The process of metropolisation is commonly and etymologically understood as the development of metropolises as a consequence of extended urbanisation processes spreading across the territory in question (Di Méo 2010: 23). At the turn of the 20^th century, this city type (i.e. primarily capital cities such as Paris and London, but also New York City, for instance) emerged as the architectural, industrial and cultural form of Western urban modernity par excellence (Hall 1998; Harvey 2003; Lees/Lees 2007). In the wake of the post-World War II period, massive decolonisation and post-Fordist deindustrialisation, paired with globalised capitalism and neoliberalism, shifted the definitional and functional nature of metropolises (Harvey 1989; Brenner 2013). As a result, metropolitan areas are now an urban showcase for the imperatives of networked global competitiveness and the world economy (Sassen 1991; Taylor 2003; Brenner/Keil 2005). Through metropolisation processes, the urban question is negotiated beyond the traditional spatial boundaries of ‘the city’: “Metropolisation is this ability and will towards global functional integration, of control through urbanisation mechanisms of ever wider spaces, placed under the authority of cities [cités], of cores operating as networks (…) to the extent that they will constitute some kind of unique and virtual (foil) entity on a global scale beyond the city [la ville] as such” (Di Méo 2010: 24; translation by the authors).

Accordingly, metropolisation processes embody the constitution of integrated urban structures that transcend and reinvent the geographic, functional, institutional and imaginative spatiality of the city. In this context, the imperative of overcoming the traditional understanding of the city as an enclosed spatial entity is contingent on understanding metropolisation as regionalised and networked urbanisation processes. This implies a redefinition and re-scaling of the urban as a category of analysis.

The unprecedented growth and extent of urban morphologies call for new urban planning strategies. Social structural imperatives of the globalised era and the regionalisation and networking of urban systems challenge strategies based on spatially confined cities. Drawing on the new regionalism debates and the global city-region model, researchers and policymakers planning the city-region focus on heterogeneous social structures, spatial and functional polycentricity, trans-regional or trans-national urban networks, as well as cross-border governance (Scott 2001). These territorial reconfigurations suggest a convergence in the meaning of ‘city’ and ‘metropolitan region’ and imply a gradual process of integration. As formulated by Meijers, Hoogerbrugge and Hollander (2014), metropolisation is essentially understood in terms of the structural and cultural integration of a territorially embedded network of cities. These developments are both the grounds for and the result of the dissolution of the binary representational system of space (e.g. centre/periphery, urban/rural). Contrary to colonial metropolitan semantics (Blanchard/Boëtsch 2005), a contemporary urban geographical and planning approach to metropolisation invokes a fluid and multi-level perspective, expressed through pluralist terminology.

There is a dichotomy between understanding metropolisation as a homogenising process of territorial integration and understanding it as a plural heterogenisation of urban space. Indeed, in the process of transcending these seemingly divergent understandings, it can be important to recognise situations in which integration advances on a macro-level, while a conflicted and non-linear metropolisation process remains visible on a micro-level. In order to engage with this differentiation, this paper seeks to examine what a conflict-oriented approach to integration might reveal about the process of metropolisation at different scales, through an analysis of the case of the Métropole d’Aix-Marseille-Provence (M-AMP), in France. The paper questions whether metropolitan policy implementations aiming at stability and integration can de facto co-exist with a fragmented metropolisation of space at smaller scales, where metropolitan integration is perceived differently by, and has a variety of impacts on, different local settings and actors.

The current territorial, political, economic and cultural reconfiguration of traditional cities into polycentric urban regions has spread globally. In France’s second-largest city Marseille, five decades of regionalisation and urban redefinition debates ultimately led to the official launch of the Métropole d’Aix-Marseille-Provence on 1 January 2016. The emergence of this new metropolis provides the impulse for this investigation of metropolisation processes in France’s second-largest city and its surrounding region. Due to its long history, the complex interaction between its stakeholders and institutions, as well as its splintered territory, the phenomenon of metropolisation expresses and impacts a plurality of geographic scales. This plurality represents both an immediate policy challenge and an analytical angle for this research.

Conceptually, we draw on the variable of scales, which we understand as “the extent or size of a phenomenon” (Dahlman 2009: 189). Based on the assumption that both empirical and theoretical foundations of contemporary metropolisation are multiple and heterogeneous, we argue that metropolisation processes in Marseille and the surrounding region manifest themselves in multiple interlocked dimensions. It is particularly relevant to engage with the case of Aix-Marseille-Provence, for in many regards this territory resists unifying tendencies and integrative models. In light of the most recent metropolisation theory, this localised process in-the-making allows for an analytical oscillation between the risk of metropolitan failure and the emergence of an innovative alternative development mode. Therefore, throughout this paper, we want to ask to what extent the fragmentation and the conflictual multiplicity of geographic scales impacted by the metropolisation process can be favourable for a consensual integration process.

The multiple case study approach is based on three projects that explicitly address, shape and contribute to Marseille’s metropolisation. The first is the major state-driven urban regeneration project Euroméditerranée (1995-2015), which was branded as the port city’s “metropolisation accelerator” (Bertoncello/Dubois 2010). The so-called Euromed targets the expansion and modernisation of the built infrastructure in Marseille’s inner city districts and competitive place branding for the sake of economic growth and international capitalisation.

Second, we analyse the impact of the European Capital of Culture (ECOC) festival ‘Marseille-Provence 2013’ (MP2013). With this European award, culture became a catalyser and driving force for urban economic development, as well as a laboratory for collaborative metropolitan governance in the regionalised territory of the ‘capital’. ‘Marseille-Provence 2013’, which has been acclaimed as “the creative and popular expression of metropolitan consciousness [sentiment métropolitain]” (Mission interministerielle pour le projet Aix-Marseille-Provence 2016: 11), embodies metropolisation processes by dint of launching festivals throughout the region, rebranding local heritage and fostering new modes of inter-municipal cooperation.

Third, we focus on the newly created metropolitan region the Métropole Aix-Marseille-Provence (M-AMP). The metropolis represents a new mode of collaborative governance and urban policymaking in a regionalised urban territory, merging 92 municipalities and almost two million inhabitants in and around the urban cores of Aix-en-Provence and Marseille. This innovative administrative categorisation of the territory envisions increased productivity and competitiveness by means of multiplying resources and sharing an increased and unified spatial dimension.

This paper first discusses how and why metropolisation processes can be comprehended through a scalar perspective. To better illustrate this argument, we have chosen to operationalise the empirical cases analytically with a differentiated lens on the various scales implicated. To begin, we consider the territorial-spatial scales at which metropolisation processes become visible. Subsequently, we discuss the different scales at which urban governance stakeholders operate. Finally, we reflect on their respective scopes of impact. Overall, we explore the challenges and opportunities of a multi-scalar metropolisation for Aix-Marseille-Provence by investigating the underlying processes of metropolitan pluralisation and integration.

This paper does not rely on a single methodology but pursues several methodological approaches, mostly grounded in geography, sociology and political science (in the sense of spatial planning and development policies). Overall, this paper is the product of ongoing exchanges, and formal and informal collaborations between both authors over recent years. Based on a comprehensive review of literature, including both theoretical aspects (in particular theories of metropolisation, governance and scales) and practical considerations (on planning for Marseille, urbanism policies, urban marketing strategies etc.), the research is based to an equal extent on thorough field study, regular and documented visits to the locations analysed herein and a qualitative content analysis of 37 semi-structured interviews with cultural policy stakeholders involved in the ECOC selection process and the realisation of MP2013 (Grésillon 2011; Grésillon 2013; De Saussure 2020), as well as with urban stakeholders in Euroméditerranée and the M‑AMP developments (Grésillon/Vignau 2020).

Approaching metropolisation from a scalar perspective reveals a series of related inquiries. On what scale(s) can processes of metropolitan development be discussed? Could a scalar approach define – and limit – metropolisation normatively? What different dimensions exist (spatial/territorial, sociopolitical, economic, temporal), and at which different scales does each dimension manifest itself?

In geographic terms (Dahlman 2009) and alongside burgeoning globalisation, the normative assumptions of scale and processes of scaling are being redefined and enjoying attention amongst scholars. Adam Moore’s article (2008) “Rethinking Scale as a Geographical Category: from Analysis to Practice” provides an insightful account into the state of research on scale theory and politics, as well as discussing shortcomings and criticism, and serves here as an orientation for the following discussion.

Within critical approaches, scholars react to the analytical limits of widespread geospatial categories used as givens and often articulated in a somewhat reductive dichotomy, such as ‘local/global’ or ‘rural/urban’. Furthermore, Brenner (2001) argues that we often evoke scales when in reality we are talking about socio-spatial processes and localisation. The ensuing confusion between spatial and social configurations has resulted in a “tension between fixity and fluidity in scale conceptualization” (Moore 2008: 205). Responding to traditional understandings, critical voices have challenged the notion of scale as a normative measurement and organisation of geographical space (Taylor 1982), arguing that scales are “socially constructed, fluid, and contingent” (Marston 2000: 222). This approach implies a political dimension, the so-called politics of scale (Cox 2002), focusing on the ideological and power-related contingence of the concept. Following this understanding, neo-Marxist and political-economic scholarship theorises scales as both expressions and results of globalised capitalism accumulation (Harvey 1982; Taylor 1982; Smith 1995; Brenner 1997; Brenner 1998). Additionally, a socio-spatial theoretical thread considers scales as “embodiment[s] of social relations of empowerment and disempowerment and the arena through and in which they operate” (Swyngedouw 1997: 169). Marston, Jones and Woodward (2005) question the scalar approach, criticising the inevitable reproduction of a hierarchical ontology of spatial politics.

While Moore (2008: 222) suggests that “despite a recent constructivist turn, the scale politics literature remains wedded to a problematic conceptualization of scale as a substantial category of analysis”, Dahlman (2009: 193) sheds a more optimistic light on these issues, suggesting that “the call to evacuate human geography of scale” essentially manifests “an anxiety to ‘get space right’”. Even if scale were a construct that does not render justice to human geography, it nevertheless provides a category as common ground and a starting point for critical discussion.

In the course of recent years, the debate on the notion of scales has been supplemented and made more complex by research on the degree of pertinence of the metropolitan scales (Gross/Gualini/Ye 2019). There is not only a compelling need “to get space right”, but also – in the light of globalisation and a new international perception of development – “to get the territory right”, which is understood to include the metropolitan area (Schindler/Kanai 2019: 40).

With these critical accounts in mind, we propose a pluralist approach, which enables the dismantling of the variety of scales inherent to given phenomena through an analytical process. An understanding tending towards distinguishing and comparing the socio-political and socio-spatial components of scales resonates with the complex nature of the metropolisation processes in Aix-Marseille-Provence. Indeed, not only are different territorial levels implicated, but these are also produced through the interplay of a complex landscape of actors with different geographical scopes. Grounded in the critical reading of scales introduced above, we therefore analyse our case studies through the perspective recently proposed by D’Albergo, Lefèvre and Ye (2018) as the outcome of an international comparison between Rome, Paris and Shenzen, a perspective in which the metropolitan scale is expressed in terms of spatial, economic and political levels. Whilst Marseille as a metropolis is not really comparable to the three cities studied by the authors, the approach proposed by D’Albergo, Lefèvre and Ye (2018) lends itself to the case study under consideration, despite its specificities. Hence, it is this three-tiered analytical approach – spatial/territorial, socio-economic and political – which guides our analysis of the Marseille metropolitan area.

Being aware of the political, economic and spatial components of scale constructs and actors, as well as how they differ and often conflict, is particularly important in complex urban regions such as Aix-Marseille-Provence. In fact, in a post-welfare-state era, the monopoly of the national government has transformed into a multi-level governance model (Hooghe/Marks 2001; Gualini 2004). In this context, heterogeneous urban regions have arguably become arenas for innovative policy negotiations (Brenner 2004). These shifts have directly involved metropolisation processes: “The crucial challenge of metropolitan policy coordination is the spatial complexity of social and economic activities in the context of institutional fragmentation and the resultant diversity of power coalitions” (Salet/Thornley/Kreukels 2003: 3).

In definitional governance debates, the notion of a shift from government to governance appears recurrently. According to Pierre (2011: 5), “Governance, unlike ‘government’, looks at the interplay between state and society and the extent to which collective projects can be achieved through a joint public and private mobilization of resources”. Accordingly, national governments and state institutions still play a crucial role. However, the policymaking hierarchies are renegotiated and the stakeholder arena is diversified (Hooghe/Marks 2001; Brenner 2004).

The levelled hierarchies and multiplication of governance actors imply a new contingency of governing power and scale (Shirlow 2009). The question ‘Who holds power?’ is asked under new premises and answered variably on multiple scales. Furthermore, the mode in which power is implemented shifts. According to Shirlow (2009: 50), “good governance is constructed around heterarchy (inclusion) rather than hierarchy (exclusion)”. A heterarchical or integrative approach suggests forms of collaborative governance that essentially emphasise cooperation between public and private sectors, as well as across institutional and civil society actors (Ansell/Gash 2008). This resonates directly with the question of scales discussed above, and affects all the phenomenon’s various scalar parameters, from territorial geography to economic and political stakeholders to social processes and cultural integration. Therefore, the governing and administrative structuring of large-scale urban settlements in the metropolisation process has not only become diversified, the multi-level processes have catalysed new modes of multi-level collaborative governance as well (Zimmermann/Galland/Harrisson 2020).

Finally, as Fricke and Gualini (2018) point out, the metropolitan area is also a heterogeneous discursive construction, within which different lines of argument confront each other or are merged in order to contribute to the creation of diverse metropolitan scales. We will see that the analysis of metropolises from the perspective of “Metropolitan Regions as contested spaces” (Fricke/Gualini 2018) proves to be particularly fruitful in the context of the Aix-Marseille metropolis, where the clash of projections and visions contributes to the deconstruction of a common metropolitan perspective and helps turn the metropolitan area into “a regional problem” (Harrison/Growe 2014).

Given that the processes of perception and of metropolitan construction across several tiers (from the local to the international) are becoming ever more diversified and complex (Lefèvre/Pinson 2020), the function of metropolitan governance plays a crucial role, now more than ever. The question is not so much: ‘What is good governance?’ but rather: ‘Which mode of governance is adapted to this particular metropolitan territory?’ In the case of the Aix-Marseille metropolis, this question is essential. We examine it based on three examples of metropolitan projects. Each of the projects can be seen as typical of a particular approach to metropolitan governance: urban regeneration/renewal linked to self-marketing and competitiveness in the 1990s with Euroméditerranée; a project-based cultural approach in the early 2010s with European Capital of Culture Marseille-Provence 2013; and the renaissance of institutional approaches to metropolitan reform in the mid-2010s.

But before jumping straight into the crux of the issue, it is necessary, in particular for readers not intimately familiar with the subject, to briefly contextualise our case study within the general framework of the evolution of metropolitan policies specific to France.

Twentieth-century metropolitan governance constitutes a unique chapter in French postmodern administrative and urban political history. Developments in metropolitan policies have reflected a paradigm shift in the country’s territorial administration and post-decentralisation multiplication of (in)formal policymakers and policy- coordination/implementation bodies (Le Galès 2002; Lefèvre 2009; Pinson 2009). Among the various spheres in which the urban is being negotiated throughout France, there is a tendency for both departmental and regional authorities to lose their relevance in light of contemporary metropolisation politics. The national scale remains significant in the context of traditionally and administratively centralised France. However, the wave of territorial and functional decentralisation laws since the 1980s have led to the transformation of the Jacobin nation-state model in favour of clustered territorial areas, partly managed through new forms of metropolitan governance. Last but not least, the European Union is promoting metropolitan development politics and thereby forms yet another influential scale of urban structuration.

The history of metropolitan planning policies has shaped France’s urban governance landscape since the 1960s, with legislative milestones such as the métropoles d’équilibre (1960s-1970s), city networks and intermunicipalities (intercommunalités) (1990s) and metropolitan areas and metropolises (Métropoles) (2000s-2010s) (Deraëve 2015). French decentralisation laws in the second half of the 1900s catalysed a shift of paradigms, from a centralised elite government toward multiple localised politics (Donzel 2001; Dubois 2009). The end of the national welfare-state era and the rise of diversified and multilevel metropolitan governance could signify the political materialisation of French decentralisation. The alternative model for collaborative governance has enabled the rise of a new metropolitan age. Here, the pluralisation of power and the complexity arising from an increasing number of stakeholders represent two crucial challenges for decision-making processes. According to Ansell and Gash (2008: 547), “collaborative governance is not a ‘winner-take-all’ form of intermediation between single separated interests. In collaborative governance, stakeholders often have an adversarial relationship to one another, but the goal is to transform adversarial relationships into more cooperative ones”. The agonistic character of collaborative governance, provocatively summed up here as a “not a ‘winner-take-all’” solution, must be borne in mind while investigating the conflict-ridden, multi-scalar metropolisation processes in Marseille and its surrounding region.

On the scientific level, the case of the Marseille metropolitan region is of particular interest in that it does not correspond to established patterns or classical models of a metropolisation process. Practitioners of urban geography continue to develop new concepts (“polycentric urban regions”: Meijers 2005; “post-metropolitan territories”: Balducci/Fedeli/Curci 2017) in order to grasp how metropolises have evolved in all their complexity, with certain metropolitan areas functioning like a ‘gigapolis’ (notably, the Tokyo-Yokohama bay region), whereas others (such as the Ruhr) are now spatially, functionally and institutionally so integrated that they might appear as huge cities composed of their individual neighbourhoods. The Marseille-Aix region, however, remains in the nascent phase of metropolisation. It has yet to make substantial progress in terms of the functional integration and political cooperation now present in many large metropolitan areas. It does, however, theoretically possess all the requisite assets to form such a metropolitan area. With a population of almost two million, it is one of the most densely populated regions in France and along the Mediterranean coastline. Marseille, and secondarily Aix-en-Provence, are nexuses of political and economic power. Every single day, tens of thousands of workers, students and commuters travel to and from the main cities in the region.

On the political level and in terms of identity, however, this budding metropolis is not just complex but also a continuing source of conflict, teaming with ruptures and entrenched rivalries. Hence, “Marseille, the impossible city” (Viard 1995) or the “unfinished metropolis” (Club d’échanges et de réflexions sur l’aire métropolitaine marseillaise 1994; Chouraqui/Langevin 2000) remain topical to this day. Nonetheless, a process of metropolitan construction has recently begun, based upon certain ‘metropolitan catalysts’ that we now introduce. Three factors have decisively shaped Marseille’s metropolisation processes over the last two decades: Euroméditerranée, the European Capital of Culture Marseille-Provence 2013 (ECOC MP 2013) and the Métropole Aix-Marseille-Provence (M-AMP).

In the examples cited above, urban planning is an overriding issue. Cultural branding and city marketing, policy and politics, economic development and financial performance are further scopes of action that the willingness to embrace metropolisation seems to address. Hence, the agency and purpose of metropolisation processes are not one-dimensional. Rather, they touch on multiple and entangled developmental sectors of urbanisation. Ultimately, the combination of several complex processes and major events on different scales impacts the city, all while stimulating metropolisation. In Marseille, the combination of a process of urban and economic regeneration launched in 1995 (Euroméditerranée) and a one-off large-scale cultural event (ECOC 2013) strongly impacted not only the scale of the Marseille urban area but also the scale of the metropolis. It is no coincidence that two years after the dynamism of the European Capital of Culture had waned, the Métropole Aix-Marseille-Provence was officially inaugurated – with a more or less equivalent geographical area. And in 2020, seven years after MP 2013, another major cultural event called Manifesta-Biennal of contemporary art took place in different locations in Marseille – but only in Marseille, not in the other cities of the Métropole.

Marseille thus presents a unique case in France of a large city with a certain number of distinct metropolitan assets (economic, scientific, tourist, cultural) but one that nonetheless remains an unfinished metropolis due to centrifugal forces that are undermining it. As long as Marseille’s neighbouring and rival cities refuse to fully associate with it to jointly forge a common destiny and a new common narrative, the ‘great metropolis of southeast France’, which developers have been envisaging for the last thirty years, will not materialise.

What is at stake is a nascent metropolis in the search of its identity. While French cities of comparable size to Marseille (Lyon and Lille) have been characterised by a highly advanced process of metropolisation, the ‘lack of metropolisation’ specific to the Marseille region has heavily penalised it, if we compare it to its domestic and international rivals. Compared with Marseille-Métropole, Greater Lyon and Greater Lille have a tradition of inter-urban collaboration that is reflected across every sector involved: institutional relations, transport, waste management, the environment, economic and cultural policy, universities and research, international relations and outreach. It is challenging to draw lessons that could apply to other French cities from the case study of M‑AMP, given the fact that Marseille is a case apart, or could even be considered as a ‘poor example’, if compared to ‘best planning practice’ implemented not only in Lille and Lyon but also in Bordeaux, Nantes, Montpellier and Strasbourg (Demazière/Sykes 2020).

On the scientific level, the case of the Marseille metropolitan region is of particular interest in that it does not correspond to established patterns or classical models of a metropolisation process. Such models often bring positive economic impacts for metropolitan regions because they enable the concentration of certain metropolitan command and redistribution activities in metropolitan municipalities (Scott 2001; Taylor 2003). This twofold movement, however, has not, to date, taken place in the Marseille region. Marseille, in this respect, constitutes a counterexample because the current process of metropolisation has not yet created an integrated and powerful metropolitan area as is very often the case.

From a theoretical perspective, it seems that the Marseille metropolis is characterised by “a disadvantageous fragmentation” (Hoyler/Freytag/Mager 2006) which is of no benefit to anyone, as opposed for example to the Rhine-Ruhr region and its model of spatial and functional cooperation which benefits everyone (Knapp/Kunzmann/Schmitt 2004). If the case of ‘Marseille, impossible city and fragmented metropolis’ were to be approached from the viewpoint of other similar examples, they should probably be selected from the Mediterranean area. The city-regions of Naples or Athens certainly offer comparable contemporary developments to those in Marseille-Provence, but this hypothesis should be substantiated through in-depth analyses.

In conclusion, how can the ‘counterexample’ of Marseille provide findings of general validity for the field of metropolitan planning? Lessons can always be drawn from atypical cases. We have drawn two major conclusions from this example. Firstly, the example of the Marseille region demonstrates the extent to which the political factor remains decisive in the construction of a metropolitan perspective. For a metropolitan project to come about, there must be an unequivocal commitment on the part of the elected officials, starting with the mayors, irrespective of their political affiliations. The lack of any desire to cooperate on the part of the M‑AMP mayors cost the metropolis 30 years (Maisetti 2015: 84). It is not possible to make up for this delay, even with the interventionism of the central state since 1995. The conditions appear to be far from achieving the mandate to ‘get the territory right’ in the highly competitive international environment, as described by Schindler and Kanai (2019). But secondly, there is a more positive aspect to the metropolitan trajectory of Marseille, which can be generalised for the benefit of other metropolises. If and when the major infrastructure projects of metropolitan significance (of a top-down nature) are implemented, as in the case of the three examples we have discussed in this paper, and if and when there is a strong desire (of a bottom-up nature) on the part of the local economic, cultural and scientific stakeholders to bring the metropolis into existence, it will eventually emerge and establish its structures (Gross/Gualini/Ye 2019; Lefèvre/Pinson 2020).

Even if the metropolis does retain its historical, political and sociological specificities, it could be argued that the way in which M‑AMP will evolve over the coming years may serve as a fresh reference point for urban planners and policymakers in terms of an alternative form of classic metropolisation, perhaps more chaotic and fragmented, but a reference point closer to the realities and identities of southern European cities than to the urban models from and applied to cities in northern Europe.