The Role of Knowledge and Learning in Sustainability Transformations

At a time where we are witnessing unprecedented effects of environmental pressures, as well as growing conflict and global geopolitical instability, addressing sustainability challenges remains an urgent concern. Knowledge is what enables us to act, as it helps us anticipate events, evaluate alternatives and collectively respond. Yet, this is increasingly difficult in a post-truth era and at the advent of AI technologies, where the authority of science and the possibilities of any shared truth are increasingly contested. Against this background, this dissertation set out to understand the role of knowledge and knowledge processes, including learning, in informing decisions and actions towards sustainability transformations. We do this in a cumulative dissertation format, via three distinct articles: a systematic literature review on knowledge-related concepts, an empirical study of knowledge systems as social networks, and an exploration of learning processes with an agent based model. Our approach scrutinizes how knowledge shapes decision and action situations, interacts with uncertainty and becomes entangled in structures of legitimacy and power. We make several contributions, both to the conceptual advancement of the field and to understandings of knowledge-to-action pathways. Within sustainability science, we observe that the meaning of knowledge-related concepts varies widely across studies and scholarly communities, which impedes knowledge cumulation and theoretical advancement. In particular, this fragmentation prevents an empirical grounding of normative claims to knowledge co-production, so that the processes of how knowledge is mobilized, by whom, and to what effect, remain underexplored. Our work addresses these issues by clarifying key terms and frameworks (RQ1), uncovering the areas of theoretical scarcity around knowledge processes (RQ2), and offering new ways to analyse knowledge and learning in decision-making settings (RQ3). By examining knowledge and learning processes in two decision contexts, we uncover several mechanisms through which knowledge and action co-evolve: relational anchoring through information and knowledge exchange, trust-mediated cooperation for strategic action, and differentiated uses of knowledge and learning types aligned to specific uncertainties and decision goals (RQ4). We also highlight the importance of ecological and social feedbacks in shaping how knowledge gains legitimacy, is validated, and drives or fails to drive change (RQ5). Overall, our findings confirm that knowledge-to-action pathways are not linear, but adaptive, situated, and structurally mediated. At the science-society interface, these insights suggest that calls for co-production and stakeholder engagement must be accompanied by deeper reflection on how knowledge functions in contested, power-laden settings, including among scholarly communities within sustainability science itself. Epistemic and political pluralism must remain not just a methodological commitment within transdisciplinary processes, but a normative principle to reconcile multiple communities within sustainability science and practice. This includes upholding the distinction between knowledge and values, while recognizing that some normative claims can have rational, factual foundations. It also entails embracing the complementarities of various research modes, from positivist modelling to participatory inquiry and critical theory.