Looking at the various subjects that are discussed in this latest Issue of E:CO, I am struck by the variety of topics that are under study and therefore how complexity thinking is being applied to such a wide range of subjects. Looking back, I can see how my own areas of study have ranged over (too?) many topics. Originally, as a physicist I had the good fortune to be given a Royal Society European Fellowship to continue my PhD work in physics on disequilibrium phenomena – at Brussels University with Ilya Prigogine, a founding father of complexity ideas. This was a marvellous opportunity and I was free to explore whatever topics I wanted. Indeed, when Prigogine was given some research support from the US Department of Transportation to explore the possible mathematical modelling of the spatial co-evolution of transportation and urban structures, I was the one whose task it became to develop this research. So, every three months I had to head off to Boston to present our research. Each time I went there with our slowly developing ‘complex systems’ urban model, people there would be totally preoccupied with some new potential problem in US cities (recovering from riots; building metros; what happens if Chrysler goes belly up? Etc.). And each time I replied that we were trying to build a generic model of urban system/transportation, which would potentially be applicable to almost any problem in any city. A bold statement. But this did point out the difference between our kind of ‘meta-level’ approach and that of building a specific model of one particular case, with all kinds of detailed information for the case in hand. This also brings out the difference between 1) a ‘model’ that is really just a ‘description in data’ of what has happened, with predictions given by an extrapolation of the trends within the description and 2) a model built on the interaction of different agents, each with their own knowledge and aims, leading to possible futures for the system – including perhaps some unexpected emergent features. The original connection of this work to Prigogine and his team had been made by Dave Kahn, a researcher at the Transport Systems Centre in Cambridge, Boston. Thanks to this initial support and that of the excellent team there, we had a great opportunity to try to show that Complexity was not only fun, but also was and would be of great importance. It provided an excellent opportunity to try to see how complexity ideas, and dynamical models could be applied to important areas.
Through this initial break away of my work from physics, I was able over the next decades to work on all kinds of topics involving ecological and human systems. These included an amazing variety of topics: Urban and Regional planning; National and Regional social, economic and demographic evolution (Brussels, Belgium, USA, Senegal); River Basins (Soane, Rhone, Meuse); family therapy; Canadian Fisheries, fishing fleets and fishermen; Origami; social insects; economics and market dynamics (Airline competition, Car markets, Pharmaceuticals); ecology and evolution (evolving predator-prey, Darwin’s Finches etc.); organizational evolution (the car industry, Aerospace); supply chains (Cars and Aerospace); innovations (evolution of markets); designs(aerospace); and others I have since forgotten.
I accepted System Dynamics as a sensible way of approaching any problem area (it seemed better to try to model the connected effects of different elements rather than to study the elements separately) – but to go one step further. The key step that complexity added was to recognize that the ‘system’ itself could potentially redefine itself, evolve and change – qualitatively – creating new variables, new mechanisms and new emergent features and characteristics. In all the different topics of research listed above it is the underlying Complexity that links them all. Any system at a given moment has emerged from a past in which it was not what it is now. Complexity is about evolutionary emergence of structure and form. This involves ‘learning’ and ‘forgetting’ not just functioning – recognizing changed features and elements, requiring perhaps changed values, aims and goals. Complexity admits that the ‘functional structure’ may change – and life is not just a mechanical system running forward in time! Thank goodness! Instead of just having ‘a’ system – the initial one – we are looking at possible structural instabilities in which new variables and functionalities may emerge. Complexity can reveal possible futures into which the system could evolve qualitatively over time, and can therefore suggest how we might intervene to push the system in a good direction.
We can use urban and regional models to explore how the spatial patterns, of housing, business and retail centers could develop and look at the advantages of different pathways for different players. Somehow, we need to get some social consensus over what can be considered a ‘good’ direction? The fisheries work provided an important breakthrough. Here, we found that the actual task of fishing required two different (almost opposite) activities. Firstly, that of acquiring new knowledge of where the fish were – discovery. And secondly – exploitation – using this knowledge to guide fishing trips and fleet distribution in order to catch the most fish in the most profitable way. Gradually though, it was realized that this ‘fisheries’ approach which explicitly introduced the discovery and exploitation phases as separate entities was seen to apply to all the different modelling arenas. At exactly the same time, and quite independently, a colleague in Brussels, Jean-Louis Deneubourg, discovered that ants required exactly these two kinds of behaviour (discovery through imperfect trail following and efficient exploitation by sufficiently adequate trail following) in order to find and harvest food. So, obviously, a fishing fleet needs to know where the fish shoals are if they are to plan their routes. But also, more generally, in some technology space, a designer needs to know where the underlying ‘demand’ may be. Under which product type do the (potential) shoals of avid consumers exist? So, for different types of possible product design for example, there will be different numbers of potential consumers, just waiting for the appropriate, dream design of clothes, car, vacuum cleaner etc. But of course, this will only become clear after products have been launched and bets have been placed! For markets in general, there will be different possible characteristics of products, and it is important to discover what type (qualities and cost) possible buyers will be attractive. Such an idea integrates the whole spectrum of issues from the social realities of who may buy a product, through the technical and technological realities about how the product may perform, to the qualities and styles that may attract different potential buyers.
The important point is that the design and success/failure of a product depends not only on the existing customers and producers that are in the ‘game’, but also potential players and buyers who are not yet participating. Clearly, the number of potential customers depends on how the market develops – are some firms obtaining considerable economies of scale, or are others acquiring ‘prestige’? Firms might be wary of who has not yet entered the market, and potential customers may be trying to assess the cost/benefits of buying the product which will depend both on what they consider important, and who they see buying the product. In other words, we have ‘reflexivity’, where the market (and its model) are affected not only by what is present and rational, but also by what is not present but might happen – particularly in the light of what the model is currently indicating. So, not only is the market affected by what potential players think might happen, but also by what a model of this appears to indicate will happen. I first came across this idea when I read ‘the Alchemy of Finance’ by George Soros in 1987. He particularly focused on financial markets as being under the influence of the beliefs about values, partly at least, affected by what any model was indicating. But really this idea is completely general and applies to almost all human systems. Many people don’t buy cars on the basis of a rational analysis of the material cost/benefits of speed, comfort, economy. In reality, they are very concerned by – how do I look in this car’? Or what does my shopping basket/clothes/jewellery say about me?
Because of this remarkable interaction between what people think and what people think other people think, and a model trying to capture what is going to happen, there is a remarkable fluidity. And this situation has given rise to the enormous advertising industry which is about shaping what people think about a product, and what they think other people will think about it. For the UK the advertising spend is more than £20 billion, (internet £8Billion, TV £5Billion…). Clearly, it is really big business to influence how people perceive products and the way the product may enhance or reduce how a buyer may be perceived by others! This seems a little far from what is often called a ‘free market’!
It is in fact a measure of the real complexity of all things human.
It is this fundamentally competitive view of human beings that introduces ‘competitive’ envisioning of possible actions, and how each action is judged on how ‘good’ or ‘bad’ it makes us look. The only way out is community. If we look at systems form the point of view of a community with genuine concerns for each citizen on the part of the others, then a model, or an action might be conceived by an agent in terms of what it does for the community, including the less fortunate.
Reflexive agents within systems can invalidate the assumptions that underlie the systemic projections and changes. Instead of just having the interactions of the agents within a system we now must include their reflexivity and motivations. But how much do they weight their own gains against that of the community? This is particularly important when we note that the current political philosophy is one that says that in ‘free’ markets, selfish actions lead to the ‘best’ outcome for all – that is, it supposedly leads to maximum profits for companies and maximum utility for consumers. But that would only be true in the absence of reflexivity, of worries about self-image and of appearances. Because of this we will find that people, organizations and societies may behave in ways that we cannot anticipate even with the help of complexity science. So, the message that arises out of this is that complexity leads to the recognition of an irreducible element of uncertainty in human systems and a real limit to knowledge about the future. Instead of simply finding that we move from answers based on a mechanistic view of the world, to answers based on a complexity view, we find that there may not be answers! We therefore need to adopt an exploratory and opportunistic attitude where we shall continually need to reflect on what we thought would happen and what is actually happening. Surprisingly perhaps, instead of saying that there is no point in modelling complex systems, it really makes modelling and reflection more vital than before. Reality can break away from any predicted trajectory. This means that instead of building models in order to predict the future, we need to model in order to detect emergence and creativity within the system when our current interpretive framework fails.
The complexity of the world leads to compulsive ‘post-hoc’- ery in many people’s attempts to assign causality to what seems to have happened. Now, suddenly, people say they can ‘explain’ the BREXIT referendum result and the Trump win, in terms of people that felt they had been ‘left behind’. But the people involved in the vote are many and various, and possibly simply wanted to ‘send a sign’ of disaffection to the world in general. The voters that caused the ‘upset’ probably have no coherent visions of what needs fixing or how! Probably this UK and US experimental jump into the unknown will be messy, and will have unforeseen consequences for both voters and those who organized it. In some of my early work on social evolution I often used ‘ignorance’ as a ‘mechanism’ that lead to novel explorations. I think we are about to witness a massive experiment of this kind. Let us hope that we survive our coming ‘learning’ experience. The application of Complexity to human systems, gives us new ways of exploring possible futures, and a range of evolutionary paths that could occur. However, it is also correct in warning us that unpredictable occurrences, can and will occur.