**2. Theoretical and methodological framework**

Sociology of Science is a field of research in sociology that focuses on the science production's environment. Science and technology studies (STS) is a more interdisciplinary field of research that concentrates also on the study of science production and its relation to society, its values and politics. Both fields study the actors involved, the organization of work, the social and cognitive motivations, the communication tools, the place of objects in technology development, as other related subjects like the building of legitimacy and the power struggle in the scientific world. Number of theories and concepts sprung out of these works. Here we will be focusing on two of them: the actor-network theory (ANT) and the laboratory concept that will be translated in our field through the notion of "bricolage planning".

#### **2.1 Actor-Network Theory (ANT)**

The actor-network theory (ANT) is a large corpus of research that developed as of the eighties around the work of several authors in the field of the sociology of science and science and technology studies, mainly Michel Callon and Bruno Latour. These works

developers even civil society actors are leading their own urban development initiatives. These developments, of different scales and kinds, are in their great majority non-state

This situation poses different central questions to planning theory and to the future of planning. Is planning even necessary in this new context? Can urban development lead to

We believe that in the age of the fragmented city, planning still has an important role. In fact, as Thévenot (1995) puts it, next to being a "prolongation of intentionality", a plan is also a communicative space around the semantics of action. And in today's city shaped of an agglomerate of places under the pressure of unstable global market dynamics and egocentric NIMBY local logics, planning could bring a necessary political dimension. We believe also that planning could be constructed from the bottom-up not necessarily by

This chapter aims at investigating these questions by relying on Sociology of Science and Science Studies' concepts and tools and case studies on Beirut suburbs' municipal networks. From these case studies we can see that in a fragmented city and on a local level, urban development initiatives are not necessarily a chaotic juxtaposition of autonomous projects. Local networks including different kinds of actors may well be in action. These local networks represent laboratories experimenting governance arrangements, urban planning tools and territory building strategies, without necessary calling them so. Some of these networks may join at a time or another to face more challenging supra-local issues. They do so by relying on their experience to attempt to engage in a more formal planning at a supralocal level. In this optic, planning and territories seem "condensations" of networks and development initiatives. Finally, the study of these experimentations in the light of the actornetwork theory (ANT) may help understand what we call here a "bricolage" approach to

Sociology of Science is a field of research in sociology that focuses on the science production's environment. Science and technology studies (STS) is a more interdisciplinary field of research that concentrates also on the study of science production and its relation to society, its values and politics. Both fields study the actors involved, the organization of work, the social and cognitive motivations, the communication tools, the place of objects in technology development, as other related subjects like the building of legitimacy and the power struggle in the scientific world. Number of theories and concepts sprung out of these works. Here we will be focusing on two of them: the actor-network theory (ANT) and the laboratory concept that will be translated in our field through the notion of "bricolage

The actor-network theory (ANT) is a large corpus of research that developed as of the eighties around the work of several authors in the field of the sociology of science and science and technology studies, mainly Michel Callon and Bruno Latour. These works

initiatives and do not necessarily fit in a metropolitan urban strategic plan.

urban planning? What does that mean for planning theory and practice?

articulating development initiatives but by building on them.

**2. Theoretical and methodological framework** 

planning.

planning".

**2.1 Actor-Network Theory (ANT)** 

investigate the ways research networks develop in the fields of science. One of the main ideas of ANT is the consideration of objects (material or ideas) as actors – or actants as they call them – in these networks. Beyond social actors' power game logics, ANT draws the attention to the central role that these objects play in the construction and evolution of these networks.

In an article titled "elements for a sociology of translation", Callon (1986) introduces one of the most interesting applications of ANT. The articles uses the metaphor of "translation" and "betrayal" to explain how an innovative idea gets to be systematized after getting different kind of actors and actants to work together for its success. The translation is a four stages process. First there is the problematization the lead actor makes of a certain phenomenon transforming it into an issue that needs to be dealt with by an intervention. This problematization may well get other actors and actants interested and thinking how this may concern them, Callon speaks of interessment. This interessment is the stage where the lead actor will target these actors and actants to get them to participate in the proposed intervention. If this happens and these actors bring in their resources to be part of the intervention. Callon speaks of translation if this is the case, if not he speaks of betrayal and dissolution of the actor-network. A fourth stage is that of mobilization that questions the possible generalization of the intervention to similar phenomena. In this stage the lead actor becomes the spokesman of a certain complex reality to the outer-world.

The basic analytic position that made such a method relevant is what Callons (1986) calls "free association" indiscriminately between elements of Nature and elements of Society. This means that objects and actors are equal members in any network building. They equally can hold the network project by "translating" the project options, or equally can make it fail. They're not only instruments in the hands of the different actors, they have their own logic and modus operandi. For actors to enrol them they must "interest" them, even adapt them by extending their capacities to fulfil new tasks.

Here, actor has a somehow different definition than its common understanding. *"Actors are entities, human or otherwise, that happen to act. They are not given, but they emerge in relations"*  (Law, 2004, p 102). An actor does not exist outside of an actor-network, he's an actor because he manages to define or alter relations between other actors or actants with whom he gets to form a network. He does so by using intermediaries. *"The intermediary does not serve to merely describe a set of relations, it also manages to order the actions of others. […] Through translation the identity of actors is defined and negotiated and interaction is managed"* (Tait, 2002, p. 73).

Networks too hold a different meaning. "*For actor-network theory, networks are not stable systems of links and nodes (like a telephone system); instead they are metaphors for associations and connections between entities which may be heterogeneous in character. Furthermore, they do not have scale in the traditional sense, but are simply longer or more intensely connected (Latour, 1997, p. 3)"* (Tait, 2002, p. 73).

Interestingly, urban planning and urban development are both, somehow, processes bringing together different actors (politicians, planners, technicians, economic, associative) and objects (spaces, construction materials and tools, but also a large set of legal, administrative, managerial, conceptual, scientific, literary and negotiation tools) and connecting them in different ways. In the last ten to fifteen years we see a rising interest among urban planning and urban studies scholars in ANT and other STS concepts. In the

Bricolage Planning: Understanding Planning in a Fragmented City 99

intention is projected in time and space, a "detour" by a certain number of objects is necessary. "*This detour is generally associated with the confection of equipment transposable from a situation to another, consequently associated to the notion of investment, by opposition to the direct use of existing instrumental resources*" (Thévenot, 1995)5. Consequently, though he acknowledges the crisis of planning as an emanation of a top-down "public intentionality", he defends planning as a possible and necessary communicative constructive space; and he believes that this other version of planning is feasible by focusing on the objects of the

Tait (2002) mobilizes ANT to question central-local relations in the British planning system and the room for manoeuvre it leaves for local actors. ANT analysis shows how some actors by their presence in one network – even in an enrolled position – can *"draw on network resources to order others and construct their own (albeit limited) room for manoeuvre".* The ability for ANT to take in consideration all groupings even the temporal an informal ones help to understand not only the central and stable but also the local and unstable. Tait's article shows also how texts can be of central importance since they may define groups and enrol them in an actor-network construction initiated by others. But texts leave also space for interpretation and consequently for manoeuvre and adjustment. They're central in the

Bryson, Crosby & Bryson (2009) use ANT to understanding strategic planning and the implementation of strategic plans as "a way of knowing". The article builds on the MetroGIS experience, an organization that works on fostering geospatial information sharing and map building in the metropolitan area of the Twin Cities region in Minnesota, USA. The authors come to the conclusion that conventional understanding of strategic planning as "fixed and stabilized category of action", and strategic plan construction and monitoring as standardized objects and methods, abstract strategic planning and strategic plans from the actor-networks within which they're enmeshed. They believe that an ANT perspective can allow an understanding of the central role of objects – here maps and map making – in connecting entities with other entities, and as Latour (2005, p.119) puts it, bringing "*multiple realities [together so] that may lead to stability and unity*". Here participatory map making where realities are confronted and complemented could help to understand, in

Boelens' (2010) actor-relational approach (ARA) that we mentioned earlier is a bold analytical and action framework that relies heavily on ANT. This approach resorts to ANT because it "*sidesteps the stifling duality between macro- and micro-*" and "*offers a subtle extension to the discursive, entrepreneurial or growth management approaches, by including things and entities as autonomous (not passive) forces or (f)actors of importance*." (Boelens, 2010, p. 38). This gives to the approach the possibility to defend an endogenous (local actors) perspective to development, and at the same time an opening to external investment as long as they're inscribed in the actor-network that this planning approach helps emerging. ARA also owes much to ANT in the way it operates: Boelens clearly identifies links between his seven steps scheme and Callon's four stages of translation. However Boelens identifies some "imperfections" in ANT and tries to go by them by resorting to urban regime theory and

an actor-network, the "here-and-now" and point to the "there-and-then."

"composite devices of coordination" present in all plans.

construction of power in planning.

associative democracy literature.

5 Personal translation from French

light of ANT different authors tried to understand different urban phenomena. Here we present some of these works.

The conception of the City as a technical object is not new2, however the new orientations in STS and the development of sociotechnical concepts had an important repercussion on urban studies. One of the main entries developed with the reflections around the impact of new technologies and most importantly the need to adapt the existing infrastructures to new functions and usages. One of the most influential works in this line is the book "splintering urbanism" of Graham & Marvin (2001). It identifies a linkage between infrastructure evolution and social and economical evolution leading to further fragmentation. The theoretical framework of "splintering urbanism" is partially founded on ANT literature. Another interesting insight is that of Hommels (2005) reflecting on urban "obduracy". In this article the author builds on two concepts of STS, the Social Construction Of Technology theory (SCOT) and ANT3, to explain the stability of sociotechnical systems, mainly infrastructures, despite the rapid change of technological innovation. Other authors apply ANT in the study of the implementation of different technologies in the urban environment or their management (i.e. Bowler (1998) on recycling urban waste, Martin (2000) on the use of GIS and Beckmann (2004) on questions of safety and mobility).

Actor-network theory is also mobilized in global and world cities literature. Smith (2003a, 2003b) uses ANT to rethink and refold space and time and redefines world cities as "bodies without organs" stretching like an actor-network through space and time. Smith's writings defy the dominant portrayal of world cities as conceived in the political-economy approach through works like those of Saskia Sassen (2001).

On another level, the work of Jonathan Murdoch (1997a, 1997b, 1998, 2006) largely appropriating ANT has been very influential, and represents an interesting reflection on its application in geography, governance and planning. In this regard the traditional split between society and nature, actors and factors is abolished, only networks remain. Space itself is reconfigured : "*there exists no absolute time-space – just as there is neither absolute nature nor absolute society – but only specific time-space configurations, which are conditioned by motives and relations in networks*" (Boelens, 2010 reporting on Murdoch, 1997a).

In planning STS concepts and especially ANT has also had an important impact. For Thévenot (1995)4, A plan's efficacy is proportional to that of the objects he's mobilizing, objects that give power to the planner over a situation. In the production of a plan, where an

<sup>2 &</sup>quot;In 1979 the Journal of Urban History published the first special issue on the city and technology. A new research agenda emphasized the importance of examining the "intersection between urban processes and the forces of technological change" (Tarr 1979, 275). More precisely, the main purpose of these urban historians was to study the effects of technology on urban form. Researchers studied the role of technologies like street lighting, sewage, or the telegraph in the processes of geographical expansion of cities and of suburbanization. Technology was analyzed as a force that shaped society and the cities, but its own character and development were regarded as rather unproblematic and even autonomous; this new trend in urban history was similar to the early work in technology studies." (Aibar & Bijker, 1997)

<sup>3</sup> To which she adds another entry she develops under the name of persistent traditions to stress the weight of socially interiorized practices

<sup>4</sup> Thévenot in this article stresses some of the limitations of ANT but overall his "regime of familiarity" concept could well be sought as an STS concept.

light of ANT different authors tried to understand different urban phenomena. Here we

The conception of the City as a technical object is not new2, however the new orientations in STS and the development of sociotechnical concepts had an important repercussion on urban studies. One of the main entries developed with the reflections around the impact of new technologies and most importantly the need to adapt the existing infrastructures to new functions and usages. One of the most influential works in this line is the book "splintering urbanism" of Graham & Marvin (2001). It identifies a linkage between infrastructure evolution and social and economical evolution leading to further fragmentation. The theoretical framework of "splintering urbanism" is partially founded on ANT literature. Another interesting insight is that of Hommels (2005) reflecting on urban "obduracy". In this article the author builds on two concepts of STS, the Social Construction Of Technology theory (SCOT) and ANT3, to explain the stability of sociotechnical systems, mainly infrastructures, despite the rapid change of technological innovation. Other authors apply ANT in the study of the implementation of different technologies in the urban environment or their management (i.e. Bowler (1998) on recycling urban waste, Martin (2000) on the use

Actor-network theory is also mobilized in global and world cities literature. Smith (2003a, 2003b) uses ANT to rethink and refold space and time and redefines world cities as "bodies without organs" stretching like an actor-network through space and time. Smith's writings defy the dominant portrayal of world cities as conceived in the political-economy approach

On another level, the work of Jonathan Murdoch (1997a, 1997b, 1998, 2006) largely appropriating ANT has been very influential, and represents an interesting reflection on its application in geography, governance and planning. In this regard the traditional split between society and nature, actors and factors is abolished, only networks remain. Space itself is reconfigured : "*there exists no absolute time-space – just as there is neither absolute nature nor absolute society – but only specific time-space configurations, which are conditioned by motives* 

In planning STS concepts and especially ANT has also had an important impact. For Thévenot (1995)4, A plan's efficacy is proportional to that of the objects he's mobilizing, objects that give power to the planner over a situation. In the production of a plan, where an

2 "In 1979 the Journal of Urban History published the first special issue on the city and technology. A new research agenda emphasized the importance of examining the "intersection between urban processes and the forces of technological change" (Tarr 1979, 275). More precisely, the main purpose of these urban historians was to study the effects of technology on urban form. Researchers studied the role of technologies like street lighting, sewage, or the telegraph in the processes of geographical expansion of cities and of suburbanization. Technology was analyzed as a force that shaped society and the cities, but its own character and development were regarded as rather unproblematic and even autonomous; this new trend in urban history was similar to the early work in technology studies."

3 To which she adds another entry she develops under the name of persistent traditions to stress the

4 Thévenot in this article stresses some of the limitations of ANT but overall his "regime of familiarity"

of GIS and Beckmann (2004) on questions of safety and mobility).

*and relations in networks*" (Boelens, 2010 reporting on Murdoch, 1997a).

through works like those of Saskia Sassen (2001).

(Aibar & Bijker, 1997)

weight of socially interiorized practices

concept could well be sought as an STS concept.

present some of these works.

intention is projected in time and space, a "detour" by a certain number of objects is necessary. "*This detour is generally associated with the confection of equipment transposable from a situation to another, consequently associated to the notion of investment, by opposition to the direct use of existing instrumental resources*" (Thévenot, 1995)5. Consequently, though he acknowledges the crisis of planning as an emanation of a top-down "public intentionality", he defends planning as a possible and necessary communicative constructive space; and he believes that this other version of planning is feasible by focusing on the objects of the "composite devices of coordination" present in all plans.

Tait (2002) mobilizes ANT to question central-local relations in the British planning system and the room for manoeuvre it leaves for local actors. ANT analysis shows how some actors by their presence in one network – even in an enrolled position – can *"draw on network resources to order others and construct their own (albeit limited) room for manoeuvre".* The ability for ANT to take in consideration all groupings even the temporal an informal ones help to understand not only the central and stable but also the local and unstable. Tait's article shows also how texts can be of central importance since they may define groups and enrol them in an actor-network construction initiated by others. But texts leave also space for interpretation and consequently for manoeuvre and adjustment. They're central in the construction of power in planning.

Bryson, Crosby & Bryson (2009) use ANT to understanding strategic planning and the implementation of strategic plans as "a way of knowing". The article builds on the MetroGIS experience, an organization that works on fostering geospatial information sharing and map building in the metropolitan area of the Twin Cities region in Minnesota, USA. The authors come to the conclusion that conventional understanding of strategic planning as "fixed and stabilized category of action", and strategic plan construction and monitoring as standardized objects and methods, abstract strategic planning and strategic plans from the actor-networks within which they're enmeshed. They believe that an ANT perspective can allow an understanding of the central role of objects – here maps and map making – in connecting entities with other entities, and as Latour (2005, p.119) puts it, bringing "*multiple realities [together so] that may lead to stability and unity*". Here participatory map making where realities are confronted and complemented could help to understand, in an actor-network, the "here-and-now" and point to the "there-and-then."

Boelens' (2010) actor-relational approach (ARA) that we mentioned earlier is a bold analytical and action framework that relies heavily on ANT. This approach resorts to ANT because it "*sidesteps the stifling duality between macro- and micro-*" and "*offers a subtle extension to the discursive, entrepreneurial or growth management approaches, by including things and entities as autonomous (not passive) forces or (f)actors of importance*." (Boelens, 2010, p. 38). This gives to the approach the possibility to defend an endogenous (local actors) perspective to development, and at the same time an opening to external investment as long as they're inscribed in the actor-network that this planning approach helps emerging. ARA also owes much to ANT in the way it operates: Boelens clearly identifies links between his seven steps scheme and Callon's four stages of translation. However Boelens identifies some "imperfections" in ANT and tries to go by them by resorting to urban regime theory and associative democracy literature.

<sup>5</sup> Personal translation from French

Bricolage Planning: Understanding Planning in a Fragmented City 101

Consequently, in order for their planning strategy to get through, they're forced to enrol in their actor-network a lot more social actors (in a participatory approach) or technical actants (in a regulatory approach), and most of the cases, both. And they must make a central extra effort to present their case as an emanation of public interest. On the other hand, this actornetwork not only needs to secure agencies between different actors and actants at the moment of plan making, it also needs to do it on the long run, and sometimes with actors and actants that are not yet present. It is clear that in this perspective there is wider possibility for "betrayals" than in the case of local urban development initiatives that can be put together by few actors who do not necessarily claim public interest and usually are more delimited in time. We will address this stabilization issue through the notion of laboratory that will translate,

Laboratories had been an important field of study in the Sociology of Science and in Science and Technology Studies since the seventies. A large number of authors contributed to their study. There is no single model of what is a laboratory and how it functions. However, since we're in the ANT perspective, we propose here to adapt this broad definition of a laboratory that is relevant to our argument: laboratory "*is a typical form of organization of the 'society of knowledge'. Its capacity to act on the world of objects and its dynamism are related to its know-how and to its capacity to reconfigure the entities of the natural and social world*"6 (Vinck, 2007, pp.161). Laboratories are "unstable environments" in different aspects. On one hand, Vinck (2007) stresses that in the case of the laboratory, organization is somehow different than in traditional bureaucracy or production environments. It is more fluid and open. Researchers of a laboratory come in and out of a laboratory, their contribution may be punctual and they're even not necessarily working in the same physical place. On the other, in a laboratory, researchers work with objects (physical or ideal) whose forms, statuses and boundaries are not stabilized and pose interesting questions to science. The way a laboratory operates is by reconfiguring resources: *"Laboratories fusion and reorient existing sociotechnical entities, disconnect and transform them to set them up in new phenomenological and* 

This experimental aspect of the laboratory is it's real strength, it makes it do things (innovations, discoveries) that cannot develop in other types of organizations. However, it is also its major Achillies' heel. This openness and instability can easily lead to the disruption of the laboratory. Good communication and mutual understanding is here central. It is common in laboratories to have people from different disciplinary backgrounds, with different methodologies and sometimes values. To get these researchers working together, there is a great deal of communication issues to be stabilized. That's why in laboratories protocols, classification methodologies, definitions and other conceptual objects are central

In this fragmented world where diversity and disruption render intelligibility more difficult, we believe that the laboratory model – where experimentation and accumulation are the

in the field of planning, through the idea of "bricolage planning"

**2.2 "Bricolage planning"** 

*relational universes"*7 (Vinck, 2007, pp. 162).

for its functioning.

6 Personal translation from French 7 Personal translation from French

The main critiques of Boelens to the adoption of ANT as a framework for the analysis of urban planning could be summarized in three points. First, the absence of a normative dimension in ANT while planning is much a discipline where intention has a central place. Second, the secondary role of objects compared to human-actors in planning where objects are always intermediary and rarely actors, which makes a strong focus on objects not always productive. Third, the agnostic – if not cynical position (Webb, 2010) – of ANT towards central values like democracy and sustainability.

Boelens' critiques have received interesting responses from other authors (Rydin, 2010; Webb, 2010). We rally those responses and believe that ANT has much more potential than that assigned to it by Boelens. For instance, as Rydin (2010) argues, the objects we should focus on in applying ANT framework to planning are less the objects of the planning procedures than the objects that make part of sociotechnical systems that planning is trying to affect. In this case we might well find objects "authoring" networks by problematizing a social situation and getting some human groups to take positions and change their relations with other actors and objects. ANT has also a particular relevance in the study of informality and manoeuvring spaces (Tait, 2002). By mapping the actor-networks we can identify the nodes where some actors could hold a position giving them the opportunity to widen the possibilities by enrolling actors and resources in other networks or by forming their own networks. As for Boelens argument about the lack of normative dimension in ANT, we believe effectively that ANT puzzles: it cannot be classified neither as a normative theory nor as a critical theory – the standard two categories of planning theory. However this makes much the point of ANT: "*Planning activity involves a range of actors and actants interacting, engaging with each other (a rather neutral term), enrolling each other (a less neutral term) and producing outcomes which are a mix of the desired and un-desired, the intended and the unintended consequence. Thus ANT itself challenges the simple distinction critical and normative planning theory*" (Rydin, 2010).

Another recurrent question that poses much of a challenge to the sociology of translation in ANT is its difficulty in analysing stability. In fact, the sociology of translation was developed to understand innovation and the formation of innovation networks. Callon's (1986) identifies a fourth and last step in a translation, the "mobilization of the allies", to get the maximum support to the translation and wider its scope, but Callon does not say anything of the aftermath of the translation, how will this actor-network stabilize itself on the long run. It might be more or less easy to get actors to work together to achieve a certain translation at a certain moment but it is more difficult to maintain this cooperation. Time will bring other problematizations – to use Callon's terminology – causing other actornetworks to emerge and probably hinder the stability of this actor-network.

This distinction between the short-term and long-term is central to our argument. For an urban development to go through, a certain agency between different objects and actors must be stabilized as an outcome of one actor translating others for a certain project very limited in space and time. The will of this "author", as long as he manages to mobilize the necessary resources for his urban development, is the only thing that counts. In the light of ANT, city-level urban planning seems more complex and difficult to achieve than urban developments. In fact, it faces two important challenges. On one hand, urban planning tradition has always been linked to a certain conception of public interest. The authorities in charge of planning, in lot of cities, are still conceiving their role in that perspective. Consequently, in order for their planning strategy to get through, they're forced to enrol in their actor-network a lot more social actors (in a participatory approach) or technical actants (in a regulatory approach), and most of the cases, both. And they must make a central extra effort to present their case as an emanation of public interest. On the other hand, this actornetwork not only needs to secure agencies between different actors and actants at the moment of plan making, it also needs to do it on the long run, and sometimes with actors and actants that are not yet present. It is clear that in this perspective there is wider possibility for "betrayals" than in the case of local urban development initiatives that can be put together by few actors who do not necessarily claim public interest and usually are more delimited in time.

We will address this stabilization issue through the notion of laboratory that will translate, in the field of planning, through the idea of "bricolage planning"
