Manuscript of the talk given at the launch of Koi Tū: The Centre for Informed Futures, Wellington, 19 March 2020
First, I want to express my gratitude to be part of launching Koi Tū, particularly here at Te Papa, which I perceive as being the very heart of Aotearoa. Second, I think that the topic which I am given to talk about is indeed a central topic when it comes to deliberate about our informed futures. In our modern world we cannot hope to run a country or to devise futures for us and our children unless we are attentive to the voice of science. But what or who exactly is that voice, what weight shall it be given, and why is it so difficult to figure out what it says?
Let me start – rather immodestly I admit – by quoting myself from a publication in 1995: “Our theories are simply not designed to deal with complex singularities, but rather with simple generalities.”  In 1995 I wrote this about the relationship between basic and applied science, but now I would claim this to apply to the relationship between science and policy in particular. It is one of the few things I would still claim today. It is perhaps an awkward sentence, but it captures, I believe, the essence of what I want to say about science for policy: the science may be the very best, but still not fit to provide solutions to problems out there in the real world. The crucial term here is complexity, and this also relates to post-normal science . Let me explain.
Once science believed that if we just can dissect reality into its parts and analyze their relationship with each other (reductionism), then we can have control of the whole, we can manage reality. Now we have learned that this works only for some aspects of reality, though we use this extensively in technology. But many real systems are characterized by an inherent complexity. Complexity is not the same as being just complicated. Complicated we can normally sort out if just use enough time and effort to understand complicated relationships. But reality is full of complex systems even where we did not expect them. Shallow lakes  are an often cited example from ecology of systems that, for instance, under nutrient loading may suddenly change their state into eutrophication without a predictable tipping point (nonlinearity), and that recovery from the turbid state requires far more resources than was needed for avoiding that state in the first place. Cities are seen as examples of complex systems, which is systems where the autonomous elements, i.e. people, organize from bottom-up which may result in new and emergent patterns on a different scale. This is how inner-city dynamics can create unexpected property values, hot-spots for criminal activity, or ghettos. A consequence of this is that top-down management and planning has clear limitations or is even impossible in the last instance. Complex systems have a self-organization from the bottom up. The whole system or parts thereof may change to a different state of emergent property at unpredictable thresholds.
What I want take from this is that many if not most of the systems we are interested in and which define important parts of our life are essentially full of uncertainty. We do not have – and assumedly cannot have – the insights which would give us robust tools to plan the future of the system. We have to live with these imperfections of our knowledge. The system uncertainty is, of course, the first important element in the post-normal framework of science for policy. A corollary of this uncertainty is that even in science there will be disputes about what some perceive as facts and others as mere speculation or hypotheses. This dimension of not being able to resolve all matters of fact by science is another dimension of post-normal science.
Now, what happens if we combine the natural or artificial systems with human systems (as we already indicated when mentioning cities)? To what extent are we and our actions, preferences, values and desires complex? Are we really this unpredictable?
The perhaps troubling news is that we are surprisingly predictable in many ways. Basically, this is how pre-election polls make a living. They actually predict quite well impressively often. Collective action is often well predictable, e.g. in panic situations. Some social scientists  studying mobility via mobile phone data concluded that humans are in general 93% predictable. Algorithms on internet platforms are pretty good in predicting individual user’s likes and preferences. Spontaneity and novelty happen not so often in our lives.
But the good news is that it does sometimes! People step out of their routines and predefined pathways, and chose a different track. We decide we need change and we opt for a different act, preference or even life. These choices can be the small but hard choices in life, as e.g. to stop smoking, or the really existential choices, like leaving all our past behind and starting afresh. Most of us know examples of this from the world literature. Tolstoy’s Anna Karenina did the unexpected and left everything behind for pursuing her true love. So did Nora in Ibsen’s Dollhouse. Or take Hans Castorp’s decision to leave the sanatorium in order to join the war in Mann’s The Magic Mountain. Like Hamlet it can even be the choice between life and death, i.e. suicide. The upshot is that humans are essentially unpredictable because we assume free will at the bottom. And at this point it is all about meaning and content in our lives. We need to be able to identify with the goals we are pursuing in our daily lives, with the values that we embrace. When those goals and values clash too much with our reality, we may decide to step out. If we do not do this, we suffer. We want our values to guide our actions.
This goes in reality back all the way to antiquity. Even in Greek philosophy it was noted that our actions can be explained as the combinations of three different factors:
It is this world of multiple values that provides a further challenge to our planning of society and the socio-economic developments in it. Individuals display what I call different value landscapes, and the diversity of these value landscapes in society is not harmonized, often they are in outright conflict with each other. In fact, we developed democracy in order to manage these different values landscapes. Respect for the autonomy of the individual values is fundamental for our modern democracies. In civil society (and in so-called deliberative democracies) these values need to negotiated in the public arena. And this now brings us to the next important element of post-normal science, the value dimensions.
Let me then introduce what is seen as the core mantra of post-normal science:
This view has been called post-normal because it implies stepping out of the usual academic quality assessments (peer reviews) of normal sciences. No longer can we leave it to a narrow range of disciplinary experts to provide us with the sole guidance on a pressing societal issue. The issue is complex and the solution lies within democratic society rather than top-down expertise. Typically the problem is also “wicked”  in the sense that there is not even an agreed upon problem formulation for policy, and there is no stopping rule for when we have found a solution.
Let me now address the central question of this talk: who are the experts? Obviously, the arena of people and voices who claim to be experts is rather crowded. Even those who are travelling with fake news and misinformation normally claim to have some special expertise in the field they are talking about. What we cannot deny is that all of those voices, ours included, are experts on our own value landscapes, and thus we have a democratic right to be heard.
It is, though, a different matter with claims to any factual insights. Here we need to ask how robust this claim is, and if its validity has been subjected to intersubjective tests of reliability. We do not want to accept all claims at face value. I want to stress that this is not necessarily looking for the big truth with a capital T. It is about finding out what works. This often weeds out a lot of noise, also noise that is embraced by politics, for instance a US president. But it also may turn up some alternative epistemologies, for instance indigenous belief systems, which offer insights, and some of them we may find useful even if we do not share the whole belief system. This is the same with science: we need to identify those widely shared robust insights which work, while we often can suspend judgement on their larger abstract and theoretical underpinnings.
As socially responsible scientists, what are we supposed to do then? Well, we are not to impose ourselves as the sole and only trustworthy experts on a subject, simply because we normally never are. Remember that even the best expert is a non-expert of most issues most of the time. And most pressing issues are complex, implying that they always transcend the boundaries of our neatly organized scientific disciplines. Yet, we should identify the pressing societal issues and the dominant framing of these issues. Then we should assess where our special insights and expertise fit in, and contribute this, possibly adjust it according to demands. If we can contribute a good draft of the system we need to understand, and the (causal) relations in it, this is mostly of great help. This is then part of what we now call a transdisciplinary approach. It is then crucial that we not only communicate what we think we know, but also what we do not know, and where the underlying uncertainties are in what we believe we know. Communicating uncertainties is crucial not only for the decision maker who may have to rely on the knowledge, but is also crucial for building up the trust among recipients and publics. Our integrity and trust that we can build up over time is key to how our contribution will be received. In addition, we need to open up for the implicit value dimensions related to the issue at hand. Scientists often withdraw from this discussion by saying “oh, this is not my field.” I believe this is wrong, as the scientists as everybody else are guided by values. Science is, in the end, not value-free. We need to make those values more explicit, and in fact this is a task where for instance some social science or philosophy could help.
But an important aspect of the post-normal science view is that we are sitting around the same table, with all other societal actors being present as well. Here we have to demonstrate the strength of our insights, and we have to show that we indeed can listen and learn from the others as well. We have to be self-reflective and realize our own (science’s) constraints and shortcomings. There will always be many constraints on what constitutes good solutions, resulting from different considerations, and there will ongoing disputes. Thus, my plea about expertise is double sided but without ambiguity: we from science need to acknowledge that we do not have automated priority claims when it comes to expertise; we have to earn the trust and we need to engage in the transdisciplinary effort, with all the uncertainties that follow with it. But we also need to prepare and package the insights that we do have, so that our knowledge becomes a constructive element in the common effort to meet our big societal challenges. Perhaps we can afford to leave some of us behind to the comfort of the ivory tower and the disciplinary silos. Many of them have done good work there. But as a whole, the scientific community needs to take up the challenge and become a part of our common effort to design informed and sustainable good futures for us and those who come after us. This is a moral appeal.
1. Kaiser, M. (1995). “The Independence of Scientific Phenomena”, in: Theories and Models in Scientific Processes, eds.: William Herfel et al., 179-200, Amsterdam: Atlanta.
2. Funtowicz, S. O. & Ravetz, J. R. (1993). “Science for the Post Normal age”, Futures 25: 739-755. https://doi.org/10.1016/0016-3287(93)90022-L
3. Scheffer, M. & van Nes, E.H. (2007). “Shallow lakes theory revisited: various alternative regimes driven by climate, nutrients, depth and lake size”. Hydrobiologia 584, 455–466 (2007). https://doi.org/10.1007/s10750-007-0616-7
4. Barabási, A.-L. (2016). Network Science. Cambridge: Cambridge University Press.
5. Rittel, H. W. J. & Webber, M. M. (1984). “Planning problems are wicked problems”, in: Developments in Design Methodology, ed.: Nigel Cross, 135-144, Chichester: John Wiley.
Professor Matthias Kaiser is an international expert in how science, society and policy interact. He is the former Director of the Centre for the Study of the Sciences and Humanities at the University of Bergen, Norway. He has studied at the Universities of Munich, Oslo, Stanford and Frankfurt. Matthias is an expert in technology assessment, the ethics of science and the philosophy of science, and has served as director of the Norwegian National Committee for Research Ethics in the Sciences and Technology, advising government, parliament and the scientific community.
Koi Tū Director Sir Peter Gluckman has known Matthias for many years and offered to host him (generously funded by a Norman Barry Trust visiting fellowship). Matthias will spend a significant part of his two year sabbatical during 2020 and 2021 at Koi Tū. He wants to contribute to the development of Koi Tū using his expertise on the science-society relationship, in particular science and ethics, and food ethics, and share his experience with participatory science methods. In return, he expects to learn more about the complexities in science-policy relations, and about the integration of Māori and Pasifika cultures and epistemologies in New Zealand.
In his spare time, Matthias has written a thriller and plays electric blues bass and is a self-described hard-core liberal European.