Claes Andersson

Associate professor, Department of Space, Earth and Environment, Physical Resource Theory.

My research focus lies on the long-term and large-scale evolution and organization of human societies. Presently, I work mainly with the deep human past, opinion dynamics and economic geography.

We work on several levels – from specific case studies up to conceptual, modeling and methodological development. The basis is the presence of dynamical systemic features that are common to just about any adaptive dynamical systems: animal culture, early hominin culture, modern culture, biological organic evolution – even how we distribute our activities today across geographical and social space.

The origins and evolution of human societies is key to understanding the world of today. Our cultural and biological features evolved together over periods of time that are vast compared to historical times. Culture makes us human – but we are also very specifically adapted to a cultural life, with a unique suite of genetic adaptations that permits us to operate in such a setting.
Building an improved understanding of this is key to understanding things like how nature and nurture should be delineated and how their interplay should be understood. But also to address very fundamental and even existential questions about what – if anything – it is to be human as opposed to just an animal. Although the empirical evidence is growing dramatically, we still lack a unified understanding of how humans evolved.

I also take an interest in complex systems more fundamentally: What type of system are we talking about? What features do problems and solutions in such a system have with regard to finding, understanding and dealing with them? Research has become more and more tightly integrated with policy over the past decades, but we are still quite inexplicit about how systems and problems under study are or should be conceived. Society is complex – nobody would object to that. But what is it to be complex? How does the complexity of societies compare with the way in which, say, bird flocks and computers are complex systems? What does this tell us about what methodologies and tools we should be using or developing? About the types of goals that we may have when trying to control societal system? What about unknown by-effects, for example?

What I see as key to the type of complexity we see in modern large-scale human societies is a general lack of clear scale separation. The dynamics of generated higher levels of organization unfold on timescales that overlap with those on lower levels. There is in the general case no way to make a clear cut across cultural sub-systems and isolate them for controlled study (such as by means of models). The rules of the game change on similar times scales that the game itself changes.

We call such systems, and others like them, “Wicked Systems” (see Societal Systems: Complex or Worse below) – and to understand them we think we need a constructively critical attitude to current approaches in science, one where one takes care not to throw the baby out with the bathwater!"
Please visit the database Chalmers research, to see which research projects Claes participates in. 
Like so many others at our department we are interested in what has been characterized as “sustainability transitions”: how we can bring about transitions in society – technological, economic, social and so on – to practices that do not undermine their own existence by their very operation. But this is a highly challenging task. Compared with “naturally occurring” transitions, sustainability transitions are economically and politically “uphill reactions”. Their benefits are diffuse, hard to quantify and reside in the future. Often their benefits are even negatively defined: they consist in averting some disaster that we have not and should not produce an example of. Their costs on the other hand are focal, immediate and highly quantifiable.
Are present policy instruments – taxes, subsidies, certifications, regulations and so on – up to this task? They may or may not be. In any case they were never designed to tackle this sort of problems, and they were developed under a paradigm where societal systems have been described as relatively intuitive and well-behaved systems, responding in regular ways to our attempts at steering them. But it is easy and cheap to complain – and no doubt we can do much better with existing tools than we are doing today. But to really solve these problems, we think that we will need tools that better reflect how the system that we aim to control really works. If you want to control a system it is generally a good idea to understand how it works, even if you in the end have to make strong assumptions about it.

Complexity science is widely seen as bringing some hope in this regard. It allows us to consider dynamical systems in much more detail and it has added a number of new concepts to our vocabulary. But we think that complexity science must be developed in some new directions, and that is what we are mainly focusing on. We start from the observation that although complexity science contributes a crucial new set of tools, it has proven very hard to move from the social to the societal level.  Why is it so hard to understand societies from a complexity perspective? Why is it so hard to understand them from any perspective really? This question is the red thread through our work. Why do different approaches work in certain settings and not in others? What are societal systems “really like”; if we allow this philosophically naïve way of putting it to pass?

When the scope of complexity science is described, this is often done by introducing a distinction between “complexity” and “complicatedness”: if a flock of birds is a complex system, a space shuttle is a complicated system. These two system qualities are also often referred to, respectively, as dynamical and structural complexity. So is society a complex or a complicated system? We think it is of utmost significance that both standpoints are equally easily argued. Societies are both complex and complicated. What happens when these qualities are mixed? Can we understand such systems by combining tools for understanding complicatedness (e.g. systems theory) with tools for understanding complexity (e.g. agent-based simulation)? We think this is not a trivial question. The combination between two qualities is by no means necessarily a simple concatenation. It must be expected to be emergent; i.e. new and qualitatively different from its components. This observation, or its potential significance, appears to have escaped the community.

We refer to this combined quality as Wickedness, seeing it as the culprit of what has since long been known as “wicked problems” in systems theory and operational research. We also see it as a crucial bridge concept that can teach us more about that elusive but strong similarity between ecological and societal systems; ecological systems are also clearly both complex and complicated and are very challenging to understand.

Key to understanding wickedness we believe is to understand how they evolve! The qualities of complexity and complicatedness mutually maintain one another in Wicked Systems. The multi-level systemic structure is realized and maintained by complex dynamics, involving large numbers of entities, and this structure all the time acts as a scaffold for this complex dynamics. They are characterized by radical innovation and its dynamics effects, which includes both powerful lock-ins and violent cascades. They also lack proper beginnings and they are not cyclical (at least not in this sense). When did society begin? Regardless of when we arbitrarily choose to draw the line, the beginning will be from a contingent, complicated and very specific initial state. By contrast, an engineered artifact, or an organism, has a clear “assembly phase” where it develops, followed by a stage where its makeup does not change dramatically. Also, the types of systems that complexity science has excelled at understanding begin from a simple initial state.

Our primary approach is to combine elements of theory and models that bring out this Wickedness; many of which are to be found in evolutionary developmental approaches to biology. Our work is mainly of a basic science nature, in the sense that we are striving to set the stage for better design and policy tools. In the process we are striving to also contribute directly, but we think there is a lot of foundational work to be done.  Among the most directly useful immediate consequences is probably that we can better explain why certain approaches would work and not work in certain situations.

New articles:

  • An evolutionary developmental approach to cultural evolution”, Claes Andersson, Anton Törnberg, Petter Törnberg. Accepted for publication in Current Anthropology, 2013
  • Societal systems – complex or worse?”, Claes Andersson, Anton Törnberg, Petter Törnberg. Under review with Research Policy, 2013
  • Fidelity and the emergence of stable and cumulative sociotechnical systems”, Claes Andersson. Accepted for Publication in PaleoAnthropology, 2013





Published: Thu 09 Jan 2020.