an interface. In the course of actualisation (from virtual to actual, see above) a set of potential connections is actualized while other connections remain potentials. Actualised ‘governance-structures’ thus refer to realised (or actulised) conversions between the above domains. On the other hand, ‘governance’ also refers to an event of de/re-territorialisations, which modulates the actualisation of both domains in respect of each other. How can we address those conversions, which modulate the relations on an interface between the domains? In the next part we will introduce the concept of transduction in order to engage with conversions on the basis of non-metaphorical complexity and we address temporality as a crucial problem when dealing with eventful de/re-territorialisations or conversion respectively. This brings us one step closer to posing the question of urban governance and strategic planning differently in the aftermath of the invention of complexity.
19.5 Governance, transduction and heterochrony
How can we conceptualise ‘governance’ within a non-metaphorical complexity approach? On the basis of the above ontogenetic conceptualisation of the relevant scales of urban governance it is not adequate to limit the analysis to (linguistic) negotiations. Besides the linguistic roles (referred to in network governance as negotiation) other expressive as well as material roles have to be considered. Furthermore, assemblage theory makes it clear that there cannot be any entity, which is ‘above’ or ‘beyond’ assemblages. There is no bird’s eye perspective and no master process; the only position to view, affect, or even think an assemblage is another assemblage. Therefore planning organisations, social movements, specialised associations etc are component parts of the processes of self-organisation. ‘Governance’ becomes a problem of affecting and being affected. This perspective of being one element in a relational socio-technical collective of spacing actors legitimises a pro-active role of planners as catalysts (as in Albrechts, 1999) and the use of all available means in order to affect. In order to grasp governance in a non-metaphorical approach to social complexity we moot the concept of transduction (see below).
The relational fundament of assemblage theory makes clear that assemblages are secondary to the relations from which they emerge. Therefore by beginning from the relation individual things (people, interpersonal networks, cities) can be understood as effects of relationality. Now an operation, in general terms, modulates the relations between elements of a set. If a process specifically occurs at some kind of limit or interface between different orders, the operation is called a transduction. E.g. a microphone converts sound into electrical energy; a computer screen converts electrical energy into rays of light; an eye converts rays of light into neuro-electrical signals. Transductions thus convert between different domains; a particular domain undergoes a kind of ontogenetic modulation, which means that the way it is actualised changes. A transduction is an event, which cuts through existing lines of actualisation and folds together previously unconnected elements across domains. Something new actualises, something that was pre-individual or virtual. Doing research on urban governance means in a first step tracing the topologies of transductions within spacing.
If we view the population of socio-technical collectives in knowledge-economy and government/non-government hybrids in the field of strategic spatial development as two relatively separate domains, urban governance (as a real event) means a transduction between them. As we stick to our ontogenetic design transduction is no metaphor. Since in assemblage theory a thing is what it does, urban governance is what transduces between domains in spacing.
What does this mean in terms of actualisation and thus the ontogenetic perspective as outlined in this chapter? In the case of regional studies the interface is topologically, but also temporally, complicated. Ongoing actualisation occurs since transductive operations are nested with each other (on the basis of exterior relations assemblages relate outward to other assemblages). Petersson (2005) discusses time in non-linear processes. With Deleuze (in DeLanda, 2002, 84) he views actual time as the coexisting plurality of time-formations which – as assemblages do – relate outwards (to other time-formations). A time-formations refers to the intrinsic time of a system. A straightforward example may be the individuation cycle of an organism (duration necessary for replacing the cells of an individual without affecting the organism’s identity). But business cycles, project cycles, electoral cycles, planning cycles, innovation cycles etc can also be understood as the heterochronous time-formations of the respective assemblages. This means that their positions in their intrinsic time-cycles usually will be different. They are not governed by a shared interval of time in the sense of a synchronisation of cycles, of phases of higher and lower (in)stability or if you like changing degrees of freedom. If we recall that the degree of freedom alters with the connections in heterogeneous entities (assemblages) and if we also mind our non-linearity foundation it becomes evident that the same disturbance produces a different effect in relation to varying positions of assemblages in a time cycle. Thus a governance network displays nested cycles of time-formations (plurality of intrinsic times in the assemblages which give rise to a governance network).
Assemblages therefore display a high sensitivity to the positions, which they occupy in their time-cycles. This means that the same modulation will have different effects according to the relative temporal positions in the cycles of the two domains; the same event thus can produce a set of various consequences. This corresponds to the notion that explanations in an assemblage approach must include non-linear causality. Think e.g. of representations in a planning process. They do produce different consequences according to their relation to the time cycle of the planning procedure. Or as (DeLanda, 2002) puts it: “A process may change too slowly or too fast in relation to another process, the relationship between their temporal scales determining in part their respective capacities to affect one another. Even when two processes operate at similar scales, the result of their interaction may depend on their coupled rates of change.”
Petersson argues that if an evolutionary process (as in spacing) occurs in parallel development of relatively independent processes, then an altering of the duration of one process relative to another creates new problems and triggers the actualisation of new designs. The actual ‘heterochrony of becoming’ thus fuels processes of creation. Since the intrinsic time of a system always relates outwards, alterations in one system may produce radical change (a bifurcation) in a different system (Petersson, 2005). This may cause a phase transition among a set of nested cycles of time-formations. This means that cycles alter their intrinsic rhythm in relation to each other (since they relate outwards). Transductions thus are likely to alter the individuation of time-formations in affected domains. E.g., one outcome of urban governance processes involving organisations and associations from the building industry and governmental and non-governmental organisations is the adjustment of planning cycles with the time-formations in real estate industry, which, again, tends to synchronise with market cycles. A conversion between domains thus also means a re-actualisation of the time-formations. In a topologically and temporally complex field such as urban governance the modulation of one time-formation will almost certainly trigger another problem in respect to different time-formations. In complicated topologies and temporalities a synchronisation in one domain may trigger the potential of heterochrony in related assemblages and further introduces change.
Why bother with social complexity? In this chapter we have tried to explore how the invention of complexity as an event may impact on the field of urban governance. In order to clear the field we followed Manuel DeLanda in his assemblage approach. With this planners can clarify the ontogenetic status of the things they deal with and also with their own role in the actualisation of places. We introduced the theoretical framework into the field of urban governance by referring to social movements, hybrid government/non-government assemblages, political feed forward-processes, socio-technical collectives, firms and networks, spatial individuals and assemblages of location-networks. This may proof that there are many connections between complexity sciences and more mainstream research in terms of relevant issues and as addressed by various theories.
The relationality foundation of the assemblage approach on the basis of exterior relations shows that collectives emerge from elementary relations. Therefore understanding of how relationality can trigger different actual solutions to one virtual problem is crucial for urban governance. The inclusion of forms of causality in combination with linguistic and non-linguistic expressivity transforms the view on ‘governance’ – interaction cannot be limited to negotiation but must include direct expressive and material roles which component parts of assemblages play. In general terms assemblage theory urges us to put our focus on real processes of inclusion and exclusion rather than reify terms and typologies (this leads to homogenisation which is no longer adequate on the basis of non-linearity). It puts the term creation into a very prominent though non-metaphorical position (move beyond social contructivism).
The ontogenetic design of assemblage theory helps to clarify the role of planning. There is no god’s eye perspective. Each (planning) intervention is a potential effect between externally related elements – self-regulation and autopoiesis are a direct consequence of the ‘flat ontology’ of radical relationality of this view on social complexity. Furthermore the ontogenetic perspective on socio-technical assemblages and the material roles that components play move the focus of research and theorising away from an anthropocentric view and manoeuvres analysis around the cliffs of social and technological determinism. The origin of events is as relational as individual things.
Furthermore the outlined approach opens up into a number of potentially useful domains of contemporary theory-building. Concepts such as transduction or heterochrony seem potentially relevant to spatial planning and governance research. We believe that urban governance as a field of research provides a useful empirical fund for empirical research in social complexity – the illustrations in this chapter may underline this. However, the next step in this line of research comprises of careful empirical tracings of the roles component parts play and of the consequences, which (modulations of) relations produce. However, it is not adequate to merely trace actual things and thus fall back on describing things instead of engaging with their dynamics and their potentials for difference. Since in our ontogenetic perspective the actualisation of things is problematic, ‘history’ has to be doubled with a becoming. Thus the tracing has to be put back on the map (Deleuze and Guattari, 1987, 99-100).
Albrechts, L. (1999). Planners as Catalysts and Initiators of Change. The New Structure Plan for Flanders. European planning studies, 7(5), 587-604.
Amin, A. (2002). Spatialities of Globalisation. Environment and Planning A, 34, 385-399.
Amin, A. (2004). Region Unbound: Towards a New Politics of Place. Gegrafisker Annaler, 86(B), 33-44.
Bathelt, H. (1994). Die Bedeutung der Regulationstheorie in der wirtschaftsgeographischen Forschung [The relevance of Regulation Theory for Research in Economic Geography] – Geographische Zeitschrift, 2, 63-90.
Bourdieu, P. (1979). La distinction. Critique sociale du jugement. Paris: Ed. De Minuit. Case of Economic Development. International Social Science Journal/UNESCO, 155.
Cooke, P. (1996). Reinventing the region: firms, clusters and networks in economic development. In P. Daniels & W. Lever (Eds.), The global economy in transition. Harlow: Longman.
Davies, E. (N.Y.). DeLanda Destratified. techgnosis.com. Retrieved 20.12.2007, from http://www.techgnosis.com/delandad.html
De Roo, G. (2003). Environmental Planning in the Netherlands: Too Good to be True. London: Ashgate.
DeLanda, M. (1997). A Thousand Years of Nonlinear History: Zone Books.
DeLanda, M. (2002). Intensive Science and Vitrual Philosophy. London: Continuum.
DeLanda, M. (2006). A New Philosophy of Society. Assemblage Theory and Social Complexity. London, New York: Continuum.
DeLanda, M. (2006a). A New Philosophy of Society. Assemblage Theory and Social Complexity. London, New York: Continuum.
DeLanda, M. (2006b). Deleuzian Social Ontology and Assemblage Theory. In M. Fugslang & B. Meier Sorensen (Eds.), Deleuze Connections (pp. 250-266). Edinghburgh: Edinghburgh University Press.
DeLanda, M., Protevi, J., & Thanaem, T. (2006). Deleuzian Interrogations: A Conversation with Manuel DaLanda, John Protevi and Torkild Thanem. [Electronic Version]. Journal of Critical Postmodern Organisation Science. Retrieved 15.01.2007 from http://www.tamarajournal.com.
Deleuze, G. (1990). The Logic of Sense (trans. Mark Lester and Charles Stivale, orig. 1969). New York: Columbia University Press.
Deleuze, G. (1991). Empiricism and Subjectivity. New York: Columbia Press.
Deleuze, G., & Parnet, C. (2002). Dialogues II. New York: Columbia Press.
Dosi, G. (Ed.). (1988). Technical Change and Economic Theory. London: Pinter.
Dosi, G., & Nelson, R. R. (1994). An Introduction to Evolutionary Theories in Economics. Journal of Evolutionary Economics, 4, 153-172.
Giddens, A. (1984). The Constitution of Society. Outline of the Theory of Structuration. In. Berkely: University of California Press.
Granovetter, M. (1973). The Strength of Weak Ties. American Journal of Sociology, 78(6), 1360-1380.
Healey, P. (2004a). Creativity and Urban Governance. Policy Studies, 25(2), 87-102.
Healey, P. (2004b). Relational Complexity and the Imaginative Power of Strategic Spatial Planning. European Planning Studies, 14(4), 525-546.
Healey, P. (2006a). Network Complexity and the Imaginative Power of Strategic Spatial Planning. In L. Albrechts & S. Mandelbaum (Eds.), The Network Society: The New Context for Planning. London: Routledge.
Healey, P. (2006b). Urban Complexity and Spatial Strategies: Towards a Relational Planning for our Times. London: Routledge.
Hillier, J. (2005). Straddling the post-structuralist abyss: between transcendence and immanence? Planning Theory, 4(3), 271-299.
Jessop, B. (1997a). Survey article: the Regulation Approach. The Journal of Political Philosophy (5), 287 – 326.
Jessop, B. (1997b). Twenty Years of the (Parisian) Regulation Approach: the Paradox of Success and Failure at Home and Abroad. New Political Economy(2), 503 – 526.
Jessop, B. (1998). The Rise of Governance and the Risks of Failure. The
Jessop, B. (1999). The Governance of Complexity and the Complexity of Governance: Preliminary Remarks on some Problems and Limits of Economic Guidance. Retrieved 20. May, 2006
Law, J., & Urry, J. (2004). Enacting the social. Economy and Society, 33(3), 390-410.
Lundvall, B.-A. (Ed.). (1992). National Systems of Innovation. Towardsa theory of innovation and interactive learning. London: Pinter.
Mackenzie, A. (2002). Transductions: bodies and machines at speed. London: Continuum.
Massey, D. (2005). For Space. London: Sage.
Massumi, B. (2002). Parables of the Virtual. Movement, Affect, Sensation. Durham: Duke University Press.
Mayr, E. (1978). Evolution und Vielfalt des Lebens. Berlin: Springer.
Moulaert, F., & Mehmood, A. (2007). Analysing Regional Development: From Territorial Innovation to Path Dependent Geography. In W. Dolfsma & J. Davies (Eds.), The Elgar Handbook of Socio.Economics: Edward Elgar Publishing.
Moulaert, F., & Swyngedouw, E. (1989). A Regulation Approach to the Geography of Flexible Production Systems. Environment and Planning D: Society and Space(7), 327-334.
Moulaert, F., & Swyngedouw, E. (1989). A regulation approach to the geography of flexible production systems. Environment and Planning D: Society and Space(7), 327-334.
NN. (2006). Gilles Deleuze. The Internet Encyclopedia of Philosophy. Retrieved 08-08-2006, from http://www.iep.utm.edu/d/deleuze.htm.
Oehler, S. (2005). Entwicklungsstrategien und Kooperationsformen im mittleren Glattal. [Development Strategies and Cooperation in the mid Glattal]. University of Zürich, Zürich.
Pease, A. (2005). Manuel DeLanda’s Art of Assemblage. electronic book review Retrieved 17.10, 2007
Petersson, D. (2005). Time and Technology. Environment and Planning D, 23, 207-234.
Sonntag, R. E., & Van Wylen, G. J. (1991). Introduction to Thermodynamics: John Wiley & Sons.
Sorensen, E., & Torfing, J. (2004). Making Governance Networks Democratic. Working Paper Series, Centre for Democratic Network Governance, 2004(1).
Stengers, I. (2000). The Invention of Modern Science: University of Minnesota Press.
Thierstein, A. (2006). Raumentwicklung im Verborgenen. [Spatial Development in Secrecy]. Zürich: NZZ Verlag.
Thrift , N. (1999). The Place of Complexity. Theory, Culture & Society, 16(3), 31-69.
Thrift , N. (2000). Afterwords. Environment and Planning D: Society and Space, 18, 213-255.
Thrift, N. (1996). Spatial Formations. London: Sage.
Tilly, C. (1999). Durable Inequality. Berkeley: University of California Press.
Van Wezemael, J. E. (2000). Verändert die Marktmiete die residenzielle Segregation? Die Marktmiete aus sozial- und wirtschaftsgeographischer Sicht. [Does Uncontrolled Rent Change Residential Segregation? Uncontrolled Rent from a Social and Economic Geography Perspective]. Geographica Helvetica, 4, 251-261.
Van Wezemael, J. E. (2006). The Contribution of Assemblage Theory and Minor Politics for Democratic Network Governance. Conference on Democratic Network Governance in Europe. Centre for Democratic Network Governance, Roskilde.
Zuidema, C., & de Roo, G. (2004). Integrating Complexity Theory into Planning: Truth or Dare? Paper presented at the 18th Annual Aesop Conference at Grenoble. Retrieved Access 24.10.2007 from http://www.ruimte-rijk.nl/index/ publicaties/publicaties/DeRoo%20Zuidema%202004%20Aesop.pdf
 Let us illustrate this ‘event’ with this anecdote quoted from Manuel Delanda: “If you read the essays by the first guy who saw spontaneously oscillating chemical reactions, you find out he was unable to publish his essays. This was in the 50’s, not long ago. The idea that orderly behavior could arise spontaneously from matter was so counter-intuitive. At that time, the only two ways they could see stable things arising in nature was through rational perfection – the best possible outcome – or heat-death. What nonlinear science brings about is a complete new range of structurally stable forms of behavior, which has absolutely nothing to do with rationality or the heat-death of entropy. Now attractors are appearing all over the place. We’ve discovered a whole new reservoir of forms of stabilization. It’s a paradigm warp.” (Davies, N.Y.).
 With some noteworthy exceptions like Hillier (2005) or De Roo (2003).
 They do so largely on the basis of mechanism-independencies. The most popular example may be the Lorenz attractor.
 One family of solutions to the linkage of the micro (‘level of the individual subject) and the macro (level of society as a whole) has been a strategy to reduce one to the other. Classical sociology in a Durkheimian or Parsonian tradition as well as many forms of Marxist social theory do not pretend that individuals would not exist, but they assert that society makes the individual. On the basis of socialisation and thus the internalisation of the societal rules – traditional regulations and cultural values – individual persons are conceptualised as an epiphenomenon to social structures (DeLanda 2006a). Social structures, then, are the agents. This macro-sociological reductionism was heavily criticised since in 1960’s and after by what we (with the necessary simplification) can call a micro-sociological reductionism; it views society as an aggregate of social action in the sense described above. Similar arguments apply to neoclassical economics. In the view of micro-reductionism society as a resulting macro entity thus does not have emergent properties of its own; therefore we can break down its complexity to the action of the individual members of society. Society then has no agency and it becomes an epiphenomenon from the point of view of phenomenological experiences. A third reductionist position has gained popularity more recently in governance related topics such as planning (Healey, 2004a; Healey, 2006b). Authors like Giddens (1984) or Bourdieu (1979) point out that both agents and structure are mutually constructed by practice – they are both sides of a single coin. The constitution of society thus can be reduced to social practice as ultimate reality, or to the habitus as a ‘master process’.
 In respect of metropolitan governance Kübler (2003, 535-537) identifies the disintegration of urban areas in socio-economic, political and cultural aspects as a driver in the quest for new organisational forms in metropolitan regions.
 For a conceptualization of democratic network governance on the basis of assemblage theory and minor politics see Van Wezemael (2006).
 In this chapter we will treat the terms of urban, regional and metropolitan governance as synonyms and use the term ‘urban governance’.
 As Bathelt (1994) argues regulation theory was an important step since it aimed at drawing together the interplay of economic-technical and socio-institutional structures in a national economy and to put them into a development context with an attempt of learning how they steer themselves (learning regions etc) as well as how to intervene into their dynamics (mostly in order to increase regional competitiveness). For a careful discussion of regulation approaches see (Jessop, 1997a, 1997b; Moulaert & Swyngedouw, 1989).
 The most important ones are: abstract machines, quasi-causal operator, (de/re)territorialisation, diagrams, individuation/morphogenesis.
 This is clearly the case in functionalism, however it is also true in less obvious cases such as in Structuration Theory (Giddens 1984): agency and structure mutually constitute each others and produce a seamless whole (DeLanda 2006, 10).
 Since both individualism and collectivism are reductionist and essentialist concepts, the antipole to both is relationalism.
 In his thesis that ‘minds are what brains do’, Minsky (1986) views the human mind and any other naturally evolved cognitive system as a vast society of individually simple processes known as agents.
 ‘Scale’ is not used in a strictly geographical sense. It rather follows a topological sequence of emergence.
 For further information visit http://www.are.admin.ch/themen/raumplanung/02114/index.html?lang=de
 Specialised expressive media can be seen as a form of interiorisation of intensive individuating factors.
 For instance, many organisations are closely associated with e.g. a building (Petronas Tower, Chrisler Building), cities with a skyline (skyline of Manhattan, Paris with the Eiffel tower), firms with colours (Ferrari with Maranello-red).
 This argument refers to the before mentioned ‘society of mind’ approach.
 This may be illustrated in the struggles about the distribution of resources for metropolitan regions versus transit roads in Switzerland. For further information see http://www.parlament.ch/do-avanti.
 Political feed-forward processes can be dealt with as ‘autocatalytic dynamics’ (DeLanda 1997, 49-52), which foster new development on the basis of their self-stimulating dynamics.
 Autocatalytic systems refer to coupled, looped reactions that lead to conditions in which the component parts of the system depend on one another in more than proportional ways (forward loop).
 Amin argues that material formations such as a region must be summoned up as temporary placements of ever moving material and immanent geographies, as ‘hauntings’ of things that have moved on but left their mark as situated moments in distanciated networks, as contoured products of the networks that cross a given place.
 This means that they mainly trigger or ease processes like a catalyst.
 The two basic assumptions of linearity are: large causes will produce large scale effects (and vice versa), and the same cause always produces the same effect. Non-linear science systems display multiple determinations, which is why the two basic assumptions of linearity do no longer apply. Complexity thus implies a move in social sciences away from the linear analysis of structure or agency. The assumptions of complexity are that (1) there is no necessary proportion between ‘causes’ and ‘effects’, that (2) the individual and statistical level are not equivalent, and that (3) systems do not result from simple addition of individual components. (Law & Urry, 2004))
 The argument of high and low intensity can be related to questions of temporality. Problems are posed differently in the course of a planning procedure, and the procedure is more or less sensitive to changing demands, to new demands or to alterations of the (supposed) conditions. We will return to this line of reasoning towards the end of the chapter (see ‘heterochrony’).
 This could be linked to Christensen’s concept of ‘degrees of complexity’ (see also (Zuidema & de Roo, 2004).
 This means that firms need (knowledge and other) resources which they either cannot produce in one place or which they have to buy from or exchange with other forms.
 In general terms a main expressive component in organizations is legitimacy, the material one refers to the enforcement of authority structures.
 Since resource dependency is a general problem to which all firms will develop a ‘solution’ we can refer to these topological invariants as universal singularities. The specific solutions of an assemblage (a firm or a network of firms) can be referred to as individual singularities. See DeLanda (2006b, 28-29).
 Mind that the equation ‘same reason = same effect’ is only true in linear relations (Law & Urry, 2004).
 For instance the above mentioned association ‘glow. Das Glattal’ deals with the recent breathtaking dynamics of economic, traffic, building etc in the area between the city of Zürich and the Airport in Kloten/Rümlang. Its effort to create shared projects e.g. in traffic infrastructure and the like may provide provisory solutions. However, if we put the tracing of the relations in connection with knowledge-economy back on the map, it becomes clear that processes with vectors working on other scales pose more singular problems.
 Switzerland is often called a Verbandsdemokratie (‘association-democracy’) because powerful stakeholders form very dense connections (assemblages) on interpersonal and organizational scales with government organizations and build a hybrid para-state domain, which has more effective power than parliament (see Hotz-Hart 1995). The federal parliament, however, plays an important expressive role. It has a strong ceremonial role in Swiss democracy.
 Assemblage theory can also be embedded in an evolutionary trajectory of space-related research. It moves from an ontological perspective to ontogenesis, from space to spacing. Space (as in relativity theory) can be seen as the emergent product of populations of relational networks. Space therefore is a becoming, and the processes, which produce space are those very processes that create all the individual assemblages.
 We must distinguish actualized governance structures (which are effects of actualized relations), governance as an event (which de/re-territorialises both domains) and governance as the potential of all conversions between domains in the process of their reproduction.
 The domain refers to the assemblage, which emerges from the relation of sets of lower-scale assemblages.
 Transduction refers to the Deleuzian abstract machines of a ‘probe head’ that searches among neighbouring systems for relations of loose singularities or “singularities of resistance, ready to modify these relations, overthrow them and change the unstable diagrams” (Deleuze, 1990, 130).
 He refers to a set of concepts which are all implicitly or explicitly founded on the work of Deleuze. We keep ontogenetically sound by introducing his elaboration on temporality into the assemblage approach.
 For a detailed discussion see DeLanda 2002, chapter 3.
 Individual things are effects of relationality.