4.1 Interventional nature and a theory of change (ToC) focus

At the heart of transformative research is the desideratum to go beyond observing and analyzing current states of the world, to enacting change toward desirable states of the world (Fazey et al., Reference Fazey, Schäpke, Caniglia, Hodgson, Kendrick, Lyon, Page, Patterson, Riedy, Strasser, Verveen, Adams, Goldstein, Klaes, Leicester, Linyard, McCurdy, Ryan, Sharpe and Young2020; Schneidewind et al., Reference Schneidewind, Singer-Brodowski, Augenstein and Stelzer2016). A transformative sustainability research project implies the existence of a knowledge co-production interventional stage (Lam et al., Reference Lam, Horcea-Milcu, Fischer, Peukert and Lang2019; Wiek & Lang, Reference Wiek, Lang, Heinrichs, Michelsen, Martens and Wiek2016). This can take the form of sustainability pathways building (Chambers et al., Reference Chambers, Wyborn, Klenk, Ryan, Serban, Bennett, Brennan, Charli-Joseph, Fernandez-Gimenez, Galvin, Goldstein, Haller, Hill, Munera, Nel, Österblom, Reid, Riechers, Spierenburg and Rondeau2022), guided deliberative nature valuation (Stålhammar, Reference Stålhammar2021), or the formulation of new narratives of sustainable societies (Linnér & Wibeck, Reference Linnér and Wibeck2020). Sustainability-oriented labs in all their forms (see McCrory et al., Reference McCrory, Schäpke, Holmén and Holmberg2020) are one of the main formats or settings for the contextualization of transformative research, as they are often composed by an sequence of knowledge co-production interventions (Charli-Joseph et al., Reference Charli-Joseph, Siqueiros-García, Eakin, Manuel-Navarrete, Mazari-Hiriart, Shelton, Pérez-Belmont and Ruizpalacios2022, see also the transformative transdisciplinary interventions and living labs subtype in Busse et al., Reference Busse, Zscheischler, Zoll, Rogga and Siebert2023) or experiments (Caniglia et al., Reference Caniglia, Schäpke, Lang, Abson, Luederitz, Wiek, Laubichler, Gralla and von Wehrden2017, see also 4.2).

In comparison with transdisciplinary sustainability research, the knowledge co-production process in itself (see 4.2) is the main intervention, in addition to expecting that the co-produced knowledge will generate societal and scientific impact by feeding other external policy, governance, or technological interventions. These knowledge co-production interventions hold transformative potential for all participants involved, from researchers to community members who become themselves changed in the process (Leventon et al., Reference Leventon, Duse and Horcea-Milcu2021, see also 4.4). Looking at knowledge co-production as an intervention affects the research design, especially when planning for monitoring and evaluation (see 4.6), and when formulating the research question. Answering hypothesis testing or ‘what is’ questions, diagnosing problems and their causes may not necessarily translate into actionable knowledge or provide solution-options to solving a problem. Maintaining humility regarding the capacity of transformative research to offer solution-options is necessary, seeing the entanglements between science, politics and the economy (Spiering & Barrera, Reference Spiering and Barrera2021; van der Hel, Reference van der Hel2018).

To operationalize the solution-orientation and interventional nature of a transformative project, a growing scholarship employs a ‘theory of how and why a certain intervention will be successful’ (Deutsch et al., Reference Deutsch, Belcher, Claus and Hoffmann2021; Schneider et al., Reference Schneider, Giger, Harari, Moser, Oberlack, Providoli, Schmid, Tribaldos and Zimmermann2019a; Wanner et al., Reference Wanner, Hilger, Westerkowski, Rose, Stelzer and Schäpke2018). To explain how change might happen within respective research projects, generic elements of theories of change (ToC) such as interventions (knowledge co-production activities), pathways, impacts, and the roles of scientists need to be articulated (Oberlack et al., Reference Oberlack, Schneider, Herweg, Messerli, Tribaldos, Breu, Giger, Harari, Mathez-Stiefel, Moser, Ott, Providoli and Zimmermann2019). Specifying from the onset a ToC with an appropriate level of detail can reveal flaws in the underlying assumptions or management of change, and facilitate planning for additional activities needed for the intended impact (Belcher & Claus, Reference Belcher and Claus2020). In addition, a ToC can serve as a living monitoring, evaluation and learning tool (Belcher & Claus, Reference Belcher and Claus2020). However, an over-specified theory of change may hinder responsiveness to emergent challenges. Adopting a systems approach that allows for the modularity and adaptability of the ToC may be more suitable (Fritz et al., Reference Fritz, Schilling and Binder2019, see also 4.3). Without the involvement of non-academics and the willingness to learn, ToC runs the danger of becoming a non-reflected top-down box ticking exercise. Transformative researchers have experimented with combining a ToC with other design tools (e.g. Theory U) to counteract its mechanistic side (Moriggi, Reference Moriggi2021).

4.2 Collaborative modes of knowledge production, experimentation and learning

A commitment to collaboration between science, civil society, governments, and industries is necessary when weaving knowledge production to processes of social change (Cvitanovic & Hobday, Reference Cvitanovic and Hobday2018; UN Environment, 2019). To minimize the barriers between science, society, policy, and action, new collaborative forms of knowledge production and use are employed (Irwin et al., Reference Irwin, Culligan, Fischer-Kowalski, Law, Murtugudde and Pfirman2018; Jacobi et al., Reference Jacobi, Llanque, Mukhovi, Birachi, von Groote, Eschen, Hilber-Schöb, Kiba, Frossard and Robledo-Abad2022). The co-modes of co-design of research, co-production and co-implementation of knowledge have been highlighted as suitable to address complex, wicked problems of planetary boundaries, human institutions and behavior (Chambers et al., Reference Chambers, Wyborn, Ryan, Reid, Riechers, Serban, Bennett, Cvitanovic, Fernández-giménez, Galvin, Goldstein, Klenk, Nel, Österblom, Bednarek, Bennett, Brandeis, Charli-Joseph, Chatterton and Pickering2021; Hakkarainen et al., Reference Hakkarainen, Mäkinen-Rostedt, Horcea-Milcu, D'Amato, Jämsä and Soini2022; Moser, Reference Moser2016; Norström et al., Reference Norström, Cvitanovic, Löf, West, Wyborn, Balvanera, Bednarek, Bennett, Biggs, Bremond, De, Campbell, Canadell, Carpenter, Folke, Fulton, Gaffney, Gelcich, Jouffray, Leach and Österblom2020; UNESCO, 2017). Other authors point to the ambition of co-production to build capacities, strengthen social networks and relationships, and transform norms and institutional structures within science and society (Jagannathan et al., Reference Jagannathan, Arnott, Wyborn, Klenk, Mach, Moss and Sjostrom2020, see also 4.6). In this vein, an essential requirement is for the co-design of the research process to start from problem framing that is from the creation of a joint understanding of the sustainability problem to be addressed (Lang et al., Reference Lang, Wiek, Bergmann, Stauffacher, Martens, Moll, Swilling and Thomas2012). Researchers and stakeholders as problem co-definers have been also emphasized by other interventional research communities including action research (Bradbury, Reference Bradbury and Bradbury2015). In fact, prior to problem framing, to maximize the transformative potential of knowledge collaborations, researchers deploy substantial resources to, for example, initiate a transdisciplinary process. Neglecting this initiation phase risks glossing over the tangible challenges that arise in real-world problem constellations, and prevents them from being fully addressed (Horcea-Milcu et al., Reference Horcea-Milcu, Leventon and Lang2022).

Beyond co-production, another collaborative knowledge creation process is experimenting with co-produced knowledge (Fazey et al., Reference Fazey, Schäpke, Caniglia, Patterson, Hultman, van Mierlo, Säwe, Wiek, Wittmayer, Aldunce, Al Waer, Battacharya, Bradbury, Carmen, Colvin, Cvitanovic, D'Souza, Gopel, Goldstein and Wyborn2018, see also 4.1). Epistemological attention has recently focused on observing experiments in real-world settings, often conceptualized as real-world laboratories (McCrory et al., Reference McCrory, Schäpke, Holmén and Holmberg2020; Schäpke et al., Reference Schäpke, Stelzer, Caniglia, Bergmann, Wanner, Singer-Brodowski, Loorbach, Olsson, Baedeker and Lang2018b; Stojanovic, Reference Stojanovic2021). Transformative experiments produce empirical evidence through iterative prototyping, evaluating and learning from outcomes (Caniglia et al., Reference Caniglia, Schäpke, Lang, Abson, Luederitz, Wiek, Laubichler, Gralla and von Wehrden2017). Unlike classical experiments in the natural and social sciences, transformative experiments offer the opportunity to link co-produced knowledge to real-world contexts, enhance interactive learning, and generate outcomes for innovative solutions (Weiland et al., Reference Weiland, Bleicher, Polzin, Rauschmayer and Rode2017). Experimental settings with potential for upscaling may also accelerate social innovation. To build the capacity to link knowledge with action for sustainability, scientists encourage the creation of organizational safe spaces fit for experimentation (Clark & Harley, Reference Clark and Harley2020; Pereira et al., Reference Pereira, Frantzeskaki, Hebinck, Charli, Scott and Dyer2020).

The collaborative modes of knowledge production and experimentation collectively emphasize the importance of learning together along the whole of knowledge co-design, co-production or co-implementation processes (Hakkarainen et al., Reference Hakkarainen, Mäkinen-Rostedt, Horcea-Milcu, D'Amato, Jämsä and Soini2022). Co-learning is seen as a necessary premise for reflexive collaboration to occur or to produce evidence within transformative experiments. A transformative research design actively creates and fosters opportunities for learning not as a by-product of research, but as an intentional outcome, while action research plans for learning as an integral part of the research process (Bradbury, Reference Bradbury and Bradbury2015). More so since some practitioners of change frame transformations in terms of learning at the level of values and paradigms (Waddell, Reference Waddell2016), while in natural resource management they regard co-learning as supporting positive change (Hakkarainen et al., Reference Hakkarainen, Mäkinen-Rostedt, Horcea-Milcu, D'Amato, Jämsä and Soini2022).

4.3 Systems thinking literacy and contextualization

There is an increasing demand for a systems perspective on sustainability transformations from both policy and academia (Voulvoulis et al., Reference Voulvoulis, Giakoumis, Hunt, Kioupi, Petrou, Souliotis, Vaghela and Rosely2022). Systems thinking is well placed to comprehend the non-linearity of wicked sustainability problems, and allows for the generation of transformation knowledge, such as the identification of points of intervention or leverage points, supporting the interventional nature of transformative research (Leventon, Reference Leventon2021, see also 4.1). As a holistic cognitive framework, ‘thinking in systems’ (Meadows, Reference Meadows2008) shifts the focus of analysis from elements to interrelationships and stimulates the ability to identify and deal with uncertainty. Especially the notion of social-ecological systems has highlighted the constantly evolving relationships between human societies and their environment (Partelow, Reference Partelow2018). Despite sometimes being criticized for a mechanistic approach, systems thinking notions, such as leverage points, have been largely taken up as a metaphor, while knowledge about system properties such as emergence and hierarchical organization has been integrated as everyday heuristics, practical wisdom (Caniglia et al., Reference Caniglia, Freeth, Luederitz, Leventon, West, John, Peukert, Lang and von Wehrden2023; Fazey et al., Reference Fazey, Schäpke, Caniglia, Hodgson, Kendrick, Lyon, Page, Patterson, Riedy, Strasser, Verveen, Adams, Goldstein, Klaes, Leicester, Linyard, McCurdy, Ryan, Sharpe and Young2020), or cognitive skill (Chowdhury, Reference Chowdhury2023). The UN Educational, Scientific and Cultural Organization (UNESCO) cites systems thinking among the eight key sustainability competencies needed to think and act for sustainable development (Leicht et al., Reference Leicht, Heiss and Byun2018).

Alongside this theoretical literacy, sustainability problems and solution options are context-dependent, situating the practice of transformative research in ‘real-life’ circumstances, which demand attention to bounded political, institutional, socio-economic, and ecological contexts. Context dependencies, their implications for method selection, and corresponding societal and scientific effects, are still insufficiently understood in co-modes of research (4.2). Neglecting contextualization, through generalization and aggregation, can obscure system inequalities and perpetuate power asymmetries and colonial structures in science: who transforms what for whom? (Lahsen & Turnhout, Reference Lahsen and Turnhout2021, see also 4.5; Lam et al., Reference Lam, Hinz, Lang, Tengö, Wehrden and Martín-López2020). Translating transformation pathways or sustainable development goals to specific contexts can be done relative to a place or system at a certain scale (e.g. national energy system or regional mobility system) (Luederitz et al., Reference Luederitz, Abson, Audet and Lang2017; Schneider et al., Reference Schneider, Kläy, Zimmermann, Buser, Ingalls and Messerli2019b). Place-based interdisciplinary research has made valuable theoretical and methodological contributions to addressing cross-scale social-ecological dynamics and global drivers of change while maintaining a focus on locally relevant issues (Martín-López et al., Reference Martín-López, Balvanera, Manson, Mwampamba and Norström2020). However, place-based contextualization should not be done at the expense of overlooking non-linearities and scale interdependencies (e.g. telecoupling). Hence, transformative research aspires to alternate between zooming in on the small-scale system for contextualization and zooming out on the level of the embedding system for coherence, permanently managing the tension between place-based and wider systems transformations.

4.4 Reflexivity, normative and inner dimensions

Reflexivity is a way to bring transparency to the normative dimensions of transformative research. Reflexivity expands and challenges the conventional role of researchers to interrogating the assumptions they bring to the transformative research inquiry (Evans, Reference Evans2021; Horlings et al., Reference Horlings, Romero, Pisters and Soini2020). In the case of science seeking to steer transformative change, it is essential that researchers become more reflexive about their non-epistemic values underpinning methodological choices (Horcea-Milcu et al., Reference Horcea-Milcu, Abson, Apetrei, Riechers, Dușe, Dorninger, Lam, Freeth and Lang2019; Minna et al., Reference Minna, Jacobi, Korhonen, Jani, Anna, Peltomaa, Schneider, Tribaldos and Zaehringer2024). Added to this, the interrogation of epistemic values embedded in scientific practices is useful for bridging gaps between fields within collaborative contexts (MacLeod & Nagatsu, Reference MacLeod and Nagatsu2018; Nagatsu et al., Reference Nagatsu, Davis, DesRoches, Koskinen, MacLeod, Stojanovic and Thorén2020; Soininen et al., Reference Soininen, Raymond, Tuomisto, Ruotsalainen, Thoren, Horcea-Milcu, Stojanovic, Lehtinen, Mazac, Lamuela, Korpelainen, Vainio, Toivanen and McPhearson2022). Allocating resources for critical self and group reflection in inter- and trans-disciplinary teams could support the development of reflexivity as a common practice in academia (Borie et al., Reference Borie, Gustafsson, Obermeister, Turnhout and Bridgewater2020; Freeth & Vilsmaier, Reference Freeth and Vilsmaier2020). This calls for redefining notions of success and rethinking productivity within academic reward systems (see section 6).

The normative core of transformative research is perhaps what differentiates it the most from other forms of engaged research. Transformative research has the normative mission of orientating societal development toward human values (Lang et al., Reference Lang, Renn, Rehm and Ernst2021) such as empathy (Brown et al., Reference Brown, Adger, Devine-Wright, Anderies, Barr, Bousquet, Butler, Evans, Marshall and Quinn2019). Normativity and values-thinking in transformative research includes making values in research explicit, but also building upon and promoting certain values (Redman et al., Reference Redman, Wiek and Barth2021). Recent work by IPBES focuses on ‘sustainability-aligned values’, such as justice and stewardship (IPBES Reference Pascual, Balvanera, Christie, Baptiste, González-Jiménez, Anderson, Athayde, Chaplin-Kramer, Jacobs, Kelemen, Kumar, Lazos, Martin, Mwampamba, Nakangu, O'Farrell, Raymond, Subramanian, Terma and Vatn2022), and advocates for their ‘unleashing’ (IPBES, Reference Díaz, Settele, Brondízio, Ngo, Guèze, Agard, Arneth, Balvanera, Brauman, Butchart, Chan, Garibaldi, Ichii, Liu, Subramanian, Midgley, Miloslavich, Molnár and Obura2019) by multiple actors across sectors including academia. The IPBES Values Assessment refers to ‘values-centered, system-wide transformations’ and recognizes the values underpinning different sustainability pathways (e.g. green economy or degrowth). For example, green economy is underpinned by instrumental values, emphasizing the role of nature as an asset, while degrowth is underpinned by values of sufficiency and egalitarianism for shaping people's balance with nature (IPBES, Reference Pascual, Balvanera, Christie, Baptiste, González-Jiménez, Anderson, Athayde, Chaplin-Kramer, Jacobs, Kelemen, Kumar, Lazos, Martin, Mwampamba, Nakangu, O'Farrell, Raymond, Subramanian, Terma and Vatn2022). These underlying values should be scrutinized and iteratively re-evaluated against other criteria such as justice and equity, and against the community values where the co-production intervention takes place (see 4.2). Theories of change and associated interventions need not only be articulated (as described in 4.1), but also unpacked to reveal their normative assumptions and value orientations, for example with the help of ethicists.

The transformative way of doing science may also change interiorities (Wamsler et al., Reference Wamsler, Osberg, Osika, Herndersson and Mundaca2021) including those of researchers. It invites deep reflection regarding the relationship between the researched and the researcher (Wittmayer & Schäpke, Reference Wittmayer and Schäpke2014), and between knowledge and underlying values (Hansson & Polk, Reference Hansson and Polk2018), constantly challenging gained perspectives. Future knowledge systems need to be capable of actively working with values and emotions (Fazey et al., Reference Fazey, Schäpke, Caniglia, Hodgson, Kendrick, Lyon, Page, Patterson, Riedy, Strasser, Verveen, Adams, Goldstein, Klaes, Leicester, Linyard, McCurdy, Ryan, Sharpe and Young2020) so that proposed solution-options can be acknowledged as both personal and political (Pereira et al., Reference Pereira, Frantzeskaki, Hebinck, Charli, Scott and Dyer2020; Spiering & Barrera, Reference Spiering and Barrera2021; Vogel & O'Brien, Reference Vogel and O'Brien2022; Wamsler & Osberg, Reference Wamsler and Osberg2022). Tackling the complexity and unpredictability of steering transformative change requires not only new ways of producing knowledge (4.2), but also the capacity to make collaborative knowledge processes, sustainability-aligned values (Daedlow et al., Reference Daedlow, Podhora, Winkelmann, Kopfmüller, Walz and Helming2016), and inner dimensions to resonate (Wamsler & Osberg, Reference Wamsler and Osberg2022).

4.5 Local agency, decolonization, and reshaping power

Transformative research recognizes the limitations of the traditional linear science-society relationship in addressing sustainability challenges (Figure 1). It dismantles existing hegemonies such as the neoliberalisation of academia, or the Trust in Numbers imperative of quantitative research (Porter, Reference Porter1995), and problematizes dominant framings of science (Hölscher et al., Reference Hölscher, Wittmayer, Hirschnitz-Garbers, Olfert, Walther, Schiller and Brunnow2021; Temper et al., Reference Temper, McGarry and Weber2019). It recognizes the ideal to fundamentally change existing knowledge systems toward epistemically diverse ones, equipped to work with complex systemic issues (see 4.3) and with values (see 4.4) (Fazey et al., Reference Fazey, Schäpke, Caniglia, Hodgson, Kendrick, Lyon, Page, Patterson, Riedy, Strasser, Verveen, Adams, Goldstein, Klaes, Leicester, Linyard, McCurdy, Ryan, Sharpe and Young2020). Existing ways of creating and applying knowledge are often a product of the same system and power dynamics driving the sustainability problems that transformative research is seeking to address in the first place (e.g. climate change, biodiversity loss). Critically inquiring knowledge (co-)production for whom and with whom (Fritz & Binder, Reference Fritz and Binder2020; Manuel-Navarrete et al., Reference Manuel-Navarrete, Buzinde and Swanson2021; Turnhout et al., Reference Turnhout, Metze, Wyborn, Klenk and Louder2020) becomes central to the organization of the transformative scientific process. In order to build socially inclusive spaces and prevent marginalization of voices, breaking ingrained power relations is needed (Temper et al., Reference Temper, Walter, Rodriguez, Kothari and Turhan2018; Vogel & O'Brien, Reference Vogel and O'Brien2022). As are decolonizing social relations in the practice of knowledge creation (Gram-Hanssen et al., Reference Gram-Hanssen, Nicole and Julia2022) and celebrating plural ways of knowing that challenge narrow scientific framings (Latulippe & Klenk, Reference Latulippe and Klenk2020; Nightingale et al., Reference Nightingale, Gonda and Eriksen2022). Within the conservation field, the ‘transformative turn’ prompted conservation researchers to promote inclusivity (of underrepresented voices) and plurality (of ways of knowing) when revisiting biodiversity research and action (Colloff et al., Reference Colloff, Lavorel, van Kerkhoff, Wyborn, Fazey, Gorddard, Mace, Foden, Dunlop, Prentice, Crowley, Leadley and Degeorges2017; Raymond et al., Reference Raymond, Cebrian-Piqueras, Andersson, Andrade, Raymond, Cebria, Schnell, Romanelli, Filyushkina, Goodson, Horcea-Milcu, Plieninger, Van Riper, Verburg and Wiedermann2022; Wyborn et al., Reference Wyborn, Montana, Kalas, Clement, Davila Cisneros, Knowles, Louder, Balan, Chambers, Christel, Forsyth, Henderson, Izquierdo Tort, Lim, Martinez-Harms, Merçon, Nuesiri, Pereria, Pilbeam and Ryan2021).

Transformative research moves from considering those taking part in the research process as design participants (see 4.2) to considering them agents of change. The agenda of transformation needs to be shared with local people (Nightingale, Reference Nightingale2017), recognizing the centrality of human agency in responses to global crises (Mehta et al., Reference Mehta, Srivastava, Movik, Adam, Souza, Parthasarathy, Naess and Ohte2021). Applications in local decision-making, as opposed to the global level, present conditions necessary for a higher proportion of human-nature relationships to be represented instead of aggregated. Transformative research acknowledges existing efforts, experiences and agency in a given place (Lam et al., Reference Lam, Horcea-Milcu, Fischer, Peukert and Lang2019). Its goal is to support and enable sustainability transformation processes according to the local target knowledge. Initiatives driven by local actors and communities can lead the pathway to sustainability according to an explicit or implicit context-specific vision of social-ecological well-being (Fischer et al., Reference Fischer, Horcea-Milcu, Lang, Thale-Bombien, Abson, Apetrei, Clarke, Derwort, Dorninger, Duse, Freeth, Jager, Klaniecki, Lam, Leventon, Newig, Peukert, Riechers and Schaal2019). Failing to explore, understand, and capitalize on such ongoing efforts is a missed opportunity and risks disempowering individual and group agency. For example, to favor ownership and empowerment, the notion of sustainability and its associated target vision needs to be locally elicited or collectively co-created (McPhearson et al., Reference McPhearson, Iwaniec and Bai2016). Weaving together place-based conceptualizations of sustainability and of transformative change can mitigate the risk of continued colonial imaginaries and dependencies (Lam et al., Reference Lam, Hinz, Lang, Tengö, Wehrden and Martín-López2020; Tengö et al., Reference Tengö, Hill, Malmer, Raymond, Spierenburg, Danielsen, Elmqvist and Folke2017).

4.6 New quality criteria and rethinking impact

Traditionally, the scientific production of knowledge is evaluated through criteria such as reliability, replicability, and validity. As sustainability challenges have become more pressing, sustainability scholars feel increasingly obligated toward society, exploring options such as scholar-activism (Sandover, Reference Sandover2020), where researchers take an explicit political standpoint in their work, and have a social change agenda to influence policy or public opinion. Thus, to adequately capture and assess the impact and quality of transformative research, classic quality criteria must be revisited and complemented. In turn, implementing new or other standards of how we define quality of research may shift the focus of knowledge production to incentivize research that catalyzes sustainability transformation (see 4.1, 6.3). Similarly, transformative research demands rethinking impact in terms of more diverse and expanded metrics of success encompassing social outcomes such as improved community relationships, social learning or trust building (Bergmann et al., Reference Bergmann, Schäpke, Marg, Stelzer, Lang, Bossert, Gantert, Häußler, Marquardt, Piontek, Potthast, Rhodius, Rudolph, Ruddat, Seebacher and Sußmann2021; Davies et al., Reference Davies, Putnam, Ainsworth, Baum, Bove, Crosby, Côté, Duplouy, Fulweiler, Griffin, Hanley, Hill, Humanes, Mangubhai, Metaxas, Parker, Rivera, Silbiger, Smith and Bates2021; Fischer et al., Reference Fischer, Horcea-Milcu, Lang, Thale-Bombien, Abson, Apetrei, Clarke, Derwort, Dorninger, Duse, Freeth, Jager, Klaniecki, Lam, Leventon, Newig, Peukert, Riechers and Schaal2019; Karcher et al., Reference Karcher, Cvitanovic, Colvin, van Putten and Reed2021; Schneider et al., Reference Schneider, Giger, Harari, Moser, Oberlack, Providoli, Schmid, Tribaldos and Zimmermann2019a; Spiering & Barrera, Reference Spiering and Barrera2021) and personal outcomes such as consciousness expansion (Woiwode et al., Reference Woiwode, Schäpke, Bina, Veciana, Kunze and Parodi2021) and mindfulness (Wamsler et al., Reference Wamsler, Osberg, Osika, Herndersson and Mundaca2021).

To expand the notion of quality of scientific knowledge production, we are drawing on the following previous suggestions on the issue: In 2003, Cash et al. (Reference Cash, Clark, Alcock, Dickson, Eckley, Guston, Jäger and Mitchell2003) offered a first starting point for evaluating the effectiveness of scientific information in mobilizing action through three knowledge usability parameters: credibility (scientific adequacy), salience (relevance compared to needs) and legitimacy (fair treatment of divergent values and interests). Credibility and legitimacy were also brought forward by Hansson and Polk (Reference Hansson and Polk2018). Daedlow et al., Reference Daedlow, Podhora, Winkelmann, Kopfmüller, Walz and Helming2016 emphasized the role of social robustness or socially responsible research, while Shaw Reference Shaw2022 adds practical feasibility and moral justifiability. In the work of Mertens (Reference Mertens2009), quality criteria in transformative research include among others: authenticity (a fair presentation of all perspectives), deep understanding of the community, and critical reflexivity (high level of awareness concerning self and others, see also 4.4). Taken together, it becomes evident that transformative research is characterized by an understanding of quality that (a) focuses on whether and how research results are usable to contribute to sustainability transformation, and (b) puts special emphasis on the process of knowledge production (such as inclusivity and researchers' accountability [Hölscher et al., Reference Hölscher, Wittmayer, Hirschnitz-Garbers, Olfert, Walther, Schiller and Brunnow2021]).

5.1 Process- and output-orientation

The process-oriented nature of transformative research, with its place-based focus at the science-society interface often conflicts with the output-oriented and productivity-based reward system in academia typically seeking global relevance, high citation indices and number of publications (Fam et al., Reference Fam, Clarke, Freeth, Derwort, Klaniecki, Kater-Wettstädt, Juarez-Bourke, Hilser, Peukert, Meyer and Horcea-Milcu2019; König, Reference König2015). For example, the laborious stage of initiating knowledge processes in co-modes (see 4.2) is seldom acknowledged and documented, perpetuating an academic system that does not allow it space and time. Similarly, trying to fit evaluation and monitoring costs within the strict bounds of a research project proves to be challenging, despite opportunities for mutual learning. Tensions arise due to the process-oriented research strategy primarily aiming at (co-)modeling and inciting societal change, with the academic output being treated as secondary. In contrast, the output-oriented strategy primarily aims at publishable results, with societal change as a side effect. Although the two orientations can be complementary (Schacter & Toonen, Reference Schacter and Toonen2010), researchers often encounter the paradoxical situation where process-orientation hinders academic outputs because of administrative or bureaucratic constraints (Reid et al., Reference Reid, Fernández-Giménez, Wilmer, Pickering, Kassam, Yasin, Porensky, Derner, Nkedianye, Jamsranjav, Jamiyansharav, Ulambayar, Oteros-Rozas, Ravera, Bulbulshoev, Kaziev and Knapp2021), while output-orientation drives low quality engagement (Kirchherr, Reference Kirchherr2023). Transformative researchers are torn between needing to prove adequate contributions to science according to the expectations of their academic institutions and system, and wanting to engage in processes of creating change (Mitchell et al., Reference Mitchell, Cordell and Fam2015). Interestingly, funding context was found to affect achieving both academic and societal outcomes (Newig et al., Reference Newig, Jahn, Lang, Kahle and Bergmann2019).

This is a well-documented tension for many forms of collaborative knowledge production such as those employed by transformative research (Chambers et al., Reference Chambers, Wyborn, Klenk, Ryan, Serban, Bennett, Brennan, Charli-Joseph, Fernandez-Gimenez, Galvin, Goldstein, Haller, Hill, Munera, Nel, Österblom, Reid, Riechers, Spierenburg and Rondeau2022; see 4.2, Reid et al., Reference Reid, Fernández-Giménez, Wilmer, Pickering, Kassam, Yasin, Porensky, Derner, Nkedianye, Jamsranjav, Jamiyansharav, Ulambayar, Oteros-Rozas, Ravera, Bulbulshoev, Kaziev and Knapp2021). The specific form this tension takes does not solely derive from the co-modes of knowledge production (4.2), but from their interactions with other characteristics such as normativity (4.4) and reshaping power (4.5), which push researchers to tailor their methodology to the situated on the ground knowledge and values assumptions. While creating additional strain on the researchers, such science-society collaborations do not enable proper academic compensation and recognition, often forcing them to choose between creating either academic or societal impact. Collaborating with non-academic actors has a positive influence on societal outcomes and a negative one on academic outputs and citations (Newig et al., Reference Newig, Jahn, Lang, Kahle and Bergmann2019). Conversely, Chambers et al. (Reference Chambers, Wyborn, Ryan, Reid, Riechers, Serban, Bennett, Cvitanovic, Fernández-giménez, Galvin, Goldstein, Klenk, Nel, Österblom, Bednarek, Bennett, Brandeis, Charli-Joseph, Chatterton and Pickering2021) demonstrated that within science-society collaborations solely pursuing knowledge production negatively correlates with obtaining outcomes that inspire action, such as empowerment, building networks, or process learning. However, carefully designing collaborative process phases and sequenced methods of knowledge integration has the potential to weaken the trade-off between academic and societal impacts (Newig et al., Reference Newig, Jahn, Lang, Kahle and Bergmann2019).

5.2 Accountability toward society and accountability toward science and the scientific method

Transformative sustainability research problematizes the question of researchers being accountable to society or to the institution of science, invigorating the debate about the aim of science in society (Gibbons et al., Reference Gibbons, Limoges, Nowotny, Schwartzman, Scott and Trow1994; Kitcher, Reference Kitcher2011). Some argue for strict accountability toward the scientific method, against adapting the presented data and models to socio-political realities (Geden, Reference Geden2015). Others point to the novel roles opening for researchers when embedding knowledge production in processes of social change, such as ‘honest brokers’ of knowledge (Goodrich et al., Reference Goodrich, Sjostrom, Vaughan, Nichols, Bednarek and Lemos2020; Karcher et al., Reference Karcher, Cvitanovic, Colvin, van Putten and Reed2021; Pielke, Reference Pielke2007), process and dialogue facilitators (Bulten et al., Reference Bulten, Hessels, Hordijk and Segrave2021; Horlings et al., Reference Horlings, Romero, Pisters and Soini2020), or even change agents (Wittmayer & Schäpke, Reference Wittmayer and Schäpke2014) and co-conspirers (Temper et al., Reference Temper, McGarry and Weber2019). Similarly, researchers now have the option to publish their results with open access, following the proliferation of ‘transformative’ journals. Although salutary, such efforts toward open science pose equity problems when considering how diverse funding availability and institutional requirements are for the various researchers across the globe. These intricate responsibilities, limited options, yet diversified roles may lead to identity conflicts (Kump et al., Reference Kump, Wittmayer, Bogner and Beekman2023), and leave researchers in uncharted territory where mainstream academic training is not always helpful (Nature Editorial, 2023).

The accountability issue is also tightly connected with perspectives on the roles of values in research (see 4.4). Transformative modes of research explicitly recognize that values play a role in the ethical assumptions underlying research and researchers' epistemic choices (Douglas, Reference Douglas2000; Kincaid et al., Reference Kincaid, Dupré and Wylie2007), and in the normative scenarios guiding a system's trajectory (Stojanovic, Reference Stojanovic2021, see also 4.4; Weiland et al., Reference Weiland, Bleicher, Polzin, Rauschmayer and Rode2017). This stands in radical opposition to the traditional academic adherence to the ideal of scientific objectivity (i.e. splitting facts from values) and commitment to normatively neutral academia (Schneider et al., Reference Schneider, Kläy, Zimmermann, Buser, Ingalls and Messerli2019b). The new roles proposed for sustainability and especially for transformative researchers, ultimately based on a normative orientation of their work and an arguable diminishing of the empirical part of their research, exemplify this tension starkly. The accountability tension leads to a paradoxical situation where the engagement of transformative researchers in brokering social processes reduces the role of empirical methods in the scientific practice and increases the importance of other competences and skills (Leicht et al., Reference Leicht, Heiss and Byun2018; Redman et al., Reference Redman, Wiek and Barth2021). Yet, their focus on accountability toward science risks producing socially irrelevant knowledge (Wuelser, Reference Wuelser2014). The two-fold accountability toward science and society is a challenge also known to other forms of engaged research, especially in the transdisciplinary community (Sellberg et al., Reference Sellberg, Cockburn, Holden, Lam, Sellberg, Cockburn, Holden, Lam, Cockburn and Holden2021). Transformative or transdisciplinary researchers often need to ‘walk a tightrope’ between these two extremes in order to survive and prosper in university institutions.

5.3 Methodologies rooted in scientific traditions and post-normal methodologies

Both previous tensions are tightly correlated with the methodological tension between traditional scientific methodologies, based on a Cartesian specialization and compartmentalization of knowledge, and post-normal systemic approaches characterized by contested facts, urgent decisions, and high stakes (Funtowicz & Ravetz, Reference Funtowicz and Ravetz1993). Indeed, in most pressing sustainability problems emergent properties and non-linear system dynamics cannot be reduced to (nor deduced from) its subunits (Mitchell, Reference Mitchell2009). This challenges a reductionist view and the idea that science seeks to explain phenomena by accounting for its micro-structure, where explanations ‘flow from below’. In transformative research, methodological tensions thereby arise from the necessity of a paradigm shift, leaving the scientific methodology built according to disciplinary rules and inherited model templates validated across generations of scholars, to test methods appropriate for contemporary societal problems. The problem with not making this paradigm shift is twofold: (i) often researchers want to accomplish the aspirations and aims of transformative research with the traditional tools and methods of mode-1 science or by overlooking some of the transformative characteristics (4.1–4.6); and (ii) transformative research is assessed and evaluated using the quality criteria suited for mode-1science (see 4.6). However, researchers engaging in transformative research also have to strike a balance between being interdisciplinary by blurring disciplinary boundaries, and epistemic trespassing into methodologies or topics they are not equipped to deliver (Ballantyne, Reference Ballantyne2019).

This multilayered methodological tension contributes to another paradox intensified in the case of transformative sustainability research by the systemic integration of post-normal features into a methodological framework which, unlike other socially engaged approaches, challenges the traditional science in all six key features (section 4) ideally at the same time. Although there is a growing need for and proliferation of transformative research approaches, many of them are difficult to evaluate according to traditional quality criteria or fall short when scrutinized against established scientific standards (see 4.6). Consequently, researchers grapple with aligning the implementation of transformative research designs with the existing institutionalized science and its proven reliability, sometimes at the expense of purposeful innovation. Conversely, attempts to stay within the validated institutionalized science are proving increasingly unreliable and inappropriate to tackle sustainability problems (Mitchell, Reference Mitchell2009), and to meet expectations of ‘solving them’ which society increasingly puts forward as a new main aim of the institution of science (Saltelli et al., Reference Saltelli, Ravetz, Funtowicz, Benessia, Funtowicz, Giampietro, Pereira, Ravetz, Saltelli, Strand and van der Sluijs2016). In sum, transformative research seeks to transform knowledge systems (Fazey et al., Reference Fazey, Schäpke, Caniglia, Hodgson, Kendrick, Lyon, Page, Patterson, Riedy, Strasser, Verveen, Adams, Goldstein, Klaes, Leicester, Linyard, McCurdy, Ryan, Sharpe and Young2020), and therefore may not be rewarded by an academic and funding system designed around the knowledge systems that need to be transformed.

6.1 Moving forward together: researchers

First, in line with the collaborative ethos of transformative research, the formation of communities of practice of transformative researchers (e.g. the Transformative Learning Hub at the Wageningen University, the Working Group Transdisciplinarity for Transformation of the Programme on Ecosystem Change and Society) can help them share their own experience regarding the allocation of scarce resources between output and process as per tension 5.1. Setting such communities of practice encourages peer learning by comparing transformative work across case studies, approaches and teams (Balvanera et al., Reference Balvanera, Daw, Gardner, Martin-Lopez, Norstrom, Speranza, Speirenburg, Bennett, Farfan, Hamann, Kittinger, Luthe, Maass, Peterson and Perez-Verdin2015; Cundill et al., Reference Cundill, Roux and Parker2015). Coming together around transformative research facilitates access to insights on practical tools and steps used by peers in this still emerging research mode. Expanding these communities to welcome disciplinary academics and non-academic participants interested in transformative research builds bridges between transformative research and other research modes. Examples of spaces and communities that are already developing a transformation science norm are the Sustainability Transitions Research Network, Action Research Plus (AR + ), and the Transformations Community (2023), with some of them also providing resources for conferences to bring people together. These communities can become change agents of the current academic institutions and funding schemes by creating spaces that co-exist parallel to more traditional forms of knowledge creation. Further, communities such as the TransformERS (COST, 2023) network specifically seek to bring together the ‘old’ (traditional) and the ‘new’ (transformative) toward a common aim of sustainability transformation.

Second, making transformative research more reliable and accountable toward science and society also lies within the hands of researchers and project managers (tension 5.2). The recent eroding of quality control mechanisms across sciences, specifically of replicability, are strong arguments against methodologically loose disciplines and research modes (Peterson, Reference Peterson2021). What rigor and quality mean in doing transformative research is not yet established (see 4.6). However, researchers can look into, carefully select, and adapt those suitable criteria available from various research designs, disciplines, or modes. Emerging efforts, such as the Coalition for Advancing Research Assessment (COARA, 2023) working group on transdisciplinarity, can contribute to creating evaluation criteria for transformative research. For rigor in transformative research, Mertens (Reference Mertens2009, p. 195) suggests starting from the notions of internal validity and borrows criteria used to evaluate qualitative research, such as ‘prolonged and substantial engagement and persistent observations’, third party debriefing, or self-reflection on own subjectivity. Maintaining a design and protocol for planning, tracking and documenting phases in the transformative process (see also Newig et al., Reference Newig, Jahn, Lang, Kahle and Bergmann2019) refers to external validity (Mertens, Reference Mertens2009). A transformative project used a formative accompanying researcher as a way to study the quality of the collaborative modes of knowing and learning (Freeth & Caniglia, Reference Freeth and Caniglia2020). Such protocols could be published using platforms with transparent peer review processes (e.g. Open Research Europe).

Third, becoming aware that to do transformative research, a methodological paradigm shift toward mode-2 and post-normal methodologies, away from mode-1 science is critical. Being able to distinguish and depending on the task at hand chose between, for example mode-1 and mode-2 science, is another step forward in addressing tensions between methodologies rooted in scientific traditions and post-normal methodologies (Horcea-Milcu, Reference Horcea-Milcu2022, see also tension 5.3).

6.2 Moving forward together: universities

Seeing universities as actors and places of and for transformative change illustrates their potential manifold role: As the place where research is undertaken and new knowledge is being produced and critically discussed, as the place where future change agents are educated, and also as a place with its own impact and sustainability performance. Tackling the three tensions mentioned above touches upon universities' multiple roles: Tension 5.1 between process- and output-orientation can be addressed by universities through a recognition system adapted to transformative research. While the need to improve how researchers and their work are evaluated beyond the productivity-based reward system was already recognized in 2012 by the San Francisco Declaration on Research Assessment (DORA), more and more bright spots of innovation are emerging (Chambers et al., Reference Chambers, Nel and Hille Ris Lambers2020). Notably, Utrecht University has recently abandoned easy-to-measure citation metrics such as impact factors and h-indexes to evaluate its researchers. Researchers are now assessed based on other standards, including societal impact and efforts to promote open science (Woolston, Reference Woolston2021). While some researchers hail this change in the culture of the evaluation system, others feel more uncomfortable and are, for example, concerned that researchers at Dutch Universities may not be competitive internationally (Open letter ScienceGuide, 2021). Importantly, such a change in the reward system provides room for transformative researchers to be recognized based on the process of transformative research rather than on output only.

Tensions regarding the double accountability and methodological choices (tensions 5.2 and 5.3) can be addressed by deliberately incorporating the developments of transformative research into the curriculum of higher education, thus responding to a shift in requirements and expectations of universities as active actors (Evans et al., Reference Evans, Jones, Karvonen, Millard and Wendler2015) within societal transformation. Promising approaches can be found on the level of individual courses and curriculum development as well as at the university-wide level. A whole-institution approach (Kohl et al., Reference Kohl, Hopkins, Barth, Michelsen, Dlouhá, Razak, Abidin Bin Sanusi and Toman2022), as put forward by the UNESCO (2020) Education for Sustainable Development roadmap, aspires to transform the learning environment as a whole, which opens up a space for experimentation and empowerment on campus that also brings sustainability teaching closer to the science-society interface. Project- and experience based learning settings, as e.g., developed within the Glocal Curriculum (John et al., Reference John, Caniglia, Bellina, Lang and Laubichler2017), support the development of students as informed and responsible citizens in an interconnected world. Finally, international collaboration networks among universities are generally posited to accelerate the spread of transformative characteristics concerning both teaching and research (Withycombe et al., Reference Withycombe, Arnim, Lang and Yokohari2016).

6.3 Moving forward together: funding agencies

We envision a science landscape where the transformative and conventional forms of research co-exist, and where the diversity of the funding and institutional mechanisms reflects this co-existence. The various modes of knowledge production and use (e.g. Table 1, mode-1, mode-2 science) although different in terms of knowledge processes, are all needed to inform and complement each other to support sustainability transformation. However, in the current academic landscape, system and descriptive-analytical knowledge, as well as producing knowledge for the purpose of accumulating knowledge is disproportionately represented compared to transformation knowledge and producing knowledge for intentional change. Correspondingly, mainstream funding requirements dominate the academic scene, with less arising funding options for transformative research.

National and EU funding schemes now formalize the need to work with actors beyond academia to shape the research focus. For example, research funded through the Dutch Research Agenda (NWA) takes place and is co-designed in consortia consisting of researchers and societal partners from the public and private sector. Similarly, in the current EU funding program (Horizon Europe), there is an expectation that practice partners have been involved in shaping the research need and designing impact into the project. Such approaches fulfill the purpose of normalizing a multi-actor or participatory approach to research. However, considering tensions in 5.1 and 5.2, they also ingrain the expectation that this level of consultation and co-design occurs outside of the funded project. Moving the initiation of a multi-actor research intervention or collaboration outside of the funded project carries three risks: (1) that it is turned into a box-ticking formality that is rushed and therefore not meaningful; or the opposite (2) that it becomes exploitative of the time and energy of researchers and stakeholders who are not compensated or recognized for their efforts; (3) taking into account success rates, a very likely failure to secure funding after a laborious unfunded initiation phase may discourage researchers and stakeholders from engaging in future transformative endeavors. To counteract these risks and tensions step-change approaches are available. To address tension 5.1, seed funding may be suitable to cover the transaction costs of initiating knowledge processes in co-modes that can be fully deployed in future applications. The seed funding time period would also allow to link to established theories and methodologies and thus to tailor proposals to existing funding panels. This would perhaps increase the perceived legitimacy of submitted transformative projects, but could also dilute or co-opt the transformative elements as they are taken over by the dominant scientific paradigms or discipline-related industries. To tackle tension 5.2, there is an opportunity to organize in addition to scientific panels, supplementary mixed panels consisting of citizens, practitioners, private stakeholders, and early-career researchers, which could evaluate proposals for relevance to their lived experiences.

The methodological tension 5.3 highlights the lack of (pre-identifiable) fit of transformative researchers to evaluation panels. Transformative research usually transcends disciplinary boundaries and responds to emerging knowledge, rather than testing theories or hypotheses within existing disciplines. The standard of most funding agencies (e.g. German Research Foundation, European Research Council) is to evaluate proposals within disciplinary panels. Thus, a transformative research proposal addressing biodiversity loss would be evaluated by either an ecology panel or a sociology panel, both of which will often deem it as being outside of the scope of the discipline and/or not tied to the cutting edge of their field. Changes are being made in this direction, such as the establishment in 2021 of the ERC evaluation panel SH7 (‘Human Mobility, Environment, and Space’) which includes sustainability science. Sustainability science was included under one of the ERC evaluation panels as of 2014. A draw-back to this approach is that it ‘others’ transformative research into a single panel, rather than creating space and allowing transformative research to be a central component in various research communities. Further, this ‘othering’ disconnects transformative research from the other disciplines rather than positioning its contribution with other theoretical, methodological, and empirical disciplinary knowledge contributions. Such holistic bringing together of knowledge represents what is truly transformative and meets the calls for knowledge plurality (Caniglia et al., Reference Caniglia, Luederitz, von Wirth, Fazey, Martín-López, Hondrila, König, von Wehrden, Schäpke, Laubichler and Lang2021; Turnhout et al., Reference Turnhout, Metze, Wyborn, Klenk and Louder2020). Ways forward to tackle tension 5.3 may be represented by separate funding streams for transformative spaces (Pereira et al., Reference Pereira, Frantzeskaki, Hebinck, Charli, Scott and Dyer2020) or for living labs (Schäpke et al., Reference Schäpke, Bergmann, Stelzer and Lang2018a). These spaces would enable long-term collaborations between transformative researchers and other actors and act as a potential remedy against project short-termism, research biases, and academic output focus. Moreover, based on their research design, proposals submitted to disciplinary panels could fulfill the expectation of being relevant to these transformative spaces.