Balancing ambiguous understandings of when systems tip: Concepts,agency,and analogies

Abstract

We have followed with great interest the responses to our article in Dialogues on Climate Change concerning societal tipping points in public debate, as well as the reflections offered on their relationship to climate tipping points and their implications for society. In this response, we underscore the strengths of the four responses our article received and critically assess whether early opportunity signals can be applied to transformations in social systems, and whether such systems can be modelled in ways analogous to physical climate systems. Finally, we outline potential directions for future research on societal tipping points.

We have followed with great interest the response to our article in Dialogues on Climate Change (Hansson et al., 2025) on societal tipping points in the public debate and reflections on their relation to climate tipping points and their impact on society. Our starting point was the observation that relatively little empirical research exists on how these concepts are used analytically, how they are communicated to the wider public, and how they are ultimately interpreted and shape individual behaviour and public policy. By analysing how tipping points were discussed in public debate between 2020 and 2024 across several major international news outlets—particularly with regard to how the relationship between natural and societal tipping points was portrayed—we provided a foundation for examining the concepts’ ambiguous relationship to societal change, agency, politics, and the means for transformation.

The use of these concepts has accelerated in both the social sciences and the natural sciences over the past 15–20 years. We were not the first to point out that, in several contexts, there is considerable confusion concerning both the interpretation and analytical application of the concepts (e.g. Kopp et al. 2025), nor the first to observe an inflation in what the concepts are used to describe (e.g. Milkoreit, 2023). As pointed out by both Nadeau (2026) and Russil (2026), our main contribution was instead the empirical observation and reflection on how the balancing of natural tipping points with societal tipping points is portrayed as a race—and the political consequences of such a framing and the apocalyptic and eschatological language that is often part of tipping point discourse. Of particular importance, according to Nadeau (2026), was ourobservation that both despair and a sense of inevitability can be just as effective as denial in eroding the conditions for meaningful action. In other words, societal tipping points serve a rhetorical function to balance negative natural tipping points, which is a finding also supported in our forthcoming paper on tipping points in the Swedish public discourse (Hansson et al., under review).

In her response, Nadeau (2026) builds on our reasoning about whether the discussion and growing attention to tipping points in natural systems leads to resignation and a sense of doomism or fatalism, or whether it instead fosters confidence in what are described as positive societal tipping points that may counterbalance the negative ones. Nadeau (2026) emphasizes the importance of empirically studying the consequences of the ‘race metaphor’ and how it shapes public understanding, particularly the balance between notions of urgency and agency, which our paper does not address empirically. However, Russil (2026) further delves into how controllable societal tipping points are, and he advances the discussion by warning that authoritarian actors may mislead the public into believing that they alone can identify and manage supposedly desirable positive tipping points.

Similar to what Lederer (2025) emphasizes in his response, Nadeau (2026) also explains that reducing this social dynamic to a metaphor suggesting that there is an almost linear and ‘automatic or self-driving process’ or race between positive and negative tipping points, obscures a proper understanding of what is happening. It is precisely on this point that Lederer (2025) makes his major contribution. We agree with him that the public debate on tipping points provides few, if any at all, insights into the dynamics of societal change, and that the concept of social tipping point is not particularly useful as an analytical tool. As an explanatory model, he instead proposes shifting the analytical focus onto institutions, defined as ‘a set of rules, formal or informal, that actors generally follow /…/’. In line with our view, he also points out that institutionalization is a core aspect of transformative societal change. It is not a linear process, and experience shows that it can also encounter reversals as it is highly politicized, and the feedback mechanism can, as Geels and Ayoub (2023) explain in more detail, involve reactive sequences and feedback because actors can deliberatively respond with countermoves that change orientations. Lederer (2025) provides support for the argument by illustrating the complexity of India's, Brazil's and Indonesia's highly divergent developments in photovoltaics. In contrast to how Lederer (2025) interprets our paper and perspective, we actually share the view that the idea of a single tipping point triggering an irreversible trend presents a misleading picture of societal change. Thus, we find it worthwhile placing our shared perspective in conversation with that of Boulton (2025) instead, which is representative of research seeking to model social systems in line with physical climate systems, with the aim of turning the models into predictive decision-support tools.

Boulton (2025) makes a helpful distinction between the temporalities involved in what he terms early warning signals, which apply to natural and physical systems, and early opportunity signals (EOSs) that apply to social systems. But we find the same ontological slipperiness we named in our initial contribution to this dialogue in the portrayal of such EOSs. This slipperiness makes them difficult to engage with critically. On the one hand, in the example made by Boulton, EOS appear as a visualisation of a market response to external shocks and policy interventions. As such, as primarily a policy analysis tool, it is hard to disagree with its potential utility. On the other hand, though, Boulton also states the premise for such analyses as the ‘belief that social systems can be modelled in a similar way to the physical climate systems’ (p. 3). It is with this view that we must take issue, since we believe that the approach to social systems misrepresents the potential and importance of politics for sustained and profound societal change. We have two fundamental objections, which we assume also align with Lederer's (2025) perspective, to the latter view of EOS modelling.

First, in this portrayal, social systems appear as linear evolution, which is the case in the example given by Boulton (2025) from the UK electric vehicle secondhand market. There is a steady if somewhat uneven progress towards the supposed endpoint of 100% electric vehicle uptake, with spikes marking external events and the introduction of policies that accelerate that development. Certainly, such analyses could be informative for policymakers interested in the efficacy of their policies, but, contrary to what is implied in the example, they don’t reveal anything about the future. If we read Boulton correctly, the temporalities involved here are directly comparable to those of drought recuperation in the Amazon, foreboding an approaching tipping point. But as political events of the last year show with undesired clarity, there is no comparable accumulation process driven by inevitable social forces on the European or US electric vehicle markets (see also Lederer (2025) on photovoltaic development). The EOS framing is inherently reactive in its view of society, and doesn’t offer any substantial advice for what types of political interventions are needed to sustain or accelerate the change processes it identifies.

This relates to our second objection. The example given by Boulton (2025) is of a market that already exists. In other words, the ‘early opportunity signals’ that can be given from such analyses are not early at all, but rather post fact. The models used for that analysis would not have been able to detect an opportunity on, for example, the Chinese vehicle market in the 1990s. The ‘tipping point’ on that market came about through sustained industrial policy, or political institutionalization, if using Lederer's (2025) terminology, over decades that made the market come true, which is political action of a different order of scale than the policy interventions mentioned in Boulton's example (Thurbon et al., 2023).

We submitted our article to Dialogues on Climate Change not to present final results, but to contribute to the understanding of important social phenomena and processes, and above all to help foster a dialogue that will continue for a long time to come. In this continuing dialogue, one avenue to explore is those virtuous feedback loops that can accelerate transitions through early industrial policy interventions on immature technologies (Collett et al., 2026). Another point we could have pursued further in our original contribution is that while the ‘race’ metaphor implies equality between contestants/stakes, there is a gawping scale difference between the grand phenomena described as climate tipping points and those most commonly identified in the societal realm, for example, national case-studies of electric vehicles and photovoltaics. Finally, in our forthcoming paper (Hansson et al., under review), we argue that societal tipping points are narrative devices that highlight the power of social forces—both the combined impact of many individual actions and the momentum of collective movements—in driving responsibility and agency in climate action. However, this focus on the social also carries a risk: it may weaken the pressure on political actors to take responsibility, which could be the subject of further studies.

The responses have placed several of our arguments in new contexts and, in some cases, challenged our interpretations and thereby taken the analysis further. Taken together, we believe this collection of articles offers a more nuanced understanding of societal tipping points than we could have achieved on our own.