Perspectives on urban transformation research: transformations in, of, and by cities

The narrative of ‘urban transformations’ epitomises the hope that cities provide rich opportunities for contributing to local and global sustainability and resilience. Urban transformation research is developing a rich yet consistent research agenda, offering opportunities for integrating multiple perspectives and disciplines concerned with radical change towards desirable urban systems. We outline three perspectives on urban transformations in, of and by cities as a structuring approach for integrating knowledge about urban transformations. We illustrate how each perspective helps detangle different questions about urban transformations while also raising awareness about their limitations. Each perspective brings distinct insights about urban transformations to ultimately support research and practice on transformations for sustainability and resilience. Future research should endeavour to bridge across the three perspectives to address their respective limitations.


Science highlights
We outline three perspectives on urban transformations for explaining, structuring and integrating the emerging urban transformations research field. Transformation in cities focuses on unravelling the diverse factors, processes and dynamics driving place-based transformations in cities. This perspective represents research that aims to examine and explain why transformations occur and are supported in some places and not others. Transformation of cities examines the outcomes of transformative changes in urban (sub-)systems. It serves to understand and evaluate the emergence of new urban functions, new interactions and their implications for sustainability and resilience. Transformation by cities looks at the changes taking place on global and regional levels as a result of urbanisation and urban development approaches. The

Introduction
The notion of 'urban transformation' has been gaining ground in science and policy debates. Urban transformations to sustainability and resilience are enshrined in the 2030 United Nations Sustainable Development Goals (SDGs) (UN 2016) and the New Urban Agenda (UN-Habitat 2016a). A rich research field around questions of urban transformations has started to emerge, combining multiple scientific disciplines, ontologies and methods (Elmqvist et al. , 2019Wolfram et al. 2017;Vojnovic 2014). Key to these debates is the aim to put cities on a central stage for accelerating change towards local and global sustainability and resilience.
Urban transformation narratives have been driven by the recognition of the need and opportunity for radical change towards sustainable and resilient cities. Cities constantly experience changes, but contemporary urban change processes are unparalleled. Cities grapple with a variety of interrelated challenges, including pollution, poverty and inequality, ageing infrastructure and climate change UN-Habitat 2016b;Seto et al. 2017). Urbanisation in its current form causes significant changes in land use, energy demand, biodiversity and lifestyles and raises questions about the contribution of cities to global environmental change Alberti et al. 2018;Elmqvist et al. 2013;Seto et al. 2017). At the same time, cities concentrate the conditions and resources for realising the fundamental changes in energy, transportation, water use, land use, housing, consumption and lifestyles that are needed to ensure liveability, wellbeing and sustainability of our (urban) future Koch et al. 2016;Elmqvist et al. 2018). The potential and momentum in cities is visible in for example the 'climate emergency' declarations of local governments that call for accelerated climate action in view of international stalemate.
The notion of urban transformation guides and formulates a better understanding of urban change. On the one hand, 'transformation' serves as an analytical lens to describe and understand the continuous, complex and contested processes and dynamics manifesting in cities, as well as how these dynamics alter urban functions, local needs and interactions between cities and their surroundings (McCormick et al. 2013;Iwaniec et al. 2019). On the other hand, the transformation perspective provides a normative orientation that emphasises the need for radical and systemic change in order to overcome persistent social, environmental and economic problems and to purposefully move towards sustainable and resilient cities in the long-term (Hölscher et al. 2019;Kabisch et al. 2018). Accordingly, sustainability and resilience are complementary concepts to asses and orient urban transformation processes (Elmqvist et al. 2019;Pickett et al. 2014;Simon et al. 2018).
In this paper, we distinguish three perspectives on urban transformations to structure and guide research and practice on urban transformations. Urban transformation research is an emergent, loosely connected interdisciplinary field combining urban studies and complex system studies. Various research fields and disciplines converge in urban transformation research; the multitude of disciplines has been systematically reviewed in Wolfram et al. (2017) and Wolfram and Frantzeskaki (2016). This diversity engenders multiple entry points and provides complementary concepts, theories and insights. However, the diversity causes ambiguities in ontologies, use of concepts and fragmented knowledge about how urban transformations unfold and can be supported.
Urban transformation research would benefit from "gradual interconnection, and the articulation of a certain range of research perspectives" (Wolfram and Frantzeskaki 2016: 2). To facilitate this, we distinguish and describe three perspectives on urban transformations that provide areas of convergence across diverse research approaches. Each perspective provides distinct starting points to generate, structure and integrate knowledge along certain questions. Ultimately, the perspectives outline an agenda for advancing theory and practice on urban transformations for sustainability and resilience: they generate implications for urban policy and practice and a way forward to bridge across the perspectives to address the respective limitations.

Perspectives on transformations in, of and by cities
We distinguish between perspectives on urban transformations in, of and by cities. The perspectives provide entry points for formulating and structuring research questions on urban transformations, integrating research approaches and knowledge, and deriving implications for practice.
The three perspectives start from similar assumptions about cities and urban transformations. They focus on urban transformations as complex processes of radical, systemic change across multiple dimensions (e.g. social, institutional, cultural, political, economic, technological, ecological) (Hölscher et al. 2018;Frantzeskaki et al. 2018a;McCormick et al. 2013). Cities are understood as complex, adaptive and open systems (Alberti et al. 2018;McPhearson and Wijsman, 2017;Ernstson et al., 2010;Collier et al. 2013). This implies that urban transformations are not spatially limited, and driven by and driving cross-scale and cross-sectoral dynamics: cities are "local nodes within multiple overlapping social, economic, ecological, political and physical networks, continuously shaping and shaped by flows of people, matter and information across scales" (Wolfram and Frantzeskaki 2016: 143; see also Hansen and Coenen 2015;Chelleri et al. 2015). To describe, explain and evaluate urban transformations, cities are increasingly approached as social-ecological-technical systems (SETS), including (1) socio-economic, political and institutional dimensions (social); (2) natural resource flows and physical phenomena (ecological); (3) as well as the manmade surroundings (technological) (McPhearson 2020; Alberti et al. 2018;Bai et al. 2017). Actors have a central position within urban systems, influencing how cities are organised and resources are produced and consumed. Given the open character of urban systems, actors are diverse and include household members, local governments, and entrepreneurs also regional and national governments, international bodies and multinational companies, amongst others (Glaas et al. 2019;Webb et al. 2018).
Urban transformations can be desirable or undesirable (Elmqvist et al. 2019;Hölscher 2019). A shared aim across urban transformation research perspectives and approaches is to generate actionable knowledge to intervene in urban transformation processes and support radical change towards sustainable and resilient urban systems (cf. Wittmayer and Hölscher 2017).
Despite these shared starting points and aims, the perspectives ask distinct questions about transformations vis-à-vis urban systems. They look at systemic change dynamics taking place in cities ("in"), the outcomes of systemic change of cities ("of"), or systemic change on global and regional levels driven by cities ("by"). These entry points and corresponding questions manifest in differences along key descriptors of urban transformations (cf. Hölscher et al. 2018). The differences are not contradictory: they generate complementary insights for understanding and supporting urban transformations given the different level of aggregation, analysis and understanding of system dynamics and points of intervention (Table 1).
The main aim of the perspectives is to facilitate structuring of urban transformation research along shared themes and questions. Specifically, in articulating these, we show the actionable knowledge generated through each perspective to support urban transformations for sustainability and resilience. We also show that the perspectives offer bridges across knowledge to strengthen research and practice.

Transformation in cities: cities as places of transformations
Transformation in cities focuses on unravelling the diverse, local, regional and global factors, processes and interactions that converge in cities as places of transformations, thus driving or constraining place-based transformations.
The perspective zooms in on cities as spaces and places. Cities are geolocated in an objective, abstracted point, i.e. space, which is for example demarcated by geographical and administrative boundaries. Cities as places are defined by the physical (i.e. urban form) and philosophical (i.e. imagination and representation) relationships between people and place (Roche, 2016;Knox 2005). Thus, cities as places are both "a centre of meaning and the external context of people's actions" (Knox 2005: 2). As spaces and places of transformations, cities harbour specific potentials, driving forces and barriers (Hansen and Coenen 2015).
Place-based transformations are the result of the social construction by people responding to the opportunities and constraints of their particular locality (Fratini and Jensen 2017;Späth and Rohracher 2014). Endogenous conditions and developments include geographic location, climate, local economic structure, population dynamics and the built environment. For example, urban segregation and inequality result from and are reinforced by interactions between residential choices, personal preferences, job markets, land and real estate markets and public policies (Alberti et al. 2018). The construction of place-based transformations does not take place independently of societal norms and representations of the world. Economic and cultural globalisation and the resulting 'network society' becomes manifest in cities and shape place-based transformation dynamics (Roche, 2016). Scholars seeking to understand the 'geography in transitions' emphasise that cities are positioned within cross-scale spatial and institutional contexts that influence local change dynamics (Hansen and Coenen 2015;Truffer et al. 2015;Coenen et al. 2012;Hodson et al. 2017;McLean et al. 2016). Along similar lines, Loorbach et al. (2020) show the translocal character of social innovations that are locally rooted but globally connected.
This perspective positions transformative agency as deeply embedded in socio-spatial contexts. A central research focus is on urban niches that experiment with and scale

Agency and governance
Urban transformations are multi-actor, contested processes. Urban governance targets cities as geographical and administrative entities, but is embedded within multi-level governance structures and networks.  (Frantzeskaki and Rok 2018). Urban experimentation or real-world laboratories have become process tools to facilitate co-creative and innovative solution finding processes that empower actors to deal with urban problems, for example related to mobility, regeneration, community resilience or green job creation ( A key value of this perspective lies on its embedded research inquiry into the 'black box' of a city, including social, economic and ecological situated and contextual knowledge. A main implication for urban policy and planning practice is to facilitate placebased innovation by going beyond sectoral infrastructuring and top-down masterplanning towards situated and cross-sectoral place-making. Experimental and co-creative governance approaches help recognise and mobilise place-specific capacities. The need for place-based innovation further calls for higher-level policies to be centred on the local dimension. For example, the current European Union Cohesion Policy puts a place-based approach into practice that recognises place variety (Solly 2016) and further extends it to a governance capacity building programme that engages with cities on the ground through the URBACT program (www.urbact.eu).
A limitation of this perspective is that knowledge about and actions instigating transformations in a specific city context are very entrenched in context dynamics. This can limit transferability or scaling other than 'scaling deep' pathway (Moore et al. 2015;Lam et al. 2020) if not connected with mechanisms for global and transnational learning and knowledge transfer (Section 2.3). In (Moore et al. 2015;Lam et al. 2020) addition, neighbourhood-level interventions need to be connected to knowledge about city-level outcomes. This calls for critical evaluations of systemic outcomes in urban systems (Section 2.2).

Transformation of cities: outcomes of transformation dynamics in urban systems
Transformation of cities examines and evaluates the outcomes of transformation dynamics in urban (sub-)systems in terms of new urban functions, local needs and interactions and implications for sustainability and resilience.
This perspective focuses on urban (sub-)systems defined by specific functions (e.g. economy, energy, transport, food, healthcare, housing). Compared to the other perspectives, it most explicitly applies socio-technical and social-ecological, and increasingly SETS, frameworks to describe urban (sub-)systems. Urban transformations are the outcome of radical changes of dominant structures (e.g. infrastructures, regulations), cultures (e.g. values) and practices (e.g. mobility behaviours) of such urban (sub-)systems.
As a result of these changes, what kind of and how system functions are delivered is fundamentally altered (Ernst et al. 2016).
The main aim of this perspective is to explain and evaluate how transformation dynamics affect urban systems' functions. Frameworks and models to investigate how transformation dynamics influence urban (sub-)systems pay attention to the complex processes and feedback loops within, across and beyond urban systems and the accumulated effects on the urban system level. For example, studying social-ecologicaltechnical infrastructure systems in cities advances understanding of urban structurefunction relationships between green space availability, wellbeing, biodiversity and climate adaptation (McPhearson 2020). Similarly, urban metabolism analysis and ecosystem studies seek to understand energy and material flows (Bai 2016;Dalla Fontana and Boas 2019). An emerging perspective on cities as 'multi-regime' configurations investigates dynamics across different functional systems (e.g. energy, water, mobility, food) (Grin et al. 2017;Irvine and Bai 2019). This provides opportunities to unveil interactions across multiple urban systems and scales. For instance, rapid changes in electricity systems can have knock-on effects for urban mobility or heat systems (Chen and Chen 2016;Chelleri et al. 2015). The relational geography perspective puts forth a differentiated view of urban systems: it zooms in on different boroughs, districts or neighbourhoods and raises questions such as how innovation and change in one location affects neighbouring locations (Wachsmuth et al. 2016).
This perspective most explicitly addresses prescriptive, 'goal'-driven and recently mission-driven orientations for reinventing cities to be more sustainable, resilient, inclusive, attractive, prosperous, safe and environmentally healthy Kabisch et al. 2018;Rudd et al. 2018). Researchers and urban practitioners and planners employ concepts like 'sustainability' and 'resilience' as frames to evaluate the state of urban systems and to inform urban planning and regeneration programmes (Elmqvist et al. 2019). The systemic focus and application of such concepts also helps to identify synergies and trade-offs across urban systems and goals. For example, the sustainability paradigm of maximising efficiency in mobility or energy systems might result in vulnerability to natural disasters when systems lack parallel or redundant back-up systems (ibid.). Similarly, scholars point to the risks of green gentrification: while urban greening interventions have multiple benefits for the environment and climate adaptation, if not planned and governed inclusively, they can create unintended dynamics of exclusion, polarisation and segregation (Anguelovski et al. 2019;Haase et al. 2017).
This perspective takes a meta-level view on the agency and governance in cities, highlighting strategic partnerships and interventions based on desired system-level outcomes. From this perspective, cities may act as coherent strategic entities based on systemic understandings of city-specific and long-term effects to pursue managed transitions of their large-scale (sub-)systems (Jensen et al. 2016;Hodson et al., 2017). Urban transformation governance needs to facilitate alignment, foresight and reflexive learning to recognise, anticipate and shape transformation dynamics and leverage points (Hölscher et al. 2019b). Key starting points are shared definitions of what 'desirability' means in specific contexts. Orchestration can align priorities and connect emerging alternatives, ideas, people and solutions (ibid.; Hodson et al., 2017). Shared and long-term visions re-orient short-term decisions and interventions that create synergies across multiple priorities. For example, Galvin and Maassen (2020) analyse Medellín's (Columbia) mobility transformation that also contributed to inclusiveness and public safety. Transition management is a practice-oriented framework to co-develop shared visions, pathways and experiments in an ongoing learning-by-doing and doing-bylearning way (Frantzeskaki et al. 2018b;Loorbach et al. 2015).
In summary, this perspective provides a view on interpreting transformation dynamics and developing orientations and practical guidance for intervention. It becomes visible in urban planning and policy practice through the development of systemic urban concepts as 'anchor points' or attractors for urban transformations such as 'sharing cities', 'circular cities', or 'renaturing cities'. Cities like Rotterdam in the Netherlands and New York City in the USA are using such concepts to formulate long-term climate, sustainability and resilience agendas and establish cross-cutting city-level partnerships for their implementation (Hölscher et al. 2019a). A main implication of this perspective is about the need to institutionalise and prioritise such long-term agendas into policy and planning across sectors and scales (ibid.).
A limitation of this perspective is that it overlooks place-specific implications and can nuance or be agnostic to politics and contestations at local sub-system level. Strategically linking place-based initiatives (Section 2.1) with systemic urban concepts and visions provides a powerful tool to align the multitude of activities taking place in cities and to coordinate urban transformations on (sub-)system scale. Additionally, this perspective requires explicit attention to the relationships between urban systems and their hinterlands or other distant territories, which affect and are affected by urban system's functioning (Section 2.3).

Transformation by cities: cities as agents of change at global scale
The third perspective on transformation by cities draws attention to the changes taking place on global and regional levels as a result of urbanisation and urban development.
The main emphasis is here placed on cities as "agents of change at global scale" (Acuto 2016). As open systems, cities are not just influenced by developments outside their spatial boundaries (see Section 2.1). Urban transformations also have implications on global resources, environmental conditions, commodities and governance.
On the one hand, citiesincluding their social-ecological-technological configurations and the diversity of actors influencing themcan be viewed as culprits driving global high emissions, resource depletion and unsustainability. This raises critical questions about the relationship between current and unprecedented urbanisation and global sustainability (Seto et al. 2017;Haase et al. 2018). For example, the expansion of cities will triple land cover by 2030, compared to 2000, with severe implications on biodiversity (Alberti et al. 2018;Elmqvist et al. 2013). Different frameworks and concepts are employed to describe and assess the linkages between cities and their hinterland and other distant territories, including 'urban land teleconnections' (Seto et al. 2012), 'regenerative cities' (Girardet 2016) and 'urban ecological footprint' (Folke et al. 1997;Hoornweg et al. 2016;Rees and Wackernagel 2008).
On the other hand, cities have become key loci for trialling sustainable approaches and solutions that inform the global sustainability agenda (UN-Habitat 2016b; Seto et al. 2017;Bai et al. 2018). Citiesespecially local governmentsplay key roles in shaping global sustainability programmes and discourses and in developing and sharing knowledge and best practices. Local governments have also become celebrated for taking action when the national government is not (van der Heijden 2018; Acuto 2016). Governance strategies such as experimentation, best practices or imaginaries have been taken up globally (Haarstad 2016;McCann 2011;van der Heijden 2016). This raises questions about how the experiences and best practices showcased in cities become knowledge to be diffused and shared, as well as how transformations travel between places and across scales (Lam et al. 2020).
This perspective supports a polycentric and multi-level approach to global environmental governance. Global environmental governance is becoming increasingly decentralised and polycentric, which is visible for example in climate governance (Ostrom 2014;Jordan et al. 2018;Hölscher and Frantzeskaki 2020) and the urban SDG (UN 2016). The recent 'city charters' of global organisations such as the IPCC Cities and Climate Change, the Convention on Biological Diversity and Cities and Future Earth Urban Knowledge Network, showcase the recognition of 'cities' as key players on a global level. While urban sustainability governance has often proliferated without leadership at national levels, the nestedness of local governance in legal and institutional frameworks at regional, national and international levels requires alignment of priorities and legislation across governance levels (Hughes et al. 2017;Keskitalo et al. 2016).
In summary, this perspective creates knowledge about the role of cities in contributing to global change and what it means for governance, policy and planning at global, national, metropolitan and regional levels. It provides and requires big data from cities and their resource footprints, flows and dynamics so as to draw on patterns and pathways for change that can inform and reinforce global agendas for action. A key mechanism for urban practitioners is to strengthen policy knowledge exchange across frontrunning cities (Hölscher et al. 2019a). Transnational city networks such as the International Council for Local Environmental Initiatives (ICLEI), C40 and 100 Resilient Cities facilitate knowledge exchange and inter-city learning, foster the creation of collective goals, lobby for international attention, and enable the transplantation of innovative, sustainable and resilient policy and planning approaches (Acuto et al. 2017;Lee 2018;Mejía-Dugand et al. 2016;Frantzeskaki et al. 2019;Davidson et al. 2019).
A danger of this perspective is that this global discourse is mainly focused on 'global cities'. Medium-sized and middle-income cities are leaders in terms of actual sustainability performance and need to be actively acknowledged and considered (Vojnovic 2014). Florida (2017) criticises how "winner-take-all cities" reinforce inequality, while many cities stagnate and middle-class neighbourhoods disappear. This requires more research into how resources and opportunities are distributed and made accessible across different cities, for example 'global' cities, metropolitan cities and developing countries' cities (Coenen et al. 2012;Gavin et al. 2013). Additionally, cities are not necessarily a united front: priorities and interpretations differ across cities (Growe and Freytag 2019). To address these issues, this perspective would benefit from a more critical and contextual research approach on place-based transformations (Section 2.1), questioning why transformations occur and are supported in some places and not others. Comparative analyses into the factors and dynamics influencing place-based transformations can facilitate transnational knowledge transfer and upscaling of placebased initiatives.

Conclusions
We offer three perspectives on urban transformations research as a means to cherish and celebrate, but also structure the diversity of the growing urban transformations research field. Our paper is a first attempt to distinguish these perspectives, by discussing key questions, entry points, practical implications and limitations. We show that the perspectives help converge research approaches and clarify how different perspectives provide evidence for urban policy and planning.
The perspectives are not merely conceptual devices: they show up in cities' agendas, programmes and approaches and give guidance to practitioners. The 'transformation in cities' perspective asks practitioners to experiment with collaborative place-making approaches like urban living labs to integrate local knowledge and strengthen a sense of place and empowerment. The 'transformation of cities' perspective appears as underlying integrative systems' approach for core urban strategies such as climate change and biodiversity strategies. The 'transformation by cities' perspective highlights the need to invest in policy knowledge exchange between cities, for example through transnational city networks.
The three perspectives on urban transformation do not exist in isolation from one another. We have shown how the perspectives can feed into and complement each other to address respective research gaps and practical challenges. The main future research direction we put forth is to bridge across the perspectives to address their respective limitations and generate comprehensive actionable knowledge. This means to formulate integrative research questions bridging across perspectives: How do placemaking initiatives in a specific neighbourhood affect urban systems' functioning? How can place-based transformation knowledge be transferred to other city contexts? How can place-based experiments and transformation initiatives or projects inform policy at city and city-network level? What are the conditions for downscaling strategic initiatives from global levelfor example, post-Aichi biodiversity targetsconsidering capacities of urban sub-systems? Ethical approval and consent to participate Not applicable.
Authors' contributions KH conceived of the conceptual structuring approach. Both authors contributed equally to the literature review and drafted the manuscript. All authors read and approved the final manuscript.

Funding
The research leading to this article has received funding from the European Community's Framework Program Horizon 2020 for the Connecting Nature Project (grant agreement no. 730222; www.connectingnature.eu) and the European Community's Framework Program FP7 for the IMPRESSIONS project [grant number 603416, www.impressions-project.eu].

Availability of data and materials
Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.

Consent for publication
Not applicable.