Social versus natural complex systems
By Greg Fisher
Last week I referred to the differences between natural and social complex systems in a blog article about FuturICT. I received an email from an esteemed colleague who questioned (paraphrasing) whether this distinction was useful and whether in so doing I was putting humans on a pedestal. In this article I want to expand on what I meant because I do think there are fundamentally important difference between these two types of system.
My argument centres on two main points: emergent domains & principles and imagination.
Emergent Principles
My first point is about the emergence of new domains and principles as the universe self-organises in various ways. In his book, Re-Inventing the Sacred, Stuart Kauffman described how new domains have emerged over the universe’s history (philosophers might use the term ontological domains). For example, the properties of molecules emerge from the combination of constituent atoms but, importantly, these properties cannot be traced back to those individual atoms – the properties are due to the combination of the atoms and they are genuinely new. Equivalently, the properties of organs emerge from constituent cells; and the properties of organisms emerge from constituent organs; etc. So some properties of human beings will be genuinely novel, and some properties of social systems will be too.
Kauffman emphasised that such genuinely new, emergent principles will not offend the laws of the known universe. Rather we might view them as additional to natural law and they will be authentically novel. Kauffman presented this argument to get away from the clockwork view of the universe and human nature, which would lead us to think that physics was the only real discipline and everything else merely applied physics. If this view of emergent domains and novel principles is true then the social sciences should be viewed as a genuinely important study of a particular ontological domain in its own right. It’s not just applied physics.
Imagination
My second point is about human cognition, which is an example of an emergent feature of human beings. The word that best captures this point is imagination. For those who might think this a little soft or vague, I am referring to the capacity of people to process information cognitively in order to consider hypothetical circumstances or scenarios, across space or time (or something like that). Imagination is an emergent property of our whole physiology (largely involving our central nervous systems). Let me drill in to this a little deeper and then discuss some of the implications of imagination for groups of people.
Perhaps the two most powerful functions or advantages of imagination are (i) the ability to consider being in different situations, including the capacity to empathise with other people (both intellectually and emotionally); and (ii) and the ability to project in to the future i.e. foresight. I am not being exhaustive: these are two of probably many advantages of our imagination.
Moreover, these two advantages can manifest themselves in terms of “we” as individuals and also “we” as whole groups of people. Imagination therefore gives us the capacity to contemplate collective action in the future. A simple example of this was when my colleagues Paul Ormerod and Bridget Rosewell built a model to consider the impact of Crossrail on London’s economy. This was in order to inform the decision of whether or not Crossrail should be built. What they did was imagine a model of London that included feedback effects (making it different in significant ways to the models used previously).
We might think of imagination as an emergent property of individual human beings. It is probably fair to say that what we imagine will be socially constructed but our ability to imagine results from our DNA and therefore evolution. One of the important phenomena that arises collectively i.e. within social systems, as a result of imagination is reflexivity. This concept is less commonly understood and discussed than it ought to be (in my opinion): it is a highly problematic and prevalent concept in social systems.
Reflexivity
Reflexivity arises when, for example, my decisions are contingent on my expectations of your future actions, and vice versa, simultaneously. Consider me trying to anticipate you anticipate me anticipate you anticipate me anticipate you… etc, to infinity. More concretely, picture bumping in to someone on the street – they move to their right and you move to your left, so you bump in to each other again, etc. You will need to communicate to resolve the reflexivity problem, allowing co-ordinated movement.
This (potential) problem holds not only bilaterally but in groups of people too e.g. my decision about how much to spend and save today will depend on my expectations of the future state of the economy, which will be determined by mine and everyone else’s actions. And everyone else is having the same thought, creating uncertainty all round. As I wrote in an article Reflexivity and Narratives, in many circumstances reflexivity is resolved by the emergence and perpetuation of narratives, as well as by norms.
Reflexivity is extremely important because in the face of uncoordinated action, it can give rise to less than ideal outcomes, which we might call collective action problems. When faced with reflexivity (even if we don’t know it), we might make decisions that end up being sub-optimal for a whole group, which then feeds back negatively on ourselves. This might happen even when our choices seemed sensible at the time. We have seen these types of outcome in Game Theory, notably in the Prisoners’ Dilemma game, so this is a well-understood point. (I do find it puzzling, however, that this has not enticed scholars to develop a well-considered, coherent theory of collective action. If I am wrong about this, please let me know.)
In the blog Reflexivity and Narratives I also noted that narratives can be socially damaging. The best example of what I am referring to here is a recession: reflexivity means that an economy will not necessarily gravitate to full employment equilibrium (there may be other reasons for this) as the neoclassicals would have us believe, and if a dominant recession narrative emerges then it will tend to be self-fulfilling.
Social vs Natural Systems
So I am arguing that natural systems do not include constituent parts that can process information in the same way that humans do. Some animals appear to have the ability to imagine to some extent but it seems much more limited than it is for humans. And the self-organising behaviour of groups like ant colonies can make it look as if the colony is able to imagine and anticipate but the process by which this is achieved is different to that of our central nervous systems. In any case, it is clearly not true that atoms and molecules imagine in the same way that we humans do. Positrons don’t stand around wondering what their neutron friend Bob is going to do next.
This is what I mean by social systems being different to natural systems. In writing such things it is important to be dispassionate and not to put humans on a pedestal for the purpose of self-flattery. But that does not mean social systems aren’t fundamentally different to natural systems. Avoiding self-flattery should not get in the way of us studying social systems in their own right. Of course, I should go further: different ontological domains should be treated with the same respect because they will all have their own unique, emergent features.
The Complexity Sciences in the Social Sciences
An interesting and important question arises from this point. If humans are different, should we apply concepts from the complexity sciences to social systems? Some people have argued that concepts like self-organisation and co-evolution emerged from the natural sciences, so by using them in the social sciences we will be mapping metaphors from one domain to another. This is an important critique because metaphor mapping can be extremely dangerous e.g. it is not too great a stretch to say that orthodox economics maps Newtonian mechanics in to the social sciences. (For an excellent read on the role of imagination in economics, I would recommend Richard Bronk’s The Romantic Economist, the sub-title of which is in fact “Imagination in Economics”).
My own view is that many of the concepts in the complexity sciences are equivalent to highly abstract patterns, which we might call general principles. And, furthermore, while different ontological domains involve different principles as I described above, they also involve common patterns too, in a fractal sort of way.
Let me use an example to illustrate what I mean. Suppose I defined self-organisation as being when spontaneous order emerges within a complex system. This definition can apply to atoms, ants and nation states. The definition is so abstract that it can apply very broadly and I believe we observe the same principle in multiple domains. So, yes, I think that the complexity sciences can be usefully used in the social sciences but we must tread carefully by being clear about the meaning of our terms.
Conclusion
Social systems are different to natural systems because they have their own genuinely novel features, one of which is how we process information. This leads to social systems behaving in fundamentally different ways. I imagine most people would agree.
The new(ish) field of ‘hybrid psychology’ (Rom Harre and others) offers a simple and very effective way to think about these distinctions. The starting point is that there is nothing in the universe save molecules and meanings. Molecules just are. Meanings are socially, interactionally and publically constructed. Confusion arises when one is mistaken for the other. Which is not to say that the two domains don’t overlap – of course they do.
I have blogged a little about this in reviewing Rom Harre’s latest book ‘Psychology for the Third Millenium’ at http://sfworkblog.wordpress.com/2012/11/30/book-review-psychology-for-the-third-millenium/
I think you’ve put your finger on a major issue here that is often not made explicit when ideas from complexity are applied to social systems. This has created a lot of confusion and miscommunication and we need more discussion at this level to clarify the issue. I think it was Popper who noted that since our actions in the world are strongly affected by our view or model of the world – and this changes with experience – then strongly predictive theories of the social world are logically inconsistent since they imply we can predict now how we will view the word tomorrow. Or to put it another way: if we can know now how we will view the world tomorrow then why are we not viewing the world like that now? Or to put it even more starkly – if you could produce a model that predicted the next theoretical breakthrough in physics then you’ve already made that breakthrough.
On the subject of coherent theories of collective action – well, there was the old classic by Mancur Olson “the logic of collective action” but that was from a purely rational action perspective. I would say that the more recent classic by Eleanor Ostrom “governing the commons” is the good start on this enterprise – based as it is on empirical work and self-organisation principles. What is true of both through is that importance of the group, how it is defined, where its boundaries are and how the behaviour of some relate others.