Ecology and SocietyEcology and Society
 E&S Home > Vol. 26, No. 4 > Art. 19
The following is the established format for referencing this article:
Mascarenhas, A., J. Langemeyer, D. Haase, S. Borgström, and E. Andersson. 2021. Assessing the learning process in transdisciplinary research through a novel analytical approach. Ecology and Society 26(4):19.
Research, part of a special feature on Holistic Solutions Based on Nature: Unlocking the Potential of Green and Blue Infrastructure

Assessing the learning process in transdisciplinary research through a novel analytical approach

1Humboldt-Universität zu Berlin, Department of Geography, Landscape Ecology Lab, Berlin, Germany, 2Institute of Environmental Science and Technology (ICTA), Universitat Autònoma de Barcelona, Barcelona, Spain, 3Helmholtz Centre for Environmental Research (UFZ), Department of Computational Landscape Ecology, Leipzig, Germany, 4Department of Sustainable Development, Environmental Science and Engineering, KTH Royal Institute of Technology, Stockholm, Sweden, 5Stockholm Resilience Centre, Stockholm University, Stockholm, Sweden, 6North-West University, Unit for Environmental Sciences, Potchefstroom, South Africa


Inter- and transdisciplinary research projects bring with them both challenges and opportunities for learning among all stakeholders involved. This is a particularly relevant aspect in social-ecological research projects, which deal with complex real-world systems and wicked problems involving various stakeholders’ interests, needs, and views, while demanding expertise from a wide range of disciplines. Despite its importance in such research efforts, the learning process is often not the primary focus of investigation and therefore the knowledge about it remains limited. Here, we put forward an analytical framework that was developed to assess the learning process of both the research team and other participating stakeholders within the scope of an international transdisciplinary project dealing with urban green and blue infrastructure. The framework is structured around five dimensions of the learning process: “Why learn?” (the purpose of knowledge generation and sharing); “What to learn about?” (the types of knowledge involved); “Who to learn with?” (the actors involved); “How to learn?” (the methods and tools used); 'When to learn?' (the timing of different stages). We developed an interview protocol to operationalize the framework and tested our approach through interviews with project researchers. Based on our empirical results, we draw main lessons learned that can inform other transdisciplinary projects. These include capitalizing on what already exists, addressing trade-offs inherent to different types of knowledge, fostering inter- and transdisciplinarity, engaging stakeholders, supporting a learning environment and fostering reflexivity. Besides the empirical insights and the lessons we present, the main contribution of this research lies in the analytical framework we developed, accompanied by a protocol to apply it in practice. The framework can capture the learning process taking place in transdisciplinary research more comprehensively than similar existing frameworks. The five intertwined dimensions it covers are essential to understand and plan such learning processes.
Key words: interdisciplinarity; knowledge; learning; reflexivity; stakeholders; transdisciplinarity


Mutual learning and self-reflexivity are key for transdisciplinary knowledge production (Polk and Knutsson 2008, Jahn et al. 2012, Wittmayer and Schäpke 2014), which in turn is an important process underlying the resilience and sustainability of social-ecological systems (Brandt et al. 2013, Clark et al. 2016, Hoffmann et al. 2017; Evely et al. 2012, unpublished manuscript, Transdisciplinarity is, at its core, “both critical and self-reflexive: It not only systematically scrutinizes in which ways knowledge is produced and used by different societal actors in support of their concerns; it also methodically challenges how science itself deals with the tension between its constitutive pursuit of truth and the ever increasing societal demand for the usefulness of its results” (Jahn et al. 2012:9). A greater recognition of the different ways of understanding and working with knowledge is thus needed. So is moving away from merely technical approaches to knowledge exchange, limited to uni-directional, linear exchanges (Reed et al. 2014).

Knowledge that supports action toward sustainable development should be perceived by stakeholders as salient (relevant to their needs), credible (scientifically adequate), and legitimate (unbiased, fair, and respectful of stakeholders’ divergent values and beliefs; Cash et al. 2003). Existing research suggests that the attributes of knowledge co-production processes—tightly linked with knowledge legitimacy—are important determinants of whether that knowledge leads to action (Posner et al. 2016). Approaches to assess such attributes are therefore needed, in line with calls for monitoring, reflecting on, and continuously refining knowledge exchange as a flexible process (Reed et al. 2014). The learning process often refers to the production of knowledge as a joint process among stakeholders, including scientists (Walter et al. 2007, Vilsmaier et al. 2015), building on the notion of mutual learning, defined as “the basic process of exchange, generation, and integration of existing or newly developing knowledge in different parts of science and society” (Scholz 2001:118).

Despite its importance in transdisciplinary social-ecological research efforts, the learning process is often not the primary focus of investigation and therefore the knowledge about it remains limited. Literature presenting self-reflections by researchers on the learning taking place in transdisciplinary efforts is rare, while empirical studies of learning often remain implicit regarding who learns about what and why (van Mierlo et al. 2020). Empirical evidence from the different parties involved in transdisciplinary research is needed to improve the existing body of knowledge and better support guidance for knowledge exchange (Reed et al. 2014). Hence, several authors stress the need for more studies focusing on learning, for example in the context of sustainability transitions research (van Mierlo et al. 2020).

In this article we put forward an analytical framework that was developed to assess the learning process of both the research team and other participating stakeholders within the scope of an international transdisciplinary project dealing with urban green and blue infrastructure (GBI). Drawing from existing literature, the framework is structured around five dimensions of the learning process, covering (i) the purpose of knowledge generation and sharing, (ii) the actors involved in the learning process, (iii) the knowledge, insights, ideas, and perspectives involved in the learning process, (iv) the methods and tools used in the learning process, (v) the timing of different stages in the learning process. Because knowledge can be seen as context-dependent and strongly related to an individual’s perceptions and worldview (Gibbons et al. 1994; Evely et al. 2012, unpublished manuscript), we developed an interview protocol to operationalize the framework and tested our approach through interviews with project researchers. Our empirical results from a multi-case study research project offer insights into a learning process taking place in different contexts but under a common overarching conceptual framework. Such an international perspective across local contexts is seen as increasingly important in urban research (Hurley et al. 2016). We conclude by drawing main lessons learned and take-home messages, which can inform other transdisciplinary projects.


Analytical framework

Our approach covers the different components of a knowledge system (Posner et al. 2016): the knowledge co-production process, the participants in the process, and the resulting knowledge itself. However, we refer to the process as a “learning process” instead of knowledge co-production or knowledge exchange process (Enengel et al. 2012, Reed et al. 2014). This is in line with a shifting understanding of knowledge, from “knowledge as a thing” that can be produced, given, and received, toward “knowledge as a process” that is evolving and context-specific (Raymond et al. 2010, Reed et al. 2014; Evely et al. 2012, unpublished manuscript). It also aligns with our interest not only in knowledge sensu stricto (which can be interpreted in a more formal sense, related to education), but also on ideas, insights, or perspectives that the different participants in the learning process might gain. In this sense, we follow a definition of learning process from the transitions literature, as “the process of acquiring and generating new knowledge and insights, and of meaning-making of experiences in communicative interaction, in a reciprocal relationship with the social, (bio-)physical and institutional context. Moreover, it is a non-linear, iterative process in which ideas and possibilities for collaborative action are being developed, experimented with and pursued in a diversity of networks” (van Mierlo et al. 2020:253).

We drew on existing literature to develop our analytical framework for assessing the learning process taking place in transdisciplinary research. As noted by Hoffmann et al. (2017), various frameworks have been developed to structure evaluations (ex ante or ex post; formative or summative) of transdisciplinary research. However, most of them are unsuitable or too limited for our purpose. For example, very few differentiate types of actor involvement at different stages of the research (Hoffmann et al. 2017). Our framework draws more heavily on the works of Enengel et al. (2012), Hoffmann et al. (2017), and Roux et al. (2017). On a study about the specific challenges for implementing co-production of knowledge in doctoral studies, Enengel et al. (2012) developed an analytical framework to compare transdisciplinary case studies, consisting of the following elements: (1) typology of actor roles: Who?, (2) research phases: When?, (3) objectives and forms of actor integration: Why?, and (4) types of knowledge: What? Hoffmann et al. (2017) adapted the framework by Enengel and colleagues to compare transdisciplinary integration across four synthesis processes regarding different types of generated knowledge (what?), different types of involved actors (who?), and different levels of actor involvement (how?) at different stages of the processes (when?). The study by Roux et al. (2017) is the most aligned with our purpose because it focuses on mutual learning within transdisciplinary research, specifically on three aspects that could guide researchers in designing and facilitating such learning: “who to learn with,” “what to learn about,” and “how to learn.” The development of the analytical framework was supported by joint reflections and feedback from all consortium researchers in internal workshops throughout the ENABLE project (described below).

Our analytical framework is structured around five dimensions of the learning process (Fig. 1):

The purpose of knowledge generation and sharing (Why learn?) is linked to the applicability of knowledge. This can take various forms, but two general purposes are particularly relevant: (i) knowledge to develop policy options, strategies, decisions, practices (applicability for policy or society); (ii) knowledge to develop theories, methods, and models (applicability for science). These can be regarded as poles in a gradient, which reflect a life-world approach vs. an inner-scientific approach to transdisciplinary efforts (Jahn et al. 2012), linked with local or context-specific knowledge vs. generalized knowledge (Raymond et al. 2010, Enengel et al. 2012). This dimension differs from the “Why?” by Enengel et al. (2012), which was focused on different types of stakeholder involvement.

The knowledge, insights, ideas, and perspectives involved in the learning process (What to learn about?) can be limitless, so several authors have developed typologies of knowledge. We consider three mutually dependent types of knowledge inherent to transdisciplinary research as particularly relevant (Hadorn et al. 2008): (i) systems knowledge of empirical processes and interactions of factors, addressing questions about the origin, development, and interpretation of life-world problems; (ii) target knowledge concerning questions related to determining and explaining the need for change, desired goals, and better practices; (iii) transformation knowledge dealing with questions about technical, social, legal, cultural, and other means of action to transform existing practices and introduce desired ones. We were also interested in “knowledge on how to create knowledge,” which refers to a reflection about the research approach (e.g., how different methods can be combined to generate knowledge on GBI-related issues in different case study cities; Dunford et al. 2018).

Regarding the participants of the learning process (Who to learn with?), we identified two main groups: (a) the research team; (b) other project stakeholders. These can be further detailed, following the five categories developed by Ritter et al. (2010) and adopted by Enengel et al. (2012): (i) core scientists: the main scientific actors throughout the course of a project; (ii) scientific consultants: academic experts who support the core group; (iii) professional practice experts: practitioners who are often very familiar with the practical and political aspects of the issues investigated, but not necessarily with the specific local case context; (iv) strategic case actors: practitioners at case level with a specific formal or informal responsibility, or professional competence; (v) local case actors: all other actors involved in the processes at the case level. It can be relevant to consider alternative ways of categorizing participants, for example according to the types of knowledge they represent (Roux et al. 2017) or to their level of interest and influence on the project (Reed 2008).

The methods and tools used in the learning process (How to learn?) will vary across projects and case studies, but two concepts are important here (Opdam et al. 2015, Roux et al. 2017): (i) boundary concepts: a special case of boundary objects (e.g. models, indicators, and maps). Co-production of these objects can establish shared interest and bridge understanding across multiple knowledge domains. Similarly, boundary concepts, which are non-material, can play a mediating and translating role in a transdisciplinary context, by creating a discursive space in settings with a common urgency, but without consensus or a common knowledge base; (ii) “third places”: in a transdisciplinary sense, a third place represents a learning space at the interface between academia and practice, where academics and non-academics can have an equal voice when they engage to find common ground regarding particular social-ecological issues. We consider that third places refer not only to physical spaces, but more widely to settings that can promote such a learning space (e.g., through a set of rules of engagement).

The timing of different stages in the learning process (When to learn?) can have several implications, like influencing the policy uptake of scientific knowledge, according to policy windows (Rose et al. 2020). The key stages of a knowledge co-creation process can be categorized as (1) problem history, (2) problem identification and structuring, (3) research design and selection of methods, (4) data collection, (5) data analysis and triangulation, (6) reflection/interpretation and synthesis, and (7) dissemination of results (Pohl and Hadorn 2007, Enengel et al. 2012).

The framework’s dimensions do not necessarily follow a specific sequence; there can be different and multiple entry points, depending on the features of a specific application (like aims, scope, or time of assessment). For example, if applying the framework ex ante to support the design of a learning process, it might make sense to start with what motivates the learning process (why learn?), followed by who should be involved (who to learn with?) and what knowledge, insights, or perspectives are expected or wanted from participants (what to learn about?), and finally considering methods and tools to support the process (how to learn?), always keeping in mind the time dimension (when to learn?).

The ENABLE project

ENABLE was a research project funded under the 2015–2016 call from BiodivERsA, a network of national and regional funding organizations promoting pan-European research on biodiversity and ecosystem services. The project aimed at enabling GBI potential in complex social-ecological regions using a systems perspective and engaging local actors in five case studies: Barcelona (Spain), Halle (Germany), Łódź (Poland), Oslo (Norway), and Stockholm (Sweden). New York City (USA) was also included as an external node for benchmarking. The research reported in this article targeted the five European case studies.

As a transdisciplinary project, ENABLE represented an opportunity to foster learning among all participants, including members of the consortium and the different stakeholders who were engaged in the process, mainly in each case study. Project partners developed approaches tailored to each urban region to achieve that aim (as illustrated by many of the articles in this Special Feature) under ENABLE’s common overarching conceptual framework (Andersson et al. 2019).


We developed an interview protocol to operationalize (ex post) the analytical framework (Appendix 1). Joint reflections and feedback from all ENABLE researchers in internal workshops throughout the project supported the development of the interview protocol (in line with the analytical framework). Some questions differed depending on the main group of participants (the research team or the project stakeholders). To contextualize the results regarding the five dimensions of the framework, it was also important to gather information on what participants found interesting and useful about the process. This can be used to identify relevant aspects that influence learning (Restrepo et al. 2018). To capture that information we added questions drawing on the Most Significant Change technique (Serrat 2017), a story-based, qualitative method for uncovering most significant project impacts experienced by individuals. The main guiding question used to open a conversation through this technique was: “What did you find most interesting and useful from the project? What were the main “take-home messages”?” Two other questions included in this technique drew on the work by Cvitanovic et al. (2016), one on barriers preventing knowledge exchange and one on suggestions for improving knowledge exchange.

Ten members of ENABLE’s research team were interviewed per online call toward the end of the project, between June 2019 and April 2020. We aimed for individual perspectives (as opposed, for example to having a spokesperson per case study) because we see them as most relevant in an inter- and transdisciplinary learning process, where researchers within the same team have had different roles. Because of practical constraints it was not possible to conduct the interviews with project stakeholders across case study cities. The first author of the present article conducted all the interviews and was not included as interviewee, whereas the remaining co-authors were. Interviews were conducted in English and lasted between 30 and 60 minutes. Interviewees signed an informed consent form (Appendix 2). The first author transcribed and manually coded the interviews supported by the software MAXQDA Plus 2020, release 20.0.8 (VERBI Software 2019). The remaining co-authors have verified the coding in a subsequent stage, to identify potential inconsistencies or deviations in interpretation. Interviews were transcribed, in a close way to what the interviewees said, but not fully verbatim, because it was the content of what was being said that was of interest and not the wording (Kuckartz and Rädiker 2019). We anonymize interviewees when presenting results in this article, using identifiers composed of the initials of the case study city (BAR: Barcelona; HAL: Halle; LOD: Łódź; OSL: Oslo; STO: Stockholm; CC: cross-case) followed by an ordinal number (e.g., BAR1). This retains the identification of different case studies and interviewees, which is relevant for the analysis of results.


We present results according to, first, the different dimensions of our analytical framework (see Table 1 for a summary), and second, topics cutting across dimensions. Because the analytical dimensions are closely interrelated, we cross-reference dimensions along the text when pertinent, for example by flagging content that is relevant for other dimensions with “→[dimension’s short designation].” Given the qualitative nature of this research, we have tried to highlight recurring topics from the interviews, while capturing the diversity of topics brought up by interviewees. However, it is not possible to cover all points raised by interviewees, so we refer readers to the coded interviews’ transcriptions in Appendix 3.

Why learn? Applicability for policy, society, and science

Findings on the usefulness of the knowledge, insights, or perspectives resulting from ENABLE varied across case study cities while covering the applicability for policy, society, and science. In terms of applicability for policy and society, because of the scope of ENABLE, most of its outputs and outcomes were aiming to be relevant for GBI planning and management in the case study cities, or in other words, to be salient (Cash et al. 2003). Overall, the project has raised general awareness about GBI benefits, enhanced the focus on the social dimension (distributional issues) of GBI, and it provided planning authorities with data and analyses that they probably could not accomplish themselves because of time constraints or lack of technical capacities. For example, in Oslo, three tools were developed that are already being taken up in practice: a model-based tool to prioritize where green roofs fill demand gaps most effectively, which supports planning and zoning decisions; a Nordic standard for tree valuation, which can equip Oslo’s municipality with an up-to-date tree damage compensation assessment that includes ecosystem services; a blue-green factor standard that can be used as a policy instrument to integrate GBI in new property developments (Horvath et al. 2017). In Łódź, research on children’s exposure to green spaces while walking to school and the production of a digital sociotope map (a map of social functions of public green spaces; Łaszkiewicz et al. 2020) are among the outputs that have “started to inform the local authorities on different green space availability and accessibility standards” [LOD1]. In Stockholm, through the resilience assessment, researchers have promoted “more of a systems understanding” that GBI is not only about the infrastructure itself, “but very much a question of how you think about the city and its inhabitants, around those green and blue spaces” [STO1] (see Borgström et al. 2021; Andersson et al., in press). That process also raised stakeholders’ awareness that GBI “will change and be impacted by change - demographic, economic, governance changes, climate change, environmental change” [STO2] (see Borgström et al. 2021). ENABLE has started a discussion (and provided supporting knowledge) about how to move beyond the dichotomy of conservation only in natural areas vs. densification only in urban areas. In Barcelona, among other efforts that are aligned with policy concerns, a direct contribution to the new municipal resilience strategy (De Luca et al. 2021) was highlighted as a relevant ENABLE outcome for the planning and management of GBI.

Both in the Stockholm and Barcelona cases the joint learning process itself was noted as useful instead of knowledge as such, which “is very intangible in a way, but we speak now the same language, we understand each other in these forums, and I understand the city’s needs and they understand where we are heading, this is very critical and a fundamental way of bringing in new concepts, new critical ideas into the discussion” [BAR1]. The learning process provided a platform for stakeholders “to meet and discuss things that they normally do not have room for discussing in their daily work-life context” [STO2] (→How). This focus on the learning process itself supports the notion that knowledge is not a package that can simply be transferred from producers to users; instead it is better seen as “a process of interaction characterized by multiple changing meanings and interpretations about what the knowledge is about, and how relevant, challenging, or good it is considered to be” (Tuinstra et al. 2019:135). Related to this, we argue that the saliency of the knowledge produced, apparent in several of the ENABLE cases, was tightly linked to its legitimacy, i.e., respectful of stakeholders’ values and beliefs in an unbiased and fair way (Cash et al. 2003), which in ENABLE was actively sought through its transdisciplinary approach (→How, →Who, →When). Nevertheless, we note the transitory nature of solutions to societal problems, inherent to transdisciplinary research (Jahn et al. 2012). It also became apparent that differences across ENABLE cases in terms of knowledge applicability for policy and society reflect the notion that actors in the learning process “enter a setting that has already been shaped by previous experts and past advisory practices, including formal and informal rules and codes of working, as well as a certain understanding of what counts as authoritative knowledge” (Tuinstra et al. 2019:128).

Concerning the applicability for science, across cities ENABLE was seen as useful for considerations of “how to build a comprehensive approach to both understanding and actively engaging with green and blue infrastructure and its functionalities and benefits” [STO1]. The mixed- and multi-methods approaches used within ENABLE (Andersson et al. 2021) “were quite useful to think about how can we look at and address a specific issue through multiple lenses and still combine the insights from them” [STO1]. Similarly, interviewees from Halle and Barcelona highlighted the thinking around filters through ENABLE’s conceptual framework (Andersson et al. 2021), together with the concepts of availability, accessibility, and attractiveness of GBI (Biernacka and Kronenberg 2019) in relation to environmental justice (Langemeyer and Connolly 2020) as useful for science but also for policy and society, which underlines their potential to act as boundary concepts (→How). All these insights speak to the notion of integration, considered to be the main cognitive challenge of transdisciplinarity and defined as “the cognitive operation that establishes a novel, hitherto non-existent connection between distinct entities of a given context” (Jahn et al. 2012:7). Considering the insights reported throughout this article (see also Andersson et al. 2021), it becomes apparent that the ENABLE learning process entailed the three levels of integration suggested by Jahn et al. (2012): epistemic (understanding the methods, notions, and concepts of other disciplines and recognizing and explicating the limits of one’s own knowledge); social-organizational (explicating and connecting different interests or activities of participating researchers, subprojects, and larger organizational units); communicative (establishing some kind of common language that advances mutual understanding and agreement). In this regard, ENABLE’s outcomes could be useful for funding bodies because they show “what interdisciplinary research can be about and what different parts are needed, ... other capacities than [ordinary] research projects” [STO2]. Finally, interviewees also noted that the learning from ENABLE can support the writing of new research proposals and how they conduct future similar research (“why it worked or why it didn’t work” [CC1]), teaching and writing of scientific publications, work as experts in other processes, the ability to engage with emerging topics, like the role of GBI during the COVID-19 crisis (see Barton et al. 2020), or promoting further collaboration with local stakeholders.

Interviewees also reflected on the usefulness of knowledge for themselves. Most answers referred to expanding one’s conceptual understanding or methodological toolbox, related to: the concept of filters, “quite useful ... for the way you engage with the benefits of green and blue infrastructure” [HAL2]; the framework (and assessment methods) of GBI availability, accessibility, and attractiveness; having “a more operational idea of the actual design of transdisciplinary science, ... what are the critical things that need to be integrated, how can they be integrated and how can I describe how to do that and the resources needed” [STO1]; a deeper understanding of preferences, values, and perceptions of citizens concerning the design of green infrastructure; the Blue-Green Factor assessment; or thinking about “the beneficial overlaps between different techniques and methods” [HAL2] (→What). Expanding one’s conceptual understanding can support individuals in adapting mental models and promote double-loop learning (Fazey et al. 2005). These self-reflections can stimulate individual researchers to orient their work toward favoring learning over knowing, which is one of the ways to help build improved capacity for social learning in a sustainability context (Clark et al. 2016).

What to learn about? Different types of knowledge

The most recurring topic emerging from the interviews regarding this dimension was related to insights on working with stakeholders. These included (i) the benefits, challenges, and limitations implied (“For the first time we were doing this exercise with stakeholders ... and I think this is something that we learned is very useful and that we would like to do in the future as well” [LOD1]; “I was reminded of the challenges of working with stakeholders, in terms of problem understanding, the time budget and capacity in total” [HAL2]; “I see much more the limitations linked to that and the bias that the selection of stakeholders brings with it” [BAR1]); (ii) better understanding of stakeholders (“Knowing who the actors are, how they view the system, how they think about other actors” [STO1]); (iii) how to better align the research with stakeholders’ needs (“Getting the research from the lab to the end-users and practitioners, that is definitely what we have learned a lot about” [BAR1]). These insights represent target knowledge as well as “knowledge on how to create knowledge.” Related to the latter but also with systems knowledge, another topic that emerged was learning on applying different methods to specific issues or contexts, which is closely related to the scope and goals of the project (“How different aspects can be studied using different research methods and how manifold methods have been applied to different extents in the different cities and also with different outcomes” [CC1]; see also Andersson et al. 2021). A third emerging topic that links with different types of knowledge concerned governance issues with a spatial expression. In one case this had to do with a disconnect between the city-wide scale of planning and the problems at neighborhood scale (related to transformation knowledge and systems knowledge). The other case concerned the surprisingly large impact of formal administrative boundaries in how people talk about values (more related to target knowledge).

The researchers gained further trans- and interdisciplinary “knowledge on how to create knowledge” through ENABLE. They learned new terms, which can act as boundary concepts (Opdam et al. 2015; →How). These were mainly the concept of filters (infrastructural, perceptual, institutional) mediating the benefits flowing from GBI, put forward in ENABLE’s conceptual framework (see Andersson et al. 2019, 2021); flows (of benefits) and barriers, both closely associated with the filters (Wolff, Mascarenhas, Haase, et al., unpublished manuscript); and the triad of availability, accessibility, and attractiveness to or of GBI (see Biernacka and Kronenberg 2019). Several interviewees stressed that it was not so much about learning new terms per se, rather trying to operationalize them and “having a deeper understanding of what the terms could mean” [STO1], particularly in the different contexts of each case study city. This happened for example with the concepts of environmental justice, nature-based solutions, sustainability, and resilience. In line with a process perspective of learning (Beers and van Mierlo 2017), several interviewees identified not only knowledge, ideas, insights, or perspectives as such, but referred to learning opportunities that the project offered them, often related with the conceptual approaches and different methods that were applied in the project (→How). For example “approaching the green infrastructure planning and the benefits of green infrastructure under a framework of resilience and environmental justice” [BAR2], “looking more in-depth into the mapping of preferences and values ... try and test and adjust the Q-methodology for the first time on our own” [CC1], or more generally “learning by doing, learning by mistakes in trying to develop tools for discussing these things along the way” [STO2]. The latter challenges the fear to fail, one of the most critical shortcomings that transdisciplinary sustainability research has to navigate (Lang et al. 2012).

Who to learn with? Diversity of perspectives

ENABLE researchers engaged with various stakeholders throughout the project and drew different learning insights from that engagement. Because of the scope of ENABLE, focused on the benefits flowing from GBI in urban areas, partners engaged mainly with local authorities, especially their planning, environmental, green space management or similar departments. Engaging with those stakeholders was seen as particularly beneficial to learn about “what is going on in terms of policy” [BAR2], “how processes actually work, what are the real obstacles” [STO1], “the realities and challenges of planners” [BAR1]. Another type of stakeholder involved in several of the case study cities were initiatives or organizations at very local scale, e.g., of the neighborhood. This was considered useful to learn, for example, about the multiple perspectives of residents in a neighborhood facing several social challenges like unemployment or poor integration of migrants [HAL2]. In some cases, stakeholders also included citizens in general, who were “there on their free time just because they cared about the area or had a specific interest in the area” [STO2]. Engaging with stakeholders generally provided an opportunity for critical reflection among the researchers and gaining a better understanding of how to design participatory processes in a transdisciplinary research context (including insights on requirements or different degrees of inclusiveness) or how to apply methods coming from research to specific contexts, “so that it is still understandable and can also create meaningful results” [CC1] (→What, →Why).

Interactions with colleagues within the project consortium promoted learning on a more abstract level. This included conceptual development of aspects related to the ENABLE framework, like the notion of barriers (Wolff, Mascarenhas, Haase, et al., unpublished manuscript), learning how to conduct integrated research or work with different epistemologies, ontologies, and different researchers’ backgrounds, or stimulating reflexivity to extract lessons from what worked or not in each city (see Andersson et al. 2021). Learning also took place through discussions with other scientists, e.g., in conferences or case study workshops, where insights and experiences from ENABLE can be compared with those from similar projects [STO1] (→What). This provides support to the notion that mutual learning among the researchers during a research process needs to be actively established and learning processes beyond the boundaries of individual projects must take place for a comprehensive embedding of the own case and contributing to extant knowledge (Lang et al. 2012).

Interviewees identified stakeholders who could have been beneficial to the learning process, but who were not engaged. Private actors were mentioned several times, for example, “stakeholders from private housing companies ... who actually have quite decisive impact on GBI benefits” [HAL2]. Politicians were noted as a type of stakeholder with similarly high influence. Difficult-to-reach stakeholders were also mentioned, namely marginalized groups representing a specific kind of GBI users who influence “the functionality and perceptions of green and blue infrastructure” [STO1]. Other stakeholders included grassroots groups or neighborhood associations, as well as the general public, which included people who might be engaged in societal issues but not necessarily through organized groups. Insufficient contact with stakeholders (mainly decision makers and practitioners) from case study cities other than one’s own was also noted. Related with this, “also maybe direct interaction between cities could be beneficial for the project” [LOD2]. Engaging with other projects running under the same funding scheme was also seen as potentially beneficial, “to exchange, see what is their research focus and if there may be some overlaps or similarities” [CC1].

How to learn? Framings, boundary concepts, and third places

The project partners promoted a variety of events or opportunities to foster learning within ENABLE. Across case studies, this included workshops with local stakeholders, participation in expert groups, thematic meetings with individual stakeholders, or training events. Additionally, consortium workshops in each case study city brought together the project partners, allowing them to internally discuss different aspects of the project (including self-reflections on the transdisciplinary process itself), as well as getting to know each case study better through field trips and direct interactions with local stakeholders.

Common across case study cities was an effort to meet project needs through the events and learning opportunities promoted, while aligning them with ongoing “real” local GBI planning and management processes and challenges for the planning and management of GBI (→When, →Why). This has guided the framing of each event and the choice of appropriate boundary concepts, around which to focus discussions. For example, in Barcelona the concept of nature-based solutions (linked to GBI) served as an overarching boundary concept. Then, each event was framed around specific topics related to it, such as the evaluation of effective green roof strategies (Langemeyer et al. 2020) or the resilient flow of ES (De Luca et al. 2021). In Stockholm, a resilience assessment process provided an overarching framing, with each event serving as a stepping stone in the process (Borgström et al. 2021). Researchers there made an effort to “find a language and commonalities, common boundary objects to talk about. We’ve had to work very hard to find something that they could start their dialogue about” [STO2]. They also conducted “constant framing exercises that we had to do to explain what we were doing and also for us to learn about the system. The framing was everything from writing invitations, writing documentation, having the first presentation at all the workshops that we had ... all these meetings have a very careful thinking about how we start them, how we talked about the system that we wanted to discuss with the actors. So using words that we know that they know about but also then linking them to the conceptual framework within the project, that was a very tricky part” [STO2]. The Oslo case offered an example of another kind of approach. There, the leading ENABLE researcher engaged with ongoing processes as a member of expert groups.

These sorts of collaborative approaches can promote a genuine bridging of research and practice, hence addressing a critical challenge for knowledge exchange, that of providing access to research knowledge in ways that meet stakeholders’ needs and constraints (Hurley et al. 2016), and enhancing knowledge utilization (Hoffmann et al. 2019). This is aligned with the notion of problem solving organized around a particular application, an attribute of transdisciplinary knowledge production (Gibbons et al. 1994). Framing issues persuasively is an integral part of responding to policy windows, increasing the chances that the research is taken up by policy (Rose et al. 2020). Boundary concepts such as the ones described here can help finding shared interests and bridge understanding across multiple knowledge domains (Opdam et al. 2015, Roux et al. 2017).

Across different framings, goals, and formats, several interviewees stressed the fact that the events described here promoted learning both for researchers and other stakeholders (“It’s also learning for us, because we always use these forums for giving key stakeholders the opportunity to present and discuss their work ... There’s also a learning process in two directions” [BAR1]; “we had a nice exchange [with a local stakeholder], which I would count as a learning event for both sides. For us as researchers as well as the local stakeholders” [HAL2] (→Who). This illustrates the efforts from ENABLE partners in promoting third places (Roux et al. 2017), and is aligned with the notion that collaboration between individuals is needed to gain a fuller understanding of dynamic social-ecological systems (Olsson et al. 2004, Fazey et al. 2005). In an urban planning context like the one in ENABLE, planning practice benefits from new perspectives and improved understanding of problems and solutions from research, while research benefits from being informed by practice problems and practical knowledge (Hurley et al. 2016). This also helps building informal and formal linkages between the project team and other stakeholders, which can play a key role in enhancing the use of knowledge coming from the project (Hoffmann et al. 2019).

When to learn? Key stages, temporal alignment

The most relevant topic emerging from the interviews, related with this dimension, was the temporal alignment of the research project with ongoing processes in each case study city, in order to maximize the relevance of the former to the latter (→How). This shows recognition that timing influences both the extent to which research findings are likely to be perceived as relevant by decision makers, and the way that knowledge from research is used in the decision-making process (Reed et al. 2014), aligned with the notion of “policy windows” (Rose et al. 2020). It played a relevant role to guide the “research design and selection of methods” (one of the key stages introduced in the analytical framework), and it seemed to play a bigger role in the cases where stakeholder engagement was more extensive. For example, in Barcelona, with stakeholder workshops taking place around three times a year, the topics of the meetings varied “depending on the needs of the project at some point, at the same time we try also to talk about topics that are relevant for the stakeholders” [BAR2]. However, aligning project and others’ timelines involved some trade-offs: “At times the two timelines did not align too smoothly, so we tried to bring in ENABLE inputs at specific times that we thought were relevant. So trying to address different stakeholders’ needs and desires in terms of outcomes, which has sometimes maybe detracted from the more pedagogical design of the process” [STO1].

The time preceding the project’s beginning often played an important role for aligning the project with the needs and interests of local stakeholders, thereby increasing its relevance. In most cases, ENABLE was part of broader, pre-existing processes involving the researchers and local stakeholders. There were also consultations with stakeholders in the project’s preparation phase, “about their needs, what are the priority questions, what are the key topics they want to work on through this process and also thinking about key areas in the city for interventions” [BAR1]. This kind of setting the scene and determining what was relevant for the city was seen as a “critical phase” and “a very useful approach in making the entire stakeholder engagement process worth the effort for the stakeholders” [BAR1] (→Why, →How). This illustrates the key stage of “problem identification and structuring” (Pohl and Hadorn 2007, Enengel et al. 2012), being analogous to the “problem transformation” process, the first phase in Jahn et al.’s model of an ideal transdisciplinary research process, whereby societal and scientific problems are linked to form a common research object (Jahn et al. 2012).

The time following stakeholder engagement events was also stressed, particularly in the Barcelona and Stockholm cases, as important to contact stakeholders, requesting feedback from them, and for focused internal reflection: “We test our ideas and approaches with the stakeholders in the individual meetings. And then we have the reporting back phase, where we presented results to the stakeholders and asked for additional feedback. Depending on the study this is more or less intensive” [BAR1] (→What, →How). This is more related with the stages of “data analysis and triangulation,” “reflection/interpretation and synthesis” or assessing new knowledge, and also “dissemination of results/new knowledge” (Pohl and Hadorn 2007, Enengel et al. 2012, Hoffmann et al. 2019). The two latter stages were also the main focus of stakeholder workshops organized across cities, toward the end of the project.

Cross-cutting topic: barriers to learning

Several barriers to learning within the project have been pointed out. Concerning interactions between the project team and other stakeholders, barriers included the following: different “cultures of participation” and different starting points across cities (in some cities, there were previous collaborations between the ENABLE researchers and local stakeholders, in others not, or the general willingness to participate was low); reaching stakeholders “who do not see themselves as stakeholders” [STO1]; conflicts in scheduling, particularly relevant for stakeholders like grassroots groups, neighborhood associations, or NGOs (→Who); ENABLE’s level of abstraction, making it hard for stakeholders to grasp its conceptual framework and demanding extra effort to make it more concrete through illustrative examples. Some stakeholders who could have been beneficial to the learning process were not engaged (→Who). Reasons included changes in personnel within local organizations, which demand renewing contacts and rebuilding trust with researchers, bad or unwanted relationships between researchers and stakeholders, issues of trust among stakeholders (“If you involve people with very strong and very different opinions ... it could take a long time just to find common ground and start to build trust” [STO1]), lack of time from stakeholders like politicians or businesses, and different schedules (e.g., between stakeholders participating on a professional vs. voluntary basis). In this respect, one interviewee noted that “[w]e do have a gap in cooperating with stakeholders from the private sector, that would be in theory and in practice I am not really sure if that would have been helpful for this stakeholder process to learn more. Obviously we could have learned different things, but probably we would have missed out others” [BAR1]. This reflects the need to consider the best form, level, and scale of participation, tailored to the research topic and the preferences and capacities of different stakeholders, instead of assuming that more participation is always better (Enengel et al. 2012, Lang et al. 2012).

Within the consortium, the parallel evolution of a common theoretical framework during the project was thought to have negative implications for the design and integration of empirical methods. A similar issue has been experienced by other authors, for example, in the context of transdisciplinary synthesis projects (Hoffmann et al. 2017). The level of consistency between case studies was often mentioned as not satisfactory. There was the feeling that different teams were working using different approaches “and because of this the opportunities for mutual learning are not as big as they could have been had everyone worked on much more similar things” [LOD1], or if there had been “a more joint comparative analysis” [LOD2]. For one interviewee there was a tension between trying to understand the system and then also adding the aspects of change. There was a focus on the former, which left the researchers with little capacity to address the latter. Finally, time and resource constraints (both from researchers and other stakeholders) were also seen as a barrier. We hypothesize that the barriers described here can be associated with the explorative nature of the project, and the different research teams iteratively working toward a joint understanding of it, making the end goal less clear.

Difficulties related to the use of terms or jargon, including different interpretations thereof, also posed a barrier to learning, mainly within the consortium, but also in engaging with stakeholders. “Sometimes we managed to reach some sort of consensus, in other cases we just had to step back and leave the differences where they were” [STO1]. The triad of GBI availability, accessibility, and attractiveness was mentioned most often. Some partners struggled with the exact definition of each one of those concepts and to some extent different teams used the concepts differently, posing a challenge when it came to cross-case integration. Similar issues of coherence in interpretation were noted for the concepts of perceptions, institutions, governance, or justice. These are known communicative integration challenges in transdisciplinary research (Lang et al. 2012). Regarding possible reasons underlying such difficulties, not putting enough effort into discussing terminology and differences in how different people express their ideas was mentioned. One partner who works in applied research felt there was an overload of complex theoretical terms. In relation to stakeholder engagement, the language also needed adjustments according to stakeholders’ backgrounds. For example, in Barcelona, stakeholders were concerned about the concept of nature-based solutions, because they were more familiar with the concepts of ecosystem services or environmental services and green infrastructure. Although the difficulties described above posed barriers to learning, discussions on finding common ground for definitions were “particularly insightful for all” [LOD1] and they have resulted in “a deeper understanding of what the terms could mean” [STO1]. This is a positive learning outcome and is aligned with the idea that a “learning zone” can emerge out of a situation of discomfort (beyond the comfort zone), as conceptualized by Freeth and Caniglia (2020). Establishing some kind of a common language that advances mutual understanding and agreement also supports integration in transdisciplinary research (Jahn et al. 2012).

It is also useful to identify unmet expectations and the reasons behind them. In ENABLE’s learning process these were mainly related to four issues:

(i) Several interviewees were expecting more comparative work (using joint approaches like common scenario development) to be conducted during the project than it did. Reasons for this included the constellation of disciplines and expertise in the project, different interests across research partners, or the need to be pragmatic in face of the existing amount of work. This provides an alternative expression of the concern that “transdisciplinary settings allow for mutual learning but not for joint research” (Maasen and Lieven 2006:406);

(ii) The balance between a more theoretical or empirical approach. Whereas one researcher thought that ENABLE ran too much as a scientific project, thereby missing more contact with stakeholders from other cities to learn “from those who deal with realities” [LOD2], another researcher would have wanted “more in-depth discussion on how do we best connect methods, theories, frameworks” [STO1]. This mirrors the two contrasting approaches to transdisciplinarity found in the literature: a life-world approach vs. an inner-scientific approach (Jahn et al. 2012), which are linked with a tension between local or context-specific knowledge vs. generalized knowledge (Raymond et al. 2010, Enengel et al. 2012). Hoffmann et al. (2019) regard these as two processes of knowledge production, which transdisciplinary research processes strive to combine: a societal one, where stakeholders address a particular sustainability problem, and a scientific one, where researchers develop research on that particular problem;

(iii) Not being able to conduct some analyses, or at least reaching as far as desired. This was noted, for example, for system and agent-based modeling, as “data gathering was so hard” [HAL1], or learning about justice and resilience together, which was not entirely possible, because “it has been so much work just to link green-blue infrastructure just to these two dimensions” [STO2]. Related with this, one interviewee noted that possibly researchers have tried to address too many topics and that “we might have gotten further if we focused on fewer issues” [OSL1];

(iv) There were difficulties in implementing a planned mobility scheme for young researchers across the cities. This was seen by some as a missed opportunity because it “is a very fruitful way of learning and understanding and exchange” [STO2]. It is a very concrete example of an effort to foster conditions for collaborative learning, in line with suggestions by Freeth and Caniglia (2020). One interviewee noted that expectations have changed several times over the course of the project, which is not necessarily negative, as illustrated by the Barcelona case, where most of the studies conducted were carried out as they emerged as relevant during the project’s lifetime.

Cross-cutting topic: role of context

The role of context, in a project like ENABLE analyzing real complex urban social-ecological systems, became apparent in several responses. Different cities are in different stages in terms of capacities, existing data, and knowledge. The starting point in each city determines to a smaller or greater extent how far one can go in terms of testing new ideas or approaches. “Maybe ecosystem services and green infrastructure are two examples for that: Barcelona has incorporated that already, other cities have not, so if you now come up with new concepts and you elaborate further on this, but the baseline is not given to work with these concepts, then obviously that is much more difficult” [BAR1]. As another interviewee put it, “I would love to be advanced but first I need to have a basic database” [LOD2]. There are also different cultures of participation shaped by the levels of trust and interest in such participatory processes. This became apparent when comparing the stakeholder engagement that took place for example in the Nordic cities (Oslo, Stockholm) represented in the project and in post-socialist cities (Halle, Łódź). Other contextual factors inherent to stakeholders, like cultural differences, e.g., different languages, or different interests, had to be dealt with when engaging with them. Political changes or changes in personnel within stakeholder organizations, like local authorities, can imply contextual changes in perspectives or attitudes and demand building new relationships between project researchers and other stakeholders. Even among project researchers, “your personal background and legacies play a role how you see things and how you understand progress, conflicts, dependence, weakness, success,” so that it becomes relevant “to see how previous learning shapes recent learning” [HAL1]. These insights corroborate the notion that “[t]ransdisciplinarity is a context-specific negotiation” (Klein 2004:521)

Study’s limitations and strengths

A relevant limitation of our application was the inclusion of only the consortium partners, or core scientists (Enengel et al. 2012). Including the views of other stakeholders involved in the project would allow us to assess the learning process more comprehensively. It would also contribute to our approach’s ability to, at least partly, assess social learning, a change in understanding in the individuals involved, and how did the process occur through social interactions and processes between actors within a social network (Reed et al. 2010). However, this was not possible for practical reasons. In Appendix 1 we provide the interview protocol developed specifically for that purpose, for future applications.

The double role of the co-authors also as researchers in the ENABLE project demands some clarification and reflection. The first author was part of the research team leading the case study for the city of Halle (Saale) in Germany. This allowed him to be more actively involved in, and consequently gain deeper insights about, the project activities taking place in that city, than for the remaining case studies. However, he took more of a secondary role in his involvement on most of the activities specific to the Halle case study, allowing for a rather more distanced perspective. Nevertheless, it is impossible to equate this to a situation where the first author would be external to a specific case study or even to the whole project consortium. In principle that would allow for a more distanced perspective, but it could also carry disadvantages with it, most notably a lower level of trust between interviewer and interviewees, with negative impact on the (quality of) information given by interviewees or on their willingness to be interviewed at all by someone external to the project. Aware of the limitations inherent to this study’s context, we took some precautions. The first author strove to draw his analysis solely from the material resulting from the interviews. He also wrote the draft manuscript of the article, while the remaining co-authors contributed at a later stage and were not involved in processing interview data. This was important because they were also interviewed for the study. By appending the coded interview transcripts to the article (Appendix 3), we also give readers the opportunity to make their own judgments on our findings and claims, in face of the underlying data.

The analytical framework developed in this research proved useful to us for capturing the learning process. It enables a broader analysis than each one of the frameworks adapted for its development (Enengel et al. 2012, Hoffmann et al. 2017, Roux et al. 2017) because it covers more dimensions. For example, “including the ‘who’ and ‘when’ may lead to a more sophisticated conceptualization of knowledge that goes beyond simply categorizing different types of knowledge and instead emphasises knowledge as more as a process that can be modelled” (Evely et al. 2012:7, unpublished manuscript). Also, the questions developed to guide the interviews have elicited from the interviewees the information needed to operationalize the framework. We argue that our approach can be useful for future transdisciplinary research projects with similar scope and in different geographic contexts, not only for ex post analysis as we did, but also ex ante, to consider the different aspects of the learning process explored here at a planning stage. As one interviewee put it: “One thing that could be very beneficial for us researchers who are aiming at these very complex research and knowledge processes is to find tools for ourselves to capture this, like having this interview got me thinking about things that I would not necessarily have time or room or acknowledged that I would need to reflect upon. Because if I have that self-reflexive routine that would make this transfer of experiences and insights between projects and processes more clear and visible for me and maybe for others as well” [STO2]. This statement is aligned with the notion that learning outcomes may lead to increased reflexivity, but they can also result from reflexivity changes (Beers and van Mierlo 2017). Applying our approach in other projects would allow gathering additional empirical data to build a more robust body of evidence regarding the findings of this exploratory research.

Whereas the analytical framework supporting our analysis can be used in different stages of a learning process, the interview protocol we developed to operationalize the framework is suitable for an ex post analysis. Nevertheless, we acknowledge the importance of continuous reflexivity throughout transdisciplinary research efforts (Polk and Knutsson 2008, Lang et al. 2012). In the ENABLE project, this was pursued in different ways, for example in meetings among case study teams, or through time slots in project workshops dedicated to joint reflection. However, reporting on the whole reflection process is beyond the scope of this article.

Fostering a learning process within transdisciplinary research projects: take-home messages

Interviewees have reflected on what were the main take-home messages from the project. Based on their answers and further reflection among the authors, we present a set of lessons learned, aiming to support future similar transdisciplinary research projects. Regarding their validity, we acknowledge the exploratory nature of this research. Nevertheless, one should note that transdisciplinarity is “problem solving capability on the move,” so it is hard to predict “where this knowledge will be used next and how it will develop” (Gibbons et al. 1994:13). The following emerged as main lessons learned (clustered around six themes), which can be helpful for future similar initiatives. With these take-home messages we aim at supporting similar efforts:

  1. Capitalizing on what already exists: (a) Assess what sort of systematic learning can be gained from already existing data and knowledge, e.g., feeding it into dynamic models, before collecting new data. There is often the tendency to add more data rather than learn from what already exists. (b) Take advantage of opportunities to engage with ongoing policy-related processes, instead of designing stakeholder engagement processes from scratch that do not have a policy-driven purpose or relevance.
  2. Addressing trade-offs inherent to different types of knowledge: (a) Find a balance between addressing local stakeholders’ concerns and conducting comparative research. Transdisciplinary urban research should be relevant for stakeholders, building on their needs if it is to be impactful. Nevertheless, comparing problems across cities helps put the magnitude of local problems in perspective and in context, and sorting out priorities. It also helps thinking about future scenarios, because one can see alternative states that a given city could be in. Approaching different case studies with a common approach is particularly useful for learning among scientists. These goals can be achieved for example by establishing cross-case working groups targeting specific sets of issues and promoting interactions between researchers and local stakeholders from other cities. Being part of a multi-city endeavor can also leverage stakeholder engagement (higher willingness to participate if people know the same effort is being conducted in other cities, especially “model”/frontrunner cities. (b) Take into account the important role of context in real complex urban social-ecological systems. This relates to the previous point and is particularly relevant when trying to draw more general insights from different case studies.
  3. Fostering inter- and transdisciplinarity: (a) For integrated research running in multiple case studies, promote a continuous (as possible) dialogue between the different research teams. In ENABLE, conducting a deeply integrated transdisciplinary project over a dispersed network proved challenging in this regard. Having a mobility scheme in place, which allows extended stays from researchers in partner organizations, might be helpful. ENABLE had such a scheme but it was not fully realized, so reflecting on its potential was part of the learning process. (b) Embrace different views, expectations, the variety of knowledge people have, and the way they use this knowledge. Accept that there are multiple possible pathways toward a certain desirable state or goal. This might require stepping out of one’s comfort zone, e.g., in terms of one’s academic background, which can be useful to stimulate learning in interdisciplinary collaborative research (Freeth and Caniglia 2020). Paying attention to how one frames issues and looking for ways to find a common ground can prove useful to deal with such differences. This demands being aware of and assuming certain researcher roles, like that of a process facilitator (facilitating the learning process), knowledge broker (mediating between different perspectives), or self-reflexive scientist (being reflexive about one’s positionality and normativity, as part of the system or process under study; Wittmayer and Schäpke 2014). (c) Assign different roles within the team promoting the learning process. This can enable different team members to have different perspectives on the same process. This requires the respective human resources, for example, one person will in most cases not be enough to cover all the different needs of the process, like facilitating and being an observer. Constant reflection on researchers’ roles is also advisable; see previous point and Wittmayer and Schäpke (2014) for additional roles.
  4. Engaging stakeholders: (a) Consider the pros and cons of different stakeholder engagement formats when designing the engagement process. For example, smaller focus groups bring less perspectives together than a larger stakeholder workshop, but they can create a safer space for discussion among stakeholders, while they can also free the researchers from other roles (like being more a facilitator), with benefits for the learning process in both cases. A mix of different formats in different stages of the project, targeting specific objectives, can be most useful for the learning process. Choosing the best mix should take into account the distinct interests, roles, and practices of communication brought by stakeholders. (b) Accept that virtually no one participatory process is perfect. Every project has its limitations, leading to trade-offs in terms of who is involved and what is learned. It might not always be needed and suitable to involve stakeholders in all phases of the project, because different stakeholders contribute differently to different stages of the research process. Participation is shaped by the research aims and should consider stakeholders’ values, preferences, interests, power levels, or constraints. (c) Be explicit about what is on the agenda in terms of stakeholders or processes exerting pressure on GBI, underlying conflicts, or factors hindering research or initiatives to promote GBI.
  5. Supporting a learning environment: (a) Promote exploration and researchers’ own learning within the research team. This was seen as a very positive experience from ENABLE because of its flexibility, and as something that is not taken for granted, when compared to other projects with a more rigid approach. (b) Acknowledge that different kinds of learning opportunities can be important to foster learning, each contributing with its own benefits to the whole learning experience. ENABLE researchers identified various activities in this regard, for example, the writing of scientific articles as an interdisciplinary learning process, internal workshops providing a safe-to-fail environment, or workshops in other case study cities giving insight into other contexts. (c) Encourage learning also beyond the boundaries of the project. Strive for sharing the project’s products and knowledge with stakeholders at different levels, enabling a sustained communication channel between the researchers and other stakeholders. (d) Acknowledge the importance of failure in both process and outcomes. Analyzing non-success can reveal the weak points of a system, which can put it onto an undesired pathway. Reflecting on failing efforts can be insightful not only for the internal learning process but also for others to avoid making the same mistakes. In ENABLE, having safe-to-fail opportunities was seen as beneficial for learning, in line with the notion that a “learning zone” can emerge by going beyond an understimulating comfort zone (Freeth and Caniglia 2020).
  6. Fostering reflexivity: Develop tools and routines to capture the learning process taking place in the project. Having a self-reflexive routine can facilitate the transfer of experiences and insights between projects and processes. Several ENABLE researchers found the exercise reported in this article as useful, to trigger thinking about issues for which they would not necessarily have the time or acknowledged they would need to reflect upon.


Our analytical framework for capturing the learning process taking place in transdisciplinary research projects covers different dimensions of the learning process (Why, What, Who, How, When). It draws inspiration from and expands existing similar frameworks, and has been operationalized through an interview protocol across five European urban regions. The framework helped us distill a set of recommendations for future similar transdisciplinary research projects. These include capitalizing on what already exists, addressing trade-offs inherent to different types of knowledge, fostering inter- and transdisciplinarity, engaging stakeholders, supporting a learning environment, and fostering reflexivity. More generally, the case application also provided empirical insights for each of the framework’s dimensions, and identified cross-cutting issues concerning barriers to learning and the role of context. Further research is needed to test and develop the framework’s applicability for more diverse groups of stakeholders; the case only drew on the experiences of the researchers in the project consortium. Finally, while ours was an ex post application, the framework can also be used ex ante to plan transdisciplinary projects that enhance learning in its multiple dimensions, and throughout projects to identify and engage with barriers to learning and make best use of evolving insights.


Responses to this article are invited. If accepted for publication, your response will be hyperlinked to the article. To submit a response, follow this link. To read responses already accepted, follow this link.


We would like to thank three anonymous reviewers for their constructive comments on an earlier version of this article. We also thank the colleagues who have given their time for the interviews supporting this research. This research was funded through the 2015-2016 BiodivERsA COFUND call for research proposals, with the national funders the Swedish Research Council for Environment, Agricultural Sciences, and Spatial Planning; Swedish Environmental Protection Agency; German Aerospace Center (DLR); National Science Centre (Poland); the Research Council of Norway; and the Spanish Ministry of Science, Innovation and Universities. We acknowledge support by the German Research Foundation (DFG) and the Open Access Publication Fund of Humboldt-Universität zu Berlin.


The data supporting the findings of this study are available as appendix to the article.


Andersson, E., S. Borgström, D. Haase, J. Langemeyer, A. Mascarenhas, T. McPhearson, M. Wolff, E. Łaszkiewicz, J. Kronenberg, D. N. Barton, and P. Herreros-Cantis. 2021. A context-sensitive systems approach for understanding and enabling ecosystem service realization in cities. Ecology and Society 26(2):35.

Andersson, E., S. Borgström, D. Haase, J. Langemeyer, M. Wolff and T. McPhearson. In press. Urban resilience thinking in practice: ensuring flows of benefit from green and blue infrastructure. Ecology and Society.

Andersson, E., J. Langemeyer, S. Borgström, T. McPhearson, D. Haase, J. Kronenberg, D. N. Barton, M. Davis, S. Naumann, L. Röschel, and F. Baró. 2019. Enabling green and blue infrastructure to improve contributions to human well-being and equity in urban systems. BioScience 69(7):566-574.

Barton, D., D. Haase, A. Mascarenhas, J. Langemeyer, F. Baro, C. Kennedy, Z. Grabowski, T. McPhearson, N. Hjertager Krog, Z. Venter, and V. Gundersen. 2020. Enabling access to greenspace during the COVID-19 pandemic: perspectives from five cities. The Nature of Cities, 4 May.

Beers, P. J., and B. van Mierlo. 2017. Reflexivity and learning in system innovation processes. Sociologia Ruralis 57(3):415-436.

Biernacka, M., and J. Kronenberg. 2019. Urban green space availability, accessibility and attractiveness, and the delivery of ecosystem services. Cities and the Environment (CATE) 12(1):5.

Borgström, S., E. Andersson, and T. Björklund. 2021. Retaining multi-functionality in a rapidly changing urban landscape: insights from a participatory, resilience thinking process in Stockholm, Sweden. Ecology and Society 26(4):17.

Brandt, P., A. Ernst, F. Gralla, C. Luederitz, D. J. Lang, J. Newig, F. Reinert, D. J. Abson, and H. von Wehrden. 2013. A review of transdisciplinary research in sustainability science. Ecological Economics 92:1-15.

Cash, D. W., W. C. Clark, F. Alcock, N. M. Dickson, N. Eckley, D. H. Guston, J. Jäger, and R. B. Mitchell. 2003. Knowledge systems for sustainable development. Proceedings of the National Academy of Sciences 100(14):8086-8091.

Clark, W. C., L. Van Kerkhoff, L. Lebel, and G. C. Gallopin. 2016. Crafting usable knowledge for sustainable development. Proceedings of the National Academy of Sciences of the United States of America 113(17):4570-4578.

Cvitanovic, C., J. McDonald, and A. J. Hobday. 2016. From science to action: principles for undertaking environmental research that enables knowledge exchange and evidence-based decision-making. Journal of Environmental Management 183:864-874.

De Luca, C., J. Langemeyer, S. Vaňo, F. Baró, and E. Andersson. 2021. Adaptive resilience of and through urban ecosystem services: a transdisciplinary approach to sustainability in Barcelona. Ecology and Society Vol. 26 Iss. 4 in press.

Dunford, R., P. Harrison, A. Smith, J. Dick, D. N. Barton, B. Martin-Lopez, E. Kelemen, S. Jacobs, H. Saarikoski, F. Turkelboom, W. Verheyden, J. Hauck, P. Antunes, R. Aszalós, O. Badea, F. Baró, P. Berry, L. Carvalho, G. Conte, B. Czúcz, G. Garcia Blanco, D. Howard, R. Giuca, E. Gomez-Baggethun, B. Grizzetti, Z. Izakovicova, L. Kopperoinen, J. Langemeyer, S. Luque, D. M. Lapola, G. Martinez-Pastur, R. Mukhopadhyay, S. B. Roy, J. Niemelä, L. Norton, J. Ochieng, D. Odee, I. Palomo, P. Pinho, J. Priess, G. Rusch, S.-R. Saarela, R. Santos, J. T. van der Wal, A. Vadineanu, Á. Vári, H. Woods, and V. Yli-Pelkonen. 2018. Integrating methods for ecosystem service assessment: experiences from real world situations. Ecosystem Services 29:499-514.

Enengel, B., A. Muhar, M. Penker, B. Freyer, S. Drlik, and F. Ritter. 2012. Co-production of knowledge in transdisciplinary doctoral theses on landscape development - an analysis of actor roles and knowledge types in different research phases. Landscape and Urban Planning 105(1-2):106-117.

Fazey, I., J. A. Fazey, and D. M. A. Fazey. 2005. Learning more effectively from experience. Ecology and Society 10(2):4.

Freeth, R., and G. Caniglia. 2020. Learning to collaborate while collaborating: advancing interdisciplinary sustainability research. Sustainability Science 15(1):247-261.

Gibbons, M., C. Limoges, H. Nowotny, S. Schwartzman, P. Scott, and M. Trow. 1994. The new production of knowledge : the dynamics of science and research in contemporary societies. SAGE, London, UK.

Hadorn, G. H., S. Biber-Klemm, W. Grossenbacher-Mansuy, H. Hoffmann-Riem, D. Joye, C. Pohl, U. Wiesmann, and E. Zemp. 2008. The emergence of transdisciplinarity as a form of research. Pages 19-39 in G. H. Hadorn, H. Hoffmann-Riem, S. Biber-Klemm, W. Grossenbacher-Mansuy, D. Joye, C. Pohl, U. Wiesmann, and E. Zemp, editors. Handbook of transdisciplinary research. Springer, Dordrecht, Netherlands.

Hoffmann, S., C. Pohl, and J. G. Hering. 2017. Exploring transdisciplinary integration within a large research program: empirical lessons from four thematic synthesis processes. Research Policy 46(3):678-692.

Hoffmann, S., J. Thompson Klein, and C. Pohl. 2019. Linking transdisciplinary research projects with science and practice at large: introducing insights from knowledge utilization. Environmental Science and Policy 102:36-42.

Horvath, P., D. N. Barton, E. A. Hauglin, and H. W. Ellefsen. 2017. Blue-green factor (BGF) mapping in QGIS. User Guide and Documentation. NINA Report 1445. Norwegian Institute for Nature Research, Oslo, Norway.

Hurley, J., C. W. Lamker, E. J. Taylor, D. Stead, M. Hellmich, L. Lange, H. Rowe, S. Beeck, P. Phibbs, and A. Forsyth. 2016. Exchange between researchers and practitioners in urban planning: achievable objective or a bridge too far?/The use of academic research in planning practice: who, what, where, when and how?/Bridging research and practice through collaboration: lessons from a joint working group/Getting the relationship between researchers and practitioners working/Art and urban planning: stimulating researcher, practitioner and community engagement/Collaboration between researchers and practitioners: Political and bureaucratic issues/Investigating Research/Conclusion: Breaking down barriers through international practice?. Planning Theory and Practice 17(3):447-473.

Jahn, T., M. Bergmann, and F. Keil. 2012. Transdisciplinarity: between mainstreaming and marginalization. Ecological Economics 79:1-10.

Klein, J. T. 2004. Prospects for transdisciplinarity. Futures 36(4):515-526.

Kuckartz, U., and S. Rädiker. 2019. Analyzing qualitative data with MAXQDA. Springer, Cham, Switzerland.

Lang, D. J., A. Wiek, M. Bergmann, M. Stauffacher, P. Martens, P. Moll, M. Swilling, and C. J. Thomas. 2012. Transdisciplinary research in sustainability science: practice, principles, and challenges. Sustainability Science 7(1):25-43.

Langemeyer, J., and J. J. T. Connolly. 2020. Weaving notions of justice into urban ecosystem services research and practice. Environmental Science & Policy 109:1-14.

Langemeyer, J., D. Wedgwood, T. McPhearson, F. Baró, A. L. Madsen, and D. N. Barton. 2020. Creating urban green infrastructure where it is needed - a spatial ecosystem service-based decision analysis of green roofs in Barcelona. Science of The Total Environment 707:135487.

Łaszkiewicz, E., P. Czembrowski, and J. Kronenberg. 2020. Creating a map of the social functions of urban green spaces in a city with poor availability of spatial data: a sociotope for Lodz. Land 9:183.

Maasen, S., and O. Lieven. 2006. Transdisciplinarity: a new mode of governing science? Science and Public Policy 33(6):399-410.

Olsson, P., C. Folke, and F. Berkes. 2004. Adaptive comanagement for building resilience in social-ecological systems. Environmental Management 34(1):75-90.

Opdam, P., J. Westerink, C. Vos, and B. de Vries. 2015. The role and evolution of boundary concepts in transdisciplinary landscape planning. Planning Theory and Practice 16(1):63-78.

Pohl, C., and G. H. Hadorn. 2007. Principles for designing transdisciplinary research. oekom, Munich, Germany.

Polk, M., and P. Knutsson. 2008. Participation, value rationality and mutual learning in transdisciplinary knowledge production for sustainable development. Environmental Education Research 14(6):643-653.

Posner, S. M., E. McKenzie, and T. H. Ricketts. 2016. Policy impacts of ecosystem services knowledge. Proceedings of the National Academy of Sciences of the United States of America 113(7):1760-1765.

Raymond, C. M., I. Fazey, M. S. Reed, L. C. Stringer, G. M. Robinson, and A. C. Evely. 2010. Integrating local and scientific knowledge for environmental management. Journal of Environmental Management 91(8):1766-1777.

Reed, M. S. 2008. Stakeholder participation for environmental management: a literature review. Biological Conservation 141(10):2417-2431.

Reed, M. S., A. C. Evely, G. Cundill, I. Fazey, J. Glass, A. Laing, J. Newig, B. Parrish, C. Prell, C. Raymond, and L. C. Stringer. 2010. What is social learning? Ecology and Society 15(4):r1.

Reed, M. S., L. C. Stringer, I. Fazey, A. C. Evely, and J. H. J. Kruijsen. 2014. Five principles for the practice of knowledge exchange in environmental management. Journal of Environmental Management 146:337-345.

Restrepo, M. J., M. A. Lelea, and B. A. Kaufmann. 2018. Evaluating knowledge integration and co-production in a 2-year collaborative learning process with smallholder dairy farmer groups. Sustainability Science 13(5):1265-1286.

Ritter, F., A. Muhar, and M. Fiebig. 2010. Transdisziplinärer Dialog : Fachwissen und Erfahrungswissen im Austausch über Sommer-Bergtourismus und Klimawandel [Transdisciplinary dialogue: expert and experiential knowledge in a discourse on summer mountain tourism and climate Change]. GAIA - Ecological Perspectives for Science and Society 19(3):194-203.

Rose, D. C., N. Mukherjee, B. I. Simmons, E. R. Tew, R. J. Robertson, A. B. M. Vadrot, R. Doubleday, and W. J. Sutherland. 2020. Policy windows for the environment: tips for improving the uptake of scientific knowledge. Environmental Science and Policy 113:47-54.

Roux, D. J., J. L. Nel, G. Cundill, P. O‛Farrell, and C. Fabricius. 2017. Transdisciplinary research for systemic change: who to learn with, what to learn about and how to learn. Sustainability Science 12(5):711-726.

Scholz, R. W. 2001. The Mutual learning sessions. Pages 117-129 in J. T. Klein, R. Häberli, R. W. Scholz, W. Grossenbacher-Mansuy, A. Bill, and M. Welti, editors. Transdisciplinarity: joint problem solving among science, technology, and society: an effective way for managing complexity. Birkhäuser Verlag, Basel, Switzerland.

Serrat, O. 2017. Knowledge solutions: tools, methods, and approaches to drive organizational performance. Springer, Singapore.

Tuinstra, W., E. Turnhout, and W. Halffman. 2019. Usable knowledge. Pages 126-140 in E. Turnhout, W. Tuinstra, and W. Halffman, editors. Environmental expertise: connecting science, policy and society. Cambridge University Press, Cambridge, UK.

van Mierlo, B., J. Halbe, P. J. Beers, G. Scholz, and J. Vinke-de Kruijf. 2020. Learning about learning in sustainability transitions. Environmental Innovation and Societal Transitions 34:251-254.

VERBI Software. 2019. MAXQDA 2020. VERBI Software, Berlin, Germany.

Vilsmaier, U., M. Engbers, P. Luthardt, R. M. Maas-Deipenbrock, S. Wunderlich, and R. W. Scholz. 2015. Case-based mutual learning sessions: knowledge integration and transfer in transdisciplinary processes. Sustainability Science 10(4):563-580.

Walter, A. I., S. Helgenberger, A. Wiek, and R. W. Scholz. 2007. Measuring societal effects of transdisciplinary research projects: design and application of an evaluation method. Evaluation and Program Planning 30(4):325-338.

Wittmayer, J. M., and N. Schäpke. 2014. Action, research and participation: roles of researchers in sustainability transitions. Sustainability Science 9(4):483-496.

Correspondent author:
André Mascarenhas
Jump to top
Table1  | Figure1  | Appendix1  | Appendix2  | Appendix3