Disagreements are the bread and butter of science. What’s more, they tend to have positive consequences: scientific controversies usually motivate scientific rivals to gather new evidence, refine their theories, and reassess long-held assumptions.
But while controversies are normally good for science, it is no less true that, left on their own, many of them easily evolve into pointless exchanges where meaningful communication breaks down. After all, scientists are humans, and as such, we are prone to biases that ultimately influence the mood with which we approach disagreements. Particularly, emotional attachments to theories and motivated reasoning can favor what psychologist and Nobel Prize winner Daniel Kahneman called “angry science”, namely, a confrontational approach where proving the opponent wrong sidelines rational theory assessments. Angry science is the malaise of scientific controversies, contributing to their stagnation and preventing the attainment of their scientifically desirable outcomes. So how can we prevent it?
A promising solution is adversarial collaboration, that is, the approach where disagreeing scientists collaborate, with the help of a neutral arbiter, to move their controversy forward. In practice, this entails agreeing on the views each side espouses, collaboratively designing a study to test their competing hypotheses, and jointly publishing the resulting manuscript.
So far, adversarial collaboration has only been applied in psychology and neuroscience, but data suggest that it can indeed bring angry science to an end and ensure disputes remain constructive. So why don’t we apply it to evolutionary biology, where controversies abound?
This seems an a priori great idea, but the initial enthusiasm is hampered by the realization that evolutionary biology is not psychology, and that several of its characteristics may make the implementation of adversarial collaboration harder. Therefore, if this approach is to get a proper shot in our field, we should first identify what these problematic characteristics are and adapt the approach so that it can successfully deal with them.
A recent paper by Juan and Tobias (Gefaell & Uller, Trends in Ecology and Evolution, 2025) seeks to do exactly that, using for this conceptual tools drawn from the philosophy of science.
According to their analysis, to be successful, adversarial collaboration in evolutionary biology should effectively deal with three potential hindrances: the diversity of organisms studied in the discipline, the complexity of evolutionary change, and the conceptual nature of many of their disagreements. Each of these poses a problem for adversarial collaboration inasmuch as they threaten three of its core principles.
The diversity of organisms studied in evolutionary biology, for instance, hinders epistemic peerhood. For collaborations between rivals to succeed, participants should know more or less the same about the topic of dispute. Otherwise, the collaboration can be unbalanced, setting the competing hypotheses on unequal footing. The fact that disagreeing researchers use different organisms to investigate the question upon which they disagree can make ensuring epistemic peerhood difficult. A way to remedy this, Juan and Tobias suggest, is, for instance, seeking a suitable species that neither of the rivals has studied before to test their competing hypotheses.
The complexity of evolutionary change, for their part, makes it difficult to refute hypotheses once and for all, even if the collaboration unambiguously finds evidence against one of them. After all, a process might be important in one species, yet irrelevant in another. This feature does little to alleviate disagreements and goes against the falsificationist logic in which controversies are typically framed. To counteract this, Juan and Tobias argue, adversarial collaborations in evolutionary biology should explicitly be modest: rivals should agree beforehand that their collaboration would not constitute the final word on the matter, but only add relevant evidence for future discussions and assessments.
Finally, and perhaps most worrisomely, many controversies in evolutionary biology do not seem amenable to direct empirical resolution. For example, as Tobias has tried to show in previous works (e.g., Uller et al., 2020, 2024), the controversy surrounding the role of development in evolutionary theorizing is unlikely to be solved by empirical evidence alone, as it revolves around basic decisions concerning the way we model the evolutionary process. The problem with conceptual disagreements is that the adversarial collaboration approach was initially designed for empirical disputes only. As the arbiter of the first adversarial collaboration explicitly put it, for this approach to work properly, “the differences between antagonists cannot be too deep or too much at the philosophical level” (Latham, Erez, & Locke, 1988, p. 770).
But as Juan and Tobias argue, there is no convincing reason for not applying adversarial collaboration to conceptual disputes as well. For this, though, the key adjustment is shifting its aim from empirical advancement to dispute clarification. That is, “theoretical adversarial collaborations”, as they call them, should primarily seek to identify where rivals agree and disagree. To help figure this out, they suggest rivals focus their attention on three potential meta-scientific domains of disagreement: the ontological domain (that concerning very general beliefs about the nature of reality), the epistemological domain (regarding beliefs about the qualities of good scientific theories), and the semantic level (dealing with how we define concepts). Figure 1 shows a sketch of how a theoretical adversarial collaboration might work, according to Juan and Tobias.
Approaching conceptual disagreements from an adversarial collaboration perspective cannot only shed light on where the conflict really lies, but also, Juan and Tobias argue, promote mutual respect between rivals, thus helping establish the foundations for a healthier scientific pluralism.
Figure 1. Three-step model of theoretical adversarial collaboration. First, rivals jointly explore their disagreements at the ontological, epistemological, and semantic domains. Second, they reflect on the connection of these domains to their scientific hypotheses. Third, they articulate a defense of their respective hypothesis by arguing why their underlying meta-scientific bases are more warranted. (Taken from Figure 1 in the paper).
In sum, adversarial collaboration can effectively be applied to evolutionary biology, provided that some adjustments are made to the approach. Now it is up to the evolutionary biology community to decide whether it is worth it to put it into practice. But, irrespective of the outcome, this paper shows how the philosophy of science can inspire us to think of ways to improve the way we do science.
Read Juan and Tobias’ publications here:
Gefaell J & Uller T. 2025. From rivals to partners: adversarial collaboration in ecology and evolution. Trends in Ecology & Evolution: S0169534725003192. https://doi.org/10.1016/j.tree.2025.11.001
Other references:
Latham GP, Erez M & Locke EA. 1988. Resolving scientific disputes by the joint design of crucial experiments by the antagonists: Application to the Erez–Latham dispute regarding participation in goal setting. Journal of Applied Psychology 73: 753–772.
Uller T, Feiner N, Radersma R, Jackson ISC & Rago A. 2020. Developmental plasticity and evolutionary explanations. Evolution & Development 22: 47–55.
Uller T, Milocco L, Isanta-Navarro J, Cornwallis CK & Feiner N. 2024. Twenty years on from Developmental Plasticity and Evolution: middle-range theories and how to test them. Journal of Experimental Biology 227: jeb246375.